Gentoo Forums
Gentoo Forums
Gentoo Forums
Quick Search: in
please new forum _Advanced_Gentoo_without_KIT_
View unanswered posts
View posts from last 24 hours

Goto page Previous  1, 2, 3, 4, 5  Next  
Reply to topic    Gentoo Forums Forum Index Gentoo Forums Feedback
View previous topic :: View next topic  
Author Message
depontius
Advocate
Advocate


Joined: 05 May 2004
Posts: 3509

PostPosted: Wed Dec 04, 2013 1:53 am    Post subject: cgroups, etc Reply with quote

At this point it's worth noting that Tejun Ho has begun reworking the cgroup code. I believe this showed up in Linux Weekly News of November 21. Heck, here... https://lwn.net/Articles/574317/ Basically he said it was done by someone with very deep C skills, but not a lot of experience with the Linux kernel. Mistakes were re-made that were fixed in other kernel subsystems long ago.

Unfortunately, because cgroups have already made it into the wild, it means that there is going to be a Version 2, and the existing stuff is going to have to hang around a while for backward compatibility.

Armchair analysis...

Miguel de Icaza long ago talked about "Windows done right" being his goal. Maybe I'm wrong, but I get the feeling that that's what's really happening with freedesktop.org and its associates. However I feel that it's all going more or less wrong, because they give every indication of not truly understanding underlying Unix philosophy.

The net result is rather, "Windows done better". Sometimes I'm not even sure if Windows can be done right - there are problems right down at the base of Windows.

Back to concrete suggestions...

1 - Separate "do-without" groups, some of us are more purist, some of us more sullied. I for one don't plan on doing without pam, though I know that some do. I'm also willing to stick with udev for the time being, even if it is the spawn of systemd.

2 - For each group we need a few basics - package.mask, USE flags, and probably package.keywords because the jump to gnome-3.8 (for instance) may have skipped an in-between release that is quite servicable and perhaps better in appropriate ways.

3 - We might need an overlay - for when some of these packages get dropped from portage. I ran some sort of qt3 legacy overlay for quite a while, until mythtv and "make xconfig" moved to the newer qt.

Anyway, I'm interested - I started the "assimilation" thread, basically on this topic, less general
_________________
.sigs waste space and bandwidth
Back to top
View user's profile Send private message
mv
Watchman
Watchman


Joined: 20 Apr 2005
Posts: 6747

PostPosted: Wed Dec 04, 2013 9:12 am    Post subject: Reply with quote

ulenrich wrote:
cgroups are not that simple as claimed there

Sure they are (for what concerns systemd's usage of them; the implementation in the kernel is a different thing, but in contrast to systemd this appears to be done by experts): In systemd it is a nothing but a slightly more secure chroot used as a security feature for subdaemons and a way to make daemons steal slightly less time. Which becomes kind of a joke when you make the system insecure and steal time by having permanently unnecessary root daemons running for tasks for which no daemon would be necessary at all in a proper implementation.
The real advantage of cgroups is only when you need certain virtualization for a particular purpose which has nothing to do with L.P. loudmouth "application".
Back to top
View user's profile Send private message
Anon-E-moose
Watchman
Watchman


Joined: 23 May 2008
Posts: 6095
Location: Dallas area

PostPosted: Wed Dec 04, 2013 10:19 am    Post subject: Reply with quote

mv wrote:
ulenrich wrote:
cgroups are not that simple as claimed there

Sure they are (for what concerns systemd's usage of them;


Indeed the myth that they aren't simple has long been busted.
But it's one of those things that sycophants and trolls love to trot out even though it's a lie.
They've not done any real research on their own, they simply take L.P's or Kay's or someone else's myth as fact.

In the scheme of things the number of people that actually use cgroups is extremely small.
But it's pushed as one of the great things that systemd does.
It's really a solution in search of a problem.
_________________
PRIME x570-pro, 3700x, 6.1 zen kernel
gcc 13, profile 17.0 (custom bare multilib), openrc, wayland
Back to top
View user's profile Send private message
Anon-E-moose
Watchman
Watchman


Joined: 23 May 2008
Posts: 6095
Location: Dallas area

PostPosted: Wed Dec 04, 2013 10:25 am    Post subject: Re: cgroups, etc Reply with quote

depontius wrote:

Back to concrete suggestions...

1 - Separate "do-without" groups, some of us are more purist, some of us more sullied. I for one don't plan on doing without pam, though I know that some do. I'm also willing to stick with udev for the time being, even if it is the spawn of systemd.


I'm one of those who doesn't use pam at all. I once had it on my system,
but that was because it was installed with gnome when I had it on my system.
It's not that I have a problem with pam, I just don't see the need on my system.
I don't have a problem using sudo.

I still use udev, but it's the 171-r6 version. And it still works perfectly fine.

