- luv-main - luv.asn.au

Wifi performance problem
by Russell Coker 03 Nov '13

03 Nov '13

My home Wifi network used to allow me to FTP files to my Android phones at rates exceeding 1MB/s, I think that 1.9MB/s was my record. With that sort of speed transferring a 200MB or 300MB TV show to a phone was reasonably quick. Performance wasn't always great, sometimes it was as low as 400KB/s, but even that was usable. In the past I had even watched TV shows on my phone over HTTP (from a local server) with no performance problems. In the last few days performance has dropped to below 60KB/s with rates like 40KB/s (that would be 320Kb/s) being common. Because Android doesn't let me know what's happening I connected my Thinkpad T420 via Wifi and saw the following result: # iwconfig wlan0 wlan0 IEEE 802.11abgn ESSID:"abcdef" Mode:Managed Frequency:2.412 GHz Access Point: AA:AA:AA:AA:AA:AA Bit Rate=1 Mb/s Tx-Power=15 dBm Retry long limit:7 RTS thr:off Fragment thr:off Encryption key:off Power Management:on Link Quality=70/70 Signal level=-30 dBm Rx invalid nwid:0 Rx invalid crypt:0 Rx invalid frag:0 Tx excessive retries:286 Invalid misc:11 Missed beacon:0 What's going on here? I ran "iwconfig wlan0 rate auto" which made it briefly claim a rate of "130 Mb/s" but then it reported 1Mb/s again. 1Mb/s rate for the Wifi connection would be expected to give something like 40KB/s FTP transfer speed so it seems that the phones are connecting in the same way as my laptop. It says that "abgn" is supported which should give lots of options for rates higher than 1Mb/s, especially when the quality is 70/70 and the Wifi access point (a Netgear DGN 2200 ADSL modem) is about 2M away from my laptop with nothing of note in between. Any suggestions for what I can do to improve performance? -- My Main Blog http://etbe.coker.com.au/ My Documents Blog http://doc.coker.com.au/

6 6

I386 V AMD64 Debian part 2
by zlinw＠mcmedia.com.au 02 Nov '13

02 Nov '13

During a recent discussion on i386 V AMD64 linux a number of people said stay away from Physical Address Extension PAE kernels, althougth no reasons/evidence was given to back up the warning, What is the problem with PAE. A search on the net showed no major problems ascociated with this extension apart from the obvious limited amount of memory avaible for a single process. As stated I have run PAE kernels now for some time with no problems what so ever. On Debian AMD64 I have managed to get wine running but as yet I cannot compile wine itself, the build complaining of no 32 bit compiler (which is correct). Sadly Debian sid will not let me install one without removing alot of packages including xorg, this certainly not what I want. A search on the net came up with a good bit of info on both running and compiling wine on Debian AMD64 unfortunately none of it was recent and some of the info conflicked. So I am going to have a crack at Debian 7.0.0 i386 which I have a set of DVD's for, having obtained them "just in case" . If its required re intallation of Sid is a non issue with apt-move and saving the package selections. Although there's one minor gripe apt-move will store the AMD64 packages OK but will not handle the i386 items (on the system for wine). The actual amount of data invloved is only around 100meg so its not a real major issue. Lindsay

3 2

Re: I386 V AMD64 Debian part 2
by zlinw＠mcmedia.com.au 02 Nov '13

02 Nov '13

Thanks for the reply, it is apreciated. Russell Coker said, > The limit on per-process address space (which isn't just a limit on memory > use) is the main thing. Also there are performance issues with PAE. The problem with claims of performance differences is that an awefull lot of things can effect this and I have only seen one report (of 32 v 64 bit) giving full details and test results of said tests. These showed some things were faster and some were not. There are plenty of sites claiming performance enhancements of 64 bit, almost none (only a single site) gives any real evidence, most sites __clearly__ cherry picking data. This sort of thing makes me __very__ sceptical as if something really is faster one will see benchmarks published all over the place. > Have you enabled multi-arch? Yes, the last remaining problem is getting the closed src driver working everything I have tried so far leads to the same thing. The X server crashing after start up from a memory violation to do with a font library. This is also with Debian Stable AND sid, four different kernels, two of Debian's the others my own and of course two differnt vserions of xorg. I am going to try i386 to see what will happen, I may add I have never had the current graphics card an NVidia GTX 680 working on a closed src driver under linux, works __mostly__ great under the open src driver. The card works great under Windows XP. Lindsay

