Is Gentoo dying?

[srunni]

Why focus on poaching users from anywhere? We're not selling a product. We just try to make this distro the best one for ourselves and our core user community. It's all about scratching an itch. People with the same kind of itches as us will feel at home with Gentoo, and others will not. And that's just fine. Linux is about freedom and the right tool for the job. If people want something that is easy to set up and works out of the box, that doesn't need a lot of tweaking, they will feel much more at home with Ubuntu. And that's fine. Others do want that tweakability and low-level configuration, and will find that Gentoo fits better.

But don't you think that there are people out there who would use Gentoo if they knew more about its benefits? Also, don't you think that more users will result in more bug reports, more support for other users, and more developers, as well as more recruiters of even more users?

Where did you get that idea? All that is needed is to find a dev who is interested in the package, and is willing to maintain it. Or find a dev who is willing to do proxy-maintenance.

I believe it was poly-p man who said he's tried to get ebuilds into the Portage tree quite a few times, and they've never gotten in.

This is sweeping generalisation, most of us simply don't have that.

Again a sweeping generalisation. I don't come across this attitude in our developer community much. It is the exception rather than the rule.

I didn't say that it's true; I'm saying that it appears to be, which is as much of a problem as there actually being a sense of elitism among the developers, as it drives current & potential users away.

Again, where did you get that idea? I became a dev last February and found the process relatively straight-forward. Yes, you are tested on your knowledge of making and maintaining ebuilds. But that is to be expected. You wouldn't want someone working on portage that doesn't know what he's doing, right? There has to be a certain level of quality control, and vetting new devs is part of that.

From what I understand, you have to be recommended by a current dev, there is a mentorship process, etc, etc. Instead, people should just be allowed to submit patches and know that they won't be ignored just because they're not a dev. I understand if you don't want to give someone access to *.gentoo.org servers, but why not have dev's review submissions and integrate them into Gentoo if they're good?

[desultory]

From what I understand, you have to be recommended by a current dev, there is a mentorship process, etc, etc.

In practical terms, yes, a recommendation from a current developer is required if for no other reason that to have someone to mentor the prospective developer and explain to them how and why things are handled in the way they are.

The recruiting process for new developers, in brief:

1. A prospective developer does something helpful or generally useful to the project, be it fixing bugs, submitting new ebuilds, testing or any of various other things as determined by the relevant existing developers.
2. They are asked if they are interested in joining the project.
3. If they accept an, existing developer shows them how things are to be done, commit messages, change logs, and various other points.
4. The prospective developer answers some questions which cover aspects of ebuilds and some information about how Gentoo is structured all of which, by this point, they should be familiar with by mere exposure.
5. The prospective developer is then given commit access to the main ebuild repository, during their first month all of their commits are to be reviewed by their mentor.
6. The new developer is given an e-mail address, shell access and web space on a server (the latter two tend to get minimal use), increased access to Bugzilla, an IRC cloak, a scarlet brand on the forums and a throng of users eager to devour their work.

Time constraints aside, do you seriously consider that to be a barrier to anyone who is actually interested?

Instead, people should just be allowed to submit patches and know that they won't be ignored just because they're not a dev.

Patches are solicited, and accepted, on a regular basis via bugs.gentoo.org.

I understand if you don't want to give someone access to *.gentoo.org servers, but why not have dev's review submissions and integrate them into Gentoo if they're good?

That is done now, and is by no means a recent development, in point of fact there is a project dedicated to that very task.

[lysergicacid]

Do you have specific examples of packages that you have to maintain in Gentoo but are available in the default repositories of the big distros (e.g., Fedora, Ubuntu, SuSE)?

If so, what do you propose to do about it? Mind you, Gentoo does not have a multi-millionaire cutting paychecks, nor a large corporation ...

mythtv / freevo / mupen64 among others...

mythtv svn been working just fine for me till very recently when something messed up the widescreen stuff :-/ ive seen that portage has ebuilds for limited 22 stuff in portage if u unmask it, u get just mythtv no plugins :-/, freevo is on release 1.8.2 yet even tho it functions fine its not in portage, mupen64 shouldn't even be in portage no more as its development seems to of stopped why not add http://code.google.com/p/mupen64plus into the tree ? at least its still active ?