Quote:
2 - For each group we need a few basics - package.mask, USE flags, and probably package.keywords because the jump to gnome-3.8 (for instance) may have skipped an in-between release that is quite servicable and perhaps better in appropriate ways.


I agree, but I think that the response from ebuild maintainers, devs, etc will be "no, too much extra work".

Quote:
3 - We might need an overlay - for when some of these packages get dropped from portage. I ran some sort of qt3 legacy overlay for quite a while, until mythtv and "make xconfig" moved to the newer qt.

Anyway, I'm interested - I started the "assimilation" thread, basically on this topic, less general


An overlay is probably the best given the response that I think will happen re. point 2
_________________
PRIME x570-pro, 3700x, 6.1 zen kernel
gcc 13, profile 17.0 (custom bare multilib), openrc, wayland
Back to top
View user's profile Send private message
depontius
Advocate
Advocate


Joined: 05 May 2004
Posts: 3509

PostPosted: Wed Dec 04, 2013 1:10 pm    Post subject: Re: cgroups, etc Reply with quote

Anon-E-moose wrote:
depontius wrote:

2 - For each group we need a few basics - package.mask, USE flags, and probably package.keywords because the jump to gnome-3.8 (for instance) may have skipped an in-between release that is quite servicable and perhaps better in appropriate ways.


I agree, but I think that the response from ebuild maintainers, devs, etc will be "no, too much extra work".

I wasn't suggesting anything for the maintainers - I was thinking of posting/editing entries on a sticky thread.
"To prevent gnome-3.8 add this to package.mask", and then a list from package.mask.
"To prevent systemd add this to package.mask, plus the gnome-3.8 entries", and then another list.
Etc, etc, it would end up being a series of lists, simply maintained on a forum thread. Preferably the information would be at the top, and kept current with edits, discussion and contributions below.
Anon-E-moose wrote:
depontius wrote:

3 - We might need an overlay - for when some of these packages get dropped from portage. I ran some sort of qt3 legacy overlay for quite a while, until mythtv and "make xconfig" moved to the newer qt.

Anyway, I'm interested - I started the "assimilation" thread, basically on this topic, less general


An overlay is probably the best given the response that I think will happen re. point 2

My thought of the overlay is for packages becoming obsolete in mainline portage, to keep them around. It seems to me that this could be entirely user-maintained, no developer load, unless a developer grew interest.

Call it "Gentoo Old Codger Linux".
_________________
.sigs waste space and bandwidth
Back to top
View user's profile Send private message
Anon-E-moose
Watchman
Watchman


Joined: 23 May 2008
Posts: 6095
Location: Dallas area

PostPosted: Wed Dec 04, 2013 1:24 pm    Post subject: Re: cgroups, etc Reply with quote

depontius wrote:
I wasn't suggesting anything for the maintainers - I was thinking of posting/editing entries on a sticky thread.
"To prevent gnome-3.8 add this to package.mask", and then a list from package.mask.
"To prevent systemd add this to package.mask, plus the gnome-3.8 entries", and then another list.
Etc, etc, it would end up being a series of lists, simply maintained on a forum thread. Preferably the information would be at the top, and kept current with edits, discussion and contributions below.


Ah, that would be workable.

Quote:
Call it "Gentoo Old Codger Linux".


LoL

I do keep all old ebuilds, but that's just part of the solution, the packages themselves src tarballs and patch tarballs need to be kept too.

Anyway sounds workable.
_________________
PRIME x570-pro, 3700x, 6.1 zen kernel
gcc 13, profile 17.0 (custom bare multilib), openrc, wayland
Back to top
View user's profile Send private message
Anon-E-moose
Watchman
Watchman


Joined: 23 May 2008
Posts: 6095
Location: Dallas area

PostPosted: Wed Dec 04, 2013 2:05 pm    Post subject: Reply with quote

NeddySeagoon wrote:
Anon-E-moose,

I suppose we could split off the 'bloatware' into its own forum, leaving the traditionalists to the main forms to enjoy the way its always been.


Judging by the few threads I've seen lately about problems with gnome 3.8
and the lack of competent help from the cheerleaders to fix those problems
that might not be a bad idea. LoL

Edit to add:
Seems one of the devs, Leio, actually has some idea of how to help, that's a start
_________________
PRIME x570-pro, 3700x, 6.1 zen kernel
gcc 13, profile 17.0 (custom bare multilib), openrc, wayland
Back to top
View user's profile Send private message
olek
Apprentice
Apprentice


Joined: 22 Oct 2011
Posts: 173

PostPosted: Wed Dec 04, 2013 6:35 pm    Post subject: Reply with quote

Well, I kind of am trying to maintain a suckless Gentoo as much as possible, while keeping every usual functionality.
I'm quite pleased with the outcome so far.
If this is the sort of thing this thread (or the wished forum) is asking for, I'll be glad to help/ contribute.
_________________
https://plaintext.blog
Back to top
View user's profile Send private message
NeddySeagoon
Administrator
Administrator


Joined: 05 Jul 2003
Posts: 54098
Location: 56N 3W

PostPosted: Wed Dec 04, 2013 7:44 pm    Post subject: Reply with quote

olek,

I like your philosophy.

