Gentoo Users Assemble!

One of my good friends, Josh (aka tsunam) is a Trustee for Gentoo. He has been fervently pointing out on his blog that the whole community is in a state of flux and Gentoo really needs its community to step up and help out however possible. If any of you have any words of encouragement or are willing to volunteer I'm sure Josh would really love to hear from you.

a good mate of mine (Neddy) was made one as well.
Gentoo isn't in so much of a bad shape, the fact that the legal side (to keep US law happy) was allowed to slip was bad but gentoo on the whole is in good shape

it doesn't help you have the likes of ciaranm and co just sniping (still!) at teh core of gentoo and now underhand tactics like kde overlay needing paludis if you want to use kde4 overlay
 
it doesn't help you have the likes of ciaranm and co just sniping (still!) at teh core of gentoo and now underhand tactics like kde overlay needing paludis if you want to use kde4 overlay

Political problems aside, paludis has turned out to be an excellent package manager. I started using it on new installs about six months ago, after portage pooched my /var/db/pkg database while I was playing with some external repositories on my main workstation. Now I'm using paludis on nearly all of my machines, with zero problems so far. I like it better than pkgcore, the only other alternative to portage.
 
Political problems aside, paludis has turned out to be an excellent package manager. I started using it on new installs about six months ago, after portage pooched my /var/db/pkg database while I was playing with some external repositories on my main workstation. Now I'm using paludis on nearly all of my machines, with zero problems so far. I like it better than pkgcore, the only other alternative to portage.
++
When portage was conceived back in the day they were toying with new ideas and capabilities so the code was young and stitched together. Well, they never really went back and architected the code correctly when it was needed with a major rewrite. Now portage is a total wreck of spaghetti code which makes it nearly impossible to add new features. In addition, often "stable" features of portage rely on the many quirks in behavior that portage has developed over the years. Therefore, if you were brave enough to try and actually fix some of portage's problems you end up breaking 20 things that should have nothing to do with what you changed.

Pkgcore/Paludis are the answers to this problem, and in my opinion, paludis is the better of the two. Give it a shot, and take the time to learn the slightly different ways of configuring and installing packages. It's worth it.
 
++
When portage was conceived back in the day they were toying with new ideas and capabilities so the code was young and stitched together. Well, they never really went back and architected the code correctly when it was needed with a major rewrite. Now portage is a total wreck of spaghetti code which makes it nearly impossible to add new features. In addition, often "stable" features of portage rely on the many quirks in behavior that portage has developed over the years. Therefore, if you were brave enough to try and actually fix some of portage's problems you end up breaking 20 things that should have nothing to do with what you changed.

Pkgcore/Paludis are the answers to this problem, and in my opinion, paludis is the better of the two. Give it a shot, and take the time to learn the slightly different ways of configuring and installing packages. It's worth it.

Yup Portage is basically getting patched to keep it viable but it is very hard to work, the code is true spagetti code. What portage does now and what portage was originally written todo are not the same thing (quite alot to take into account STUPID upstream tar-balling!!!!!). Portage while being written in python didn't take advantage of alot of pythons power (OOP and co) and as more and more features were required to make ebuilds work nice the state of the code became apparent.

Pkgcore actually started as a branch in the portage version control system (mainly for testing dependencies and new features) and the codebase is alot more workable with. Pkgcore main dev took a break for quite some time and basically for 8months it had issues that ment that it couldn't be used with some types of package (in the 2weeks he has been back coding all is good again with pkgcore).

The issue with Paludis (besides the political aspect where the lead-dev has been kicked from being a gentoo-dev TWICE, he trolls on the -gentoo-dev ml and de-rails it far too many times... let alone his deputies doing the same kinda stuff... spb been perma-banned from f.g.o) you have their underhand tactic of forcing EAPI-1 extra stuff into the ebuilds of kde4-svn repository (thus if ppl want to use kde4-svn they MUST use paludis)

funny that take something then extend it so only ppl using yr software can take advantage.. sounds familiar w.r.t. Microsofts: Embrace,extend,exterminate mantra (and with an ex-gentoo dev who was kicked is BAD!!!).
Yes some of the things that the paludis lot proposed for EAPI-1 and have implemented in paludis are nice (the src => redirect for downloading is very nice, again to get around upsteam stupidity) but others are pointless. And in true Paludis style rather then discussing it to weed out the useless "enhancements" and implement the beneficial ones the coders implement it all themselves and then push such enhancement into the wild EVEN before portage-dev's have had a change to even concider which ones are good BAD!!!!

in the long term portage will be replaced (the dev's that work on it get burnout really quickly with the portage codebase) but whether it is replaced (by default) with paludis or pkgcore all depends on whether the core-dev's have learn't anything from when they were kicked/left the gentoo team (if they still can't work in a big team then no chance!)
portage still does things that neither pkgcore or paludis can do (mainly do to blackmagic being added to ebuilds to get around portage issues ;) )

also paludis has a nice list of situations where it wont work and it doesn't work on all arch's. portage/pkgcore since it just needs a toolchain and python works on alot more arch's and for a major release of python and proven out both work fine