[Hypnos]

* Seems like it takes a while before a mythtv version stabilizes

* As for freevo and mupen64plus, you have to get a developer interested enough to maintain in it. This is a volunteer distro, after all. There is an open bug for the latter.

[steveL]

Time constraints aside, do you seriously consider that to be a barrier to anyone who is actually interested?

I thought that during that process the prospective dev could be turned down at any point with no reason given?
"Gentoo Recruiters may reject new developers during this time if they feel it is appropriate."
That bit comes right at the end of the whole process.

I'm not trying to make trouble, I just want to know if it is indeed the case that someone could do all the preparation, and have come to the notice of other devs who want to bring the recruit in, but be turned down with no explanation, right at the end of it all.

[tanderson]

Time constraints aside, do you seriously consider that to be a barrier to anyone who is actually interested?

I thought that during that process the prospective dev could be turned down at any point with no reason given?
"Gentoo Recruiters may reject new developers during this time if they feel it is appropriate."
That bit comes right at the end of the whole process.

I'm not trying to make trouble, I just want to know if it is indeed the case that someone could do all the preparation, and have come to the notice of other devs who want to bring the recruit in, but be turned down with no explanation, right at the end of it all.

Have you ever gone through the process? It's really not that difficult at all.

Also, where did you get the part that he could be turned down with "no explanation"? All I can see is that recruiters may reject developers in this process(probably due to major problems like QA or personal problems), as is their right. No where does it state that these rejected users can't readmitted if they fix whatever they were doing wrong.

I really think you're blowing that out of proportion.

[steveL]

Have you ever gone through the process? It's really not that difficult at all.

Obviously not.

Also, where did you get the part that he could be turned down with "no explanation"?

I'm not sure where I heard it, on IRC sometime I think. That's why I was asking for clarity, since the document given seemed to indicate that it is a possibility, but only at the end of the whole process.

All I can see is that recruiters may reject developers in this process(probably due to major problems like QA or personal problems), as is their right. No where does it state that these rejected users can't readmitted if they fix whatever they were doing wrong.

Sure that's their right, the question is whether the recruit would get any explanation or even the chance to retry. If as you say it could be about "personal problems" then that doesn't sound so good; the recruit might simply not get on with some of the people who have a say (whoever that group might be, clearly not whoever mentored them or their peers) and it's kind of hard to "fix" that, as well as it being irrelevant to how good their work is. IOW it starts to sound like the political, "brown-nosing" process that the other poster was questioning.

I really think you're blowing that out of proportion.

Yeah that's your right. Thing is you don't sound like you actually know that much about it, since your response is based simply on what you've just read and your own recruitment, and not experience of being a recruiter. If you don't mind, I'd rather hear from someone who works on that process, if possible.

[yngwin]

Sure that's their right, the question is whether the recruit would get any explanation or even the chance to retry.

Of course they would get an explanation, and yes, the chance is given to retry later.

If as you say it could be about "personal problems" then that doesn't sound so good; the recruit might simply not get on with some of the people who have a say (whoever that group might be, clearly not whoever mentored them or their peers) and it's kind of hard to "fix" that, as well as it being irrelevant to how good their work is. IOW it starts to sound like the political, "brown-nosing" process that the other poster was questioning.

There is nothing like that going on in the recruitment process. Of course the Code of Conduct needs to be adhered to, as is the case everywhere in Gentoo. But the recruitment process is all about assessing (and where necessary improving) the recruit's knowledge and abilities as to the technical and organisational aspects of working as a Gentoo developer.

[Gankfest]

S**t if Gentoo died I would too. /cry with a tear drop!

Paradox(>")>

[-Craig-]

I'd just say: *plonk*.

[d2_racing]

S**t if Gentoo died I would too. /cry with a tear drop!

Paradox(>")>

If Gentoo died, I think that funtoo.org would be an official alternative.

[AllenJB]

S**t if Gentoo died I would too. /cry with a tear drop!

Paradox(>")>

If Gentoo died, I think that funtoo.org would be an official alternative.

Yes, Daniel Robbins is going to maintain all the packages in the tree on his own - a task that ~250 odd developers struggle with. Not to mention continued development of the Gentoo toolset (portage and all related tools).