Indeed, I can remember a time when you could not add memory to a computer, so if your program was too big to fit, you had to make it smaller.
This lead to lean mean code everywhere, or the code cold not be run. It also bred its own nightmares, like self modifying code.
_________________
Regards,

NeddySeagoon

Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail.
Back to top
View user's profile Send private message
saellaven
l33t
l33t


Joined: 23 Jul 2006
Posts: 646

PostPosted: Wed Dec 04, 2013 9:03 pm    Post subject: Reply with quote

My suspicions that this thread was started purely to try to troll aside...

At some point, all of the package masks and everything else will become rather cumbersome. Just look at the mask for gnome3 right now. As has also been intimated by others in this thread, packages updates have a tendency to suck in other packages we don't want. Add in that there are several Gentoo devs that I no longer trust to NOT put their political agenda before the overall health of Gentoo, and I'm not sure Gentoo's sanity is something that can be taken for granted going forward.

Ultimately, I think the proper answer isn't a list of masks since it will be impossible to blacklist every potential package update that's bad in a reasonable time frame (particularly since ANY dev can change an ebuild of a package they aren't the maintainer of), it's to fork Gentoo and regularly pull in what is known good and sane (much like libreoffice/openoffice) as time permits.
_________________
Ryzen 3700X, Asus Prime X570-Pro, 64 GB DDR4 3200, GeForce GTX 1660 Super
openrc-0.17, ~vanilla-sources, ~nvidia-drivers, ~gcc
Back to top
View user's profile Send private message
Anon-E-moose
Watchman
Watchman


Joined: 23 May 2008
Posts: 6095
Location: Dallas area

PostPosted: Wed Dec 04, 2013 9:52 pm    Post subject: Reply with quote

saellaven wrote:
My suspicions that this thread was started purely to try to troll aside...


You wouldn't be wrong, but as I said things take on a life of their own.


Quote:
Ultimately, I think the proper answer isn't a list of masks since it will be impossible to blacklist every potential package update that's bad in a reasonable time frame (particularly since ANY dev can change an ebuild of a package they aren't the maintainer of), it's to fork Gentoo and regularly pull in what is known good and sane (much like libreoffice/openoffice) as time permits.


That's an interesting thought.
_________________
PRIME x570-pro, 3700x, 6.1 zen kernel
gcc 13, profile 17.0 (custom bare multilib), openrc, wayland
Back to top
View user's profile Send private message
John R. Graham
Administrator
Administrator


Joined: 08 Mar 2005
Posts: 10587
Location: Somewhere over Atlanta, Georgia

PostPosted: Wed Dec 04, 2013 9:56 pm    Post subject: Reply with quote

saellaven wrote:
Ultimately, I think the proper answer isn't a list of masks since it will be impossible to blacklist every potential package update that's bad in a reasonable time frame (particularly since ANY dev can change an ebuild of a package they aren't the maintainer of), it's to fork Gentoo and regularly pull in what is known good and sane (much like libreoffice/openoffice) as time permits.
Yeah, that'll take less time and energy than maintaining a profile and some masks.

- John
_________________
I can confirm that I have received between 0 and 499 National Security Letters.
Back to top
View user's profile Send private message
saellaven
l33t
l33t


Joined: 23 Jul 2006
Posts: 646

PostPosted: Wed Dec 04, 2013 11:07 pm    Post subject: Reply with quote

John R. Graham wrote:
saellaven wrote:
Ultimately, I think the proper answer isn't a list of masks since it will be impossible to blacklist every potential package update that's bad in a reasonable time frame (particularly since ANY dev can change an ebuild of a package they aren't the maintainer of), it's to fork Gentoo and regularly pull in what is known good and sane (much like libreoffice/openoffice) as time permits.
Yeah, that'll take less time and energy than maintaining a profile and some masks.

- John


At any given moment, a package can be checked in that does something harmful... whether that's renaming/moving a package that breaks existing systems since the devs didn't test as much as they should have (as recently happened with python-exec), or changing system profiles/useflags/depends, or even on the more extreme end, pulling in malware, etc.

Look at GNOME 2. How do you stay on GNOME 2 when GNOME 3 goes stable? Well, it's time to go off masking dozens of packages. Maybe you miss one that seemed innocuous enough, but that eventually ends up pulling in stuff that you don't want in a later revision at some random time. How do you pre-blacklist everything that may potentially go bad without being so aggressive as to totally stagnate the system? There's currently a thought on gentoo-dev that maybe we should forgo networking in stage 3 because the default option isn't favored by those that want to eliminate choice in Gentoo. I mean, who needs functioning networking to, say, install networking? Staring at your system long enough will magically download the packages you need to download packages.

