Go to:
Gentoo Home
Documentation
Forums
Lists
Bugs
Planet
Store
Wiki
Get Gentoo!
Gentoo's Bugzilla – Attachment 79297 Details for
Bug 114944
Problems with ciaranm atagonizing other devs!!
Home
|
New
–
[Ex]
|
Browse
|
Search
|
Privacy Policy
|
[?]
|
Reports
|
Requests
|
Help
|
New Account
|
Log In
[x]
|
Forgot Password
Login:
[x]
02-08-06 #-dev log
log (text/plain), 39.10 KB, created by
Brian Harring (RETIRED)
on 2006-02-08 21:39:45 UTC
(
hide
)
Description:
02-08-06 #-dev log
Filename:
MIME Type:
Creator:
Brian Harring (RETIRED)
Created:
2006-02-08 21:39:45 UTC
Size:
39.10 KB
patch
obsolete
>17:46 <@ferringb_> ciaranm: actually, jstubbs does not agree with you last I looked. >17:46 <@brad[]> Certain proprietary operating systems start requisite services as part of the desktop environment >17:46 <@ferringb_> nor does genone >17:46 <@ciaranm> jstubbs implemented it for me >17:46 <@ferringb_> think I was the only one who thought it was a half way decent idea to jam it in there; the proposed solution was addition of metadata, not abusing existing. >17:46 <@ferringb_> ciaranm: clarify. >17:46 <@ciaranm> he wanted to find out whether it would break anything >17:46 <@ciaranm> it didn't >17:48 <@ferringb_> ciaranm: again, clarify. >17:48 <@ferringb_> bug #, explicit statement of what he actually implemented, etc. >17:48 <@ciaranm> certain people were spreading sneaky rumours that changing USE_EXPAND would make portage explode >17:49 <@ferringb_> ciaranm: just stick to the facts please >17:49 * SuperLag waits for the mushroom cloud >17:49 <@spock> blackace: AFAIK Fedora uses Xvesa for rhgb >17:49 <@ferringb_> ciaranm: what are you specifically referencing that he implemented; simple question >17:49 <@spyderous> ajax: as i recall, it actually caused problems to start kdrive, then stop it, then start regular X >17:49 <@ciaranm> ferringb_: wouldn't you rather infer things from vaguely worded truths? >17:49 <@ferringb_> ciaranm: I'd rather de-op you and throw you out of here. >17:50 <@ferringb_> but right now I'm going through the proper channels. >17:50 <@ferringb_> so play nice, as I've attempted. >17:50 <@ciaranm> ferringb_: the details were provided in the channel where this was under discussion. you weren't there, and this doesn't concern you, so stop poking around >17:51 <@ferringb_> gah... do I have to sit in here to watch your statements? >17:51 <@ferringb_> just answer the question >17:51 <@ciaranm> no no, no need to join if you'd rather not. you could not get involved if you prefer >17:52 <@ferringb_> bleh, still working well with others I see. >17:52 <@Halcy0n> ferringb_: you aren't helping. >17:52 * antarus sighs >17:52 <@ferringb_> Halcy0n: sorry, I respond to sarcasm with sarcasm. >17:52 <@ferringb_> meanwhile I'll just ask jason specifically, bit easier. >17:52 <@ciaranm> sure. i'm working just fine with the qa team. who are in another channel, discussing the issue properly, rather than relying upon forwarded half-statements from a peon >17:53 <@antarus> ciaranm, to be clear I pasted the entire conversation >17:53 <@antarus> so if they were half statements you made then :) >17:53 <@antarus> s/then/them >17:53 <@ferringb_> ciaranm: and I'm asking because no such changes were made, if they were I want to know when/how >17:53 <@ciaranm> yes, but the people involved in said conversation already knew about what i was talking >17:53 <@ferringb_> and since you're stating it, you presumably know when it was modified. >17:54 <@ciaranm> you've had enough bug and -dev mail about it >17:54 * ferringb_ sighs >17:54 <@antarus> you jokingly(?) said you would modify the source of another project of which you aren't a member >17:54 <@ciaranm> antarus: uh, no i didn't >17:55 <@ciaranm> what's the deal with certain people inventing things and claiming i said them? >17:55 <@antarus> That was my interpretation >17:55 <@antarus> my apologies if thats not what you meant >17:55 <@ciaranm> now read what i actually said >17:56 -!- zmedico [n=zmedico@cpe-66-27-157-183.socal.res.rr.com] has joined #gentoo-dev >17:56 -!- mode/#gentoo-dev [+o zmedico] by ChanServ >17:56 <@antarus> <antarus> ciaranm: b) i can do this quite easily without portatge commit access >17:56 <@ciaranm> yupyup >17:56 <@ferringb_> antarus: key thing is that there is no indication he requires portage modification for it. >17:56 <@ferringb_> word games. >17:56 <@rane> who can fix a typo in man euse? >17:56 <@ciaranm> i don't require portage modification for it >17:57 <@ferringb_> or misinterpretation, either way it's a stupid thing to get the knickers in a twist over. >17:57 <@rane> ferringb_: could you? >17:57 <@ferringb_> rane: not right now (no cvs access) >17:57 <@antarus> rane, I can >17:57 <@ciaranm> and i don't require portage modification because jstubbs changed portage to not require modification for it >17:57 <@rane> antarus: good >17:57 <@ciaranm> is all now becoming clear? >17:58 <@antarus> see that only took like 15 minutes of arguing ;) >17:58 <@rane> antarus: there's line "[- ] moznocompose (net-www/mozilla):.br Disable building of" - just press an enter before .br :) >17:58 <@rane> antarus: thanks in advance :) >17:58 <@ferringb_> antarus: annoying thing is, still need to find out what the change exactly was. :) >17:58 <@ciaranm> antarus: dude. Halcy0n understood it just fine straight off >17:58 <@antarus> good for him, some of us are "peons" and take a bit longer >17:59 <@rane> like me >17:59 <@ciaranm> so, if you didn't understand, why did you run off crying that i was going to modify portage? >17:59 <@ferringb_> ciaranm: he dumped me the log because you were being a tool; called venting. >17:59 <@antarus> I pasted because someone asked for it >17:59 <@ferringb_> one particular chunk of your statement I wondered about, thus I popped in. >18:00 <@ferringb_> from there... the pissing fest. :) >18:00 <@antarus> after I commented that you were being your usual fun self >18:00 <@ciaranm> ferringb_: uh, no, i was having a technical discussion with someone who understood exactly what i was saying, and then the peanut gallery stepped in >18:00 <@ferringb_> oy. >18:01 <@ciaranm> now, since you asked so politely and didn't jump to any silly conclusions, the bug starts with an 8, and has USE_EXPAND in the title >18:01 <@ciaranm> if you want more than that, you'll have to wait for me to grep for it >18:03 <@ciaranm> !metadata bitpim >18:03 <+jeeves> ciaranm: Package: app-mobilephone/bitpim Herd: mobile-phone Maintainer: mobile-phone Description: BitPim is a program that allows you to view and manipulate data on LG VX4400/VX6000 and many Sanyo Sprint cell phones. This includes the PhoneBook, Calendar, WallPapers, RingTones (functionality varies by phone) and the Filesystem for most Qualcomm CDMA chipset based phones. >18:03 <@ferringb_> I don't think he added features to use_expand there... >18:03 <@ferringb_> I'd bet he added test pass through >18:04 <@ciaranm> ferringb_: no. he moved USE_EXPAND to be a profile setting rather than a make.globals setting >18:04 <@ferringb_> if features was being expanded by USE_EXPAND, it would be prefixed. >18:04 <@ciaranm> yup >18:04 * spyderous resyncs for fluxbox fix >18:05 <@ciaranm> spyderous: that's Halcy0n's fault btw >18:05 <@spyderous> ok >18:05 <@Halcy0n> Hrm? >18:05 <@ciaranm> you told me i could delete all those files :P >18:06 <@Halcy0n> I didn't tell you to break anything though :) >18:06 <@ciaranm> pffff. >18:07 <@spyderous> would people prefer to just have a VIDEO_CARDS setting for all ati cards, or to have it split out into radeon, rage128 and mach64? >18:07 <@ciaranm> mmm. fatal: app-text/ggv-2.8.5::x-/usr/portage: Couldn't generate metadata >18:08 <@ciaranm> spyderous: split it! >18:08 <@ciaranm> no point in being half-arsed about modular deps >18:08 <@spyderous> got a decent reason, or just because modularization is a cool fad >18:08 <@spyderous> well it doesn't actually change deps at all >18:08 * ciaranm has a radeon and no rage128 or mach64 >18:08 <@spyderous> just will change what's built within mesa and x11-drm >18:08 <@spyderous> so with a radeon, you'd save yourself 2 3d drivers and 2 kernel modules >18:09 <@ciaranm> incidentally, your x package count is wrong, since several packages that you think are ported actually aren't, and won't trigger the repoman check either >18:09 <@spyderous> example? >18:09 <@spyderous> any package.mask'd file i don't think will get checked >18:09 <@ciaranm> dunno. i'm filing bugs as i go. lots of people can't spell xinerama >18:09 <@ciaranm> some people can't spell X11 >18:10 -!- vanquirius_ [n=vanquiri@c906e262.virtua.com.br] has joined #gentoo-dev >18:10 <@spyderous> as it's basically testing what's emergeable on my system for whether it tries to pull in xorg >18:10 -!- vanquirius_ [n=vanquiri@c906e262.virtua.com.br] has left #gentoo-dev [] >18:12 <@spyderous> kinda wish configure scripts had a good way to define deps so they could parallelize >18:15 <@ciaranm> mmm. packages that PDEPEND upon themselves. whatever next? >18:16 -!- JoseJX [n=JoseJX@c-67-171-66-14.hsd1.pa.comcast.net] has quit ["Leaving"] >18:17 -!- vanquirius [n=vanquiri@c906e262.virtua.com.br] has quit ["bye"] >18:21 -!- wrobel [n=user@p548FBC58.dip0.t-ipconnect.de] has quit [Read error: 110 (Connection timed out)] >18:23 -!- gustavoz [n=gustavoz@gentoo/developer/pdpc.active.gustavoz] has quit ["Connection reset by beer... zzzz...."] >18:25 -!- nerdboy|off is now known as nerdboy >18:26 -!- uniplex [i=0kmF7gbB@wsip-70-168-213-195.hr.hr.cox.net] has joined #gentoo-dev >18:28 -!- tannewt [n=scott@gentoo/developer/tannewt] has quit ["Leaving"] >18:28 <@marienz> spyderous: btw, xkbdata depends on xkbcomp, xkeyboard-config doesn't. Is that intentional? >18:29 <@spyderous> marienz: sure it does, cvs up >18:29 <@marienz> heh, sorry :) >18:29 <@nerdboy> yo peeps >18:30 <@marienz> still don't know why my scrolllock light started behaving mysteriously, but meh. >18:30 <@nerdboy> just be glad you don't have one of those ibm rack/kvm keyboards >18:36 -!- nonotme [n=james@gentoo/user/nonotme] has joined #gentoo-dev >18:36 -!- JoseJX [n=JoseJX@c-67-171-66-14.hsd1.pa.comcast.net] has joined #gentoo-dev >18:36 -!- mode/#gentoo-dev [+o JoseJX] by ChanServ >18:37 <@spyderous> blackace: updated ideas.txt again >18:40 -!- Joky [i=joker@netswarm.net] has quit [Read error: 113 (No route to host)] >18:41 -!