2 1

Opera not rendering page correctly
by Terry Duell 01 Nov '13

01 Nov '13

Hello All, I am using Opera 12.16 on Fedora 19 x86_64. Recently Opera began to incorrectly render the page at <https://groups.google.com/forum/#!forum/hugin-ptx>. Here, it renders a lot of text, which looks like a lot like page source code, in the top 3/4 of the page, and the normal forum posts right at the bottom of the page. I posted a question on the Opera forum, and it was suggested that I clear the cache and restart Opera. That actually worked...for a very short time, then it was back to the same problem. I have tried clearing the cache again, but it now has no effect. The page renders OK on Firefox. I guess I could use Firefox, but would prefer to keep on with Opera if I can. Anyone have any ideas on how this might be fixed? If it is of any help, I can capture a screen and post the image on Dropbox. Cheers, -- Regards, Terry Duell

1 0

re: LUV Main November (Wednesday): MMORPGs + Parsing/Perl
by Toby Corkindale 31 Oct '13

31 Oct '13

I think you have Jacinta's surname wrong in the announcement below.. unless by a huge coincidence there are two Jacinta's doing Perl training in Melbourne. On 30/10/13 11:29, Les Kitchen wrote: > Start: Nov 6 2013 19:00 *Wednesday* > End: Nov 6 2013 21:00 > Location: > The Buzzard Lecture Theatre. Evan Burge Building, Trinity > College, Melbourne University Main Campus, Parkville. > Link: http://luv.asn.au/meetings/map > > Lev Lafayette: MMORPGs and Linux > > Jacinta Richards: Don't repeat yourself. Fun with parsing > > NOTE: Because of Cup day, we meet this month on *Wednesday*. > > Lev Lafayette: MMORPGs and Linux > > Lev will talk about Massively Multiplayer Online Role-Playing > Games and Linux. Lev is currently LUV President. > > Jacinta Richards: Don't repeat yourself. Fun with parsing > > The problem: your client gives you a document covering the > specification of some weird message format they want you to > handle. It's tab delimited, and pure ASCII, but actually some > fields are multi-part and include tabs, and nothing is > quoted. Oh, and actually there's multiple different kinds of > messages you will need to parse. > > The solution: Using Perl 5.10+ and named capture groups we'll > create a data definition. Over the rest of the talk we'll be > taking advantage of a host of tactics (including re-parsing a > regular expression) to see how we can take an interesting > problem and make it fun again. Although Perl will be used in > this talk, and code will be shown, hopefully you won't actually > need to know any Perl at all to understand the ideas I run > through. > > Jacinta says: I speak for a living. I run Perl training courses, > I speak at Perl monger meetings, user groups and numerous > conferences across the world.

2 1

CentOS 5 change from IDE to RAID after cloning
by Tony White 30 Oct '13

30 Oct '13

Hello all, I am hoping someone has done this before as I am unable to locate anything on the web to help me. I have a drive that has been cloned to a RAID controller but when it boots it only gets to to "Kernel Panic". The only reason I can see is the RAID driver is not loading. Can someone tell me how/where I can make this change without having to do a full reinstall please? TIA -- best wishes Tony White

7 8

kpartx for netbsd
by James Harper 30 Oct '13

30 Oct '13

This is kind of OT, but I'm hoping someone here is familiar with both Linux and BSD. Is there a BSD equivalent of kpartx, or is kpartx available for BSD? I'm exchanging emails with someone who is using BSD and trying to assist in solving a problem, but I know nothing about BSD. Thanks James

2 1

C library for parsing email addresses
by Russell Coker 24 Oct '13

24 Oct '13

Does anyone know of a good free library for parsing email addresses? Preferrably something that's already packaged for Debian. I want to have a library sort out the local part, the domain, and the comment of an email address. I've just discovered a bug in a program which does this. The fact that the program in question has been running for many years (last update was over a year ago and the code in question wasn't changed for some time before that) and processed probably hundreds of thousands of messages without the bug being noticed demonstrates that it's not so easy to get this right. It seems likely that there may be other bugs after I fix this one. So I'd rather use a library to ensure that it's had wider testing before I use it. -- My Main Blog http://etbe.coker.com.au/ My Documents Blog http://doc.coker.com.au/

7 9

NVidia closed src drivers crashing after startup
by zlinw＠mcmedia.com.au 22 Oct '13