The best approach, IMO, is to fork and independently review ebuilds to ensure sanity before accepting them. It's really not all that different from kernel development. Linus' lieutenants have a lot of patches in their trees that will never make it into Linus' kernel and those that do have made it there through a combination of review by those lieutenants and frequently Linus himself. In my idea, the sane fork would take the place of Linus in that scenario, with Gentoo being one of the primary testing grounds. Linus' method certainly seems easier than trying to create blacklists of every potential kernel patch out there in existence, since that blacklist will never be complete.

Gentoo has turned into an asylum with the patients running it. That's not to say it's all bad, but it's no longer possible to consider it or any direction it takes as sane. Not for nothing, but one of the Council members that voted for mandating initramfs for a separate /usr doesn't even understand the consequences or complexities of what he voted for, or why a source distro and a binary distro have different needs. Everyone else is jumping off a bridge, so we should just blindly do so too, screw the technical merits or doing a little research into whether or not it's a good idea.
_________________
Ryzen 3700X, Asus Prime X570-Pro, 64 GB DDR4 3200, GeForce GTX 1660 Super
openrc-0.17, ~vanilla-sources, ~nvidia-drivers, ~gcc
Back to top
View user's profile Send private message
olek
Apprentice
Apprentice


Joined: 22 Oct 2011
Posts: 173

PostPosted: Wed Dec 04, 2013 11:13 pm    Post subject: Reply with quote

@NeddySeagoon

Thanks. I think it was reading your posts among a few other's (I'm not regularly enough around here to remember most usernames) that made me realize what's wrong with *kit etc.
Thanks for that also.

To be not completely off-topic:
My example shows me that some prominent way to show legitimate alternatives to *kit etc. would be nice.
While watching for topics about problems which derive from the design of *kit etc. and suggesting sane alternatives is nice, there is also a lack of easy looking up how to get a simplified, stable, not bloated system by best practice.
_________________
https://plaintext.blog
Back to top
View user's profile Send private message
Hypnos
Advocate
Advocate


Joined: 18 Jul 2002
Posts: 2889
Location: Omnipresent

PostPosted: Thu Dec 05, 2013 7:25 am    Post subject: Reply with quote

Pardon my ignorance, but what does any of this have to do with Gentoo? Gentoo is just a tool for (un)installing software and setting build-time options.

Olek seems to have the right idea: if you don't want crap dependencies, don't install crap software.

Right now, I use XFCE w/ gtk2 only, eudev and all the *kit stuff disabled. If that doesn't work for XFCE in the future I'll move onto something else -- there's always fvwm .
_________________
Personal overlay | Simple backup scheme
Back to top
View user's profile Send private message
saellaven
l33t
l33t


Joined: 23 Jul 2006
Posts: 646

PostPosted: Thu Dec 05, 2013 2:38 pm    Post subject: Reply with quote

Hypnos wrote:
Pardon my ignorance, but what does any of this have to do with Gentoo? Gentoo is just a tool for (un)installing software and setting build-time options.

Olek seems to have the right idea: if you don't want crap dependencies, don't install crap software.

Right now, I use XFCE w/ gtk2 only, eudev and all the *kit stuff disabled. If that doesn't work for XFCE in the future I'll move onto something else -- there's always fvwm .


portage is the tool, Gentoo is a distribution that relies heavily on portage, but also pulls in a lot of other things to create a base system... and that process of pulling in other things is heading in a direction that will mandate that you use certain software if you want to use Gentoo. You can mask that stuff and create overlays that keep the stuff you currently use alive, but at some point, what you essentially have is a fork that becomes increasingly difficult to maintain as Gentoo diverges farther from the point you forked at.

GNOME 2 will likely be exiting the tree sometime over the next year. You can keep an overlay or local tree to store it, but there will be a constant need to tweak things because GNOME 3 packages kept all the same library names despite not necessarily being compatible with GNOME 2. And with GNOME 2 out of the tree, non-GNOME packages will likely stop offering GNOME 2 options in favor of defaulting to GNOME 3, meaning all of those ebuilds need to constantly be tweaked. Eventually, you end up putting all of your effort into just keeping your overlay and masks working to keep the stuff you don't want from forcing your hand.

