Can Ubuntu 14.04 not install on a new HD?

Downloaded ubuntu-14.04.2-desktop-i386.iso, and burnt a CD. To install on a new hard-drive, selected "Something Else" at the first "Installation type" dialogue. At the subsequent "Installation type" dialogue, the new drive was detected, and came up first on the menu of drives, already highlighted. Clicking "Install Now" gave: "No root filesystem is defined. Please correct this from the partitioning menu" Clicking "New Partition Table ..." gave: "Create new empty partition table on this device?", with options of "Go Back" or "Continue". The latter took a few moments, but did not progress to another dialogue. That left no option but to again click "Install Now", which promptly reiterated the "No root filesystem" gripe. Lacking any other buttons to click on that dialogue, I find myself in an endless loop. Is there any GUI way forward, or is it necessary to hack at the disk with parted and mkfs before an install attempt now? I recall partitioning drives during GUI installation on prior occasions (Ubuntu 5.10, 7.10, 10.04), so what can I have missed this time? It's too late in the day to try coffee. I'll have yet another go after posting - that sometimes works. Erik -- My folks came to the US as immigrants, aliens, and became citizens. I was born in Boston, a citizen, went to Hollywood and became an alien. (Leonard Nimoy)

2015-03-04 10:12 GMT+01:00 Erik Christiansen <dvalin@internode.on.net>:
To install on a new hard-drive, selected "Something Else" at the first "Installation type" dialogue.
If I get the point right, "Something else" in that context means that you want to create the necessary partitions by hand. The simplest choice in that menu should e something like "Use full disk", or something like that. I do not have a 14.04 iso at hand to try right now, thus I'm just guessing. -- Mick

On 04.03.15 10:19, Michele Bert wrote:
2015-03-04 10:12 GMT+01:00 Erik Christiansen <dvalin@internode.on.net>:
To install on a new hard-drive, selected "Something Else" at the first "Installation type" dialogue.
If I get the point right, "Something else" in that context means that you want to create the necessary partitions by hand. The simplest choice in that menu should e something like "Use full disk", or something like that.
There's nothing remotely like that, unfortunately. The dialogue given by "Something else" simply offers a selection of all the disk devices present, including the new drive. The "Partitioning ..." button does not do anything to either propose default partitioning or offer manual options. It does not even plonk a minimalist root partition, which would at least allow progression via the "Install now" button. (With the hope of further partitioning choices on the next dialogue.) At the first "Installation type" dialogue, the alternatives to "Something else" are to install over or beside the ubuntu distros found on the two old drives. That dialogue does _not_ recognise the presence of the new drive. If it did, I might not be forced down the only other path. One old drive is observably approaching end of life, and the other is also a decade old, so the new drive is very necessary.
I do not have a 14.04 iso at hand to try right now, thus I'm just guessing.
I have it, and I'm just as lost. GUIs and I don't get along terribly well, but 10.04 was much better. (Or was that 7.10 or 5.10 that was the best one? There were proper choices back then.) Erik -- Logic is the beginning of wisdom, not the end. (Leonard Nimoy)

On 04/03/15 21:22, Erik Christiansen wrote:
At the first "Installation type" dialogue, the alternatives to "Something else" are to install over or beside the ubuntu distros found on the two old drives. That dialogue does_not_ recognise the presence of the new drive. If it did, I might not be forced down the only other path.
Could I suggest that you disconnect your old disks, and ensure taht the new disk is in SATA1. It's been a while since I had a blank disk, but I thought that the something else appeared as the 3rd option (after take over disk or use blank space); but my memory may be corrupted by several different installs on VBox since. If your not wedded to Ubuntu, why not try Mint. The current version is based on Ubuntu LTS, with some updates. I don't want to start a war with this, but I have found Mint a little more user friendly (says he who has just install Debian testing on my laptop, primarily because I am tired of the upgrade process needed with Ubuntu and its clones.) All the best -- Keith Bainbridge keithrbau@gmail.com +61 (0)447 667 468

On 04.03.15 22:10, Keith Bainbridge wrote:
On 04/03/15 21:22, Erik Christiansen wrote:
At the first "Installation type" dialogue, the alternatives to "Something else" are to install over or beside the ubuntu distros found on the two old drives. That dialogue does_not_ recognise the presence of the new drive. If it did, I might not be forced down the only other path.
Could I suggest that you disconnect your old disks, and ensure taht the new disk is in SATA1.
That sounds like a simple way out. ISTR that the BIOS should allow me to select the boot priority of the disks. A consequence of installer ignorance of the older distros would be that I'd have to manually hack the grub menu to add those boot options. (I always leave the immediately prior distro conveniently accessible, because it usually takes me a month to add all the apps I need to the new one.)
It's been a while since I had a blank disk, but I thought that the something else appeared as the 3rd option (after take over disk or use blank space); but my memory may be corrupted by several different installs on VBox since.
You're right, but as stated, neither overwrite, nor use blank space on the old drives, does anything to install on the new drive. ;-) It's at the "Something else" dialogue that the new drive is first recognised, and the partitioning button there appears to do nothing useful.
If your not wedded to Ubuntu, why not try Mint. The current version is based on Ubuntu LTS, with some updates. I don't want to start a war with this, but I have found Mint a little more user friendly (says he who has just install Debian testing on my laptop, primarily because I am tired of the upgrade process needed with Ubuntu and its clones.)
I have Debian with LXDE on the laptop. It might be simplest to install the latest of that on the desktop. I think you've pushed me in the direction I should have followed from the start. I have just been looking at parted's mklabel command: LABEL-TYPE is one of: aix, amiga, bsd, dvh, gpt, mac, msdos, pc98, sun, loop and was wondering whether it is only with an msdos table that we have primary, extended, and logical partitions. This doesn't appear to elucidate: http://www.gnu.org/software/parted/manual/parted.html#Using-Parted I might look a bit further into that, because it would be nice to be able to do it on the command line again, to circumvent obstructive GUIs. Many thanks. Erik -- In the Original Vulcan: "Dif-tor heh smusma." Live long and prosper. (Spock)

Hi Erik, I came across the following "How-to" from Ubuntu support site. FYI. See if it can be of help? http://askubuntu.com/questions/343268/how-to-use-manual-partitioning-during-... Regards, Wen On Mar 4, 2015 9:22 PM, "Erik Christiansen" <dvalin@internode.on.net> wrote:
On 04.03.15 10:19, Michele Bert wrote:
2015-03-04 10:12 GMT+01:00 Erik Christiansen <dvalin@internode.on.net>:
To install on a new hard-drive, selected "Something Else" at the first "Installation type" dialogue.
If I get the point right, "Something else" in that context means that you want to create the necessary partitions by hand. The simplest choice in that menu should e something like "Use full disk", or something like that.
There's nothing remotely like that, unfortunately. The dialogue given by "Something else" simply offers a selection of all the disk devices present, including the new drive. The "Partitioning ..." button does not do anything to either propose default partitioning or offer manual options. It does not even plonk a minimalist root partition, which would at least allow progression via the "Install now" button. (With the hope of further partitioning choices on the next dialogue.)
At the first "Installation type" dialogue, the alternatives to "Something else" are to install over or beside the ubuntu distros found on the two old drives. That dialogue does _not_ recognise the presence of the new drive. If it did, I might not be forced down the only other path.
One old drive is observably approaching end of life, and the other is also a decade old, so the new drive is very necessary.
I do not have a 14.04 iso at hand to try right now, thus I'm just guessing.
I have it, and I'm just as lost. GUIs and I don't get along terribly well, but 10.04 was much better. (Or was that 7.10 or 5.10 that was the best one? There were proper choices back then.)
Erik
-- Logic is the beginning of wisdom, not the end. (Leonard Nimoy) _______________________________________________ luv-main mailing list luv-main@luv.asn.au http://lists.luv.asn.au/listinfo/luv-main

On 05.03.15 08:32, Wen Lin wrote:
Hi Erik,
I came across the following "How-to" from Ubuntu support site. FYI. See if it can be of help?
Many thanks, Wen. According to that, the secret GUI obfuscation appears to be that it is necessary to move the highlight from /dev/sda to "free space" before the GUI will do anything other than cryptically refuse to progress. I don't recall that "free space" was shown in my case, perhaps due to the lacking partition table. I could go back and check, if it now mattered much. My patience with ratruns with blank walls lacking discoverability is distinctly limited, and this just confirms my decades old antipathy towards GUIs. I'll make a note of the deficiency for future reference ... and then go back to doing it with parted, I think. That will be much less dangerous for my blood pressure, since any obstacle is discoverable, google-able, and amenable to grubbing about in manpages. Thanks again, for helping solve the mystery. Erik -- The meta-problem here is that the configuration wizard does all the approved rituals (GUI with standardized clicky buttons, help popping up in a browser, etc. etc.) but doesn't have the central attribute these are supposed to achieve: discoverability. That is, the quality that every point in the interface has prompts and actions attached to it from which you can learn what to do next. - Eric Raymond, in "The Luxury of Ignorance."

On 4/03/2015 20:12, Erik Christiansen wrote:
Downloaded ubuntu-14.04.2-desktop-i386.iso, and burnt a CD.
To install on a new hard-drive, selected "Something Else" at the first "Installation type" dialogue.
Don't do that. "Something else" should have been labelled "expert mode". Unless you have some prior partitioning knowledge, use one of the guided options.

On 04.03.15 23:41, Jeremy Visser wrote:
On 4/03/2015 20:12, Erik Christiansen wrote:
Downloaded ubuntu-14.04.2-desktop-i386.iso, and burnt a CD.
To install on a new hard-drive, selected "Something Else" at the first "Installation type" dialogue.
Don't do that.
I just tried it again, to confirm. It is the only way to gain access to the new drive. Selecting the first option: "Install Ubuntu alongside them" gives a dialogue headed by a single line device selector, showing: Select drive: SCSI3(0,1,0)(sdc) 40.0 GB ATA WDC WD400JB-00EN No type of mouseclickery on the RHS menu triangle thingy causes any of the other drives to drop down or replace the existing option on the line - 100% repeatable over many attempts. That appears to leave no alternative but to use the only alternative offered by the GUI ratrun.
"Something else" should have been labelled "expert mode". Unless you have some prior partitioning knowledge, use one of the guided options.
It's no big deal. The only problem with "Something else" is that I've been unable to make it progress to the next dialogue. It wants a root partition before proceeding, but offers no way to make one. There's nothing intrinsically wrong with "expert mode" - I've used it successfully in the past. (Though admittedly with a younger brain, and with older install software, not crippled by the intervening years of "improvement".) I've found a bit more info on parted, and might have a bit of a go with that. (But will in the end probably download a more recent Debian, and just install that. When I did that on the laptop, I had MSW blown away, and Debian up in about half an hour of bringing the lappy home.) Erik -- "Those who live by the GUI, die by the GUI" - Duncan Roe, on luv-main ML.

Just had this thought. Fire up gparted in ubuntu while still in trial mode and create whtever partitions you need. Then proceed with intallation On 5 March 2015 12:53:05 am AEDT, Erik Christiansen <dvalin@internode.on.net> wrote:
On 04.03.15 23:41, Jeremy Visser wrote:
On 4/03/2015 20:12, Erik Christiansen wrote:
Downloaded ubuntu-14.04.2-desktop-i386.iso, and burnt a CD.
To install on a new hard-drive, selected "Something Else" at the first "Installation type" dialogue.
Don't do that.
I just tried it again, to confirm. It is the only way to gain access to the new drive. Selecting the first option: "Install Ubuntu alongside them" gives a dialogue headed by a single line device selector, showing:
Select drive: SCSI3(0,1,0)(sdc) 40.0 GB ATA WDC WD400JB-00EN
No type of mouseclickery on the RHS menu triangle thingy causes any of the other drives to drop down or replace the existing option on the line - 100% repeatable over many attempts.
That appears to leave no alternative but to use the only alternative offered by the GUI ratrun.
"Something else" should have been labelled "expert mode". Unless you have some prior partitioning knowledge, use one of the guided options.
It's no big deal. The only problem with "Something else" is that I've been unable to make it progress to the next dialogue. It wants a root partition before proceeding, but offers no way to make one.
There's nothing intrinsically wrong with "expert mode" - I've used it successfully in the past. (Though admittedly with a younger brain, and with older install software, not crippled by the intervening years of "improvement".)
I've found a bit more info on parted, and might have a bit of a go with that. (But will in the end probably download a more recent Debian, and just install that. When I did that on the laptop, I had MSW blown away, and Debian up in about half an hour of bringing the lappy home.)
Erik
-- Keith Bainbridge keithrbau@gmail.com +61 (0)447 667 468 Sent from my Apad

On 05.03.15 08:45, Keith BAINBRIDGE wrote:
Just had this thought. Fire up gparted in ubuntu while still in trial mode and create whtever partitions you need. Then proceed with intallation
That is a good idea that I'll have to remember. (I'm booting 10.04 in between, though, to google up on parted, read mail, and bone up a bit on partition tables, since the msdos-ism of logical partitions within primary partitions on my old drives does seem to be an unnecessary historical artifact.) Am packing now to go bush for ten days from this afternoon, so execution of the good suggestions will be delayed a bit. I'll have a play with parted, even if I hop over to Debian. Any mess I make can be blown away by the Debian install, in the worst case, I figure. Thanks. Erik -- A computer is like an air conditioner, it doesn't work as well when you open Windows.

