IRC Logs for #crux Tuesday, 2015-02-10

*** j^2 has joined #crux00:13
*** Workster has quit IRC00:41
*** fengshaun has joined #crux01:05
*** tired890 has joined #crux01:21
Romsterkyubiko, i guess needs to be packaged01:48
Romsterbut it isn't required to run tor01:48
Romsteronly to make other /stuff/ work with tor01:48
Romsteri'd recommend privoxy and set up stuff though that.01:49
*** tired890 has quit IRC01:49
*** leo-unglaub1 has joined #crux02:02
*** leo-unglaub1 has quit IRC02:03
*** leo-unglaub has quit IRC02:03
*** tired890 has joined #crux02:11
*** Workster has joined #crux02:17
*** Workster has quit IRC02:17
*** Workster has joined #crux02:17
greduanand I bet it works way better02:51
greduanof course it does02:51
greduanI wonder if somebody could make a clone that just removes all the bugs...02:51
jaegerwouldn't be a clone, then. :)02:56
prologicI've actually tried this simple little demo03:08
prologicit's really good03:08
prologicquite performant too03:08
*** Feksclaus has quit IRC03:16
greduanjaeger: :O03:35
greduanMinecraft: The Rewrite03:35
*** greduan has quit IRC03:40
*** mavrick61 has quit IRC03:40
*** mavrick61 has joined #crux03:41
*** tired890 has quit IRC05:46
Worksteri should try this out
Worksteradd it to my and profit06:23
nwehello Workster ?07:04
Romsterhi nwe07:10
nwehow are you Romster?08:00
nwedoes someone know a simple wiki for sale ppl ? :)08:00
*** tired890 has joined #crux08:11
nweRomster: I will take a look on it08:45
*** tired890 has quit IRC08:55
prologicnwe: also powers
prologicit's simple in that it doesn't have a very complex feature set, uses standard creolwiki markup and is backed by mercurial09:02
prologicso you can directly edit the content on disk09:02
*** vlnx has quit IRC09:03
nweprologic: I mean more that the persons dont need to use <b> /<i> etc..09:09
prologicthey won't09:14
prologicjust **bold** *itatlic*09:14
prologicnormal creole wiki markup09:14
prologicsahriswiki doesn't have a wysiwig editor as such yet09:14
prologicbut those are normally client-side anyway so it would not be hard to add one09:15
prologicbut that being said there are such wikis with such features09:15
prologicalthough quite feature-rich though (feature fatigue)09:15
*** jue has quit IRC09:18
*** rauz has quit IRC09:21
*** rauz has joined #crux09:22
*** tired890 has joined #crux09:27
*** jue has joined #crux09:35
*** vlnx has joined #crux09:36
*** jue has quit IRC09:40
*** jue has joined #crux09:53
*** tired890 has quit IRC09:59
*** sh4rm4 has joined #crux10:05
*** tired890 has joined #crux10:21
*** jue has quit IRC10:28
*** jue has joined #crux10:40
*** tierd891 has joined #crux10:40
*** tired890 has quit IRC10:42
cruxbot[opt.git/3.1]: hexchat: clean footprint10:58
*** jdolan_ has joined #crux11:13
*** jdolan has quit IRC11:13
*** sh4rm4 has quit IRC11:28
*** sh4rm4 has joined #crux11:31
*** sh4rm4 has quit IRC11:41
*** sh4rm4 has joined #crux11:42
*** sh[4]rm4 has joined #crux11:54
*** sh4rm4 has quit IRC11:54
*** tierd982 has joined #crux12:00
*** tierd891 has quit IRC12:03
*** Feksclaus has joined #crux12:26
frinnstso fucking awesome12:42
frinnst"Why Ethernet cables? If you store your music on a NAS (Network Attached Storage) device, Ethernet cables are part of your hi-fi."12:43
frinnstbest quote ever12:43
tvaalenGuaranteed to make your internetz faster.12:43
Romsteri saw the lan cable but not the rest12:45
Romsterwhat a rip off12:45
frinnsti am still speechless12:45
Romsteranyone that buys those are idiots12:45
frinnstAssuming the bits are perfectly transmitted, factors such as ground plane noise could still effect the sound; cables like these might make a difference.12:47
tvaalenRomster, or a rich asshole.12:47