- codeman [n=codeman@gentoo/developer/codeman] has quit [Remote closed the connection] >18:51 -!- chutzpah [n=chutz@gentoo/developer/chutzpah] has joined #gentoo-dev >18:51 -!- mode/#gentoo-dev [+o chutzpah] by ChanServ >18:56 -!- Flameeyes [n=flame@gentoo/developer/Flameeyes] has quit [Remote closed the connection] >18:58 <@ciaranm> mmmfun. use_enable in global scope >18:58 <@ciaranm> that won't even work >18:59 -!- rizzo [n=dseiler@unaffiliated/rizzo] has joined #gentoo-dev >18:59 -!- mode/#gentoo-dev [+v rizzo] by ChanServ >18:59 <@ferringb_> for executing ebuilds it ought to. >18:59 -!- npmccallum [n=mgrnatem@nicholasville-terayon1-67-20-52-12.ironoh.adelphia.net] has joined #gentoo-dev >18:59 -!- mode/#gentoo-dev [+o npmccallum] by ChanServ >18:59 <@ferringb_> ciaranm: do not enable the features USE_EXPAND btw. >18:59 <@ferringb_> it's non incremental, and a bit horked (try FEATURES="-*") >19:00 <@ciaranm> ferringb_: congratulation, you get to reopen the bug >19:00 <@ferringb_> not really. >19:00 <@ferringb_> as I've said, wrong way of doing it. >19:01 <@ferringb_> should be seperate metadata imo... >19:01 <@ciaranm> *shrug* come up with a better solution that's usable and we'll do that instead >19:01 <@ferringb_> come up with something usable, or hack the existing crap to make it usable... >19:01 <@ciaranm> it'll need to handle SRC_URI too >19:02 <@ferringb_> TEST_* was proposed, and addresses it fine here. >19:02 <@ferringb_> ciaranm: it also allows features to adjust provides >19:02 <@ferringb_> which is part of the reason it shouldn't have been added. >19:02 <@ferringb_> same for restrict, although I'm not sure of the state of use parsing for restrict (should be there, but suspect someone removed that) >19:02 <@ciaranm> you can abuse most portage features for all sorts of things >19:02 <@ciaranm> if anyone does it, give them a wedgie >19:03 <@ferringb_> ciaranm: we're in this god awful mess cause people kept abusing portage is the thing... >19:04 <@ferringb_> anal about it, but do it right, if you have to hack it, hack it so it can be done correctly down the line. >19:04 <@ciaranm> design a system that can't be abused and you've designed a system that can't be used >19:04 <@ferringb_> introducing that injection into the use namespace just plain sucks imo. >19:04 * ferringb_ disagrees, but regardless. >19:04 <@ferringb_> either way it doesn't work, so don't use it >19:06 -!- jaervosz [n=jaervosz@gentoo/developer/jaervosz] has quit [Read error: 110 (Connection timed out)] >19:07 <@ciaranm> mmm. define "doesn't work" >19:08 -!- _Zoltan_ [n=dah@adsl-6-5-101.tys.bellsouth.net] has quit [Client Quit] >--- Log closed Wed Feb 08 19:14:04 2006 >--- Log opened Wed Feb 08 19:14:37 2006 >19:14 -!- ferringb [n=bharring@c-67-171-129-219.hsd1.or.comcast.net] has joined #gentoo-dev >19:14 -!- Irssi: #gentoo-dev: Total of 254 nicks [163 ops, 0 halfops, 20 voices, 71 normal] >19:15 -!- Irssi: Join to #gentoo-dev was synced in 26 secs >19:15 -!- WilliamH [n=wdh@c-24-0-243-195.hsd1.tx.comcast.net] has joined #gentoo-dev >19:15 -!- mode/#gentoo-dev [+o WilliamH] by ChanServ >19:18 -!- frobie [n=b@c-24-60-210-111.hsd1.ma.comcast.net] has joined #gentoo-dev >19:20 -!- dostrow [n=dostrow@gentoo/developer/dostrow] has quit [Read error: 104 (Connection reset by peer)] >19:21 -!- dostrow [n=dostrow@gentoo/developer/dostrow] has joined #gentoo-dev >19:21 -!- mode/#gentoo-dev [+o dostrow] by ChanServ >19:22 -!- nakano [n=nakano@host86-145-18-251.range86-145.btcentralplus.com] has quit ["Leaving"] >19:22 -!- jaervosz [n=jaervosz@gentoo/developer/jaervosz] has joined #gentoo-dev >19:22 -!- mode/#gentoo-dev [+o jaervosz] by ChanServ >19:22 <@SpanKY> grr flameeyes >19:22 <@SpanKY> i dont mind when you add fixes, but do em right >19:23 -!- hparker [n=hparker@gentoo/developer/hparker] has quit ["reboot"] >19:24 -!- _Zoltan_ [n=dah@adsl-6-56-249.tys.bellsouth.net] has joined #gentoo-dev >19:24 <@ferringb_> SpanKY: curious, you got the default src_install function handy? >19:25 <@ferringb_> the desired one, not what's currently in portage >19:25 <@ciaranm> wasn't the general concensus that a default src_install that does stuff would be a very bad idea? >19:26 * ferringb_ recalls the consensus the other way around >19:26 <@ferringb_> issue iirc was that introducing it without some form of protection == bad things >19:26 -!- _Zoltan_ [n=dah@adsl-6-56-249.tys.bellsouth.net] has quit [Client Quit] >19:26 -!- steev-laptop [n=steev-la@gentoo/developer/steev] has quit ["Lost terminal"] >19:27 <@SpanKY> src_install() { emake -j1 install DESTDIR="${D}" || die ; } >19:27 <@ciaranm> no [[ -f README ]] && dodoc README etc? >19:28 <@SpanKY> i think there was stuff for common autotool files >19:28 <@ciaranm> also, emake install? >19:28 <@ciaranm> we usually discourage that because of the extra variables >19:28 <@SpanKY> for x in AUTHORS ChangeLog NEWS README ; do [[ -s ${x} ]] && dodoc ${x} ; done >19:28 <@SpanKY> ciaranm: we only discourage it due to -j >19:28 <@ferringb_> SpanKY: ciaranm: figure one out >19:29 * ferringb_ doesn't particularly care what the form of it is, as long as it suits peoples needs. >19:29 <@ciaranm> SpanKY: mmm. are you sure? >19:29 <@ferringb_> removing that features crap and doing it right, going to need to force an eapi bump anyways; might as well slip it in alongside (no point in having 2 bumps). >19:29 <@SpanKY> ciaranm: yes, 99% sure >19:29 <@SpanKY> `emake -j` is very bad >19:29 <@SpanKY> but i dont know of any other make option that'd be bad >19:30 -!