On 05.03.15 08:45, Keith BAINBRIDGE wrote:
Just had this thought. Fire up gparted in ubuntu while still in trial mode and create whtever partitions you need. Then proceed with intallation
Many thanks, Keith. That worked fine. However, booting Ubuntu 14.04.2 was interminably slow, eventually presenting only a _totally_ blank desktop, devoid of any menu or icon, and no apparent way to invoke an xterm. (Mouse clickies only offered a folder or a document) So from 5.10 to 10.04 seems to be the extent of my history with Ubuntu. Debian+LXDE on the laptop is giving me no grief, so now I just needed a more recent iso for the desktop. That needed some ingenuity. At http://mirror.internode.on.net/ I was unable to find any debian iso, and selecting i386 at https://www.debian.org/distrib/netinst#verysmall then selecting CDROM presents weird stuff. The first archive has a win32.exe in it! No iso. But a quick google gave me https://www.debian.org/releases/wheezy/debian-installer/ which at "netinst CD image" actually presented an iso, instead of leading back to the same weird stuff as at the netinst link in the previous paragraph, as the naming similarity threatened. Phew! The new "improved" installation methods being pushed, with their excess of instructions and complications seem an enormous step backwards. What's wrong with "whack it in and reboot"? Erik -- Anything can be impossible, given sufficient bureaucracy.

All else aside, I recently installed Jessie onto the blank area of my disk, and generally dual boots OSX (with rEFIt's help). -- Keith Bainbridge keithrbau@gmail.com +61 (0)447 667 468 Sent from my Apad On 19 March 2015 4:19:23 pm AEDT, Erik Christiansen <dvalin@internode.on.net> wrote:
On 05.03.15 08:45, Keith BAINBRIDGE wrote:
Just had this thought. Fire up gparted in ubuntu while still in trial mode and create whtever partitions you need. Then proceed with intallation
Many thanks, Keith. That worked fine.
However, booting Ubuntu 14.04.2 was interminably slow, eventually presenting only a _totally_ blank desktop, devoid of any menu or icon, and no apparent way to invoke an xterm. (Mouse clickies only offered a folder or a document)
So from 5.10 to 10.04 seems to be the extent of my history with Ubuntu. Debian+LXDE on the laptop is giving me no grief, so now I just needed a more recent iso for the desktop. That needed some ingenuity.
At http://mirror.internode.on.net/ I was unable to find any debian iso, and selecting i386 at https://www.debian.org/distrib/netinst#verysmall then selecting CDROM presents weird stuff. The first archive has a win32.exe in it! No iso.
But a quick google gave me https://www.debian.org/releases/wheezy/debian-installer/ which at "netinst CD image" actually presented an iso, instead of leading back to the same weird stuff as at the netinst link in the previous paragraph, as the naming similarity threatened. Phew!
The new "improved" installation methods being pushed, with their excess of instructions and complications seem an enormous step backwards. What's wrong with "whack it in and reboot"?
Erik

Erik Christiansen <dvalin@internode.on.net> writes:
At http://mirror.internode.on.net/ I was unable to find any debian iso,
They're under "debian-cd", not "debian": $ rsync -hhr mirror.internode.on.net::debian-cd/ | foldr grep -- .iso$ i386 netinst -rw-r--r-- 277.00M 2015/01/11 00:32:40 7.8.0/i386/iso-cd/debian-7.8.0-i386-netinst.iso -rw-r--r-- 485.00M 2015/01/11 02:22:46 7.8.0/multi-arch/iso-cd/debian-7.8.0-amd64-i386-netinst.iso I think this is an artefact of how mirror.ion does its mirroring.
and selecting i386 at https://www.debian.org/distrib/netinst#verysmall then selecting CDROM presents weird stuff. The first archive has a win32.exe in it! No iso.
#verysmall appears to be referring to the netboot (cf. netinst) media, for CD version of netboot look for "mini.iso", e.g. http://http.debian.net/debian/dists/wheezy/main/installer-i386/current/image...
But a quick google gave me https://www.debian.org/releases/wheezy/debian-installer/ which at "netinst CD image" actually presented an iso, instead of leading back to the same weird stuff as at the netinst link in the previous paragraph, as the naming similarity threatened. Phew!
That should've been from https://www.debian.org/distrib/netinst#smallcd (cf. #verysmall you mentioned). #verysmall = netboot = ~20MB (d-i kernel, ramdisk & bootloader) #smallcd = netinst = ~120MB (that + udebs (installer modules))
The new "improved" installation methods being pushed, with their excess of instructions and complications seem an enormous step backwards. What's wrong with "whack it in and reboot"?
AFAICT your main problem was clicking "Tiny" instead of "Small" :-) All the methods you mentioned above aren't new; they've been available since at least 2009.

On Fri, Mar 20, 2015 at 11:30:49AM +1100, Trent W. Buck wrote:
Erik Christiansen <dvalin@internode.on.net> writes:
At http://mirror.internode.on.net/ I was unable to find any debian iso,
They're under "debian-cd", not "debian":
$ rsync -hhr mirror.internode.on.net::debian-cd/ | foldr grep -- .iso$ i386 netinst -rw-r--r-- 277.00M 2015/01/11 00:32:40 7.8.0/i386/iso-cd/debian-7.8.0-i386-netinst.iso -rw-r--r-- 485.00M 2015/01/11 02:22:46 7.8.0/multi-arch/iso-cd/debian-7.8.0-amd64-i386-netinst.iso
Many thanks for the trail of crumbs in the maze. At the end of a bit of a google hunt, debian-7.8.0-i386-netinst.iso was the one I ended up with at the mentioned link. It's a bit over 290 MB. ...
That should've been from https://www.debian.org/distrib/netinst#smallcd (cf. #verysmall you mentioned).
#verysmall = netboot = ~20MB (d-i kernel, ramdisk & bootloader) #smallcd = netinst = ~120MB (that + udebs (installer modules)) ... AFAICT your main problem was clicking "Tiny" instead of "Small" :-)
I have NDI - just followed links till I got an iso, then ran with it. (Doesn't seem so tiny when it got here, but it goes, mostly. I'll start another thread with its most puzzling weirdness.) Erik

Erik Christiansen writes:
$ rsync -hhr mirror.internode.on.net::debian-cd/ | foldr grep -- .iso$ i386 netinst -rw-r--r-- 277.00M 2015/01/11 00:32:40 7.8.0/i386/iso-cd/debian-7.8.0-i386-netinst.iso
debian-7.8.0-i386-netinst.iso was the one I ended up with at the mentioned link. It's a bit over 290 MB.
The size difference is because you have ten fingers and I have only two. $ truncate -s277m test $ rsync -hh test -rw-r--r-- 277.00M 2015/03/23 10:20:43 test $ rsync -h test -rw-r--r-- 290.46M 2015/03/23 10:20:43 test

On Fri, Mar 20, 2015 at 11:30:49AM +1100, Trent W. Buck wrote:
Erik Christiansen <dvalin@internode.on.net> writes:
At http://mirror.internode.on.net/ I was unable to find any debian iso,
They're under "debian-cd", not "debian":
$ rsync -hhr mirror.internode.on.net::debian-cd/ | foldr grep -- .iso$ i386 netinst -rw-r--r-- 277.00M 2015/01/11 00:32:40 7.8.0/i386/iso-cd/debian-7.8.0-i386-netinst.iso -rw-r--r-- 485.00M 2015/01/11 02:22:46 7.8.0/multi-arch/iso-cd/debian-7.8.0-amd64-i386-netinst.iso
Many thanks for the trail of crumbs in the maze. At the end of a bit of a google hunt, debian-7.8.0-i386-netinst.iso was the one I ended up with at the mentioned link. It's a bit over 290 MB. ...
That should've been from https://www.debian.org/distrib/netinst#smallcd (cf. #verysmall you mentioned).
#verysmall = netboot = ~20MB (d-i kernel, ramdisk & bootloader) #smallcd = netinst = ~120MB (that + udebs (installer modules)) ... AFAICT your main problem was clicking "Tiny" instead of "Small" :-)
I have NDI - just followed links till I got an iso, then ran with it. (Doesn't seem so tiny when it got here, but it goes, mostly. I'll start another thread with its most puzzling weirdness.) Erik -- Rear Admiral Morisetti recalled that when commanding an aircraft carrier, it took a gallon of oil to move just 12 inches (30cm), while as many as 20 tonnes per hour were burned during a period of intensive take-off and landing. - http://www.bbc.co.uk/news/science-environment-15342682

Erik Christiansen writes:
Downloaded ubuntu-14.04.2-desktop-i386.iso, and burnt a CD. [problems detecting/configuring the disk]
NFI what ubiquity (the GUI installer) looks like these days, but IIRC I did a 14.04 install recently using debian-installer, and it worked the same way it always has. That's this: http://archive.ubuntu.com/ubuntu/dists/trusty/main/installer-amd64/current/i... (Change "amd64" to "i386" if you insist.)

I only have an ISO of 14.04.1 around, not .2, but the installer is probably similar enough. With two disks, both blank, the "simple" install mode (not "Something else") gives me an option to pick from either of the two disks. After installation on the first, a second attempt to install changes this -- the non-something-else option is now "Erase Ubuntu 14.04 and reinstall". Clicking "Something else" takes me into the GUI partitioner. Using this, I was able to "Create a new partition label" for the second disk. After that I made sure I had selected the second disk, then hit the "+" button to create a new partition. I selected ext4 and a mount point of / for the new partition. Hit apply, then continue, and ubuntu was happily installed onto the second disk. Not sure what was going on differently for Erik, but I couldn't trivially reproduce the problem. However I wouldn't be surprised if the partitioner was getting confused by an existing exotic partition layout. I don't get the feeling it's very robust. Suggestions: 1) Unplug other disks, or at least move them to low-priority slots, so that you're installing to your new disk with it as the highest priority item in the boot order. You were intending on replacing them shortly anyway, right? 2) If you really must leave things as they are, then try blanking the partition table (and MBR) on the new disk before use.
From a command line prompt (in existing linux, or trial mode of the ubuntu ISO), run dd if=/dev/zero of=/dev/XXX bs=1M count=1 where XXX is replaced by the id for the disk - probably sdb or sdc, but for the love of god, check that first. You can id the disks by running "fdisk -l /dev/XXX" first. (It's safe, it'll just print out some details and existing partitions for the disk)
3) Try a later version of Ubuntu, eg. 14.10, as the installer has probably improved a little. I don't think there's any point trying alternatives such as Kubuntu or Mint, as they just use the same base with different packages loaded later. -Toby Erik Christiansen writes:
Downloaded ubuntu-14.04.2-desktop-i386.iso, and burnt a CD. [problems detecting/configuring the disk]

On 05.03.15 11:25, Toby Corkindale wrote:
Clicking "Something else" takes me into the GUI partitioner. Using this, I was able to "Create a new partition label" for the second disk.
If I could somehow reach that point, I'd be home free. At no point was that option offered, so we'll have to work around it.
After that I made sure I had selected the second disk, then hit the "+" button to create a new partition.
I saw +/- buttons on the LHS, but they were greyed out. ...
Not sure what was going on differently for Erik, but I couldn't trivially reproduce the problem.
However I wouldn't be surprised if the partitioner was getting confused by an existing exotic partition layout. I don't get the feeling it's very robust.
Even parted says: (parted) p Error: /dev/sdc: unrecognised disk label for the new drive. I.e. I very much needed a "Create a new partition label" option - if it had been presented at any stage, I'd have leapt at it. Once I decide whether to go for a bsd or sun partition table (or even flop back to msdos, if there's trouble), I suspect that the installer might be able to take it from there. But since parted will sanitised partition boundaries if I specify partition sizes in % or GB, I can't see a lot of risk in just whacking in a boot partition, a swap, and one for the rest, using parted, so the installer has little or nothing to trip it up.
Suggestions:
1) Unplug other disks, or at least move them to low-priority slots, so that you're installing to your new disk with it as the highest priority item in the boot order. You were intending on replacing them shortly anyway, right?
Yup. They're a decade old, and one is occasionally wobbly at boot.
2) If you really must leave things as they are, then try blanking the partition table (and MBR) on the new disk before use.
From a command line prompt (in existing linux, or trial mode of the ubuntu ISO), run dd if=/dev/zero of=/dev/XXX bs=1M count=1 where XXX is replaced by the id for the disk - probably sdb or sdc, but for the love of god, check that first.
You did make me chuckle there. Decades ago I did "rm -rf /XXX/YYY", but fat-fingered a space after the first '/'. I suspect it takes precisely one such experience to make a hardened unix user.
You can id the disks by running "fdisk -l /dev/XXX" first. (It's safe, it'll just print out some details and existing partitions for the disk)
That confirms what we know: $ sudo fdisk -l /dev/sdc Disk /dev/sdc: 320.1 GB, 320072933376 bytes 255 heads, 63 sectors/track, 38913 cylinders Units = cylinders of 16065 * 512 = 8225280 bytes Sector size (logical/physical): 512 bytes / 512 bytes I/O size (minimum/optimal): 512 bytes / 512 bytes Disk identifier: 0x00000000 Disk /dev/sdc doesn't contain a valid partition table (I'd used fdisk -s to quickly find the device for the new drive, to use with parted, but hadn't tried the -l.)
3) Try a later version of Ubuntu, eg. 14.10, as the installer has probably improved a little. I don't think there's any point trying alternatives such as Kubuntu or Mint, as they just use the same base with different packages loaded later.
Sorry, kan't kstand Kubuntu. Will probably go with Debian, as I have it on the laptop. (Also might then not have to hoik out NetworkManager before being able to get the network up. :) Many thanks for trying to replicate, and the useful hints. I think this one is twice conquered now, since either parted or Debian will do the trick once I have enough consecutive moments to whip up a partition table, and run through the install. I could disconnect the old drives, for amusement, but I'm not going to work on a patch to fix the installer in any event, so I'd rather spend a little time trying out parted on an empty disk, in case I ever need to use it on a populated one. Erik -- Leibowitz's Rule: When hammering a nail, you will never hit your finger if you hold the hammer with both hands.