tvaalenLike people who bought those iPhone apps for $999 that didn't do shit.12:48
Romsterhaha really12:50
*** greduan has joined #crux13:01
greduanGood morning :)13:01
teK_frinnst: are you using btrfs in production?13:03
diversetvaalen: seriously?13:05
Romsteryeah but if they are rich they don't spend it on crap like that... how do you think they get rich save and spend very little13:08
tvaalenSure, but there's a difference between being rich and being a rich asshole showing off.13:09
diversemore like difference between being smart and a dumbass. There is nothing impressive about somebody being rich + stupid13:14
diverseheck, they if really want to be loose on spending, they should spend +10,000 USD on F2P games, that's a more effective way of being stupid.13:15
Romsteryeah and benefits the game company13:18
diversewell usually it's indie developers13:18
cruxbot[core.git/3.1]: lzo: update to 2.0913:41
cruxbot[core.git/3.1]: sudo: update to 1.8.1213:41
cruxbot[opt.git/3.1]: hpcups: update to 3.15.213:43
*** teK__ has joined #crux13:44
*** teK__ has quit IRC13:44
*** tierd891 has joined #crux13:54
*** tierd982 has quit IRC13:57
*** jdolan_ has quit IRC14:08
*** SiFuh has quit IRC14:09
*** jdolan has joined #crux14:38
*** nogagplz has quit IRC14:46
*** nogagplz has joined #crux14:49
*** sh[4]rm4 has quit IRC14:53
*** sh4rm4 has joined #crux14:54
diverseRomster: I splitted the shit, it still compiles :P15:06
diverseteK_: how is your studies going?15:20
teK_I have been doing project work for the last 5 days, so no progress with the studies15:24
diverserelated to the file descriptors stuff in python?15:27
teK_na that's for my studies15:27
teK_the project is organisational work mostly. And some coding in a proprietary env/language15:28
diverseby 'organizational' you mean project management? Hmm a proprietary env/language... SmallTalk?15:30
teK_yes, management. The language is Baanscript15:34
teK_my uneducated guess is that it's something like ABAP but not for SAP :)15:35
teK_it has embedded SQL .. you can iterate over data sets just as you could with iterators/loops15:36
diversethats nice15:36
teK_actually that's a nice feature, yes. On the other hand .. it has shitty multidimensional array semantics and if you want to use recursion, the compile from 2005 and the one from 2014 will warn you that 'recursion is not yet implemented'15:37
diverseno optimization flags?15:38
teK_I could not find any documentation for the compiler. And you cannot convert every type of recursion to an iterative version.15:39
teK_let alone doing that automatically15:39
diversepretty hard to do15:40
diversefor the compiler/interpreter15:41
teK_so they chose not do implement support at all. :)15:41
teK_I got used to it. It also is better wrt. stack usage ;-)15:41
diverseright, recursion can cause stack overflowing15:42
diversewell, it either way, it sounds like baanscript would be a better alternative than ms sql15:44
teK_ooh that beast will be beneath the whole data store, once the project is done15:44
teK_we're switching from SLES+Oracle to 3xWIN+MSSQL15:44
teK_not my choice :)15:44
teK_also you usually do not have to fiddle with mssql15:45
teK_oh and babysitting my key users will be the most part of future work until the go live, I guess :)15:46
diverseah, that will be a pita15:46
teK_I will try to outsource part of the nagging to the former project leader; he has a natural talent in that15:47
teK_did so with everybody involved on our and the supplier side :D15:47
teK_the vendor project lead promised to treat me to food and beer if I take over :D15:48
diversehey, free lunches! \o/15:49
teK_I will remind of that once we're done15:49