Now multiple that through every package that could potentially pull in the monstrosity that is systemd (remember the fun when xfree86 pulled in HAL before HAL became abandoned a release later? and HAL didn't try to be the one demon to rule them all with upstream trying to force dependence into every package to spread adoption).
_________________
Ryzen 3700X, Asus Prime X570-Pro, 64 GB DDR4 3200, GeForce GTX 1660 Super
openrc-0.17, ~vanilla-sources, ~nvidia-drivers, ~gcc
Back to top
View user's profile Send private message
Hypnos
Advocate
Advocate


Joined: 18 Jul 2002
Posts: 2889
Location: Omnipresent

PostPosted: Thu Dec 05, 2013 3:45 pm    Post subject: Reply with quote

saellaven wrote:
portage is the tool, Gentoo is a distribution that relies heavily on portage, but also pulls in a lot of other things to create a base system... and that process of pulling in other things is heading in a direction that will mandate that you use certain software if you want to use Gentoo.

Gentoo is not a distribution, it's a meta-distribution. Gentoo will only pull in what upstream demands as dependencies.

The problem is upstream with certain pieces of software, such as GNOME. So long as Gentoo devs support other, non-problematic pieces of software, there is no problem with Gentoo.

I suppose the concern is that the entire upstream for the Linux base system will get contaminated, in which case Gentoo will have no choice but to pull in crap. Unfortunately, the Gentoo devs have no control over that. In that case, move to BSD; Gentoo might even support that in this eventuality.
_________________
Personal overlay | Simple backup scheme
Back to top
View user's profile Send private message
saellaven
l33t
l33t


Joined: 23 Jul 2006
Posts: 646

PostPosted: Thu Dec 05, 2013 4:38 pm    Post subject: Reply with quote

Hypnos wrote:

The problem is upstream with certain pieces of software, such as GNOME. So long as Gentoo devs support other, non-problematic pieces of software, there is no problem with Gentoo.


Exactly my point... Gentoo is more than willing to bend for upstream fads than it is to incorporate existing things that already work. There's no reason that SteveL's /usr patch to openrc rather than mandating that people use an initramfs because systemd (a competitor to openrc) no longer wants to support a separate /usr. The lead of openrc (whom is also on the systemd herd) knew about the patch, requested the elimination of a separate /usr by the Council anyway, and didn't even mention the patch to his fellow Council members. The Council themselves seemed to have done little actual investigation into things, assumed that initramfs was somehow more robust than a few line patch even though, in reality, it's more fragile, and just rammed it through anyway because it's a requirement of systemd - everything else be damned.

If Gentoo wants to support GNOME, it needs to support systemd, but that doesn't mean that everyone should be forced to adhere to systemd's requirements if they DON'T want GNOME. It's not a matter of systemd's competitors all moving their upstream to the same requirements as systemd, as Gentoo IS the openrc upstream, but openrc development and robustness IS intentionally being held back in favor of GNOME. So, today it's GNOME and systemd, tomorrow, maybe it's KDE, maybe it's XFCE, maybe systemd cripples a separate udev, etc.

The entire purpose of a Gentoo is to have choice... but the devs and Council are intent on eliminating that choice to not only follow upstream, but to hold other projects back. As the choice evaporates because projects are crippled, what, exactly, is the point of having a meta distribution with use flags?

Quote:

I suppose the concern is that the entire upstream for the Linux base system will get contaminated, in which case Gentoo will have no choice but to pull in crap. Unfortunately, the Gentoo devs have no control over that. In that case, move to BSD; Gentoo might even support that in this eventuality.


Not the entire upstream... You have projects like MATE that have already forked away from the GNOME3 silliness. But the Gentoo devs certainly seem to want to hop on bandwagons and ride fads out, so the question is, how much damage are the devs/Council willing to do to the long term principles of Gentoo in the interest of simply becoming a pointless clone of RedHat (after all, RH offers .srpms so you can build it from scratch too). That's not to construe that all devs are bad, but there are an increasing number I do not trust and I have little faith in the technical competence, thoroughness and honesty of the Council to guide the project.

Anyway, to sum up, when your system gets compromised, you rollback to a verified state and go forward from there. You don't simply patch holes since, once compromised, your system will forever be in an untrustworthy state. Masks simply hide the known damage that will continue to be done to Gentoo, it won't prevent new corruption of existing packages from creeping in.
_________________
Ryzen 3700X, Asus Prime X570-Pro, 64 GB DDR4 3200, GeForce GTX 1660 Super
openrc-0.17, ~vanilla-sources, ~nvidia-drivers, ~gcc
Back to top
View user's profile Send private message
Hypnos
Advocate
Advocate


Joined: 18 Jul 2002
Posts: 2889
Location: Omnipresent

PostPosted: Fri Dec 06, 2013 2:05 am    Post subject: Reply with quote

saellaven,

Thanks for the summary -- I was not aware of all this intrigue. If there's one place where Gentoo and "upstream" converge, it's in OpenRC and init scripts.

Perhaps we need a fork/replacement of OpenRC? Just as Paludis and pkgcore gained traction as compatible alternatives to Portage, an OpenRC alternative might as well ...
_________________
Personal overlay | Simple backup scheme
Back to top
View user's profile Send private message
TomWij
Retired Dev
Retired Dev


Joined: 04 Jul 2012
Posts: 1553

PostPosted: Fri Dec 06, 2013 5:14 pm    Post subject: Re: please new forum _Advanced_Gentoo_without_KIT_ Reply with quote

ulenrich wrote:
A new forum
Traditionalistas - without modern bloat


Can you reconsider the words?