On 05.03.15 11:06, Trent W. Buck wrote:
Erik Christiansen writes:
Downloaded ubuntu-14.04.2-desktop-i386.iso, and burnt a CD. [problems detecting/configuring the disk]
NFI what ubiquity (the GUI installer) looks like these days, but IIRC I did a 14.04 install recently using debian-installer, and it worked the same way it always has.
That's this: http://archive.ubuntu.com/ubuntu/dists/trusty/main/installer-amd64/current/i...
(Change "amd64" to "i386" if you insist.)
Not being game to assume that amd64 would work on my old intel motherboard, I did hunt out i386 for the ubuntu install. The installer which came with Debian+LXDE, a couple of years back, was smooth enough - but the laptop drive already had MSW on it, so had a partition to hack. I'm not quite so keen to buy more GUI trouble when a few minutes with parted can give me root (boot), swap, and the_rest, partitions, in a more accessible way than the GUI. (I expect) Over a decade I haven't yet filled the old 40 GB drive (the twin is a mirror), so it's hard to go far wrong in a new 320 GB paddock, I figure. (An extra GB or two of root or swap is cheap insurance) Erik -- Why make things difficult, when it is possible to make them cryptic and totally illogical, with just a little bit more effort?" - A. P. J.

Erik Christiansen writes:
http://archive.ubuntu.com/ubuntu/dists/trusty/main/installer-amd64/current/i... (Change "amd64" to "i386" if you insist.)
Not being game to assume that amd64 would work on my old motherboard [...]
Pfft, that's not hard to check. Either look up the SKU (from /proc/cpuinfo) on https://ark.intel.com, or just try to boot an x86-64 kernel. The kernel will immediately print "this hardware is crap" if it won't work.

On 6 March 2015 at 10:56, Trent W. Buck <trentbuck@gmail.com> wrote:
Erik Christiansen writes:
http://archive.ubuntu.com/ubuntu/dists/trusty/main/installer-amd64/current/i... (Change "amd64" to "i386" if you insist.)
Not being game to assume that amd64 would work on my old motherboard [...]
Pfft, that's not hard to check. Either look up the SKU (from /proc/cpuinfo) on https://ark.intel.com, or just try to boot an x86-64 kernel. The kernel will immediately print "this hardware is crap" if it won't work.
Yeah, I'd been wondering about that too.. The last 32 bit mainstream Intel CPUs were the Pentium 4 range, from around 2004-2006, which is a ridiculously long time ago in computing terms. But then, Erik did mention the hard drives were 32 GB, which would date from even earlier than then. So he's operating on 10+ year old hardware? I'm actually surprised there's enough memory in the machine to run a modern desktop environment and web browser. A 2004 era machine probably only has between 0.1 and 0.5 GB of RAM. I think? Toby

On 06/03/15 11:35, Toby Corkindale wrote:
I'm actually surprised there's enough memory in the machine to run a modern desktop environment and web browser. A 2004 era machine probably only has between 0.1 and 0.5 GB of RAM. I think?
I have an old IBM Thinkpad that Strategic Data was going to throw away but gave to me instead. It has 320M RAM (yes, 64M internal plus an extra 256M) and runs LXDE Debian Wheezy just fine, including a current web browser. Cheers, Andrew

Sorry: Late reply, due to over a week spent netless. (Laying in firewood - winter is coming.) On 06.03.15 11:35, Toby Corkindale wrote:
Yeah, I'd been wondering about that too.. The last 32 bit mainstream Intel CPUs were the Pentium 4 range, from around 2004-2006, which is a ridiculously long time ago in computing terms. But then, Erik did mention the hard drives were 32 GB, which would date from even earlier than then. So he's operating on 10+ year old hardware?
The drives: yes. The mobo turned up its toes a couple of years ago, so I whipped in another - but the drives are still bigger than I need, and still spinning. (Everything worth backing up fits on an 8 GB USB stick.)
I'm actually surprised there's enough memory in the machine to run a modern desktop environment and web browser. A 2004 era machine probably only has between 0.1 and 0.5 GB of RAM. I think?
It's now a bit newer (VIA C7), and has a whole 1 GB, so no drama there. Erik -- Australia ranks 44th for average connection speed, according to The State of the Internet Report from cloud service provider Akamai. - http://www.abc.net.au/news/2015-02-09/remote-rural-australians-to-wait-anoth...

On Thu, Mar 19, 2015 at 03:38:41PM +1100, Erik Christiansen wrote:
I'm actually surprised there's enough memory in the machine to run a modern desktop environment and web browser. A 2004 era machine probably only has between 0.1 and 0.5 GB of RAM. I think?
It's now a bit newer (VIA C7), and has a whole 1 GB, so no drama there.
that's a 32-bit CPU from Sep 2006, nearly 9 years old. are you aware that you can upgrade to a modern machine for under $170? e.g. the cheapest current parts combo i can find at MSY today is: AMD A4-7300 64-bit dual-core CPU + GPU $58 ASRock A58M-HD+ motherboard $65 4G Kit DDR3-1333 RAM (2x2GB) $46 if the RAM in your current mb is DDR3 (it probably isn't), you could put off buying the new ram for a while, but 1GB isn't really adequate for a modern desktop any more. the machine will likely be swapping to disk by the time it finishes booting and you login even with a "light" desktop like xfce or lxde, let alone start running memory hogs like iceweasel or chromium. you've already got yourself a new drive but if you hadn't, you can get a 64GB SSD for $56 these days. small but much faster than any mechanical drive. FYI, comparison of the A4-7300 with the C7: http://www.cpu-world.com/Compare/887/AMD_A4-Series_A4-7300_%28JA%29_vs_VIA_C... note that the max power usage of the A4-7300 is 65W versus 20W for the C7 - but the A4-7300 includes a built-in Radeon R3 graphics processor. if your current graphics card is as old as your mb and cpu, then it probably uses at least 30-40W anyway, and isn't anywhere near as good as the R3....9+ years is a LONG time in the evolution of GPUs. also, FYI, the Intel G1840 (a dual-core 64-bit celeron with built-in graphics) costs $55 but the cheapest Haswell Refresh motherboards start at nearly twice the price of the cheapest AMD FM2+ motherboards....and Intel, as usual, sucks at upgradability. once again you have to discard your old m/b if you want to upgrade to the new generation of CPUs. AMD OTOH generally keep the same socket for at least two generations so you can keep upgrading your CPU without having the expense and waste of a new motherboard. (the asrock m/b above can take any FM/FM2+ CPU up to the current A10-7850K CPU, a quad-core 4Ghz CPU with Radeon R7 graphics for $199, and probably several future CPUs too) this upgradability issue is the main reason I haven't switched back to Intel CPUs. the latest Intel chips are undeniably better than the current AMD CPUs, but the initial expense of switching (new mb and new CPU, maybe new RAM) is much higher and intel's history with socket changes tells me I'd have the same high expenses in future every time i upgraded. my current machine is good enough for now, so i'll just wait until AMD releases an upgraded CPU i can swap in. craig -- craig sanders <cas@taz.net.au>

Quoting Craig Sanders (cas@taz.net.au):
that's a 32-bit CPU from Sep 2006, nearly 9 years old. are you aware that you can upgrade to a modern machine for under $170?
e.g. the cheapest current parts combo i can find at MSY today is:
AMD A4-7300 64-bit dual-core CPU + GPU $58 ASRock A58M-HD+ motherboard $65 4G Kit DDR3-1333 RAM (2x2GB) $46
FWIW, I've been involved in a similar discussion on my LUG mailing list here on the USA's left coast, and we've been talking about new high-performance x86_64 motherboard / AMD SoC combos supporting up to 32GB of DDR3 RAM for ridiculously low prices. And, oddly enough, my leading example of same is a current ASRock mini-ITX motherboard bundled with a quad-core AMD A4-5000 'Kabini' SoC (aka 'APU'). And some almost-as-good similar boards (with slightly less powerful SoCs are even fanless. http://www.newegg.com/Product/Product.aspx?Item=N82E16813157574 http://www.newegg.com/Product/Product.aspx?Item=N82E16813157518 http://www.newegg.com/Product/Product.aspx?Item=N82E1681313536 Intriguing if hardware-limited tiny box (maxes out at 8GB): http://www.newegg.com/Product/Product.aspx?Item=N82E16883218035 (Prices will need to be converted to AUG$ from funny-looking green stuff.)
also, FYI, the Intel G1840 (a dual-core 64-bit celeron with built-in graphics) costs $55 but the cheapest Haswell Refresh motherboards start at nearly twice the price of the cheapest AMD FM2+ motherboards....and Intel, as usual, sucks at upgradability.
Total agreement. The 2003 AMD 'Jaguar' (Kabini/Temash) SoCs beat the hell out of most Celeron/Atom/etc. stuff except the server-grade Atom variants that are probably way overpriced. AMD's successor, the 2004 'Puma' (Beema/Mullins) SoCs are intriguing but are very nearly unobtainium. And, also, where I find them at all, for some daft reason I find 8GB RAM ceilings (cf. 32GB) -- which I find odd. Example (currently out of stock, as they sold like hotcakes) is the top end of CompuLab's Fitlet range: http://www.fit-pc.com/web/products/specifications/?model%5B%5D=FITLET-GI-C67... That uses the quad-core AMD A10-Micro 6700T, which is a 'Mullins' APU. Other Fitlet variants with less-powerful Mullins SoCs, are still in stock. AMD's 2015 successor to 'Puma' in the low-mains-draw, high-performance market, is suppoed to be an architecture callled 'Carrizo-L' , the low-power variant of the 'Carrizo' x86_64 arch. Not seen at all. It's weird that the most compelling low-power, high-RAM, fast hardware on the market is based on low-budget AMD chips from 2003. A conspiracy theorist might suppose AMD is being strangled out of the market (again). Or that they cannot get OEM uptake because the OEMs opt for cruddy ARM alternatives' bottom-dollar pricing because they assume few customers care enough to buy something that doesn't suck. Or both.

Fixing the other unfortunate typo:
It's weird that the most compelling low-power, high-RAM, fast hardware on the market is based on low-budget AMD chips from 2003. A conspiracy ^^^^ '2013' theorist might suppose AMD is being strangled out of the market (again). Or that they cannot get OEM uptake because the OEMs opt for cruddy ARM alternatives' bottom-dollar pricing because they assume few customers care enough to buy something that doesn't suck. Or both.
Sorry, I'm jetlagged -- and now recaffeinating to compensate.

Craig -- I should hasten to add: The AMD A4-7300 64-bit dual-core CPU + GPU you were talking about is magnficient but is from AMD's higher-power-consumption 2014 'Kaveri' line of mobile-targeted SoCs that (as you say) fit in FM2 sockets. Kaveri is part of AMD's 'Steamroller' architecture, 2014 successor to 2013's 'Piledriver' (Trinity/Richland) SoCs, and is on 28nm dies. Kaveri appears to be the higher-power line in parallel to the lower-power 2014 'Puma' (Beema/Mullins) line I've been writing about along with the 2013 'Jaguar' (Kabini/Temash) one. Point is, a Kaveri-based system is high-performance indeed and no slouch even as a gamer box. But Kabini-based systems (even though it's an older and less powerful SoC line) is reported to be fast enough to be home-theatre PCs, and can be extremely cool / quiet and in some cases even fanless -- and sip tiny amount of mains power. I'm considering building a 32GB RAM 'Kabini'-based system on an ASROck mini-ITX motherboard to handle a variety of roles on my house's inside LAN. Because it's 2015 and because KVM virtualisation is a thing of much wonderment.