- slip [n=vitaliy@ool-4571f1ca.dyn.optonline.net] has left #gentoo-dev [] >19:30 -!- Tester__ [i=tester@Toronto-HSE-ppp3710750.sympatico.ca] has left #gentoo-dev [] >19:30 -!- Tester_ [i=tester@gentoo/developer/tester] has joined #gentoo-dev >19:30 -!- mode/#gentoo-dev [+o Tester_] by ChanServ >19:30 <@ciaranm> ok. in that case, why not make emake force -j for install targets? >19:30 -!- slip [n=vitaliy@ool-4571f1ca.dyn.optonline.net] has joined #gentoo-dev >19:30 -!- WilliamH [n=wdh@c-24-0-243-195.hsd1.tx.comcast.net] has quit ["be back later. :-)"] >19:30 <@SpanKY> you mean -j1 >19:31 <@ciaranm> uh, yeah >19:31 <@SpanKY> i dont see why not >19:31 <@SpanKY> is there a way to sanely detect you're in src_install() ? >19:31 <@ferringb_> eh? >19:31 <@SpanKY> ferringb: shush, men are talking >19:31 <@ciaranm> SpanKY: there is... or you could check the targets >19:32 <@SpanKY> ciaranm: i dont like checking $@ >19:32 * ferringb_ reiterates, eh? >19:32 -!- NightMonkey [n=NightMon@pdpc/supporter/sustaining/NightMonkey] has quit [Remote closed the connection] >19:32 <@SpanKY> that assumes the target is "install" and that you dont want to do something funky in src_compile or src_test >19:32 <@ciaranm> SpanKY: ok. then use the phase >19:32 <@SpanKY> is that like the force ? :P >19:33 <@ciaranm> noooo! EBUILD_PHASE or something like that, give or take underscores and caps >19:34 -!- hparker [n=hparker@gentoo/developer/hparker] has joined #gentoo-dev >19:34 -!- mode/#gentoo-dev [+o hparker] by ChanServ >19:34 -!- _Zoltan_ [n=dah@adsl-6-56-249.tys.bellsouth.net] has joined #gentoo-dev >19:34 <@SpanKY> http://rafb.net/paste/results/xyX85r15.html >19:34 <@SpanKY> how's that >19:35 <@ciaranm> wfm >19:35 <@ferringb_> ugh. >19:35 <@SpanKY> ferringb_: you may speak now >19:35 <@ferringb_> SpanKY: any reason you're unwilling to just call that from the default src_install? >19:35 * ferringb_ flips off spanky in the meantime >19:35 <@ferringb_> SpanKY: special casing the tools for ebuild_phase seems like a hack >19:35 <@SpanKY> so devs can define their own src_install() >19:36 <@ferringb_> ...and define -j1 themselves, imo. >19:36 <@SpanKY> unless we go on some crusade to make sure `make install` is parallel-safe, i dont see much point >19:36 <@SpanKY> especially considering devs cant even make sure src_install is parallel-safe >19:37 <@ciaranm> if it uses autotools, install isn't parallel safe >19:37 -!- slip [n=vitaliy@ool-4571f1ca.dyn.optonline.net] has quit [Remote closed the connection] >19:37 <@SpanKY> yeah, autotools guys dont really care >19:37 <@ferringb_> use a wrapper func that tags it in then... >19:37 <@SpanKY> cause the gain is negligible >19:37 <@SpanKY> hmm, by a wrapper func, do you mean "emake" ? >19:37 <@SpanKY> that seems pretty wrapper-tastic to me >19:37 <@ferringb_> gain is that it keeps EBUILD_PHASE in ebuild.sh instead of letting it bleed into the tools >19:37 <@ciaranm> SpanKY wins >19:38 <@SpanKY> what else would you propose ferringb_ ? let the wrappers bleed in the other direction ? >19:38 <@ferringb_> define a func in ebuild.sh that tags the arg on, calling emake. >19:38 <@ferringb_> hell, stick whatever extra install crap you want into it that's useful. >19:39 <@SpanKY> both seem equivalent to me >19:39 <@SpanKY> but i prefer mine obviously >19:39 * ciaranm prefers SpanKY's too >19:39 <@SpanKY> ciaranm: that wont help convince ferringb :p >19:40 -!- rizzo [n=dseiler@unaffiliated/rizzo] has quit ["leaving"] >19:40 <@ciaranm> hasn't he learned that i'm always right yet? :( >19:40 <@ferringb_> SpanKY: bit wrong on that >19:40 <@ferringb_> if it's sound technically, fine, I just don't think sticking the hack into the tool itself is good from a maintenance standpoint. >19:41 <@ciaranm> EBUILD_PHASE is more permanent than relying upon two way interfaces >19:41 <@fox2mike> antarus: ping >19:42 * ferringb_ thinks not, obviously :) >19:42 -!- slip [n=vitaliy@ool-4571f1ca.dyn.optonline.net] has joined #gentoo-dev >19:42 <@ferringb_> ebuild_phase is a semi recent addition anyways, came in during .50 iirc. >19:42 -!- slip [n=vitaliy@ool-4571f1ca.dyn.optonline.net] has quit [Read error: 104 (Connection reset by peer)] >19:42 <@ferringb_> might've been .51_pre's actually, but besides the point. >19:43 <@ferringb_> figure out a src_configure while you're at it too. >19:43 <@ferringb_> might as well get all of that crap out in one bump. >19:43 <@ciaranm> src_patch too please >19:43 <@ciaranm> or src_prepare >19:43 <@ferringb_> bleh >19:43 <@ciaranm> user hooks >19:43 <@ferringb_> we have user hooks already >19:43 <@ciaranm> yes, and they're much nicer on src_prepare >19:44 <@ferringb_> clarify src_prepare >19:44 <@ciaranm> src_patch >19:44 <@ferringb_> we've got hooks around all of the phases now >19:44 <@SpanKY> ferringb_: using a function limits the usuablity >19:44 <@ciaranm> patch implies patch, not sed >19:44 <@SpanKY> bash funcs cant be piped/env-ed/etc... >19:44 -!