  • "Traditionalistas": There are multiple traditions, this word reflects something different than what you intend to reflect.
  • "modern": This imposes a limit in time, which can be avoided.
  • "bloat": One man's bloat is another man's toolset; avoid negatively loaded words, you can use a word like "minimal". Naming the forum in a way of what wants to see as opposed to what one doesn't want to see.


ulenrich wrote:
should discuss the recipe how to go on without using such bloat as

systemd dbus udev upower consolekit logind cgroups pulseaudio
etc ...


Why do we need a separate forum for this? If I install Gentoo, I get exactly what you describe. Gentoo is about choosing what you install. What would a separate forum for this contribute to that choice?

ulenrich wrote:
such that a valide profile can be developed. This project should
have a cool name: Zombie Gentoo


That profile would be default/linux/amd64/13.0 which we already have.

ulenrich wrote:
[Edit] changed title in favour of some calm to
please new forum _Advanced_Gentoo_without_KIT_


Same story as above:


  • "Advanced": What about people new to Gentoo? What is covered by "advanced"? What is an "expert" or is there no higher level? Where do people go with "basic" questions?
  • "without KIT": Said otherwise: minimal.


A possible title could be "Minimal Gentoo" or something along those lines, if you plan to go through with this; but I really think a separate forum is a bad idea without defining its purpose first.

wuzzerd wrote:
For openers we could post our use flags. This needs some improvement of course.


Without a definition and purpose you will get people posting random sets of USE fags; whereas these are meaningful to the person itself, the set will be different for another person. USE flag configurations are something personal; it is hard to define a stop condition for improvement here, at best you can attempt to find a common denominator which is what default/linux/amd64/13.0 already is.

ulenrich wrote:
But this results to nowhere for nearly everyone not a specialist but trying. I think of an extra forum having multiple sticky threads each of which showing one valid way.


There is no single valid way to install Gentoo, it's what Gentoo avoids; you choose the way you install, we avoid spoon feeding choices.

ulenrich wrote:
On the contrary: The herd of this new forum can count on me not writing anything there :(
I am titled the herd of sycophants, because I didn't recognize Traditionalistas don't want a single one voice with arguments pro modern bloat ...


It does not seem right to use this as a basis for creating a new forum; it rather serves as a recognition of a problem, a problem that isn't supposed to be there in the first place.

ulenrich wrote:
Seriously there are many problems Gnome-3.8 related popping up.


Now this is what might be a problem; but it response shouldn't be that it overtakes the entire forums, instead the source of the problem should be dealt with.

Why are there so many GNOME 3.8 problems? Because we are currently forcing people down a painful migration. Whereas people could install a systemd stage3 early on instead or migrate to systemd early in the installation instructions; I expect this will eventually happen especially with GNOME 3.8 becoming stable and GNOME 3.10 being around the corner.

ulenrich wrote:
These guys gonna have a hard way to sort out the helping threads as of now.


How does creating a new forum make this easier?

ulenrich wrote:
On the other hand the users, who want to avoid modern bloat, don't find a consistent way around, because it is all distributed in variously mixed threads.


Is this a real problem? They can just opt to not install it.

ulenrich wrote:
That is one way only, extremly restricted. But there could be other ways:
- using eudev
- using auto module loading
etc

And avoidance of ebuilds and needed USE is not discussed.


The Gentoo Wiki (can) cover(s) this.

ulenrich wrote:
... I stop telling here. If this new forum is needed, someone who needs will talk.


+1 Sharing thoughts is very welcome, they can lead to concepts and ideas, eventually leading to implementation or starting over at the scratch board.

NeddySeagoon wrote:
Yes there are other methods to keep some of the modern bells and whistles and yes, you do need to edit a few ebuilds to get xorg to build.


This can be solved under the form of an overlay; if you can convince someone with commit access to the Portage tree, you might be able to land a fork in the Portage tree as well.

Anon-E-moose wrote:
I'm not sure why those who eschew things they don't want on their systems
should need a separate forum designated "Traditionalistas - without modern bloat" or whatever.

Are they not supposed to post on regular forums? Are they diseased? Unclean? Unfit to mingle with other users?

*scratches head in bafflement*


+1

NeddySeagoon wrote:
I suppose we could split off the 'bloatware' into its own forum, leaving the traditionalists to the main forms to enjoy the way its always been.


Interesting to see both ways discussed, but it's going to lead to more trouble than it is trying to fix; instructing mainstream to use a separate form sounds like a rather painful task to accomplish.

NeddySeagoon wrote:
New, improved etc. does not always mean better.