On Fri, Mar 20, 2015 at 04:38:33PM -0700, Rick Moen wrote:
I should hasten to add: The AMD A4-7300 64-bit dual-core CPU + GPU you were talking about is magnficient but is from AMD's higher-power-consumption 2014 'Kaveri' line of mobile-targeted SoCs that (as you say) fit in FM2 sockets. Kaveri is part of AMD's 'Steamroller' architecture, 2014 successor to 2013's 'Piledriver' (Trinity/Richland) SoCs, and is on 28nm dies.
yep, i'm sure it's possible, even easy, to find different or better CPUs for various optimisation needs, but my purpose today was to highlight the cheapest current combination of parts....a quick and dirty exercise that took me a whole 60 seconds of "research" (looking at the MSY price list :) if i was actually considering an upgrade myself at the moment, i'd spend a lot more time than that investigating the various options and coming up with the optimum parts lists for my needs at a price I consider reasonable. personally, i'm not that interested in the FM2/FM2+ series because I prefer separate graphics cards, usually nvidia (the proprietary nvidia driver is, unfortunately, still much better than the free software nvidia or radeon or intel GPU drivers), so my motherboard is an AM3+ model (asus sabertooth 990FX). At the moment i have an AMD 1090T Phenom II 6-core CPU, which i've had for several years now. I could upgrade to the 8-core FX-8350 but IMO it's not quite worth doing that for $245, so i'll wait to see what the next generation looks like. if it's good, there's a reasonable chance i'll be able to upgrade to it without having to replace my motherboard (and if i do have to replace the m/b then my win7 games box will get the current board & CPU as an upgrade).
Because it's 2015 and because KVM virtualisation is a thing of much wonderment.
yep more ram is good. multi-core 64-bit CPUs are great and virtualisation is very useful. having reasonably recent hardware and not having to stuff around with ancient crap is fantastic. craig -- craig sanders <cas@taz.net.au>

On Sat, Mar 21, 2015 at 11:04:49AM +1100, Craig Sanders wrote:
(the proprietary nvidia driver is, unfortunately, still much better than the free software nvidia or radeon or intel GPU drivers),
i should point out that this only matters to me because: a) while i do most of my gaming on a windows box i built out of spare parts, i still play some games on linux, either native games or using wine. b) vdpau video acceleration for mythtv (and smplayer etc). if i didn't need those two things then either nouveau or the free radeon driver would be more than adequate for a desktop machine. my cheap laptop has an AMD cpu and runs the free radeon driver...i've tried running the myth client on it and it's unwatchable, the motion is jerky and irritating like stop-motion animation. OTOH, it plays some steam linux games quite well. craig -- craig sanders <cas@taz.net.au>

On Sat, 21 Mar 2015 02:42:10 AM Craig Sanders wrote:
b) vdpau video acceleration for mythtv (and smplayer etc).
if i didn't need those two things then either nouveau or the free radeon driver would be more than adequate for a desktop machine.
my cheap laptop has an AMD cpu and runs the free radeon driver...i've tried running the myth client on it and it's unwatchable, the motion is jerky and irritating like stop-motion animation. OTOH, it plays some steam linux games quite well.
I've never had a problem running mplayer on ATI video cards with the free driver. My main desktop machine has a ~2500*1400 resolution screen and I play video full screen without any problems. So far the only problem I've had with playing video on a system with an AMD64 CPU is when playing 4K video. FullHD video works fine on NVidia cards with the Nouveau driver and on ATI cards. The only video problems I have on any of the systems I run concern a system with an NVidia card that won't run steam and which crashes when running Minecraft. -- My Main Blog http://etbe.coker.com.au/ My Documents Blog http://doc.coker.com.au/

Quoting Craig Sanders (cas@taz.net.au):
a) while i do most of my gaming on a windows box i built out of spare parts, i still play some games on linux, either native games or using wine.
b) vdpau video acceleration for mythtv (and smplayer etc).
if i didn't need those two things then either nouveau or the free radeon driver would be more than adequate for a desktop machine.
Thank you for clarifying that. I read your post and was trying to think of a way to express a politely different perspective based on different use cases without accidentally sounding like I was arguing. Unfortunately, on the Internet as I'm sure you know, tone and nuance are lossy - and many's the time I've tried to say 'I'm glad that works for you, notwithstanding which other things entirely have proved better for me based on other priorities', yet, people act as if you'd challenged the other chap to a duel. Your choices are absolutely sound for your requirements list. Of course. I've been known to archly quote, to (for example) gamers telling me that Nvidia proprietary drivers are indispensible on account of frame rate and perfect streaming HD video, Mr. Lincoln saying 'That's the kind of thing that will be enjoyed by those who enjoy that kind of thing', intending to sound broad-minded and good-natured -- but start a small war for my good intentions. (You might have heard of Mr. Lincoln. Funny-looking bloke, tall, gaunt, famous teller of stories and jokes, wore black and a stovepipe hat, held the American Presidency until it turned out his retirement plan was decidedly missing.) But I digress. I approached this subject because of telling my LUG I'd like to build or buy a back-of-house, inside-LAN Linux server to do at least a few things: (a) Puppet or Chef master (configuration management), (b) network IDS, (c) online backup target. Being in sleeping space, as far from my small home server farm as possible so the same fire or burglar won't likely get both, it ought to be quiet, low-power, and small. I'm enormously fond of future-proofing by spending enough to buy something I reasonably hope to still be excellent in 5, 10, 15, maybe even 20 years if I'm lucky and plan right. And one of the earliest lessons I learned is that going heavy on RAM (or at least capacity to expand it) is the best future-proofing, followed by good I/O, followed distantly by CPU. On my home LAN, network I/O matters less than it ought, because my home aDSL (static IP) has abysmal throughput for distressingly unclear reasons but we fear it's because my beloved family residence is at maximum distance from the telco central office where the DSLAM is -- a First World instance of the proverbial Last Mile Problem. I also feel slightly guilty about running colo-type rackmount gear still left over from my old employer VA Linux Systems that draws more mains power than a well-selected small server should. (Yes, my VA gear is absurdly old. I know that.) So, I admire the mains thriftiness of, e.g., the Pi, and would aspire to approach that. Anyway, after being crystal-clear about my needs and objectives, my LUG friends gave me advice: 1. 'Get a Raspberry Pi model B.' (This was before the current 2 model B.) Me: 'Um, 512kB? Nothing better than microSD and USB2? No eSATA/SATA/mSATA. Really?' 2. 'Get a Beagleboard Black.' Me: 'Um, 512kB? Nothing better than microSD and USB2? No eSATA/SATA/mSATA. Really?' 3. 'Get a large Intel 8-core thing and put it in a colo.' Me: 'Um, did you listen to _any_ part of the target requirements?' 4. 'Get a small Atom-based board. Put an SSD on it.' Me: 'That's getting there, but why does it have an 8GB RAM ceiling. Did I slip through a time vortex, or is it not 2015?' Eventually I learned that LUG friends completely ignore what your expressed hardware needs are, and burble out a recommendation of what they're familiar with, as ticking off their own wishlist criteria and not yours were what you asked. LUG friends; Can't take them anywhere, and throwing large buckets of icewater on them is frowned upon. ;-> So, I started re-researching modern hardware, having last been paid to be expert in Linux hardware support (at Cadence Design Systems) in 2008. And a lot has changed. The Raspberry Pi 2 model B came out and one of the same friends said 'You should use one of those!' And I said '1GB RAM is better by 2x than before, but still cruddy I/O. Surely better exists.' And in the process of researching, I found disconcerting facts about Linux on ARM, and read Nathan Willis's LWN.net write-up about Stephen Arnold's SCALE talk, which reinforced that: https://lwn.net/Articles/635289/ o Every bloody ARM device requires out-of-tree kernel patches that aren't necessarily very current at all and may introduce severe pecuilarities and/or lingering security holes. You simply _cannot_, for example, just run the Debian Jessie 'armhf' ('hf'= hard float) standard image, because there's no such thing as a standard armhf kernel that runs generically on armhf CPUs. Every blessed one requires a forked kernel that may never be merged and always be in danger of being orphaned. (Needless to say, I am unenthusiastic about running a high-security infrastructure computer that, e.g., security-scans my inside LAN, on hardware whose CPU may have unfixed Linux kernel hole for long periods of time.) o Every bloody ARM device requires boot solutions with some bespoke and idiosyncratic bits. o Every bloody ARM device binary-only, proprietary BLOBs if you care about X11 (obviously not router-only devices, etc.). Open source X11, if available, is usually quite old and worsens the kernel problem. o Every bloody ARM device has other one-off bits of weirdness often involving the sort of secret-sauce obsession that pervades embedded computing. Willis's LWN.net piece has detail on the above points. I kept searching, and thought, 'What possibilities beckon if I relax the requirement for ultra-thrifty mains use by 10W or so?' Which was exactly when I discovered the portion of the SFF (small form factor) PC market AMD has rightfully owned for the last few years with amazingly powerful, amazingly power-thrifty, quiet SoCs on tiny mini-ITX motherboards in quiet, small cases. And 32GB RAM limits. And great open source everything with bog-standard x86_64 distros. No, they doubtless don't quite suffice as good gamer boxes, the heft required for that role being -- as you acknowledge -- a specialty need that is very demanding. But the hardware category is called HTPC (home theatre PC), because these little things manage with aplomb to be quiet, unobtrusive Kodi (formerly xbmc) home-entertainment video boxes running in people's living rooms and bedrooms. Both on MS-Windows and Linux. And they're not slow. Not even the one I mentioned with the fanless SoC. Read, in particular, the Newegg.com customer reviews of the ASROck / AMD A4-5000 'Kabini' SoC combo. Some very picky customers have nice things to say, even the Windows users who characteristically think ridiculously high-performance CPUs are mandatory. I'm surprised that machines like that haven't been stampeded by Linux users. Seems to me, they're exceptional for a very large range of uses outside the data centre. And that they leave typical Intel-based offerings in the same segment of the market in the dust. For less money.

On 21/03/15 18:14, Rick Moen wrote:
Quoting Craig Sanders (cas@taz.net.au):
a) while i do most of my gaming on a windows box i built out of spare parts, i still play some games on linux, either native games or using wine.
b) vdpau video acceleration for mythtv (and smplayer etc).
if i didn't need those two things then either nouveau or the free radeon driver would be more than adequate for a desktop machine. Thank you for clarifying that. I read your post and was trying to think of a way to express a politely different perspective based on different use cases without accidentally sounding like I was arguing. Unfortunately, on the Internet as I'm sure you know, tone and nuance are lossy - and many's the time I've tried to say 'I'm glad that works for you, notwithstanding which other things entirely have proved better for me based on other priorities', yet, people act as if you'd challenged the other chap to a duel.
Your choices are absolutely sound for your requirements list. Of course.
I've been known to archly quote, to (for example) gamers telling me that Nvidia proprietary drivers are indispensible on account of frame rate and perfect streaming HD video, Mr. Lincoln saying 'That's the kind of thing that will be enjoyed by those who enjoy that kind of thing', intending to sound broad-minded and good-natured -- but start a small war for my good intentions. (You might have heard of Mr. Lincoln. Funny-looking bloke, tall, gaunt, famous teller of stories and jokes, wore black and a stovepipe hat, held the American Presidency until it turned out his retirement plan was decidedly missing.)
But I digress.
I approached this subject because of telling my LUG I'd like to build or buy a back-of-house, inside-LAN Linux server to do at least a few things: (a) Puppet or Chef master (configuration management), (b) network IDS, (c) online backup target. Being in sleeping space, as far from my small home server farm as possible so the same fire or burglar won't likely get both, it ought to be quiet, low-power, and small. I'm enormously fond of future-proofing by spending enough to buy something I reasonably hope to still be excellent in 5, 10, 15, maybe even 20 years if I'm lucky and plan right. And one of the earliest lessons I learned is that going heavy on RAM (or at least capacity to expand it) is the best future-proofing, followed by good I/O, followed distantly by CPU.
On my home LAN, network I/O matters less than it ought, because my home aDSL (static IP) has abysmal throughput for distressingly unclear reasons but we fear it's because my beloved family residence is at maximum distance from the telco central office where the DSLAM is -- a First World instance of the proverbial Last Mile Problem.
I also feel slightly guilty about running colo-type rackmount gear still left over from my old employer VA Linux Systems that draws more mains power than a well-selected small server should. (Yes, my VA gear is absurdly old. I know that.) So, I admire the mains thriftiness of, e.g., the Pi, and would aspire to approach that.
Anyway, after being crystal-clear about my needs and objectives, my LUG friends gave me advice:
1. 'Get a Raspberry Pi model B.' (This was before the current 2 model B.)
Me: 'Um, 512kB? Nothing better than microSD and USB2? No eSATA/SATA/mSATA. Really?'
2. 'Get a Beagleboard Black.'
Me: 'Um, 512kB? Nothing better than microSD and USB2? No eSATA/SATA/mSATA. Really?'
3. 'Get a large Intel 8-core thing and put it in a colo.'
Me: 'Um, did you listen to _any_ part of the target requirements?'
4. 'Get a small Atom-based board. Put an SSD on it.'
Me: 'That's getting there, but why does it have an 8GB RAM ceiling. Did I slip through a time vortex, or is it not 2015?'
Eventually I learned that LUG friends completely ignore what your expressed hardware needs are, and burble out a recommendation of what they're familiar with, as ticking off their own wishlist criteria and not yours were what you asked. LUG friends; Can't take them anywhere, and throwing large buckets of icewater on them is frowned upon. ;->
So, I started re-researching modern hardware, having last been paid to be expert in Linux hardware support (at Cadence Design Systems) in 2008. And a lot has changed.
The Raspberry Pi 2 model B came out and one of the same friends said 'You should use one of those!' And I said '1GB RAM is better by 2x than before, but still cruddy I/O. Surely better exists.'
And in the process of researching, I found disconcerting facts about Linux on ARM, and read Nathan Willis's LWN.net write-up about Stephen Arnold's SCALE talk, which reinforced that: https://lwn.net/Articles/635289/
o Every bloody ARM device requires out-of-tree kernel patches that aren't necessarily very current at all and may introduce severe pecuilarities and/or lingering security holes. You simply _cannot_, for example, just run the Debian Jessie 'armhf' ('hf'= hard float) standard image, because there's no such thing as a standard armhf kernel that runs generically on armhf CPUs. Every blessed one requires a forked kernel that may never be merged and always be in danger of being orphaned.
(Needless to say, I am unenthusiastic about running a high-security infrastructure computer that, e.g., security-scans my inside LAN, on hardware whose CPU may have unfixed Linux kernel hole for long periods of time.)
o Every bloody ARM device requires boot solutions with some bespoke and idiosyncratic bits.
o Every bloody ARM device binary-only, proprietary BLOBs if you care about X11 (obviously not router-only devices, etc.). Open source X11, if available, is usually quite old and worsens the kernel problem.
o Every bloody ARM device has other one-off bits of weirdness often involving the sort of secret-sauce obsession that pervades embedded computing.
Willis's LWN.net piece has detail on the above points.
I kept searching, and thought, 'What possibilities beckon if I relax the requirement for ultra-thrifty mains use by 10W or so?' Which was exactly when I discovered the portion of the SFF (small form factor) PC market AMD has rightfully owned for the last few years with amazingly powerful, amazingly power-thrifty, quiet SoCs on tiny mini-ITX motherboards in quiet, small cases. And 32GB RAM limits. And great open source everything with bog-standard x86_64 distros.
No, they doubtless don't quite suffice as good gamer boxes, the heft required for that role being -- as you acknowledge -- a specialty need that is very demanding.
But the hardware category is called HTPC (home theatre PC), because these little things manage with aplomb to be quiet, unobtrusive Kodi (formerly xbmc) home-entertainment video boxes running in people's living rooms and bedrooms. Both on MS-Windows and Linux. And they're not slow. Not even the one I mentioned with the fanless SoC. Read, in particular, the Newegg.com customer reviews of the ASROck / AMD A4-5000 'Kabini' SoC combo. Some very picky customers have nice things to say, even the Windows users who characteristically think ridiculously high-performance CPUs are mandatory.
I'm surprised that machines like that haven't been stampeded by Linux users. Seems to me, they're exceptional for a very large range of uses outside the data centre.
And that they leave typical Intel-based offerings in the same segment of the market in the dust. For less money.
_______________________________________________ luv-main mailing list luv-main@luv.asn.au http://lists.luv.asn.au/listinfo/luv-main
Great, thank you for your earlier links and especially for this excellent breakdown of required tasks and analysis of requirements. I also am not sure why these type of motherboards are not more widely known especially in the linux community. I am preparing to replace my recently replaced mythtv box with something a bit less power hungry, as the p3/800 hardware I replaced with a spare p4/3.2G IBM desktop is both not that quiet and consumes too much power. So given the extraordinary low price of the hardware you have pointed out, it doesn't take a rhodes scholar to make that decision. And given that it can then be left on most times, it will also replace my debianized SLUG that is long overdue for replacement as a network device. It was mainly used for network backups, however with the limited 32MB of ram it had available, it was always in swap, the great thing is that the drive can simply be attached via USB to the new system for transferring bind9 configs and such. :-) Thanks Robert