- rphillips [n=rphillip@c-24-4-231-48.hsd1.ca.comcast.net] has joined #gentoo-dev >19:44 -!- mode/#gentoo-dev [+o rphillips] by ChanServ >19:44 <@ferringb_> SpanKY: err... bash funcs can be piped unless I'm missing something obvious in your statement >19:45 <@ferringb_> iirc, you can define the default redirection for a func even (never seen it used though) >19:45 <@SpanKY> env emake foo >19:45 <@SpanKY> that'll [intuitively] give you diff behavior than just 'emake foo' >19:45 <@ciaranm> ferringb_: anyway, the interface segregation principle says that i'm right >19:45 <@ferringb_> yah. >19:45 <@ciaranm> on both counts >19:46 <@ferringb_> ciaranm: that's not how you convince me... >19:46 <@SpanKY> you have to sex him up first >19:46 * ferringb_ notes "xyz says I'm right" is akin to "some other dude thinks this is the way it should be" in his reading of it >19:46 <@ciaranm> no, you get jason to do it >19:47 <@ferringb_> ciaranm: better luck with genone from a time standpoint. >19:47 <@ciaranm> no, it's "this large body of software engineering literature says it's right" >19:47 <@ferringb_> ugh >19:47 -!- _Zoltan_ [n=dah@adsl-6-56-249.tys.bellsouth.net] has quit [Client Quit] >19:47 <@ferringb_> SpanKY: a default configure phase would be nice to have. don't care about patch atm unless someone makes a damn good arguement for it. >19:47 <@ferringb_> meanwhile, throw the books around, back to work >19:47 <@ciaranm> ferringb_: how about "avoids duplicating unpack cd" >19:47 -!- ferringb_ [n=cso@gentoo/developer/ferringb] has left #gentoo-dev [] >19:48 -!- KillerFox [n=rn@killerfox.home.forkbomb.ch] has quit [Remote closed the connection] >19:48 -!- KillerFox [n=rn@killerfox.home.forkbomb.ch] has joined #gentoo-dev >19:48 -!- mode/#gentoo-dev [+o KillerFox] by ChanServ >19:54 -!- codergeek42 [n=peter@gentoo/developer/codergeek42] has joined #gentoo-dev >19:54 -!- mode/#gentoo-dev [+o codergeek42] by ChanServ >19:57 -!- SuperLag [n=aaron@gentoo/developer/SuperLag] has quit ["reboot"] >20:00 <@SpanKY> ferringb: ? >20:00 <@antarus> fox2mike, pong >20:01 <@SpanKY> ferringb: you never commented on my last DEBUG_BUILD patch btw :P >20:02 <@fox2mike> antarus: your GPG key expires in march and well, it doesn't have your @gentoo.org ID attached to it :) >20:02 <@fox2mike> antarus: could you please add that ID and upload the key to subkeys.pgp.net? >20:02 <@ciaranm> antarus: you never answered ferringb's question about how you think FEATURES in USE_EXPAND breaks >20:03 <@antarus> ciaranm, yes he requested the version >20:03 <@antarus> which was 2.1_pre4 >20:04 <@ciaranm> ok. now you get to tell me how it breaks! >20:04 <@antarus> I think he didn't note that i had removed it from incrementals >20:04 <@ciaranm> so it breaks if you break FEATURES? >20:05 * ciaranm is somewhat not astounded >20:05 * ciaranm demands a new excuse from ferringb! >20:10 <@antarus> as to what, why not to turn it on? >20:10 <@ciaranm> as to why Halcy0n can't add FEATURES to USE_EXPAND >20:12 <@antarus> I don't know of any technical reason preventing that from happening >20:12 -!- Tester_ [i=tester@gentoo/developer/tester] has quit [Nick collision from services.] >20:12 -!- Tester_ [i=tester@gentoo/developer/tester] has joined #gentoo-dev >20:12 -!- mode/#gentoo-dev [+o Tester_] by ChanServ >20:12 <@antarus> There are only semantic reasons, aka, we would like a better way of doing it >20:12 <@antarus> but you knew that already >20:12 <@ciaranm> Halcy0n: ^^ you have the goahead >20:12 <@antarus> FEATURES are portage FEATURES, not package features >20:13 <@ciaranm> ... >20:15 <@antarus> This is all in reference to TEST, is it not? >20:15 <@ciaranm> this is in reference to "we need some way to have conditional deps and src_uri components for src_test" >20:15 * antarus nods >20:17 <@antarus> any reason to just drop features = test and make it a use flag? >20:18 <@antarus> er not to.. >20:18 <@ciaranm> use flags affect the resulting binary >20:18 <@ciaranm> features affect how the binary is built but do not modify the binary >20:19 <@antarus> USE=symlink >20:19 <@ciaranm> is broken. we had this discussion >20:20 <@antarus> features=noman nodoc noinfo >20:20 <@ciaranm> also, it does modify the binary, for a sufficiently vague value of binary >20:20 <@ciaranm> doesn't modify the binary! >20:20 <@antarus> "installed files" then >20:20 <@ciaranm> plus no* is deprecated in favour of INSTALL_MASK >20:20 * antarus nods >20:20 -!- KingTaco [n=kingtaco@gentoo/developer/kingtaco] has quit [Read error: 113 (No route to host)] >20:20 -!- KingTaco [n=kingtaco@gentoo/developer/kingtaco] has joined #gentoo-dev >20:20 -!