Take python26 due out in a few more months it is ment to be a bridge between python3000 and python25 so it is backwards compat and fwd-enabled thus portage/pkgcore *should* just work with no issues (and then slowly python3 features are enabled ready for python3000)
 
The issue with Paludis (besides the political aspect where the lead-dev has been kicked from being a gentoo-dev TWICE, he trolls on the -gentoo-dev ml and de-rails it far too many times... let alone his deputies doing the same kinda stuff... spb been perma-banned from f.g.o) you have their underhand tactic of forcing EAPI-1 extra stuff into the ebuilds of kde4-svn repository (thus if ppl want to use kde4-svn they MUST use paludis)

funny that take something then extend it so only ppl using yr software can take advantage.. sounds familiar w.r.t. Microsofts: Embrace,extend,exterminate mantra (and with an ex-gentoo dev who was kicked is BAD!!!).

You're totally off the deep end here - it's not as if the KDE4 repository is the main Gentoo tree. The guys maintaining that repo use paludis, and their repo reflects that. It's no different than projects using different source control packages; some use git, others mercurial, still others use subversion. I really doubt the KDE4 thing is a political statement. If I was maintaining a repo, I'd use the paludis features too; not because of my personal politics, but because of pragmatism.

Yeah, ciaranm is an asshole, get over it already. Lets move on.
 
cmon, i cant be the only zenwalk user here! someone else needs to try it!
 
just was able to more easily configure the laptop exactly the way I wanted with arch. you start out with just a basic cmd line and its super easy from there to build up and know exactly what you are putting in.... just appeals to me more :)

that, and I was able to get my memory usage down to ~48 MB of ram when loading to the desktop in arch (in openstep)
 
I was too dumb to use Gentoo for long. I struggled for a few weeks though before giving up.
 
i know i risk looking like an arch shill, but really, if you struggled with gentoo, but still have an inquiring mind about linux and want something a bit simpler, but still very advanced, give arch a try......

I used it after becoming frustrated with Gentoo (and migrating to FreeBSD), and loved it.
 
yer Archlinux is nice (for a binary distro, but like all binary distro's it has some MAJOR BRAIN DEAD SETUPS & DECISIONS!!!!!!!!! - don't get me wrong gentoo has some as well but majority of those I can work around without the premadonna's of said distro throwing their compilers out of their sandbox) - saying that I run archlinux on one of my servers and on my eeePC (simply because I was too lazy to setup a xen setup to crosscompile gentoo for both those systems)

On the Gentoo topic, may I point you all here to: update a crazy BASH script written by a mate of mine
http://forums.gentoo.org/viewtopic-t-546828.html

it basically wraps around portage/pkgcore and automates alot of the process (and hides all the compile crap that files across yr screen). It ensure alot of the behind the scene system maintenance that we gentoo users should be doing to ensure our src-distro is in tip-top condition but majority of the time forget or don't know we need to (leading to the Expat incident - which btw I didn't get hit by cause I revdep and --depclean once a forenight)
 
All I can say is that my experience is quite different. I just built a new Gentoo system inside of a virtual machine today, and I used Paludis from stage3 to rebuild the whole thing with different CFLAGS. No problems whatsoever.

Posting links like that one aren't very helpful. It's clear that the author had an agenda just from looking at the front matter, and no configuration details were provided beyond the hardware used (who cares). Post some tests that measure relevant things, and that can be reproduced by the rest of us, and then we can all have a discussion. Otherwise, you make yourself look foolish, and waste our time and bandwidth.
 
what information do you require?
and these are just har numbers either do the test yourself of accept these.

and if you want to talk about agenda's is funny when dealing with paludis
They ALWAYS push their own agenda, their dev's are known for it

just look at what has happened with the kdesvn overlay and now ONLY usable with paludis (the non-std non-complete package manager) why else force that unless there is an agenda

or the who history of how/why paludis came about


AND if you want it right here and now just type paludis moo into your terminal to have a glimpse that agenda pushing
 
I'm being neutral and objective; I'm not pushing paludis and I'm not spreading propaganda regarding Gentoo package managers. I'm not interested in agendas at all. I recognize them when I see them, and that causes me to disregard everything said by that party - which includes any sort of benchmarks, tests, or public relations drivel. Agendas don't, however, affect my choices of free software. I use free software which suits my purpose, regardless of whether the developer is a saint or an asshat. For example, I read the LKML frequently; I don't approve of how some of the kernel developers behave, but I use Linux anyway because it is good software.

If you don't know what the requirements are for objective testing, and can't enter into a discussion without preconceptions, then I'm not going to go any further with this.
Code:
# paludis moo
Usage error: don't be silly
Try paludis --help
#
I was hoping for a cool Easter egg :confused:
 
Back
Top