On Sat, Mar 21, 2015 at 12:14:17AM -0700, Rick Moen wrote:
1. 'Get a Raspberry Pi model B.' (This was before the current 2 model B.)
Me: 'Um, 512kB? Nothing better than microSD and USB2? No eSATA/SATA/mSATA. Really?'
funnily enough, i have a similar reaction to most intel motherboards - their CPUs can be quite good, but the PCIe lines available and the I/O is minimal compared to AMD AM2/3/3+ CPUs and motherboards. like 4 sata ports on intel vs 8 or 10 on amd, 2 or 3 pcie slots (at x16,x1,x1 or something like that) compared to about 4 or 5 or more on amd (at x16,x16,x16,x4,x1 or or x16,x8,x8,x1 or similar). and often only two DIMM sockets. to get the same sort of features (like ECC RAM on AM3+ m/bs) and I/O capability that AMD has on cheap consumer chips and boards, you have to spend a fortune on intel server cpus and motherboards.
And in the process of researching, I found disconcerting facts about Linux on ARM, and read Nathan Willis's LWN.net write-up about Stephen Arnold's SCALE talk, which reinforced that: https://lwn.net/Articles/635289/
yep, i've done similar research on ARM devices and come to the same conclusions as you for pretty much the same reasons. ARM sounds nice until you look at it in detail and notice the limitations and just how much hassle working with the damn things would be.
o Every bloody ARM device requires out-of-tree kernel patches [...]
true even for consumer devices like tablets, let alone devices intended to be servers or general purpose "computers". my first android tablet was an allwinner a10 based cheapie. very good specs for the day, and only about $100. keeping it updated was a major PITA, especially after all the android devs lost interest and moved on to newer, shinier hardware. i've still got it and it still works, but i bought a nexus 7 to replace it - by the time cyanogenmod etc drop support for it, it will be years obsolete.
o Every bloody ARM device binary-only, proprietary BLOBs if you care about X11 (obviously not router-only devices, etc.). Open source X11, if available, is usually quite old and worsens the kernel problem.
and it's not just X11. the arm distros are tiny compared to the full range of software in x86 debian, so you'll be spending a lot of time compiling if you need much of that. and then you'll find that the same sort of devs who think it's OK to write linux-only *nix software also think it's OK to write x86-only code. so you wont just be compiling, you'll be spending a lot of time fixing architecture-specific bugs and incompatibilities (which is, of course, part of the reason why the arm distros are tiny. if the source packages compiled OK, distro autobuilders would build them all)
I kept searching, and thought, 'What possibilities beckon if I relax the requirement for ultra-thrifty mains use by 10W or so?' Which was exactly when I discovered the portion of the SFF (small form factor) PC market AMD has rightfully owned for the last few years with amazingly powerful, amazingly power-thrifty, quiet SoCs on tiny mini-ITX motherboards in quiet, small cases. And 32GB RAM limits. And great open source everything with bog-standard x86_64 distros.
note that, unlike the AM3 etc m/bs, FM2/FM2+ does not support ECC RAM so may not be the best choice for, e.g, a NAS server. depends on how much risk you're willing to take (fairly low for non-ECC with properly memtested/burned-in DIMMS) and how much you're willing to spend up-front and in electricity bills to eliminate that risk. personally, i don't bother with ECC RAM at home - but that's only because it's harder to get. i can't just walk in to any computer shop and buy it, you can only get it from shops specialising in corporate sales (i.e. expensive) and from online stores. the price difference on ECC vs non-ECC is only about 30% these days. other than that minor detail, the boards and cpus you linked to sounded ideal for a home server.
But the hardware category is called HTPC (home theatre PC), because
heh, my mythtv backend is an fx-8150 because that's what i happened to have lying around when i built it. i bought it to upgrade my main system but never got around to installing it. craig -- craig sanders <cas@taz.net.au>

Quoting Craig Sanders (cas@taz.net.au):
funnily enough, i have a similar reaction to most intel motherboards - their CPUs can be quite good, but the PCIe lines available and the I/O is minimal compared to AMD AM2/3/3+ CPUs and motherboards.
Here's a question that stumps me just a bit: Why are so many x86_64 motherboard / CPU combinations limited to what seem like arbitrarily low ceilings on total RAM? I don't want to seem like a spoiled child in this. Even 8GB (most Atom) is nice. But why not more? Back when x86_64 was new (and was AMD64, technically), my recollection was that we heard about the glorious new 16 exabyte (2^64) _theoretical_ linear address space that for reasons of practicality would be limited to 256 terabytes (2^48). Yet, we've never seen that, right? Mind you, I'm not talking about machines shipping _with_ 256 terabytes of RAM, only ones that could address that amount if it were available in real-world hardware. But, instead, I see a lot of 8GB total-possible-RAM limitations, such as all Intel Atom gear other than the 'Avoton' Atom CPUs that support 64GB (2^36) RAM addressing. It's not that I expect to see a motherboard with 256 terabyte support at local retailers. It's that I'm surprised the real-world limit is so dramatically _lower_ than that. Closer to the real world, AMD's 2013 'Jaguar' mobile SoC architecture, split into the Kabini (more powerful, low-power) and Temash (less powerful, ultra-low power) product lines can address up to 32GB. The following year's successor to Jaguar, 'Puma', is again segmented into higher-power Beema (15W TDP[1]) and lower-power Mullins (4.5W TDP) -- but I'm unable to find any Beema/Mullins real-world units with RAM capacity above 8GB. I'm confounded by that. What's going on? I'm aware there's a small irony in my posting that question from Silicon Valley, where I can ride my bicycle past Intel and AMD headquarters complexes by pedaling 25km down the road. But we're here and the question's been on my mind -- and I'm nobody's idea of a hardware engineer. (And it's not like I could knock on their doors and demand an answer, come to think of that.) Why wouldn't there be Beema or Mullins-compatible motherboards with at least as high RAM capacity as their Kabini predecessors a year earlier, and instead capacity declined by a factor of four? Is it because AMD is conceding the market for anything bigger than a smartphone or low-end tablet to Intel, or alternatively that few OEMs will any longer pay even small change above the cost of a low-end ARM chip, outside of the colo server market? That would be sad. I'm not even sure where the physical constraint lies, in this era of SoCs (Systems on Chips, what AMD calls APUs) that merge the former CPU, GPU, north bridge, and south bridge into a single chip. I'm guessing number of address lines from the SoC are a constraint, and there may or may not still be logic outside the SoC to decode the memory address lines. And of course there are the sockets for conventional SDRAM or SODIMM sticks, which doubtless impose some constraints. But RAM is cheap, and is (in my own use cases) the single most cost-effective place to sink money into a system to extend functionality, improve performance, and prolong useful service life. So, why are we seeing system-total address limits far below what rosy projections last millennium for x86_64 promised? And where is my hoverboard, anyway? ;-> One last data point: Note the strengths and limits of this extraordinary top-end variant of the new CompuLab 'Fitlet' nono-PC: http://www.fit-pc.com/web/products/specifications/?model%5B%5D=FITLET-GI-C67... This uses the fastest and most impressive AMD 'Mullins' SoC, the quad-core AMD A10-Micro 6700T (4.5W TDP). You can't buy that particular Fitlet. It went out of stock a short amount of time after it hit the market. But you can buy other variants with slightly slower (and cooler, even more power-thrifty) Mullins SoCs. And all Fitlets max out at 8GB RAM, one of their few notable design compromises. So, I'm wondering where that limit arises.
like 4 sata ports on intel vs 8 or 10 on amd, 2 or 3 pcie slots (at x16,x1,x1 or something like that) compared to about 4 or 5 or more on amd (at x16,x16,x16,x4,x1 or or x16,x8,x8,x1 or similar). and often only two DIMM sockets.
The motherboards with only two DIMM sockets tend to be the budget and/or SFF ones, such as the mini-ITX form-factor HTPC market -- thus driven by budget and physical space. Speaking for myself, few sockets doesn't bother me much, as long as I can have high total RAM by using dense sticks. It just means I have an incentive to use the densest supported sticks immediately, rather than have to yank one and put it in a drawer when I upgrade RAM a few years later. Basically, IMO, 'use dense RAM' is one of those basic lessons you learn through making dumb errors, like 'never fight a land war in Asia' and the only slightly less famous 'Never go in against a Sicilian when death is on the line.' In my area, people giving away five-year-old computing gear frequently, and I'm always amused to note that, upon examination, you find them stuffed with 4GB SDRAM sticks. Why? Because the donor joyfully emptied his/her drawer full of 4GB sticks and extracted the 16GB ones before donating. (These are also the same guys who were trying to _sell_ 17" hulking tube-type monitors around 2005 after the world moved to LCD -- which merely meant they were trying to clear out a pile of obsolete monitors and testing the Greater Fool Theory.)
to get the same sort of features (like ECC RAM on AM3+ m/bs) and I/O capability that AMD has on cheap consumer chips and boards, you have to spend a fortune on intel server cpus and motherboards.
Here's my perspective on ECC. I've used and worked on a very great deal of server gear with ECC RAM, and the advantages are obvious. But.... Let's assume you run Linux and are even moderately attentive to the operation of your machine. In my experience, you will become aware that you have a bad stick of RAM pretty quickly, because the signs will be unmistakeable, in the form of a pattern of segfaults and spontaneous reboots. Where I come from, if you see what even looks halfway like that sort of problem, you yank the machine for testing and stress-test the RAM with your choice of usual-suspect tools. And confirm or disconfirm your suspicion. I wouldn't want to run a mission-critical database server without ECC, because one hour of corrupton from single-bit RAM errors is an hour too much. But for, in particular, a home server running Linux, I no longer see ECC as in any way worth the very substantial extra cost for both RAM and motherboards. In passing, I'll note that it's Linux itself that appears to make bad RAM quickly noticeable -- in my experience -- from system behaviour. I have known corporate WinNT (& successors) and Novell NetWare servers where if you weren't running ECC, you would have no idea that data were being silently corrupted by passing through bad RAM.
and it's not just X11. the arm distros are tiny compared to the full range of software in x86 debian, so you'll be spending a lot of time compiling if you need much of that. and then you'll find that the same sort of devs who think it's OK to write linux-only *nix software also think it's OK to write x86-only code. so you wont just be compiling, you'll be spending a lot of time fixing architecture-specific bugs and incompatibilities (which is, of course, part of the reason why the arm distros are tiny. if the source packages compiled OK, distro autobuilders would build them all)
Aha! Thanks for clarifying that. I actually hadn't figured out why the ARM variants of distros were relatively spare on package selection, but that explains it.
depends on how much risk you're willing to take (fairly low for non-ECC with properly memtested/burned-in DIMMS) and how much you're willing to spend up-front and in electricity bills to eliminate that risk.
Speaking of burning in: At the late VA Linux Systems, Inc., we developed in-house a tool called Cerberus Test Control System (CTCS) for burn-in of both newly manufactured systems and all systems RMAed back for any reason. It's still quite good. Open source. 'Cerberus FAQ' on http://linuxmafia.com/kb/Hardware had information. My own way of finding bad RAM is iterative kernel compilation with 'make -j N' set high enough to fully saturate RAM just short of driving the system into swap. Details here: http://linuxmafia.com/pipermail/conspire/2006-December/002662.html http://linuxmafia.com/pipermail/conspire/2006-December/002668.html http://linuxmafia.com/pipermail/conspire/2007-January/002743.html This was an edge-case, in that I had foolishly accepted discarded, suspect RAM from a company's data centre machines and optimistically assumed it might be good, which was a poor wager to make. It turned out that two of the system's four sticks were bad in ways that somehow partially offset each other, making it a bit of wark to find the specific badness and isolate the cause. The punch line: All four sticks were ECC, running on a server-grade ECC-supporting Intel L440GX+ 'Lancewood' motherboard in a colo 2U chassis.
personally, i don't bother with ECC RAM at home - but that's only because it's harder to get.
I don't because I'd rather spend the money on things I think are better value. But Views Differ.[tm] [1] Thermal Design Power is the maximum rated heat emission that associated cooling for a part might be called upon to handle. Thus, it is a maximum-case measure of the power draw that the part could draw at peak loading. By implication, real-world usage will typically be much lower.