Neither does it always means worse.

NeddySeagoon wrote:
There are lots of examples where better marketing has won out against technical superiority. VHS vs Betamax for example.


Though Gentoo Linux is an opposite example; it is technical superiority, and that's what attracts the right kind of people.

Given that the rest of the forum thread either agrees or is loaded with FUD, personal opinion or team calling that isn't constructive, this is where I end my first reply...
Back to top
View user's profile Send private message
TomWij
Retired Dev
Retired Dev


Joined: 04 Jul 2012
Posts: 1553

PostPosted: Fri Dec 06, 2013 5:51 pm    Post subject: Reply with quote

Yet I want to clarify one particular train of thoughts in this thread:

saellaven wrote:
Look at GNOME 2. How do you stay on GNOME 2 when GNOME 3 goes stable?


Why would you hold on to a piece of software that has became EOL and thus unsupported over a year ago?

saellaven wrote:
Gentoo is more than willing to bend for upstream fads than it is to incorporate existing things that already work.


How do you measure this?

saellaven wrote:
The Council themselves seemed to have done little actual investigation into things


As with any organization, they investigate what was provided to them; they cannot investigate what they do not know about, thus if you want to see this then I suggest you to put that patch on their agenda now (or bring it up during the open floor on IRC) such that it can be discussed next Tuesday. Or at a later meeting, they happen once a month at least; sometimes weekly if the agenda is quite loaded. Would you try this?

saellaven wrote:
but that doesn't mean that everyone should be forced to adhere to systemd's requirements if they DON'T want GNOME.


As said in the other thread(s), there is no indication of this except for initramfs; but as said in those other thread(s), systemd is not the sole reason to that. Where do you see additional force?

saellaven wrote:
But openrc development and robustness IS intentionally being held back in favor of GNOME.


What is the indication of this? Why would they?

saellaven wrote:
But the devs and Council are intent on eliminating that choice to not only follow upstream, but to hold other projects back.


Why would they do that? Where does that happen?

saellaven wrote:
As the choice evaporates because projects are crippled


Which projects? Where? Why?

saellaven wrote:
But the Gentoo devs certainly seem to want to hop on bandwagons and ride fads out, so the question is, how much damage are the devs/Council willing to do to the long term principles of Gentoo in the interest of simply becoming a pointless clone of RedHat (after all, RH offers .srpms so you can build it from scratch too).


Which damage? Why are we questioning something that has yet to happen? Where do you see this indicated in the philosophy, agenda and news of Gentoo?

saellaven wrote:
Masks simply hide the known damage that will continue to be done to Gentoo, it won't prevent new corruption of existing packages from creeping in.


Can you please file bugs instead and make the Council aware of these matters?
Back to top
View user's profile Send private message
saellaven
l33t
l33t


Joined: 23 Jul 2006
Posts: 646

PostPosted: Fri Dec 06, 2013 6:12 pm    Post subject: Reply with quote

To TomWij:

I've already participated in previous "discussions" with you, in which you trolled me (and others) and played stupid word definition games just to further frustrate us, only to eventually flat out admit that you didn't care about what we had to say - it was solely your ego that got you involved in the discussion by thinking you were some type of bridge between the rest of the dev community and us, despite using that opportunity to ultimately harden feelings between the sides. You offered and then refused to be a liason between myself and williamh, in particular, whom is the person central to the entire /usr and openrc mess that HE INTENTIONALLY caused through unwillingness to accept patches to a project he leads, his agenda as a Council member to have the Council vote to eliminate /usr despite personaly knowing about the existence of the patches, etc.

I WILL NOT respond to you since you have repeatedly shown that you do NOT act in good faith... and just to be open, you are one of the devs I absolutely do not trust precisely because of your duplicitous actions. I'm back to installing official kernels from kernel.org directly, just to bypass you, because you've lost my trust... and that's despite the fact that just last year, I was considering joining the kernel team myself, so congrats on your stellar community relations work.

As far as filing bugs, what good is it going to do? The Council's decisions are final and they've shown that they do not care about the technical merits and did little actual research before ruling.
_________________
Ryzen 3700X, Asus Prime X570-Pro, 64 GB DDR4 3200, GeForce GTX 1660 Super
openrc-0.17, ~vanilla-sources, ~nvidia-drivers, ~gcc
Back to top
View user's profile Send private message
TomWij
Retired Dev
Retired Dev


Joined: 04 Jul 2012
Posts: 1553

PostPosted: Fri Dec 06, 2013 7:13 pm    Post subject: Reply with quote

saellaven wrote:
I've already participated in previous "discussions" with you, in which you trolled me (and others) and played stupid word definition games just to further frustrate us, only to eventually flat out admit that you didn't care about what we had to say - it was solely your ego that got you involved in the discussion by thinking you were some type of bridge between the rest of the dev community and us, despite using that opportunity to ultimately harden feelings between the sides.