- mode/#gentoo-dev [+o KingTaco] by ChanServ >20:21 <@antarus> of course in this case we are going off of your definition of what use flags and features do >20:21 <@Halcy0n> Which is what the correct definition should be. >20:21 <@ciaranm> it's how it was originally intended >20:21 <@nerdboy> you can't have a rule without the occasional exception >20:22 <@antarus> nerdboy, beer for nerdboy ;) >20:22 <@ciaranm> no need to go and add more exceptions just because you don't understand the rule >20:23 <@antarus> perhaps I merely disagree with "how the rule was originally intended" :) >20:23 <@ciaranm> in which case you would be wrong >20:23 <@antarus> No, that would be my opinion >20:23 <@ciaranm> facts aren't a matter of opinion >20:23 <@antarus> opinions cannot necessarily be wrong, merely different >20:24 <@antarus> how is an intention fact? >20:24 <@ciaranm> if your opinion is that 2 + 2 = 5, does that mean anything other than that you are wrong? >20:24 -!- RiverRat [n=me@gentoo/user/RiverRat] has quit [Connection timed out] >20:24 <@antarus> and we are off subject yet again >20:24 <@ciaranm> it's entirely relevant >20:25 <@ciaranm> stop trying to turn facts into matters of opinion >20:25 <@antarus> so do you have an e-mail or reference to back up your fact? >20:26 <@ciaranm> from four years ago? >20:26 <@ciaranm> http://dev.gentoo.org/~plasmaroo/devmanual/general-concepts/features/ >20:26 <@ciaranm> there's your reference >20:26 <@antarus> written by you >20:27 * nerdboy probably has email older than both you guys >20:27 <@antarus> thanks, I'd like what is the word I'm looking for... >20:27 <@antarus> Reputable evidence ;) >20:27 <@ciaranm> devmanual is your best source of reputable evidence >20:27 * antarus wonders why he even bothers >20:27 <@Cardoe> lies! >20:28 -!- RiverRat [n=me@gentoo/user/RiverRat] has joined #gentoo-dev >20:28 -!- mode/#gentoo-dev [+v RiverRat] by ChanServ >20:28 <@nerdboy> the separation makes perfect design sense >20:28 <@nerdboy> but that doesn't mean you can't break the norm for a special package like kernel-source >20:28 <@nerdboy> the USE="symlink" thing >20:29 <@ciaranm> my objection was more that it's a global flag on that one >20:29 <@nerdboy> any changes broader in scope probably point towards a new design rather than a retro-fit >20:29 <@antarus> the other example is probably kde-enable-delta-final >20:30 <@ciaranm> antarus: i objected to that one too >20:30 <@Halcy0n> Just because you can name a few exceptions doesn't mean we should needlessly make more. >20:30 * nerdboy is pretty sure ciaranm objects to most things... >20:30 <@nerdboy> :) >20:30 <@ciaranm> nerdboy: only the stuff that's wrong :) >20:30 <@Halcy0n> nerdboy: he didn't object to me being the QA lead. >20:30 <@ciaranm> Halcy0n: now you're just finding a few exceptions >20:30 <@antarus> nor did I :) >20:31 <@Halcy0n> ciaranm: bah :P >20:32 <@antarus> Halcy0n, the problem being we then have to special case features_test because we don't want the other features bleeding through >20:33 <@ciaranm> uh, no we don't >20:33 <@ciaranm> we do want the other features bleeding through >20:33 <@Halcy0n> And if we didn't, offer a new better idea, not abusing USE flags. >20:33 <@antarus> erm, the day you need to do something special to your ebuild because features_digest is enabled.... >20:34 <@ciaranm> so? extra entries in USE doesn't hurt >20:34 <@ciaranm> just shove through all of features and let ebuild authors do whatever they want >20:34 <@antarus> Extra USE flags = extra metadata >20:34 <@ciaranm> so? >20:34 <@antarus> lang.desc fex >20:35 <@ciaranm> so? >20:35 <@antarus> so are you volunteering to maintain it? >20:35 <@ciaranm> what, sticking FEATURES into USE_EXPAND? sure >20:36 <@ciaranm> the maintenance load on that is... uh... adding it, and then hitting anyone who removes it >20:36 <@antarus> and then the USE_EXPAND flags in IUSE, and then putting them in features.desc... >20:36 <@ciaranm> easy enough >20:36 <@ciaranm> nnnnope >20:36 * antarus chuckles >20:36 <@ciaranm> USE_EXPAND is exempt from IUSE and *.desc >20:36 <@antarus> at present >20:36 <@ciaranm> and will continue to be so, since IUSE_USERLAND makes no frickin' sense >20:37 <@antarus> Nah, I meant IUSE="kernel_linux userland_GNU" fex >20:37 <@antarus> the actual flags >20:37 <@ciaranm> yes. that makes no sense >20:37 <@antarus> and exempting them means what for IUSE? >20:37 <@ciaranm> IUSE video_cards_ makes sense, since it's something a user would want to set >20:38 <@ciaranm> IUSE kernel_ doesn't, since it's not a general user setting >20:38 <@antarus> what is IUSE then, if not a list of USE flags the ebuild uses >20:38 <@ciaranm> a setting that makes it easy for portage to tell the user about stuff they can change in packages >20:39 <@Halcy0n> Users can't change kernel_linux. They don't need to be made aware of it. >20:39 <@ciaranm> that is why we have IUSE, and it is also why certain USE flags aren't listed in IUSE >20:39 < ferringb> Halcy0n: do not enable features in use_expand. >20:39 <@antarus> Halcy0n, but they can change FEATURES, and in your method, FEATURES are in USE >20:39 -!- mode/#gentoo-dev [+o ferringb] by ChanServ >20:39 <@antarus> so you want IUSE="features_" >20:39 <@antarus> basically >20:39 <@ferringb> Halcy0n: do not enable features in use_expand. >20:39 <@antarus> heh >20:40 <@ciaranm> antarus: no. changing FEATURES is not a general build setting >20:40 <@ferringb> be patient. 2.1 will have test_* metadata, mainly because I wasn't aware jason slipped that in. >20:40 <@Halcy0n> ferringb: where did I say I was? >20:40 <@ciaranm> something like two thirds of the tree uses the bootstrap USE flag but doesn't list it in IUSE because it'd be silly for the user to change it. you know that, right? >20:40 <@ferringb> Halcy0n: stating it since ciaran was telling you to... >20:40 <@ferringb> ciaranm: same for you, don't abuse that. >20:40 <@ciaranm> ferringb: it's not abuse until you come up with a better usable solution >20:40 <@antarus> ciaranm, so how do I know if an ebuild supports testing or not? >20:40 -!- dostrow [n=dostrow@gentoo/developer/dostrow] has quit [Client Quit] >20:41 <@ciaranm> antarus: not a general user setting >20:41 <@ferringb> ciaranm: just do what I tell ya, alright? >20:41 <@ferringb> no debate here >20:41 <@antarus> ummm yes it is? :) >20:41 <@ciaranm> antarus: the average user does not care whether an ebuild needs special stuff for test >20:41 <@ciaranm> ferringb: you are wrong here >20:41 <@antarus> The user does care if an ebuild has tests or not >20:42 <@antarus> and the only way for them to know is a use flag >20:42 <@antarus> and you are suggesting we hide it >20:42 <@ciaranm> antarus: uh, ebuilds don't define whether or not they have tests >20:42 <@ciaranm> the features_test flag will only be used for cases where a package has tests that require extra stuff >20:43 <@ciaranm> the features_test flag will not be used by the majorith of packages that have tests >20:43 -!- uniplex [i=0kmF7gbB@wsip-70-168-213-195.hr.hr.cox.net] has quit [Read error: 110 (Connection timed out)] >20:45 <@ciaranm> antarus: i await a new excuse! >20:45 -!- genstef [n=stefan@gentoo/developer/genstef] has quit [Nick collision from services.] >20:46 -!- genstef [n=stefan@gentoo/developer/genstef] has joined #gentoo-dev >20:46 -!- mode/#gentoo-dev [+o genstef] by ChanServ >20:47 -!- codergeek42 [n=peter@gentoo/developer/codergeek42] has quit ["Ex-Chat"] >20:53 <@antarus> ciaranm, yes well, I have actual work to do, so maybe later >20:54 <@ciaranm> in the mean time, i've still not heard what's stopping us from fixing a couple of dozen bugs >20:56 -!- steev-laptop [n=steev-la@gentoo/developer/steev] has joined #gentoo-dev >20:56 -!- mode/#gentoo-dev [+o steev-laptop] by ChanServ >21:05 <@Ticho> good morning >21:05 <@fox2mike> morning Ticho >21:08 <@ferringb> ciaranm: back that up beyond stating it... >21:08 <@ciaranm> back what up? >21:09 <@ferringb> "you are wrong here" >21:09 <@ciaranm> come up with some justification other than "because i said so" and i will >21:09 <@ferringb> features should not bleed into use. >21:10 * ferringb has already stated that. >21:10 <@ciaranm> they don't, since they're prefixed >21:10 <@ferringb> and it's not hard to do it properly. >21:10 <@ciaranm> features_* aren't user-visible USE flags >21:10 <@ciaranm> so pollution isn't a problem >21:10 <@ferringb> bolsh. >21:10 <@ferringb> it's injected into the use namespace. >21:10 <@ferringb> that's the fundamental issue. >21:11 <@ciaranm> and it's subnamespaced via the prefix >21:11 <@ferringb> it's in the namespace, yes or no? >21:11 <@ciaranm> it's in a child namespace of the use namespace >21:11 <@ciaranm> which is a pretty common programming technique >21:14 <@ferringb> bleh, phone. >21:14 <@ferringb> ciaranm: it can be done properly, seperated metadata. >21:15 <@ferringb> jamming it in isn't valid imo, and at this point, it's a judgement call. >21:15 <@ferringb> ciaranm: finally, if you're going to post snippets of irc logs from dev, post them all please. >21:15 -!- iluxa [n=anonymou@67.115.118.49] has joined #gentoo-dev >21:15 -!- mode/#gentoo-dev [+o iluxa] by ChanServ >21:15 <@ferringb> wasting my time correcting it. >21:15 -!- iluxa [n=anonymou@67.115.118.49] has quit [Read error: 104 (Connection reset by peer)] >21:16 <@ciaranm> that was all of it, minus the random other people saying totally unrelated stuff >21:16 <@ferringb> seemed to have left out the lead in, and the statement of "I'm going via the proper channels". >21:17 <@ferringb> it's akin to "senior devrel members making accusations of drug abuse". it's bs, no point in doing the games. >21:17 <@ciaranm> read on >21:17 <@ciaranm> note the executive summary, since no-one in devrel's going to bother reading anything else if they vote >21:18 <@ferringb> ugh. >21:18 <@ferringb> ciaranm: why are you here? >21:18 <@ferringb> seriously, what is your driving force for fucking with people? >21:19 <@ciaranm> i'm here to work on making gentoo suck less >21:19 <@ferringb> why not just do it right from the start in your own little corner? >21:19 <@ciaranm> because back in the good old days, gentoo was a better starting point >21:20 <@ferringb> these aren't the good old days. >21:20 <@ferringb> these days, people are expected to do their work without being an asshole to others. >21:20 <@ciaranm> no, but it shouldn't be too much effort to get us moving back that way >21:20 <@ferringb> ciaranm: the processes and beuracracy that slows gentoo is a reaction to folk like you. >21:20 <@ferringb> and yes, folk like pvdabeel >21:20 <@ferringb> people doing stuff on their own that is damaging to gentoo as a whole >21:20 <@ciaranm> no, they're a reaction to half our devs being dead weight >21:21 <@ferringb> hardly. >21:21 <@ferringb> devrel cleansing of devs is removal of dead weight (actual dead weight). >21:21 <@hparker> ciaranm: Uhm, hate to bust your buble, but.... devrel reads all of it >21:21 <@ferringb> ciaranm: you're assuming production == worth >21:21 <@ferringb> the real calculation is production - social cost of the contributor == worth >21:22 <@ferringb> some folk get along with you fine, others do not. >21:22 <@ferringb> if they're working on something you dislike, or dislike their methods, you harass the fuck out of them. >21:22 <@ciaranm> ferringb: i am not the one doing the harassing >21:22 <@ferringb> end result is that either you do the work (unlikely), they do it regardless, or they stop. >21:22 <@ferringb> bullshit. >21:22 <@ferringb> I've put up with how many months of your potshots at portage and prefix, combined with lies and games every time it's brought up? >21:23 <@ciaranm> or, the end result is that they engage in a decent techincal discussion, come up with the best solution and then do that >21:23 <@ferringb> people don't get along, you go out of your way to be a jack ass to 'em. >21:23 <@ferringb> ciaranm: you're incapable of engaging in a decent technical discussion is the problem. >21:23 <@ferringb> you fall back to harassing/attacking. >21:23 <@ciaranm> which is what usually happens, except when people like you try to avoid the technical discussions and resort to flaming people on bugzilla >21:23 <@ferringb> you have no say in a project you're not involved in, nor responsible for. >21:23 <@ciaranm> ferringb: there're plenty of people out there who disagree. have you ever considered that you are the problem, and not me? >21:23 <@ferringb> just the same as I have no say in paludis. >21:24 <@ciaranm> ferringb: funnily enough, i am involved in Gentoo >21:24 <@ferringb> fancy that, I'm the maintainer of the pkg maintainer in gentoo. >21:24 <@ferringb> s:pkg maintainer:pkg manager: >21:24 <@ferringb> frankly, if that doesn't define me as involved, nothing will. >21:24 <@ciaranm> and you should realise that work on the package manager affects the project as a whole >21:24 <@ferringb> thank you captain obvious. >21:25 <@ferringb> ciaranm: you should realize your behaviour affects the whole of gentoo also. >21:25 <@ferringb> but... you won't respond to that one will you? >21:25 <@ferringb> you'll just deflect it. >21:25 <@ferringb> and we'll go round and round, wasting more fucking time. >21:25 <@ciaranm> ferringb: my behaviour is not the problem. i am not the one running around harassing people rather than having a proper technical discussion >21:25 <@ferringb> ciaranm: there is no technical discussion to be had. >21:25 -!- npmccallum [n=mgrnatem@nicholasville-terayon1-67-20-52-12.ironoh.adelphia.net] has quit ["Ex-Chat"] >21:26 <@ciaranm> ferringb: see, that's the problem. a technical discussion is needed >21:26 <@ferringb> ciaranm: jesus you're a hypocrite. >21:26 <@ferringb> it's no different to how you're running paludis, the difference is that our source is open. >21:26 <@ciaranm> anyone who wants to discuss technical issues with paludis is more than welcome to do so >21:26 <@ferringb> we're working on the problem in our own corner, and will share results with others when we deem it ready. if they want it, they can have it, else we can use it for our own purposes (outside of gentoo) >21:27 <@ferringb> that's the thing here, nothing is being forced- you're attacking because (frankly) you don't seem to like either the concept or those involved. >21:27 <@ferringb> which is flat out bullshit. >21:27 -!- grahl04 [n=grahl04@128.227.22.195] has quit ["-"] >21:27 <@ferringb> feel free to call it harassment when someone calls you on it. >21:27 <@ciaranm> no, what i don't like is the mess that's going to be made when you try to merge it >21:28 <@ferringb> uh huh. you're attacking because of an action we've stated we will not do, further an action that hasn't occured. >21:28 -!- zmedico [n=zmedico@cpe-66-27-157-183.socal.res.rr.com] has quit [Remote closed the connection] >21:28 <@ferringb> ciaranm: so I have the right to murder my neighboer because I think he's shady and may kill me some time in the future. >21:28 <@ferringb> good logic. >21:28 <@ciaranm> ferringb: no, i'm basing it upon past behaviour >21:28 <@ferringb> what behaviour? >21:28 <@ferringb> oh yes, pvdabeel. >21:28 <@ferringb> your usual chestnut. >21:29 <@ferringb> where is he involved? >21:29 <@ferringb> where are you even involved, able to comment on this? >21:29 <@ciaranm> ...and curtis119, and you, and drobbins, and everyone else who's ever tried to push through a large change >21:29 <@ferringb> ... >21:30 -!- ferringb [n=bharring@gentoo/developer/ferringb] has left #gentoo-dev ["this is beyond stupid."] >--- Log closed Wed Feb 08 21:30:01 2006
You cannot view the attachment while viewing its details because your browser does not support IFRAMEs.
View the attachment on a separate page
.
View Attachment As Raw
Actions:
View
Attachments on
bug 114944
:
74384
|
74687
|
78986
|
78988
| 79297 |
81121
|
81127
|
83749
|
84159