On 22 March 2015 at 10:17, Rick Moen <rick@linuxmafia.com> wrote:
Quoting Craig Sanders (cas@taz.net.au):
funnily enough, i have a similar reaction to most intel motherboards - their CPUs can be quite good, but the PCIe lines available and the I/O is minimal compared to AMD AM2/3/3+ CPUs and motherboards.
Here's a question that stumps me just a bit: Why are so many x86_64 motherboard / CPU combinations limited to what seem like arbitrarily low ceilings on total RAM?
I don't want to seem like a spoiled child in this. Even 8GB (most Atom) is nice. But why not more?
Back when x86_64 was new (and was AMD64, technically), my recollection was that we heard about the glorious new 16 exabyte (2^64) _theoretical_ linear address space that for reasons of practicality would be limited to 256 terabytes (2^48). Yet, we've never seen that, right? Mind you, I'm not talking about machines shipping _with_ 256 terabytes of RAM, only ones that could address that amount if it were available in real-world hardware.
AIUI it's an economic compromise made by CPU manufacturers and mobo manufacturers. I'm no hardware engineer either however I did study electronic engineering at uni in the halcyon days of the 8086/186. Whilst 64 bit CPUs have a theoretical 64 bit address space available to them for performance and other functionality reasons the CPU designers limit the number address lines i.e pins available to memory. For example the Kabini CPUs featured in this thread have 40-bit address space, whereas some the higher end CPUs will have 42 and 43 bit address lines available to them. Of course this is larger than some of the 32Gb limits we see on mobos. Here we hit the economic compromises made by the mobo makers. They have to balance what they can physically fit on a mobo with how easy they can be physically wired to the CPU and with how much the market is willing to pay versus their bottom line. Given the clock speeds that modern mobos and CPUs operate at, balancing the arrival of bits from RAM to the CPU etc is a critical factor in making sure it all works as expected. I'm sure if you looked around you could find a mobo that maxed out the RAM limit but I expect it would be expensive, power hungry and large (well full-sized anyway). Not the focus of some of this thread I'm sure you'll agree. Again a compromise has to be made to made amongst physical size, speed, power consumption etc. Which all contributes to the less than theoretical RAM limit. -- Colin Fee tfeccles@gmail.com

Quoting Colin Fee (tfeccles@gmail.com):
AIUI it's an economic compromise made by CPU manufacturers and mobo manufacturers.
Sounds reasonable (and this is what I expected). Part of what I'm curious about (not asking _you_ specifically, but rather just putting the question out in the air) is whether this implementation compromise is, particularly in recently discussed SoCs and motherboards, imposed by the SoC or not. And which specific component or components create(s) it. Most particularly, I am curious whether I am wasting my time looking aroudn for motherboards compatible with AMD's 2014 'Puma' series of SoCs (the Beema and Mullins series) capable of supporting more than 8GB of RAM -- because to the extent I find Beema/Mullins used, which is distressingly rare, I see 8GB limits. I am vacillating on the brink of buying a SoC / motherboard combo based on the 2013 chips taht the 2014 'Puma' architecture directly replaced, namely the 'Jaguar' architecture 'Kabini' SoCs, specifically because the older 2013 SoC maxes out at 32GB supported RAM rather than the newer, 2014 replacement SoC's (seeming?) 8GB limit. The point is, I don't want to miss better, one-year-newer, hardware out of ignorance -- if higher-RAM 'Puma' motherboards do exist. In addition, the (seeming) 4x decline in RAM capacity with the successor 2014 chip series seems awfully strange, and I'm trying to understand what happened. (FWIW, I was an EECS aka Electrical Engineering and Computer Science -- pronounced 'eeks!' -- major at a certain Ivy League university named for William Prince of Nassau, back in unfathomably ancient days, not that that does me any good in this century. I switched to maths, which Yanks call 'math', before graduation. Ridiculoously long ago, and, like essentially all teenagers, I had no clue whatsoever what I was doing.)
Whilst 64 bit CPUs have a theoretical 64 bit address space available to them for performance and other functionality reasons the CPU designers limit the number address lines i.e pins available to memory. For example the Kabini CPUs featured in this thread have 40-bit address space, whereas some the higher end CPUs will have 42 and 43 bit address lines available to them.
Doesn't 40-bit address space (2^40) equate to 1 terabyte? All the Kabini-compatible motherboards I'd seen max out at 32GB, which is something like 2^36. So, if you say Kabini has 40-bit address space, I'm not sure where that has any real-world manifestation. I often make a fencepost error in these shirtsleeve calculations, and real hardware engineers keep track of 'Well, it's actually 2^36 - 1' sorts of edge conditions, but I'm sure I'm no more than off by a factor of two, here. FWIW, if in 2015, actually _had_ ability to support 1 TB-ish RAM (within a factor of two, 2^40 addresses) on Kabini or anything similar, I'd feel extremely future-proofed indeed, and consider the decade of the 2030 able to take care of itself. I've not seen that at all -- much less on those chips' next-year replacements.
Of course this is larger than some of the 32Gb limits we see on mobos. Here we hit the economic compromises made by the mobo makers.
This is what I figure -- and is very small misfortune compared to death and taxes. But I remain curious about _how_ (where) that gets implemented, and why AMD (apparently) decreed a further 4x cutback in RAM capacity when it replaced Kabini/Temash with its direct successor.
I'm sure if you looked around you could find a mobo that maxed out the RAM limit but I expect it would be expensive, power hungry and large (well full-sized anyway).
Well, yes. In particular, some of the HP ProLiant rackmount colo boxes I've recently worked with professionally can hold very high amounts of RAM -- and I'm glad I don't pay the power bill for them (or hear their banshee screaming 24x7 in my house). -- Cheers, I'm ashamed at how often I use a thesaurus. I mean bashful. Rick Moen Embarrassed! Wait--humiliated. Repentant. Chagrined! Sh*t! rick@linuxmafia.com -- @cinemasins McQ! (4x80)

On Sun, 22 Mar 2015 01:20:47 AM Colin Fee wrote:
Whilst 64 bit CPUs have a theoretical 64 bit address space available to them for performance and other functionality reasons the CPU designers limit the number address lines i.e pins available to memory. For example the Kabini CPUs featured in this thread have 40-bit address space, whereas some the higher end CPUs will have 42 and 43 bit address lines available to them.
The amount of engineering work required to add an extra address line will usually scale linearly with the number of address lines. Once you have 32 address lines adding a few more isn't going to add much extra silicon or motherboard space.
Of course this is larger than some of the 32Gb limits we see on mobos. Here we hit the economic compromises made by the mobo makers. They have to balance what they can physically fit on a mobo with how easy they can be physically wired to the CPU and with how much the market is willing to pay versus their bottom line.
The more DIMM sockets you have the further away from the CPU and memory controller they have to be which decreases the maximum clock rate through longer signal propagation delays and greater capacitance in the lines. This can be partially addressed through greater buffering which gives new different problems. The consumer motherboards with 2 DIMM sockets fit a real market need, they are simple to design and cheap to manufacture.
Given the clock speeds that modern mobos and CPUs operate at, balancing the arrival of bits from RAM to the CPU etc is a critical factor in making sure it all works as expected.
Yes.
I'm sure if you looked around you could find a mobo that maxed out the RAM limit but I expect it would be expensive, power hungry and large (well full-sized anyway). Not the focus of some of this thread I'm sure you'll agree. Again a compromise has to be made to made amongst physical size, speed, power consumption etc. Which all contributes to the less than theoretical RAM limit.
There are server motherboards that take many DIMMs (16 or more). But they are expensive, don't fit desktop cases, and have special cooling requirements. That said Intel has a long history of not supporting as much RAM as can be physically installed. For example I have a P3 router that has 2*256M DIMMs but the 3rd DIMM socket is empty because the Intel chipset only supports 512M. It's quite possible that Intel is deliberately reducing the lifetime of it's products as Robert suggests. -- My Main Blog http://etbe.coker.com.au/ My Documents Blog http://doc.coker.com.au/

On Sat, Mar 21, 2015 at 04:17:02PM -0700, Rick Moen wrote:
Quoting Craig Sanders (cas@taz.net.au):
funnily enough, i have a similar reaction to most intel motherboards - their CPUs can be quite good, but the PCIe lines available and the I/O is minimal compared to AMD AM2/3/3+ CPUs and motherboards.
Here's a question that stumps me just a bit: Why are so many x86_64
i know this question wasn't actually addressed to me, but i just wanted to say that i'm quite interested in this conversation but unable to continue at the moment because i'm getting ready to go into hospital tomorrow - i expect to be in for about 10-14 days (having my right polycystic kidney removed...the left one was removed last year and weighed 6.5 Kg or about 1 stone in archaic units) i probably wont have internet access because dealing with the hospital's wifi authentication system (they come around with a daily changed password sometime in the mid-to-late afternoon, which stops working abruptly at midnight when the password is changed) is way more hassle than it's worth. in any case, i expect i shall be making use of the magic morphine-on-demand button (max 1 hit every 15 minutes) for the first day or two. so, anyway, sorry for dropping out of the convo so abruptly - it's been interesting and educational and has spurred me to do more reading and research than i really have time for right now :) the only detail i've managed to find about the mullins/beema memory controllers is that they're both single channel and support up to two dimms. AFAICT they're both also DDR3-L (low power variant of ddr3), which only seeems to be available in 1, 2, and 4GB sticks at the moment. so that may be the source of the 8GB limit. craig ps: i have warned my surgeon that if i cark it on the operating table, i'll turn on the spot and he'll have a zombie outbreak to deal with. (i promised the same thing last year and the very first thing i said to my partner when i woke up after surgery was "braaaaaainzzzz". fortunately she had left her katana at home) -- craig sanders <cas@taz.net.au> BOFH excuse #241: _Rosin_ core solder? But...

Quoting Craig Sanders (cas@taz.net.au):
i know this question wasn't actually addressed to me, but i just wanted to say that i'm quite interested in this conversation but unable to continue at the moment because i'm getting ready to go into hospital tomorrow - i expect to be in for about 10-14 days (having my right polycystic kidney removed...the left one was removed last year and weighed 6.5 Kg or about 1 stone in archaic units)
My goodness; I've not heard weight expressed in stone since I last played an Allan Sherman medley.[1] Very best of fortune with the surgery, Craig, and get healthy soon. I'd visit your recovery room, but the airfare's a bit brutal.
the only detail i've managed to find about the mullins/beema memory controllers is that they're both single channel and support up to two dimms.
AFAICT they're both also DDR3-L (low power variant of ddr3), which only seeems to be available in 1, 2, and 4GB sticks at the moment. so that may be the source of the 8GB limit.
Aha! Thank you; that would explain it.
ps: i have warned my surgeon that if i cark it on the operating table, i'll turn on the spot and he'll have a zombie outbreak to deal with.
(i promised the same thing last year and the very first thing i said to my partner when i woke up after surgery was "braaaaaainzzzz". fortunately she had left her katana at home)
I know what _you've_ been watching. One of the few worthwhile things to emerge from the US state of Georgia since James Earl Carter. [1] Unkind to the large, perhaps better consigned to 1963. FWIW: https://www.youtube.com/watch?v=K2Cut3XwJxM

On Sat, Mar 21, 2015 at 09:30:42PM -0700, Rick Moen wrote:
Very best of fortune with the surgery, Craig, and get healthy soon. I'd visit your recovery room, but the airfare's a bit brutal.
thanks i just got home, much sooner than expected...recovering really quickly and well this time around. i'm sore and in pain and a bit stoned from the painkillers but i'm feeling much better than i have for ages. fortunately, the surgeon has a better sense of left & right than I do because it was actually the left kidney that got removed on Monday...the right one was removed last year. let us know if you do come to australia (again?) one day, i'd like to meet you in person.
I know what _you've_ been watching. One of the few worthwhile things to emerge from the US state of Georgia since James Earl Carter.
according to SBS TV, it's "International Zombie Month" and they've been showing a lot of zombie movies from around the world for the last few weeks. btw, Toni really does have a katana (and a very nice viking sword and a few others) and knows how to use them. she's only allowed to use them on me if i turn into a zombie. Fortunately Plan B (being patient zero) wasn't needed. Plan A (surviving) has worked out nicely. craig -- craig sanders <cas@taz.net.au>

On 22.03.15 15:11, Craig Sanders wrote:
in any case, i expect i shall be making use of the magic morphine-on-demand button (max 1 hit every 15 minutes) for the first day or two.
Good stuff that - though it make the zombies less friendly after prolonged use. Take care. (Of yourself and that replacement.) Erik -- Pessimist: The glass is half empty. Optimist: The glass is half full. Engineer: The glass is twice as big as it needs to be. - Read on avr-chat ML Pragmatist: Who cares, so long as there's more in the bottle.