Of course all this is mute, because Gentoo is not going to die any time soon.

[d2_racing]

Yep, of course

[Slalomsk8er]

S**t if Gentoo died I would too. /cry with a tear drop!

Paradox(>")>

If Gentoo died, I think that funtoo.org would be an official alternative.

Yes, Daniel Robbins is going to maintain all the packages in the tree on his own - a task that ~250 odd developers struggle with. Not to mention continued development of the Gentoo toolset (portage and all related tools).

Of course all this is mute, because Gentoo is not going to die any time soon.

Do you really think, that no one would help him and that all remaining devs, exdevs and capable users would switch to ubuntu?

[djinnZ]

Gentoo is not going to die

NO, GENTOO IS DIED! We all users are dead, the devels are dead also.
But we are so stupid and unable to undertstand it, so there is an appareance of life.
(Something similar to the peaple walking on vacuum in the cartoons)
Apologize me but I cant resist.

Halloween is gone, but for the next first day of April, for a joke ofcourse, is a wrong idea to announce the dead of gentoo?! (put a false announce on the main site, change groups on the forum in something related to the undeads, as was be for the ninjas in past year)
Sometimes laught is the better way to fight the misinformation.

[NeddySeagoon]

djinnZ,

This is one of the better Gentoo April Fools
GeNToo April Fool

[kernelOfTruth]

djinnZ,

This is one of the better Gentoo April Fools
GeNToo April Fool

classic

thanks for digging that out, Neddy

[tylerwylie]

Gentoo isn't dieing, it's more or less being morphed though into an enthusiasts' OS. It's very slender and slim but if you want to install 3rd party software, support may not be there. For desktops, binary packages included in distributions are being built with heavily optimized code, and when Gentoo used to zip around, you have to prelink and preload what you can for it to come close to some of the other binary distributions, (Fedora and OpenSUSE are very quick in this regard)

I really enjoyed using Gentoo on my laptop, but for now it's been relegated to servers, as the documentation fades the users will fade, as that was Gentoo's BIGGEST bonus. If you had an issue, someone else has had it before you, and wrote down HOW to fix it and put it on a website that was easy to find. This isn't the case as much anymore, and that's when Gentoo's lost it's golden edge... the community waned.

[srunni]

Fedora and OpenSUSE are very quick in this regard

You're joking, right? yum was slow as molasses for me when I tried Fedora 9 in a VM. It is much slower than Ubuntu. Maybe Fedora 10 will have improved in that regard.

[AllenJB]

Gentoo isn't dieing, it's more or less being morphed though into an enthusiasts' OS.

What makes you think of it as an enthusiasts OS? While it is and always will be aimed at the more technically minded purely by being a from-source distro, I think Gentoo is very easy to use and work with.

I'd choose it over any other distro in most situations. Its binary package system makes it awesome for servers where you want to test on one machine and install to a live system, and lower powred laptops where compiling on the machine itself can be slow. It's awesome for a desktop because I get exactly the setup I want (pure ALSA with OSS emulation for sound, for example - none of the daemons I'd be forced to install with other distros). And unlike every other distro I've used, it's never overwritten any of my configuration without asking me first.

It's very slender and slim but if you want to install 3rd party software, support may not be there.

You mean except for the entire ebuild system which means that it's easy to find, update or modify packages produced by other people around the internet in any one of hundreds of overlays. And if you can't find one, it usually takes no more than half an hour to knock up an ebuild for most packages.

For desktops, binary packages included in distributions are being built with heavily optimized code, and when Gentoo used to zip around, you have to prelink and preload what you can for it to come close to some of the other binary distributions, (Fedora and OpenSUSE are very quick in this regard)

Because they do the prelinking and preloading too. Exactly the same techniques. The only difference is that with Gentoo you know what and how things are being done because you have to set them up yourself. This has and always will be the case.

As for heavily optimized - don't forget that with Gentoo you compile for your exact CPU - binary distros will compile for the lowest common denominator (all amd64 binary distros have to compile for generic x64, anything else is most likely i686 or worse). On Gentoo (with GCC 4.3) you can use march=native.

