Support for pkgcore

The_Document · Apprentice Joined: 03 Feb 2018 Posts: 275

Made suggestion on bug tracker and Im seriously hoping devs choose to impliment pkgcore alongside missing functionality:

https://bugs.gentoo.org/show_bug.cgi?id=649158

The Doctor · Moderator Joined: 27 Jul 2010 Posts: 2678

I'd say use whichever you prefer. Gentoo is about choice.

In fairness it should be pointed out that portage is slower because it has features pkgcore lacks.
_________________
First things first, but not necessarily in that order.

Apologies if I take a while to respond. I'm currently working on the dematerialization circuit for my blue box.

Juippisi · Developer Joined: 30 Sep 2005 Posts: 724 Location: /home

I hope (and believe) portage will become faster when the code is made python3.6 compatible.

The_Document · Apprentice Joined: 03 Feb 2018 Posts: 275

Goverp · Advocate Joined: 07 Mar 2007 Posts: 2004

If pkgcore has algorithms that perform better than those in portage, while offering the same function, I'd hope portage stole them. Even so, having an alternative development to portage is good, as it's a way to explore different algorithms and user interfaces. Evolution benefits from having two parents. Similar happens with Java runtimes, where the different vendors explore different storage management (among others).
_________________
Greybeard

Zucca · Posted: Fri Mar 02, 2018 10:52 am Post subject:

Doesn't pkgcore also run on Python?

I don't know if Python or the algorithms are to blame. Iirc Paludis is mostly coded using C++. I would think Paludis is faster.

Personally I don't like any system central component relying on Python, but it sure is better than to rely on haskell or java or php. I've heard some horror stories about haskell and I've experienced the breakages of java and php. With Python I sometimes get blockers, but that's all. Gentoo devs know how to handle Python well. So, for the moment, I'll stick with Portage.
_________________
..: Zucca :..
Gentoo IRC channels reside on Libera.Chat.
--

NeddySeagoon · Posted: Fri Mar 02, 2018 10:55 am Post subject:

The_Document,

pkgcore is feature incomplete. If you can work around it, feel free to adopt it now.

My current choice is portage, so that my package manager just works.
_________________
Regards,

NeddySeagoon

Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail.

Mr. T. · Guru Joined: 26 Dec 2016 Posts: 477

I answered "yes" because the documentation could be better. In a way, it results in "multi-speed evolution".

bunder · Bodhisattva Joined: 10 Apr 2004 Posts: 5934

I'd just be happy with portage being multithreaded (not withstanding gcc et al)

Dr.Willy · Guru Joined: 15 Jul 2007 Posts: 547 Location: NRW, Germany

asturm · Developer Joined: 05 Apr 2007 Posts: 8936

The poll question is wrong. pkgcore is supported as in being available in portage, same with paludis. But pkgcore must support latest EAPI before it can be considered a viable alternative.

Mr. T. · Guru Joined: 26 Dec 2016 Posts: 477

It is a choice between Portage and Pkgcore. In general, people avoid (note) using multiple equivalent software.

Note: a package manager may be complex. Will a user be willing to learn more than one package manager?

Ant P. · Watchman Joined: 18 Apr 2009 Posts: 6920

Less umarells, please.

Zucca · Posted: Sat Mar 03, 2018 8:27 pm Post subject:

pun_guin · Apprentice Joined: 06 Feb 2018 Posts: 204

edit, as I had misunderstood pkgcore in the first place:
I want speed, so I voted for exactly that. I doubt that a vote in this board is of any relevance for the project itself though.
_________________
I already use the new Genthree.

Mr. T. · Guru Joined: 26 Dec 2016 Posts: 477

We can learn from exchanges and possibly create a good synergy. This is an opinion survey and a bug report has been opened.

szatox · Advocate Joined: 27 Aug 2013 Posts: 3135

Mr. T. · Guru Joined: 26 Dec 2016 Posts: 477

A questionnaire can be difficult to prepare. In addition, each person can express his opinion. Finally, numbers gives effectively the number of people involved,
those who are favorable and those who are unfavorable.

Ant P. · Watchman Joined: 18 Apr 2009 Posts: 6920

Zucca · Posted: Mon Mar 05, 2018 6:38 pm Post subject:

tld · Veteran Joined: 09 Dec 2003 Posts: 1816

Hu · Moderator Joined: 06 Mar 2007 Posts: 21627

Moderators (and the original poster) can modify a poll to change the phrasing of the question. I could fix it, but I feel like I should run a poll to get consensus on the fixed wording before changing it.

szatox · Advocate Joined: 27 Aug 2013 Posts: 3135

Hu, changing the wording right now, after a bunch of people already voted is a bad idea.
Right now you could use it for a data point in an attempt to answer question like "how many people notice BS when they see it"*.
Obviously, this actually is a tricky question, so you can't answer it with a single data point. Still, as you already mentioned, changing the answers half-way through would turn a bad poll into an inconsistent one, which is even worse.

* I can say one good thing about this: the results so far were really surprising for me.

Hu · Moderator Joined: 06 Mar 2007 Posts: 21627

I agree, but I could not pass up the opportunity to suggest a poll as to whether this poll needs to be fixed. Perhaps I've simply been to too many meeting-planning-meetings lately.

The_Document · Apprentice Joined: 03 Feb 2018 Posts: 275

Come to think of it, pkgcore while fast as it is NOW, might turn out dog slow when/if it obtains full portage feature equivalency. I honestly believe it would still outperform portage by a large margin, there is just something about snake oil and the fact that pkgcore will likely use the newest python which is faster.