On 22/03/15 10:17, Rick Moen wrote:
I'm confounded by that. What's going on?
I'm aware there's a small irony in my posting that question from Silicon Valley, where I can ride my bicycle past Intel and AMD headquarters complexes by pedaling 25km down the road. But we're here and the question's been on my mind -- and I'm nobody's idea of a hardware engineer. (And it's not like I could knock on their doors and demand an answer, come to think of that.)
Why wouldn't there be Beema or Mullins-compatible motherboards with at least as high RAM capacity as their Kabini predecessors a year earlier, and instead capacity declined by a factor of four? Is it because AMD is conceding the market for anything bigger than a smartphone or low-end tablet to Intel, or alternatively that few OEMs will any longer pay even small change above the cost of a low-end ARM chip, outside of the colo server market? That would be sad.
I am quite certain that the reasons for that are economic, you have said it yourself, the main component responsible for board longevity is ram capacity... Robert

On Sat, 21 Mar 2015 04:17:02 PM Rick Moen wrote:
Back when x86_64 was new (and was AMD64, technically), my recollection was that we heard about the glorious new 16 exabyte (2^64) _theoretical_ linear address space that for reasons of practicality would be limited to 256 terabytes (2^48). Yet, we've never seen that, right? Mind you, I'm not talking about machines shipping _with_ 256 terabytes of RAM, only ones that could address that amount if it were available in real-world hardware.
It's not even necessarily down to the motherboards, even the CPUs have RAM limitations, from what I've seen -EP series CPUs don't go beyond 768GB of RAM so our 1TB and larger systems are all Westmere-EX based systems (they predate the Ivybridge-EX CPUs). Similarly the current Haswell-EP's have the same limitation, you'll need to wait for the -EX series ones to ship to get to TB's of RAM... cheers, Chris -- Chris Samuel : http://www.csamuel.org/ : Melbourne, VIC

On 21.03.15 08:32, Craig Sanders wrote:
On Thu, Mar 19, 2015 at 03:38:41PM +1100, Erik Christiansen wrote:
I'm actually surprised there's enough memory in the machine to run a modern desktop environment and web browser. A 2004 era machine probably only has between 0.1 and 0.5 GB of RAM. I think?
It's now a bit newer (VIA C7), and has a whole 1 GB, so no drama there.
that's a 32-bit CPU from Sep 2006, nearly 9 years old. are you aware that you can upgrade to a modern machine for under $170?
The drives are a good bit older than the mobo - thus a new drive for the new distro. I've in the last week replaced the CDROM with a DVD, so could casually think about a new mobo when I (probably) find that the on-board (shared memory) graphics doesn't deal with a video DVD.
e.g. the cheapest current parts combo i can find at MSY today is:
AMD A4-7300 64-bit dual-core CPU + GPU $58 ASRock A58M-HD+ motherboard $65 4G Kit DDR3-1333 RAM (2x2GB) $46
That is cheaper than the older lower performance stuff I have.
if the RAM in your current mb is DDR3 (it probably isn't), you could put off buying the new ram for a while,
ISTR it's DDR2-533 - at least it shows as 533 during boot.
but 1GB isn't really adequate for a modern desktop any more. the machine will likely be swapping to disk by the time it finishes booting and you login even with a "light" desktop like xfce or lxde, let alone start running memory hogs like iceweasel or chromium.
I'm in gnome, with 4 xterms, and running iceweasel and xpdf: $ more /proc/vmstat | grep pswp pswpin 0 pswpout 0
you've already got yourself a new drive but if you hadn't, you can get a 64GB SSD for $56 these days. small but much faster than any mechanical drive.
Rats. I've followed the discussions on this list, and noted that endurance concerns are unfounded, but plumb forgot when buying. Old habit is hard to shake.
FYI, comparison of the A4-7300 with the C7:
http://www.cpu-world.com/Compare/887/AMD_A4-Series_A4-7300_%28JA%29_vs_VIA_C...
note that the max power usage of the A4-7300 is 65W versus 20W for the C7
That's not great news. The power supply for each the two of these I run is a picoPSU-80-WI-32, providing 80W in a matchbox-sized MHz switcher which entirely mounts in the ATX power connector. With a 12v battery between that and any old 13.8v power supply, a UPS is built in. (And out on the farm, where there's no mains electricity - just a petrol generator which runs only at night, it can run off a battery and a couple of solar panels. That 80w (120W peak) is all there currently is for the mobo, three hard drives, and the DVD. The power supply is cool to the touch. (It seems crazy for a PC to consume half a horsepower.)
- but the A4-7300 includes a built-in Radeon R3 graphics processor. if your current graphics card is as old as your mb and cpu, then it probably uses at least 30-40W anyway, and isn't anywhere near as good as the R3....9+ years is a LONG time in the evolution of GPUs.
There is no off-mobo graphics at present. If I'm tempted down that road, then yes, I'll have to at least dump the old drives to power a new card.
also, FYI, the Intel G1840 (a dual-core 64-bit celeron with built-in graphics) costs $55 but the cheapest Haswell Refresh motherboards start at nearly twice the price of the cheapest AMD FM2+ motherboards....and Intel, as usual, sucks at upgradability. once again you have to discard your old m/b if you want to upgrade to the new generation of CPUs. AMD OTOH generally keep the same socket for at least two generations so you can keep upgrading your CPU without having the expense and waste of a new motherboard.
Since back in last century, I've only updated CPU when the mobo died, except for the first 170 MHz one which became pretty useless before it died, even though there was only dial-up internet in those days. I try not to waste my time upgrading even the distro more than twice per decade, so hardware just stays till it goes under, since it's still meeting my needs. Thanks for the info. It is food for thought. But if the machine won't show the DVDs I was given last May, then I can fend off further such gifts without needing to say I couldn't be bothered to spend my time in front of a video. (I haven't had a TV in the house since I brought the old one out of storage for one day in 1991 to watch what happened in the first hours of the first Gulf War. The one out on the farm packed it in recently, making visiting family much more sociable. The 8 year old city slickers were unfit, too timid to touch the spines of an echidna when I got them out for a walk, and audibly suffering gadget withdrawal.) I'm kinda fond of a minimalist approach. Most days this week I've been too busy splitting firewood for winter to watch DVDs, and when I flop, the imagery invoked by an engrossing book is more my style. Erik -- Rear Admiral Morisetti recalled that when commanding an aircraft carrier, it took a gallon of oil to move just 12 inches (30cm), while as many as 20 tonnes per hour were burned during a period of intensive take-off and landing. - http://www.bbc.co.uk/news/science-environment-15342682

On Sat, 21 Mar 2015 07:09:22 AM Erik Christiansen wrote:
that's a 32-bit CPU from Sep 2006, nearly 9 years old. are you aware that you can upgrade to a modern machine for under $170?
The drives are a good bit older than the mobo - thus a new drive for the new distro. I've in the last week replaced the CDROM with a DVD, so could casually think about a new mobo when I (probably) find that the on-board (shared memory) graphics doesn't deal with a video DVD.
In the past there have been LUV members offering high end P4 systems for free and not getting takers. If you can attend LUV meetings then you might ask if someone has such a system to spare. -- My Main Blog http://etbe.coker.com.au/ My Documents Blog http://doc.coker.com.au/

On 21.03.15 10:54, Russell Coker wrote:
In the past there have been LUV members offering high end P4 systems for free and not getting takers. If you can attend LUV meetings then you might ask if someone has such a system to spare.
Ah, now that's an ingenious way to encourage greater meeting attendance. :-) I might put up my hand at some future offer - though the crazy low price for Rick's nifty little find: http://www.newegg.com/Product/Product.aspx?Item=N82E16813157574 might just beat that to the punch. The board specs don't mention board power consumption, but some of the associated glossy bumpf does indicate a 15W CPU. Since we've been on the topic, I tried a locally bought DVD, but Debian's "Movie Player" rejected it as "encrypted". I may at some stage try to figure that out. It'd be nice to know how to do this stuff. But for what I do: vim, mutt, a bit of avr-gcc, and a look at the ABC and BBC news sites each evening, the old clunker is amply sufficient so far. I'm only replacing the old drives because they're flagging themselves as "end of life". Erik --
It may not work well for you since it sounds like you're always in terminal Vim but I thought I'd pass it along.
No, I don't believe in GUIs. - Arvid Warnecke, on vim-use ML.

On 21/03/15 22:41, Erik Christiansen wrote:
Since we've been on the topic, I tried a locally bought DVD, but Debian's "Movie Player" rejected it as "encrypted". I may at some stage try to figure that out. It'd be nice to know how to do this stuff.
This is probably the issue: https://wiki.debian.org/CDDVD Hope that helps, Andrew

On 21.03.15 23:01, Andrew Pam wrote:
On 21/03/15 22:41, Erik Christiansen wrote:
Since we've been on the topic, I tried a locally bought DVD, but Debian's "Movie Player" rejected it as "encrypted". I may at some stage try to figure that out. It'd be nice to know how to do this stuff.
This is probably the issue: https://wiki.debian.org/CDDVD
Hope that helps, Andrew
Definitely does. Many thanks, Andrew. Since www.progress-linux.org is closed for maintenance, I'll compile libdvdcss - tomorrow. ATM sleep beckons. Erik -- Abbott's Admonitions: (1) If you have to ask, you're not entitled to know. (2) If you don't like the answer, you shouldn't have asked the question. - Charles Abbot, dean, University of Virginia

Erik Christiansen <dvalin@internode.on.net> writes:
On 21.03.15 23:01, Andrew Pam wrote:
On 21/03/15 22:41, Erik Christiansen wrote:
Since we've been on the topic, I tried a locally bought DVD, but Debian's "Movie Player" rejected it as "encrypted". I may at some stage try to figure that out. It'd be nice to know how to do this stuff.
This is probably the issue: https://wiki.debian.org/CDDVD
Hope that helps, Andrew
Definitely does. Many thanks, Andrew.
Since www.progress-linux.org is closed for maintenance, I'll compile libdvdcss - tomorrow. ATM sleep beckons.
Yeaaaaaah, you don't really want that one. It just mentions that one because it's dba's distro and dba wrote that advice. This is what I've been doing for stupid DVD users, on wheezy: wget -O- http://download.videolan.org/pub/debian/videolan-apt.asc | apt-key add - echo 'deb http://download.videolan.org/pub/debian/stable/ /' >/etc/apt/sources.list.d/30decss.list apt-get update apt-get install libdvdcss2 It works for me. (PS: well, actually I'm using [trusted=yes] instead of adding the key.)

On 23.03.15 11:04, Trent W. Buck wrote:
Erik Christiansen <dvalin@internode.on.net> writes:
Since www.progress-linux.org is closed for maintenance, I'll compile libdvdcss - tomorrow. ATM sleep beckons.
Yeaaaaaah, you don't really want that one. It just mentions that one because it's dba's distro and dba wrote that advice.
This is what I've been doing for stupid DVD users, on wheezy:
wget -O- http://download.videolan.org/pub/debian/videolan-apt.asc | apt-key add - echo 'deb http://download.videolan.org/pub/debian/stable/ /' >/etc/apt/sources.list.d/30decss.list apt-get update apt-get install libdvdcss2
It works for me.
(PS: well, actually I'm using [trusted=yes] instead of adding the key.)
Many thanks for the heads up, Trent. On returning to the task yesterday, I did just go to item 1 on the list: http://download.videolan.org/debian/ and simply download libdvdcss2_1.2.13-0_i386.deb , then whack it with dpkg -i. That seems to be the same as the apt-get with [trusted=yes], without any sources.list pollution. When I saw your post, I had hoped that the above links pointed to another library, because I've tried a movie with the videolan library, and while it plays, the only colour is some faint green and red diagonal stripes on a black and white picture. If that's not due to library limitations, then perhaps it's just that the on-board graphics: $ lspci ... 01:00.0 VGA compatible controller: VIA Technologies, Inc. CN700/P4M800 Pro/P4M800 CE/VN800 Graphics [S3 UniChrome Pro] (rev 01) is OK with youtube clips, but can't shovel fast enough for a full movie. I'm not sure it's worth spending a couple of hundred bucks for a faster machine, just so more can be spent on more DVDs, few of which I'd want to watch. If there's a good linux-compatible graphics card with modest power consumption, then the mobo has exactly one PCI slot, and it's still spare. That sounds like a more proportionate investment. Erik -- "Necessity is the mother of invention" is a silly proverb. "Necessity is the mother of futile dodges" is much closer to the truth. The basis of growth of modern invention is science, and science is almost wholly the outgrowth of pleasurable intellectual curiosity. -Alfred N. Whitehead

Quoting Erik Christiansen (dvalin@internode.on.net):
I might put up my hand at some future offer - though the crazy low price for Rick's nifty little find:
http://www.newegg.com/Product/Product.aspx?Item=N82E16813157574
Reminder: Yr. humble servant being a Bloody Yank despite knowing how to spell 'colour' and having generally Pommie diction[1], and being a decent cricket batsman, said link's pricing is in US$, and I am not totally certain that Newegg is an Oz vendor. (You would know, presumably.) I do know that some (e.g.) ASRock motherboard models are available in Oz but not North America and vice-versa, as well. I chose those links mostly to call attention to the HTPC category and the amazing combination of AMD's 'Kabini' series of SoCs and compatible mini-ITX motherboards. If interested in such offerings, the following LUG thread started by former _Linux Journal_ editor Don Marti, subject header 'Quiet, Freedom-compatible NAT/firewall/misc box?', may be of interest: http://linuxmafia.com/pipermail/conspire/2015-March/007983.html Again, said LUG is in the Schwarzenegger-ian principality of California, not, alas, in Victoria. (The hosting machine, linuxmafia.com, is my patheically obsolete VA Linux Systems model 2230, in my garage, manufactured in the halcyon year of 2001, bearing 1.5GB of RAM and an awesome pair of Pentium III CPUs. ;-> ) Some of the postings to the 'Quiet, Freedom-compatible NAT/firewall/misc box?' thread have further details about power and about the differing AMD SoC product lines -- to the best of my strictly amateur and fallible ability to determine. ([insert voice of Deforest Kelley:] 'Dammit, Jim, I'm a sysadmin, not a hardware engineer.') [1] As I put it to my wife, I speak American only through translation (having grown up in Hong Kong attending the local Brit school).