22 Oct '13

I am just trying to install the NVidia closed src drivers on Debian 7.1 AMD64 and the X server is crashing imediatly after start up below is the final part of the Xorg.log, [ 464.993] (**) Option "xkb_model" "pc105" [ 464.993] (**) Option "xkb_layout" "us" [ 464.994] (II) config/udev: Adding input device PC Speaker (/dev/input/event1) [ 464.994] (II) No input driver specified, ignoring this device. [ 464.994] (II) This device may have been added with another device file. [ 465.044] [ 465.044] Backtrace: [ 465.044] 0: /usr/bin/X (xorg_backtrace+0x36) [0x7f2f4ebe5d06] [ 465.044] 1: /usr/bin/X (0x7f2f4ea67000+0x182859) [0x7f2f4ebe9859] [ 465.044] 2: /lib/x86_64-linux-gnu/libpthread.so.0 (0x7f2f4dd8f000+0xf030) [0x7f2f4dd9e030] [ 465.044] 3: /usr/lib/libXfont.so.1 (FontFileListNextFontWithInfo+0x29) [0x7f2f4d8da5b9] [ 465.044] 4: /usr/bin/X (doListFontsWithInfo+0x163) [0x7f2f4eaba473] [ 465.044] 5: /usr/bin/X (ProcessWorkQueue+0x21) [0x7f2f4eabe0b1] [ 465.044] 6: /usr/bin/X (WaitForSomething+0x62) [0x7f2f4ebe3012] [ 465.044] 7: /usr/bin/X (0x7f2f4ea67000+0x52bb1) [0x7f2f4eab9bb1] [ 465.044] 8: /usr/bin/X (0x7f2f4ea67000+0x41ec5) [0x7f2f4eaa8ec5] [ 465.044] 9: /lib/x86_64-linux-gnu/libc.so.6 (__libc_start_main+0xfd) [0x7f2f4cab6ead] [ 465.044] 10: /usr/bin/X (0x7f2f4ea67000+0x4219d) [0x7f2f4eaa919d] [ 465.044] [ 465.044] Segmentation fault at address (nil) [ 465.044] Fatal server error: [ 465.044] Caught signal 11 (Segmentation fault). Server aborting [ 465.044] [ 465.044] Please consult the The X.Org Foundation support at http://wiki.x.org for help. [ 465.044] Please also check the log file at "/var/log/Xorg.0.log" for additional information. [ 465.044] [ 465.088] (II) evdev: Logitech USB Optical Mouse: Close [ 465.088] (II) UnloadModule: "evdev" [ 465.088] (II) evdev: AT Translated Set 2 keyboard: Close [ 465.088] (II) UnloadModule: "evdev" [ 465.251] Server terminated with error (1). Closing log file. Under a normal start the lines at 464.994 are the last ones in the log after startup. So the driver appears to be starting up OK. I use a custom kernel current version is 3.9.11, NOUVEAU being blacklisted for the closed src driver. A search on the net produce no usefull information. The NOUVEAU runs perfectly except the 3D accelartion is slow and has some texture mapping problems. For normal operations I prefer the OS driver, its no good though for OpenGL development. Lindsay

6 7

pros and cons of AMD64 and i386 Linux on a home system
by zlinw＠mcmedia.com.au 18 Oct '13

18 Oct '13

The situation is this I require to be able to run wine on my systems here (Debian AMD64)and I am finding this something of a pain with the current Debian. The version of wine required is 1.6. To install the develpment enviroment for wine ineffect requires an almost complete parrellel install of a 32 bit system. It seems to me it would be easier simply to install i386 and avoid the pain. There it appears only to be a minor issue in that the maximum memory for a single process would 4 gig and I have yet to see any program even gimp editing full size images from my Nikon D700 come even close to that figure. I have checked on the net for benchmarks between i386 and AMD64 and there is no obvious overall install that is faster. It would seem that the simplest path would be to install Debian 7 i386, this would allow the latest wine to be installed without any problems and I cannot myself see any down side. The system would still have access to the all the memory the only limitation being the 4gig process limit. There may be an issue with the speed of memory access on some of the latter video cards with large amounts of memory availible as this memory will need to be bank switched. This does not appear to be a problem at this stage. Hmmmmmmmmmmm, I have two main systems both efectively identical I will install i386 on one of these this will allow wine to be installled easily and the other one can be used to allow a comparison to be done between the two versions (i386 and AMD64) using 3D intensive software (a terrain simulation) to see what if any differences exist. I do have Debian i386 readily availible. Any comments would be apreciated! Lindsay

3 4