teK_after all he's a consultant and maybe you already made contact with that species15:50
diversestill itching to do that rewrite when you have time?15:52
teK_yes. Thought about the shell-parsing problem just this morning in the bathroom :P15:53
diverseah I thought of that out already15:53
teK_in plain C?15:53
diversein Rust15:54
teK_well :D15:54
diversebut it's the idea that counts15:54
teK_I am not rusty ;)15:57
diversebasically, you want to scan the pkgfile with a buf reader of some kind line-by-line and parse the variable and header information, download the distfiles, and spawn a bash process on the pkgfile15:57
teK_the question is who downloads and builds the sources15:58
teK_I am already parsing header information15:58
diverseyou write code (like with curl) to download, extract the files in the right places, and spawn a bash process reading the pkgfile executing the build function15:59
teK_you have to implement "full" shell parsing in Rust for that16:00
diversenot really16:00
teK_source=() can contain shell substitutions and there are no limits on the extents of shell features, a Pkgfile may use16:00
teK_in fact this is a feature :)16:03
diverseright, but I want bash to do the work16:05
teK_that's the thing I thought about today :)16:05
diversesh doesn't have arrays though?16:06
teK_you are right16:07
teK_most interesting candidates for sh are bash, dash (and zsh but I am biased here)16:07
teK_they all have arrays :x16:08
diverseif moving to pure sh is the attractive approach, using `source=()` as a tag `Source: ` would be an idea16:11
diverse*`# Source: `16:11
diverseand since that would be commented, you wouldn't need to worry about shell wizardry16:12
diverseteK_: ^16:20
greduan(m)ksh could also be an option, no?16:24
greduanbash I don't think should be sh, it's quite slow compared dash or mksh. mksh being an interactive shell and almost as fast as sh.16:25
*** tilman has joined #crux16:26
diversegreduan: that's not the point of our discussion. However dash is being considered as the sh alternative for the system16:28
frinnstdiverse: ok so you'll coordinate the switch with all 3rd party repos? okthx16:28
greduandiverse: Oh OK. What is being discussed then?16:29
frinnstif/when we switch /bin/sh to dash, bash will still be required for things ro work16:29
diversefrinnst: that's not up to me, that's up to teK_ and jaeger :P16:29
frinnst3.2 *will* still require bash btw16:29
diverseof course, but the idea is, have sh symlink to dash only16:30
koriI've that here16:31
*** jdolan has quit IRC16:38
*** deus_ex has joined #crux16:54
*** Pingax has joined #crux17:08
*** lnds has joined #crux17:13
*** SiFuh has joined #crux17:19
*** sh4rm4 has quit IRC17:24
greduanatm CRUX uses gcc for compiling, right?17:29
greduanAny plans to move to clang/LLVM?17:38
*** pejman has quit IRC17:38
*** pejman has joined #crux17:43
teK_not unless upstream supports it18:28
*** hhhhhhhh has joined #crux18:37
teK_and tehre are packages that won't compile with llvm.18:37
juegreduan: bash isn't slow at all, I've made a real-life comparison with dash, see here ->
juegreduan: there is a performance gain, but only noticable with a lot of sh calls18:38
greduanI've heard so many complaints about bash being slower than sh...18:39
greduangotta confirm before you say it I guess18:39
greduanthanks :)18:39
greduanjue: How does (m)ksh compare to bash and/or sh? I'm curious now18:40
*** tilman has quit IRC18:41
juedidn't tested it in the same depth than dash, but I ran in some build problems with it18:41
juesorry, don't remeber which ports fails building with mksh18:42
juethe only port that fails building with dash I found so far is opt/cppunit18:43
juebut I don't use many ports, so there might be some more18:45
*** jdolan has joined #crux18:49
greduanI haven't ran into any problems when building with mksh, yet18:50
cruxbot[xorg.git/3.1]: xorg-xf86-video-modesetting: removed, bundled with xorg-server18:51
*** tilman has joined #crux19:27
*** Pingax has quit IRC19:37
*** Legonat0r has joined #crux19:38
jueFTR, opt/cppunit fails building with mksh as /bin/sh as well19:39
*** Legonat0r has left #crux ()19:40
*** hhhhhhhh_ has joined #crux19:44
greduanjue: the Pkgfile doesn't seem to have anything that would make it not work?19:46
greduanIs it just the software itself complaining?19:49
*** hhhhhhhh_ has quit IRC20:07
*** heroux has quit IRC20:09
*** hhhhhhhh_ has joined #crux20:13
*** hhhhhhhh_ has quit IRC21:06
*** hhhhhhhh_ has joined #crux21:12
diverseRomster: why did you remove aria2 from your repo?21:18
diverseoh, one already exists in contrib21:20
*** lnds has quit IRC21:26
*** heroux has joined #crux21:36
*** sh4rm4 has joined #crux21:51
Romsterit was a duplicate of contrib21:56
*** cinolt has joined #crux22:32
*** hhhhhhhh_ has quit IRC22:35
cinoltHi. I'm having some trouble installing CRUX 3.1 onto a VMware Player 7.0.0-2305329 virtual machine. It has no trouble booting from the CRUX installation .iso ao a virtual CD, and I have followed the instructions pretty straightforwardly from the handbook. However when I then boot from the harddisk where I installed CRUX, I get a kernel panic as shown in the following link. All other relevant information is also provided. http://paste.lis22:55
cinoltAny suggestions?22:55
teK_url does not work22:56
*** mavrick61 has quit IRC22:56
*** mavrick61 has joined #crux22:56
prologicyeah broken url22:56
prologicalso you will likely have not configured your sata/scsi/fs properly in your kernel config22:57
teK_my guess is, you either missed the needed FS option or ide/sata/scsi driver for your disk. Also note that these must not be compiled as a module [m] but statically [y] into your kernel22:57
prologicthis is the most common cause22:57
prologicditto on what tek said22:57
cruxbot[xorg.git/3.1]: [notify] xorg-server: updated to 1.17.1. Resolves CVE-2015-025523:00
frinnstholy shit. my 3.5 year old mozilla bug was just marked as RESOLVED23:01
Worksterlook for virtulisation for guest settings and lspci -k23:02
Workstergo search in make menuconfig with /23:03
Workstercinolt, be sure disk contorller is compiled in as is filesystem for root and guest virutal machine modules23:03
frinnstmy previous record bug was 7 or 8 years23:04
Workstertook em long enough23:04
teK_my bet is that gcc or glibc can beat that easily23:07
cinoltBack, sorry guys I was AFK for a bit.23:09
cinoltThats weird the link works for me. Ill reupload it somewhere else23:09
cinoltI'm sure I got the correct filesystem compiled in, I must be missing a device driver somewhere.23:11
frinnstteK_: actually crux can too :D23:12
*** hhhhhhhh has quit IRC23:16
jaegercinolt: the .config that gets installed by setup should cover the majority of hardware and filesystems... what disk controller does the VM use?23:17
cinolt@jaeger: According to lspci, "00:10.0 SCSI storage controller" LSI Logic / Symbios Logic 53c1030 PCI-X Fusion-MPT Dual Ultra320 SCSI (rev 01)"23:20
jaegerdoes lspci -k list a module for that controller? mpt2sas?23:22
jaegerok, that probably needs CONFIG_FUSION_SPI which is actually one of the ones that's NOT builtin by default, it's module23:23
*** kori has quit IRC23:24
cinolt@jaeger: thanks, ill try that right now23:25
cinoltThere we go. Now I at least got something able to boot. Thanks all.23:30
jaegergood deal23:33
*** cinolt has quit IRC23:35

Generated by 2.11.0 by Marius Gedminas - find it at!