On Sat, Mar 21, 2015 at 08:45:51AM -0700, Rick Moen wrote:
Quoting Erik Christiansen (dvalin@internode.on.net):
I might put up my hand at some future offer - though the crazy low price for Rick's nifty little find:
http://www.newegg.com/Product/Product.aspx?Item=N82E16813157574
Reminder: Yr. humble servant being a Bloody Yank despite knowing how to spell 'colour' and having generally Pommie diction[1], and being a decent cricket batsman, said link's pricing is in US$, and I am not totally certain that Newegg is an Oz vendor. (You would know, presumably.) I do know that some (e.g.) ASRock motherboard models are available in Oz but not North America and vice-versa, as well.
i don't think newegg will ship to australia. ISTR reading somewhere once upon a time that they don't...by my brain's ram is non-ecc and fallible. there are, however, several shops and online stores that sell HTPC and mini/micro-ITX gear. try a google search for "Kabini site:au" or "AM1 site:au" shopbot, for example, shows several places that have an almost identical model to the one Rick mentioned (but with built-in wifi) for around $145 http://www.shopbot.com.au/m/?m=ASRock+QC5000-ITX no-one seems to have the model without wifi.
http://linuxmafia.com/pipermail/conspire/2015-March/007983.html
thanks, some interesting reading in there. craig -- craig sanders <cas@taz.net.au>

On 22.03.15 10:00, Craig Sanders wrote:
On Sat, Mar 21, 2015 at 08:45:51AM -0700, Rick Moen wrote:
Quoting Erik Christiansen (dvalin@internode.on.net):
I might put up my hand at some future offer - though the crazy low price for Rick's nifty little find:
http://www.newegg.com/Product/Product.aspx?Item=N82E16813157574
Reminder: Yr. humble servant being a Bloody Yank despite knowing how to spell 'colour' and having generally Pommie diction[1], and being a decent cricket batsman, said link's pricing is in US$, and I am not totally certain that Newegg is an Oz vendor. (You would know, presumably.) I do know that some (e.g.) ASRock motherboard models are available in Oz but not North America and vice-versa, as well.
i don't think newegg will ship to australia. ISTR reading somewhere once upon a time that they don't...by my brain's ram is non-ecc and fallible.
The two older mini-ITX boards, which still serve, came from Jetway, so I've looked there, but their mobo page search engine knows no ASRock or Kabini. :-(
there are, however, several shops and online stores that sell HTPC and mini/micro-ITX gear. try a google search for "Kabini site:au" or "AM1 site:au"
shopbot, for example, shows several places that have an almost identical model to the one Rick mentioned (but with built-in wifi) for around $145
And altech.com.au has an identical price. If we compare the US$62 Newegg price: 145 - 62/.75 = A$62 higher locally. Allow some of that for exchange fees, international shipping, the unwanted wifi, and GST which only applies to the local one, then the feasibility of local warranty replacement/repair isn't costing much. I'll have to have a look at Rick's low power mobo thread first. (Well, second - haven't build that CSS lib yet. Well, third - need to eat.) Erik -- US surgeon Alton Ochsner recalled that when he was a medical student in 1919 his class was summoned to observe an autopsy of a lung cancer victim. At that time, the disease was so rare it was thought unlikely the students would ever get another chance. - http://www.bbc.co.uk/news/magazine-20042217

On Sun, Mar 22, 2015 at 05:51:53PM +1100, Erik Christiansen wrote:
shopbot, for example, shows several places that have an almost identical model to the one Rick mentioned (but with built-in wifi) for around $145
And altech.com.au has an identical price. If we compare the US$62 Newegg price:
145 - 62/.75 = A$62 higher locally.
Allow some of that for exchange fees, international shipping, the unwanted wifi, and GST which only applies to the local one, then the feasibility of local warranty replacement/repair isn't costing much.
yep. approx $60 for exchange rate, shipping, GST, and australian statutory warranty / consumer protection laws isn't too bad. feels wrong to have to pay the extra "australia tax" on hardware and software, but it's not too excessive considering the protections we get as consumers in this country. (of course, i'm feeling benevolently stoned on painkillers at the moment so that's probably affecting my judgement) craig -- craig sanders <cas@taz.net.au>

On 22/03/15 02:45, Rick Moen wrote:
Quoting Erik Christiansen (dvalin@internode.on.net):
I might put up my hand at some future offer - though the crazy low price for Rick's nifty little find:
http://www.newegg.com/Product/Product.aspx?Item=N82E16813157574 Reminder: Yr. humble servant being a Bloody Yank despite knowing how to spell 'colour' and having generally Pommie diction[1], and being a decent cricket batsman, said link's pricing is in US$, and I am not totally certain that Newegg is an Oz vendor. (You would know, presumably.) I do know that some (e.g.) ASRock motherboard models are available in Oz but not North America and vice-versa, as well.
The wifi version is available from scorptec here in Australia for $149 Robert

On 22 March 2015 at 02:45, Rick Moen <rick@linuxmafia.com> wrote:
Quoting Erik Christiansen (dvalin@internode.on.net):
I might put up my hand at some future offer - though the crazy low price for Rick's nifty little find:
http://www.newegg.com/Product/Product.aspx?Item=N82E16813157574
Reminder: Yr. humble servant being a Bloody Yank despite knowing how to spell 'colour' and having generally Pommie diction[1], and being a decent cricket batsman, said link's pricing is in US$, and I am not totally certain that Newegg is an Oz vendor. (You would know, presumably.) I do know that some (e.g.) ASRock motherboard models are available in Oz but not North America and vice-versa, as well.
Newegg is an Oz vendor has been for approx 12 months. Access via http://www.newegg.com/Aus The item referenced above can be found, I searched via the item number. Listed at AUD$81.03. -- Colin Fee tfeccles@gmail.com

On 23/03/15 13:54, Colin Fee wrote:
On 22 March 2015 at 02:45, Rick Moen <rick@linuxmafia.com <mailto:rick@linuxmafia.com>> wrote:
Quoting Erik Christiansen (dvalin@internode.on.net <mailto:dvalin@internode.on.net>):
> I might put up my hand at some future offer - though the crazy low price > for Rick's nifty little find: > > http://www.newegg.com/Product/Product.aspx?Item=N82E16813157574
Reminder: Yr. humble servant being a Bloody Yank despite knowing how to spell 'colour' and having generally Pommie diction[1], and being a decent cricket batsman, said link's pricing is in US$, and I am not totally certain that Newegg is an Oz vendor. (You would know, presumably.) I do know that some (e.g.) ASRock motherboard models are available in Oz but not North America and vice-versa, as well.
Newegg is an Oz vendor has been for approx 12 months. Access via http://www.newegg.com/Aus
The item referenced above can be found, I searched via the item number. Listed at AUD$81.03.
I just got a 404 on that URL, but figured it out. http://www.newegg.com/global/au/Product/ProductList.aspx?Submit=ENE&DEPA=0&O... Robert

On Sat, Mar 21, 2015 at 06:09:22PM +1100, Erik Christiansen wrote:
I'm in gnome, with 4 xterms, and running iceweasel and xpdf:
you must have a lot more patience than I do.
note that the max power usage of the A4-7300 is 65W versus 20W for the C7
That's not great news. [...]
see rick's message for info on some similar but much lower power CPUs. also, newer motherboards and other parts tend to be much more power-efficient than older ones. your newish 320G drive, for example, probably uses less than your old 40G drive, especially if it's a "green" or 5400rpm drive.
- but the A4-7300 includes a built-in Radeon R3 graphics processor. if your current graphics card is as old as your mb and cpu, then it probably uses at least 30-40W anyway, and isn't anywhere near as good as the R3....9+ years is a LONG time in the evolution of GPUs.
There is no off-mobo graphics at present. If I'm tempted down that road, then yes,
the graphics built in to your motherboard also uses power and isn't included in the 20W used by the C7 CPU itself. it doesn't really matter whether it's built-in or a separate card, it still uses the same amount of power. you could find out how much by identifying what brand/model of graphics is built-in and then looking up the details of an equivalent standalone card. running 'lspci' may tell you. or maybe your m/b manual.
I'll have to at least dump the old drives to power a new card.
another advantage of SSDs - they use much less power than hard drives. craig -- craig sanders <cas@taz.net.au>

On 21.03.15 08:32, Craig Sanders wrote:
On Thu, Mar 19, 2015 at 03:38:41PM +1100, Erik Christiansen wrote:
I'm actually surprised there's enough memory in the machine to run a modern desktop environment and web browser. A 2004 era machine probably only has between 0.1 and 0.5 GB of RAM. I think?
It's now a bit newer (VIA C7), and has a whole 1 GB, so no drama there.
that's a 32-bit CPU from Sep 2006, nearly 9 years old. are you aware that you can upgrade to a modern machine for under $170?
The drives are a good bit older than the mobo - thus a new drive for the new distro. I've in the last week replaced the CDROM with a DVD, so could casually think about a new mobo when I (probably) find that the on-board (shared memory) graphics doesn't deal with a video DVD.
e.g. the cheapest current parts combo i can find at MSY today is:
AMD A4-7300 64-bit dual-core CPU + GPU $58 ASRock A58M-HD+ motherboard $65 4G Kit DDR3-1333 RAM (2x2GB) $46
That is cheaper than the older lower performance stuff I have.
if the RAM in your current mb is DDR3 (it probably isn't), you could put off buying the new ram for a while,
ISTR it's DDR2-533 - at least it shows as 533 during boot.
but 1GB isn't really adequate for a modern desktop any more. the machine will likely be swapping to disk by the time it finishes booting and you login even with a "light" desktop like xfce or lxde, let alone start running memory hogs like iceweasel or chromium.
I'm in gnome, with 4 xterms, and running iceweasel and xpdf: $ more /proc/vmstat | grep pswp pswpin 0 pswpout 0
you've already got yourself a new drive but if you hadn't, you can get a 64GB SSD for $56 these days. small but much faster than any mechanical drive.
Rats. I've followed the discussions on this list, and noted that endurance concerns are unfounded, but plumb forgot when buying. Old habit is hard to shake.
FYI, comparison of the A4-7300 with the C7:
http://www.cpu-world.com/Compare/887/AMD_A4-Series_A4-7300_%28JA%29_vs_VIA_C...
note that the max power usage of the A4-7300 is 65W versus 20W for the C7
That's not great news. The power supply for each the two of these I run is a picoPSU-80-WI-32, providing 80W in a matchbox-sized MHz switcher which entirely mounts in the ATX power connector. With a 12v battery between that and any old 13.8v power supply, a UPS is built in. (And out on the farm, where there's no mains electricity - just a petrol generator which runs only at night, it can run off a battery and a couple of solar panels. That 80w (120W peak) is all there currently is for the mobo, three hard drives, and the DVD. The power supply is cool to the touch. (It seems crazy for a PC to consume half a horsepower.)
- but the A4-7300 includes a built-in Radeon R3 graphics processor. if your current graphics card is as old as your mb and cpu, then it probably uses at least 30-40W anyway, and isn't anywhere near as good as the R3....9+ years is a LONG time in the evolution of GPUs.
There is no off-mobo graphics at present. If I'm tempted down that road, then yes, I'll have to at least dump the old drives to power a new card.
also, FYI, the Intel G1840 (a dual-core 64-bit celeron with built-in graphics) costs $55 but the cheapest Haswell Refresh motherboards start at nearly twice the price of the cheapest AMD FM2+ motherboards....and Intel, as usual, sucks at upgradability. once again you have to discard your old m/b if you want to upgrade to the new generation of CPUs. AMD OTOH generally keep the same socket for at least two generations so you can keep upgrading your CPU without having the expense and waste of a new motherboard.
Since back in last century, I've only updated CPU when the mobo died, except for the first 170 MHz one which became pretty useless before it died, even though there was only dial-up internet in those days. Thanks for the info. It is food for thought. But if the machine won't show the DVDs I was given last May, then I can fend off further such gifts without needing to explain at length how it can be that I couldn't be bothered to spend my time in front of a video. (I haven't had a TV in the house since I brought the old one out of storage for one day in 1991 to watch what happened in the first hours of the first Gulf War. The one out on the farm packed it in recently, making visiting family much more sociable. The 8 year old city slickers were unfit, too timid to touch the spines of an echidna when I got them out for a walk, and audibly suffering gadget withdrawal.) I'm kinda fond of a minimalist approach. Most days this week I've been busy splitting firewood for winter, and when I flop, the imagery invoked by an engrossing book is more my style. Erik (Who thinks he has postfix humming on the new distro now.) -- Rear Admiral Morisetti recalled that when commanding an aircraft carrier, it took a gallon of oil to move just 12 inches (30cm), while as many as 20 tonnes per hour were burned during a period of intensive take-off and landing. - http://www.bbc.co.uk/news/science-environment-15342682
participants (14)
-
Andrew Pam
-
Chris Samuel
-
Colin Fee
-
Craig Sanders
-
Erik Christiansen
-
Jeremy Visser
-
Keith Bainbridge
-
Michele Bert
-
Rick Moen
-
Robert Moonen
-
Russell Coker
-
Toby Corkindale
-
trentbuck@gmail.com
-
Wen Lin