T2 IRC Log: 2010-09-15

This is the log as captured by an IRC bot in the channel. The statements are those of the individual people and might not neccessarily reflect the policy and legal rules as set forth by the T2 SDE Project.

« prev | next »

--- Log opened Wed Sep 15 00:00:49 2010
10:32 <@rxr> http://www.t2-project.org/packages/samba.html
10:50 <@rxr> cool, conrad has 4 pin plugs, so I can repair the headphone that just broke: http://www.conrad.de/ce/de/product/730208/KLINKENSTECKER-35MM-4-POLIG-VERGOLDET
10:58 <@rxr> heh, they have an better one: http://www.segor.de/suche.shtml?foto=1&Q=linkenstekcer&M=1
10:59 <@rxr> Segor is an insider tip anyway :-)
11:09 < CIA-32> rene * r37619 /trunk/package/filesystem/dosfstools/dosfstools.desc: * updated dosfstools (3.0.9 -> 3.0.10)
11:09 < CIA-32> rene * r37620 /trunk/package/network/samba/samba.desc: * updated samba (3.5.4 -> 3.5.5)
11:13 < CIA-32> rene * r37621 /branches/8.0/package/network/samba/samba.desc: * merged r37620 from trunk: updated samba (3.5.4 -> 3.5.5) - SECURITY
11:31 < CIA-32> rene * r37622 /trunk/package/multimedia/lives/lives.desc: * updated lives (1.3.6 -> 1.3.7)
17:52 < rogermason> Hello. I just discovered that context (alternative to latex) needs ruby to be installed. That means texlive has a dependency on ruby. How do I get that dependency into the .cache file?
18:01 <@rxr> normally it get's automatically in there by monitoring the users files in the t2 build
18:01 <@rxr> that is the var/adm/cache/texlive from the newly build package
18:08 < rogermason> ruby is not in the texlive cache. Does T2 make the cache file by moitoring what happens during the configuration of a package? So for example if texlive _failed_ because of the absense of ruby, would ruby then be listed as a dependency?
18:09 <@rxr> it monitors the whole build
18:09 <@rxr> the .cache is only useful from a successful build
18:10 <@rxr> if ruby was run it should be in the .cache from that successful build
18:10 <@rxr> unless there is something broken, ...
18:11 <@rxr> one can manually tweak the cache with [E] / [DEP] tags in the .desc, but that should be avoided, ...
18:14 < rogermason> well, there is nothing in the texlive _build_ process that requires ruby - otherwise the build would fail if ruby was not found. Ruby is a _runtime_ dependency of context, itself a component of texlive. A user will only discover that context's 'texexec' command does not work when they try to compile a context document. Unfortunately, texexec fails silently in the absense of ruby, so there is no helpful ouput.
18:15 <@rxr> oh
18:15 <@rxr> that sucks
18:15 <@rxr> ok, well, you could add the deps in the .desc with: [E] + ruby
18:15 <@rxr> IIRC
18:16 <@rxr> then the dep gets addded to the .cache during a build although ruby was not used
18:16 <@rxr> but IMHO a tex that requres ruby sucks, ...
18:16 <@rxr> I would not want to have ruby just to run tex
18:29 < rogermason> I agree. It is just context that uses ruby. The context folks are moving away from ruby towards lua.
18:31 < rogermason> Perhaps instead of adding another dependency, I could patch texexec after the installation so that it issues a message saying that ruby is required.
18:32 <@rxr> well, maybe just add the dep for now :-)
18:33 < rogermason> OK.
18:54 < rogermason> I also found that latex looks for some things in /opt/texmf and /opt/texmf-dist instead of /opt/texlive/texmf and /opt/texlive/texmf-dist. I have no idea why, given that the prefix is established right at the beginning of the build. The issue can be fixed by linking /opt/texmf -> /opt/texlive/texmf. I assume those links can be created on a post-install hook?
19:02 <@rxr> ouhm, sounds ugly
19:02 <@rxr> guess there you should hunt the real cause
19:17 < rogermason> Oddly, the problem does not manifest itself when texlive is installed using Emerge-Pkg.
20:26 -!- keinek [~leonel@host16.190-136-29.telecom.net.ar] has joined #t2
20:26 < keinek> hi :)
20:54 < ch172> I tried to build an iso with kernel 2.6.35.4, but the unionfs patch fails, is there any fix for it?
21:11 -!- keinek [~leonel@host16.190-136-29.telecom.net.ar] has quit [Ping timeout: 240 seconds]
22:20 <@rxr> ah yes we need yo update or rediff the patchh
22:20 <@rxr> can do tomorrow
--- Log closed Thu Sep 16 00:00:49 2010