Can you talk about the subject instead of about the person? Please stay on the topic of this thread. It should be clear that I am not on this forum for word or blame games. Why would I participate if I don't care? Why are you focused on feelings instead of on the subject? What about that subject and not about me really bothers you? Can you answer my questions about the subject? As that is where my interest lies, I am actively listening to you and I do care. Why do you again give me a side talk that is out of my interest instead?

saellaven wrote:
You offered and then refused to be a liason between myself and williamh


Because I want you to do so, because that side talk mismatches my interest; if you want to reach something in particular, you can take that extra step. Who are we trying to convince? Where can that be done? Who should do so?

saellaven wrote:
in particular, whom is the person central to the entire /usr and openrc mess that HE INTENTIONALLY caused through unwillingness to accept patches to a project he leads


That's at his discretion, there are other init systems available or they can be forked. Which init system do you prefer instead? Or what would you see as an ideal init system?

saellaven wrote:
his agenda as a Council member to have the Council vote to eliminate /usr despite personaly knowing about the existence of the patches, etc.


There are more council members an the council changes every year; so, things might change if you want them to. Can you make them aware of your concerns such that they know about it?

saellaven wrote:
I WILL NOT respond to you since you have repeatedly shown that you do NOT act in good faith... and just to be open, you are one of the devs I absolutely do not trust precisely because of your duplicitous actions.


What are you talking about? Why do you get personal? You appear frustrated, can you clarify why?

saellaven wrote:
I'm back to installing official kernels from kernel.org directly, just to bypass you, because you've lost my trust... and that's despite the fact that just last year, I was considering joining the kernel team myself, so congrats on your stellar community relations work.


Can you clarify what you lost trust in? Which work are you talking about?

saellaven wrote:
As far as filing bugs, what good is it going to do?


Can you try it and tell us?

saellaven wrote:
The Council's decisions are final


In hierarchy, yes; but the Council can override them at any time. Can you ask them to if you want that?

saellaven wrote:
And they've shown that they do not care about the technical merits


Where did they show that?

saellaven wrote:
and did little actual research before ruling.


Organisations work differently, you have to tell them. So, could you please do tell them? Why are you telling this to us instead of them?

I'm looking forward to your answers to my questions, thank you in advance.
Back to top
View user's profile Send private message
saellaven
l33t
l33t


Joined: 23 Jul 2006
Posts: 646

PostPosted: Fri Dec 06, 2013 7:20 pm    Post subject: Reply with quote

TomWij wrote:

I'm looking forward to your answers to my questions, thank you in advance.


Tom, I'm going to say this one last time.

I will not read your posts and I will not reply to your posts.

YOUR intentional past actions have eliminated any good faith I can attempt to offer you.

You want to talk personally, either resign as a dev or lets go to ComRel. Outside of that, please, never speak to me again.
_________________
Ryzen 3700X, Asus Prime X570-Pro, 64 GB DDR4 3200, GeForce GTX 1660 Super
openrc-0.17, ~vanilla-sources, ~nvidia-drivers, ~gcc
Back to top
View user's profile Send private message
TomWij
Retired Dev
Retired Dev


Joined: 04 Jul 2012
Posts: 1553

PostPosted: Fri Dec 06, 2013 7:25 pm    Post subject: Reply with quote

saellaven wrote:
I will not read your posts and I will not reply to your posts.


Can you then please stop replying if you show no interest in my questions and/or this discussion?

saellaven wrote:
YOUR intentional past actions have eliminated any good faith I can attempt to offer you.


Which actions are you talking about? Can you please assume good faith? Why do you perceive them as intentional?

saellaven wrote:
You want to talk personally


I said multiple times that I want to talk and am talking about the subject instead. Why are you talking personal instead of discussing the central point of the subject (and its arguments)?

saellaven wrote:
Either resign as a dev or lets go to ComRel.


ComRel requires you to talk to me personally first. Can you please decide whether you agree with me, want to agree to disagree or are uninterested? Can you be clear and answer my questions that want to clarify what you mean and are referring to? ComRel also requires you to try the forum moderators first if you couldn't talk it out with me; but given that we're still talking, we can solve this ourselves? Can you thus please get back on the topic of the subject of this thread?

saellaven wrote:
Outside of that, please, never speak to me again.


This is a public forum with freedom of speech, as that is needed for constructive discussions. Constructive discussions is what I am here for. How can I grant your request?


Last edited by TomWij on Fri Dec 06, 2013 7:38 pm; edited 1 time in total
Back to top
View user's profile Send private message
Display posts from previous:   
Reply to topic    Gentoo Forums Forum Index Gentoo Forums Feedback All times are GMT
Goto page Previous  1, 2, 3, 4, 5  Next
Page 2 of 5

 
Jump to:  
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum