12-23-05 #-dev logs, aka portage devs suck

12-23-05-aka-portage-devs-suck.log (text/plain), 12.91 KB, created by Brian Harring (RETIRED) on 2006-02-05 17:03:11 UTC

14:58 <@ferringb> !seen mcummings
14:58 <+glbt> mcummings (n=mcumming@gentoo/developer/mcummings) was last seen quitting from #gentoo-commits 2 days, 6 hours, 19 minutes ago stating ({""HoHoHo""}).
14:58 <@ReJ> lol
14:59 <@ferringb> g-cpan users... curious about the pkg names it generates
14:59 <@ferringb> anyone able to comment if ':' is used as a char in the pkg name? :)
14:59  * Cardoe pokes Battousai 
14:59 <@Cardoe> hey seemant
14:59 <@araujo> Merry Christmas everyone!
14:59 <@araujo> Anyone wanna ponche?
14:59 <@ferringb> ponche?
14:59 <@Battousai> ferringb: i think it uses underscores
14:59 <@araujo> ferringb, A ron drink
15:00 <@ferringb> Battousai: don't spose you could verify it please?
15:00 <@Battousai> no
15:00  * Battousai runs
15:00  * Battousai can try
15:00 <@ciaranm> ferringb: : as a package name will break :slot. and certain backwards operating systems
15:00  * ferringb knows
15:00 <@ciaranm> so it's probably not a good idea to allow it if it's not currently used
15:00 <@ferringb> ciaranm: why do you think I'm asking what g-cpan does?
15:00 <@ferringb> :)
15:01 <@lisa> g-cpan++
15:01 <@ferringb> ciaranm: do I have to respond to the "are portage devs are bitch servants or doing whatever they want" email also?
15:01 <@spb> yes
15:01 <@ferringb> cause it won't be nice if I do, since you're in the same boat as we are.
15:01 <@ferringb> all of you are.
15:01 <@spb> you have to reply to every email on the list
15:01 <@ciaranm> ferringb: difference is, i'm quite happy adding support for things i don't personally care about to packages i maintain
15:01 -!- beejay_ [n=benni@gentoo/user/beejay] has quit [Read error: 60 (Operation timed out)]
15:02 <@spb> ciaranm: selinux support in findutils pls
15:02  * spb hides
15:02 <@ciaranm> if someone sticks in a legit request for something vim / fluxbox / whatever related then i'll go ahead and implement it, even if i personally don't use it
15:02 <@ferringb> ciaranm: I assure you, bug fixing is really something I like to do.
15:02 <@ciaranm> spb: did that three times already :P
15:02 <@spb> well do it again !
15:02 <@ciaranm> spb: now go send that updated patch upstream
15:02 <@spb> not until it doesn't segfault
15:02 <@lisa> ciaranm: emacs for vim?
15:02 <@ferringb> ciaranm: if you're making the case portage devs only do work they're interested in, bluntly, you're a fool.
15:02 <@ciaranm> spb: 20:19 #gentoo-commits: < CIA-1> pebenito * gentoo-x86/sys-apps/findutils/ (3 files in 2 dirs):
15:03 <@spb> fun
15:03 <@ciaranm> lisa: upstream package?
15:03 <@lisa> ciaranm: just kidding :-) i'm not that bad
15:03 <@spb> i got distracted and stopped paying attention to that
15:03 <@ciaranm> ferringb: no, i'm making the case that you're hurting gentoo by ignoring certain requests
15:03 <@Battousai> ferringb: any particular module you want me to try?
15:04 <@ciaranm> lisa: actually, there is a set of maps that make vim behave like emacs somewhere. they're rather evil
15:04 <@ferringb> ciaranm: case in point?
15:04 <@ciaranm> lisa: http://www.vim.org/scripts/script.php?script_id=300
15:04 <@ciaranm> ferringb: :foo[bar]
15:05 <@Tester_> anyone has modular X and dev-util/insight ?
15:05 <@lisa> ciaranm: *shudder*
15:05  * ferringb laughs
15:05  * Tester_ should try modular X
15:05 <@ferringb> ciaranm: your words...
15:05 <@ferringb> Our 'customers' don't always know what's best for
15:05 <@ferringb> them, especially when it comes down to low level issues like this one.
15:05 <@ferringb> ciaranm: implied in that bit of arrogance is that the user doesn't always know what's involved
15:05 <@ferringb> ciaranm: you don't know what's involved to do it in stable portage.
15:05 <@ciaranm> ferringb: except that we, the people who maintain the tree, know far more than you do about what the tree needs from the package manager
15:05 <@ferringb> if it were easy, it would've been done.
15:05 <@ferringb> bullshit.
15:06 <@ferringb> ciaranm: maintaining portage requires quite a bit more knowledge about the tree then you give credit for.
15:06 <@ferringb> that's ignoring the fact I also was doing tree work long before I did portage.
15:06 <@ferringb> so.. stick to the point instead of blanket statements.
15:06 <@ciaranm> ferringb: i'm still amazed at the amount of stuff you and jstubbs don't know
15:06 <@ferringb> pardon, we're not all gods like you.
15:06  * ferringb shrugs
15:06 <@ferringb> course, we also do the work.
15:07 <@ferringb> so.. you're boned unless you help (which you won't).
15:07 -!- beejay [n=benni@gentoo/user/beejay] has joined #gentoo-dev
15:07 -!- mode/#gentoo-dev [+v beejay] by ChanServ
15:07 <@ferringb> ciaranm: get off your ass and replace us with your claimed portage replacement.
15:07 <@ferringb> do something rather then run the mouth effectively.
15:07 <@ciaranm> could do. of course, then i'd have far less time to do tree stuff
15:08  * ferringb thinks you couldn't tbh.
15:08 -!- Lothmir [n=[e-mail]] has joined #gentoo-dev
15:08 <@ferringb> but that's neither here nor there.
15:08 <@slarti> compnerd: pong
15:09 <@ciaranm> pfff, all i need is uninstall code
15:09 <@ferringb> that's pretty easy to do.
15:09  * ferringb shrugs
15:09 <@ferringb> ciaranm: point is, you're not doing anything to actually help, thus you're on our timeline.
15:09 <@ciaranm> it's also insanely difficult to prove that it works
15:09 -!- codergeek42 [n=[e-mail]] has quit ["leaving"]
15:09  * ferringb doesn't really care about the excuses tbh.
15:09 <@ciaranm> ferringb: i'm helping by maintaining the tree that makes portage relevant
15:10 <@ferringb> yes, the one man army maintaining the entire tree.
15:10 <@ferringb> err, wait, that's not the case.  
15:10 <@ferringb> you're reaching, just stop.  the point is valid whether you'll acknowledge it or not.
15:10 <@ferringb> it's just plain common sense, which aparently ain't so common.
15:10 <@ciaranm> no, common sense is that portage is there to support our primary product, the tree
15:11 <@ferringb> inverse is the tree is irrevelent without the package manager.  your point?
15:11 <@ferringb> it's not a symbiotic relationship (as it should be) due to folk like you who assume we're at your beck and call.
15:11 <@ferringb> or they decide shoving hacks into the tree rather then proper fixes is the right route.
15:11 <@ferringb> end result is y'all have to work around us, and we have to work around you.
15:11 <@ciaranm> hacks go into the tree when we need a solution now rather than in three years
15:12 <@ferringb> heh
15:12 <@ferringb> ciaranm: you've been around 3 years.  you've claimed to have written your own portage replacement
15:12 <@ferringb> why the fuck haven't you fixed this?
15:12 <@ferringb> it's more productive then bitching, no?
15:12 <@ferringb> your own admission, this is worth stepping away from the tree for a bit to get it done.
15:12 <@ferringb> so... keep bitching, rather then doing unsavory work.  your choice of how to spend your time.
15:12 <@ciaranm> what, delving into the depths of the portage code? at this point i consider fixing portage to be totally pointless
15:13 <@ferringb> ciaranm: you ever _not_ make excuses?
15:13  * ferringb shrugs
15:13 <@ciaranm> ferringb: i'm not the one making excuses for lack of :[]
15:13 <@ferringb> heh, round and round baby.
15:13 <@ferringb> what I'm pointing out is that maintaining portage is a huge job in it's own, beyond feature creep.
15:13  * spb watches in amusement
15:14 <+ajax> devfight!
15:14 <@spb> i think based purely on number of lines ciaranm is winning
15:14 <@ferringb> hah
15:14 <@ferringb> spb: well here's a question.
15:14 <@ferringb> what good does ciaranm winning a debate result in?  still no implementation.
15:15 <@spb> the winner gets to win
15:15 <@ferringb> the whole point of that thread is that people are *not* helpless in this manner.  they can chip in.
15:15  * spb shrugs
15:15 <@ciaranm> it's not about winning. it's about getting what's best for gentoo
15:16 <@`Kumba> arguing gains no one anything.  What ya'll should be doing is manning the AA guns for when that pesky fat man and 9 reindeer flies overhead
15:16 <@ferringb> ... or working on portage code.
15:16 <@Cardoe> Happy Christ-Hanu-Kwanzi-Ka
15:16 <@spb> nope, definitely manning the AA guns
15:16 <+ajax> `Kumba: i parsed that as antialiasing guns
15:16 <@`Kumba> bonus to the first one that gibs Rudolph! :P
15:17 <@vanquirius> Cardoe: Aha, that's the term I have been looking for, ever since that flash animation!
15:17 <@ferringb> ciaranm: lets put it this way ciaranm.  your view, we're incompetent tools who don't give a damn about gentoo.  your opinion, don't care.  you *could* however take a week or two off to work on this feature, no?
15:17 <@Cardoe> vanquirius: glad I could help.
15:17 <@Cardoe> I'm out.
15:17 <@vanquirius> :-) later
15:18 <@ferringb> since short term hacks suck, long term is required.  put in the short term work, gain the long term benefits.
15:18 <@ciaranm> ferringb: could do, but were i to do that, i'd just go and rewrite the whole depresolver anyway
15:18 <@ferringb> go for it.
15:18 -!- Cardoe [n=Cardoe@gentoo/developer/Cardoe] has quit ["Leaving"]
15:18 <@ferringb> further gain.
15:18 <@ferringb> ciaranm: users would love it, would benefit gentoo massively having a sane resolver.
15:18  * ferringb waits for the carefully worded "not going to do it" response
15:18 <@`Kumba> ciaranm: didn't you do that already?
15:19 <@ciaranm> then i'd have to go and remove crap like this from the tree:
15:19 <@ciaranm> [n ] sys-devel/patch-2.5.9::gentoo +build! -static
15:19 <@ciaranm> [nr] sys-devel/patch-2.5.9::gentoo -build -static
15:19 <@ciaranm> `Kumba: i have a dep resolver, yeah
15:19 <@`Kumba> but it's in C++, right?
15:19 <@ciaranm> yup
15:19 <@`Kumba> yeah, not very pluggable into current portage atm, heh
15:19  * ferringb nods
15:20  * ferringb shrugs
15:20 <@ferringb> not very useful without source anyways.
15:20 -!- georges [n=[e-mail]] has quit [Read error: 104 (Connection reset by peer)]
15:20 <@ciaranm> it's really frickin' messy with all the circular deps we have in base, too
15:20 <@ferringb> heh
15:20 <@ferringb> wait, you're making the case for why we have issues as an excuse for not doing the work...
15:20 <@ciaranm> anything using patch or autoconf makes it throw a hissy fit
15:20 <@ciaranm> heh, hardly. i've already done the work on that
15:21 <@ferringb> where.
15:21 <@ciaranm> not anywhere where you can play with it
15:21 <@ferringb> so what use is it to gentoo?
15:21 <@ferringb> you go on about benefiting gentoo, yet sit back talking about your claimed rewrite that fixes everything.
15:21 <@ferringb> put it out there, replace us, benefiting gentoo.
15:22 <@ferringb> why haven't you if you're after gentoo's best interests?
15:22 <@ciaranm> pfff, i never said it fixes everything. if it every fixes everything, i'll open it up
15:22 <@ferringb> you'll never open it up, because its impossible to be perfect.
15:22 <@`Kumba> eh, nothing can fix everything
15:22 <@ferringb> ciaranm: end result is you're not helping in that area
15:22 <@ferringb> even sharing the code is helping, since it can be looked at and have the algo translated into stable.
15:23 <@ferringb> but you prefer to snipe instead, wasting our time plus pissing us off.
15:23 <@ferringb> interest of gentoo?  bolsh.
15:24 <@ferringb> ciaranm: portage isn't perfect.  this is why release keep coming.
15:24 -!- bobwhoops [n=eric@69.37.84.125] has quit [Client Quit]
15:24 <@ferringb> we still have it out there for use, because _someone_ has to do it.
15:24 <@ciaranm> what's the point in replacing one turd with another less familiar turd?
15:24 -!- eddyMul [n=[e-mail]] has joined #gentoo-dev
15:24 <@ferringb> more options then just replacing
15:24 <@ferringb> as I said, translating resolver into stable.
15:24  * ferringb waits for the 'but'
15:25  * mbishop inserts a 'nothing is perfect'
15:25 <@ciaranm> *shrug* i'll tell you the algorithm, if you need one
15:25 <@ferringb> rather see the code.  easier to translate.
15:25 <@ciaranm> heh. no it won't be
15:25 <@ferringb> heh
15:26 -!- xmt [n=[e-mail]] has quit [Client Quit]
15:26 <+mbishop> so you're saying your code sucks? :P
15:26 <@ferringb> feel free to send the algo our direction.  it's probably pretty blindingly obvious to be blunt.
15:26 <@ferringb> if you can chuck all of the existing code out and start from scratch, most of the portage flaws are easy to fix.
15:26 -!- eddyMul [n=[e-mail]] has quit [Client Quit]
15:26 <+ajax> heh
15:26 <@ciaranm> mbishop: no, i'm saying i'm making very heavy use of multiple inheritance, templates and multiple dispatch, none of which translate obviously into python
15:27  * ajax used to think the same about X re: chucking
15:27  * ferringb snickers
15:27 <@ferringb> ciaranm: so what, work has to be done.
15:27 <@ciaranm> mbishop: the ebuild.sh interface sucks horribly, i'll quite happily admit that part
15:27 <@ferringb> waah, it might involve some sweat.  amazingly people might be willing to improve it.