On Gentoo, the only machine I compile with march=i686 is my Eee and my server, where I compile on one machine then install to another (both compile hosts are Xen VMs - the server is a Xen VM too)

I really enjoyed using Gentoo on my laptop, but for now it's been relegated to servers, as the documentation fades the users will fade, as that was Gentoo's BIGGEST bonus. If you had an issue, someone else has had it before you, and wrote down HOW to fix it and put it on a website that was easy to find. This isn't the case as much anymore, and that's when Gentoo's lost it's golden edge... the community waned.

What's all this about documentation fading? Yes - the wiki died, but the vast majority of articles from the old wiki were saved (http://gentoo-wiki.info) and a new one is being built in its place (http://gentoo-wiki.com). From what I've seen the new wiki already has some high quality articles that were never on the old wiki (I've written some of them myself) and there are more foreign language wiki's than there used to be too.

Frankly, the death of the old wiki was long overdue in my opinion - while there were some good articles on there, the vast majority of it was unindexed, out of date, incorrect, used bad practices or just badly written. Most of the good articles are already on the new wiki - rewritten and updated.

[bunder]

the vast majority of it was unindexed, out of date, incorrect, used bad practices or just badly written.

so is TLDP, but people still use it.

[gerard27]

tylerwylie wrote

I really enjoyed using Gentoo on my laptop, but for now it's been relegated to servers, as the documentation fades the users will fade, as that was Gentoo's BIGGEST bonus. If you had an issue, someone else has had it before you, and wrote down HOW to fix it and put it on a website that was easy to find. This isn't the case as much anymore, and that's when Gentoo's lost it's golden edge... the community waned.

We now have EAPI-2 and portage-2.2.
Many ebuilds won't compile without it.
Yet it's still masked/unstable and docs are hard to find.
Gerard.

[AllenJB]

tylerwylie wrote

I really enjoyed using Gentoo on my laptop, but for now it's been relegated to servers, as the documentation fades the users will fade, as that was Gentoo's BIGGEST bonus. If you had an issue, someone else has had it before you, and wrote down HOW to fix it and put it on a website that was easy to find. This isn't the case as much anymore, and that's when Gentoo's lost it's golden edge... the community waned.

We now have EAPI-2 and portage-2.2.
Many ebuilds won't compile without it.
Yet it's still masked/unstable and docs are hard to find.
Gerard.

All of the EAPI-2 ebuilds I've come across are either marked for testing (where it's not at all unusual for dependencies to be packages that are also in testing) or in overlays. If you asked around or did a quick search at all, you'd discover that the upgrade to 2.2 is trivial. The only noticeable changes for users is the move to sets (@system and @world instead of system and world, and that @world does not include @system) and the preserved-libs feature (basically designed to replace revdep-rebuild).

[gerard27]

@AllenJB,
Sure if you search you'll find info.
But it's still masked and there's no "official" howto.
I use Gentoo on a desktop exclusively so I shy away from masked/unstable.
I removed Windows completely,have no fall-back.
Gerard.

[AllenJB]

@AllenJB,
Sure if you search you'll find info.
But it's still masked and there's no "official" howto.
I use Gentoo on a desktop exclusively so I shy away from masked/unstable.
I removed Windows completely,have no fall-back.
Gerard.

For the most part "testing" packages are generally no less stable than those considered "stable" for the vast majority of people - there are some exceptions to this - baselayout 2 for example contains major changes. "Unstable" portage versions generally don't cause any issues, and when they have (I think this has only been the case on 1 occaision for me), the devs have been very helpful in sorting them out.

Information on what the testing branch is and how to use it is fully documented in the Portage section of the Gentoo Handbook. You can find the changes in Portage 2.2 by simply googling for "Portage 2.2", where you'll find plenty of dev blogs (also keeping an eye on http://planet.gentoo.org is a good way to keep up with such changes).

Gentoo generally isn't a distro for people who like all the information they need put in front of them. It expects its users to be competent is general internet usage and reading documentation. You do sometimes have to "go out of your way" to find information. The Gentoo devs have far more interesting things to do than try to serve you all the information you want in easily digestable formats. Don't like this? Find another distro or write documentation - either by contributing to the official docs or somewhere else like the wiki, alternatively you could just post to the forums, your blog (if you have one) or write a book.