| View previous topic :: View next topic |
| Should X 7.1 still be testing only because of closed source drivers (ati/nvidia)? |
| Yes! |
|
55% |
[ 254 ] |
| No! |
|
37% |
[ 172 ] |
| I don't care! |
|
6% |
[ 28 ] |
|
| Total Votes : 454 |
|
| Author |
Message |
NeddySeagoon
Administrator
Joined: 05 Jul 2003
Posts: 55190
Location: 56N 3W
|
 Posted: Sun Jul 30, 2006 11:43 am Post subject: |
 |
|
boroshan,
Its a design issue rather than a QA problem. The Xorg project made some design decisions that resulted in changes too the ABI.
The changes were well know well in advance of the release of Xorg 7.1, so its not like purveyors of binary drivers were taken by surprise when Xorg 7.1 came out.
Masking Xorg 7.1 because binary drivers have not kept pace with the Xorg ABI change is a bit like the tail wagging the dog.
The FOSS should go where its design decisions take that it. It should be up to users to choose and use binary drivers (or not).
However, this masking does reduce the support load on the forums and in IRC - thats the only thing in its favour.
Oh and I run a full ~x86 system with Xorg 7.1 in my package mask so I can still use the nvidia binary blob.
Releasing binary drivers to preserve IPR is a design decision too. There are ways of preserving IPR and releasing GPL drivers. There are several precedents in the kernel today.
_________________
Regards,
NeddySeagoon
Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail. |
|
| Back to top |
|
 |
boroshan
l33t
Joined: 16 Apr 2003
Posts: 730
Location: upside down
|
| Posted: Sun Jul 30, 2006 12:27 pm Post subject: |
|
|
| NeddySeagoon wrote: |
Its a design issue rather than a QA problem. The Xorg project made some design decisions that resulted in changes too the ABI.
The changes were well know well in advance of the release of Xorg 7.1, so its not like purveyors of binary drivers were taken by surprise when Xorg 7.1 came out.
|
Understood.
| NeddySeagoon wrote: |
Masking Xorg 7.1 because binary drivers have not kept pace with the Xorg ABI change is a bit like the tail wagging the dog.
The FOSS should go where its design decisions take that it. It should be up to users to choose and use binary drivers (or not).
However, this masking does reduce the support load on the forums and in IRC - thats the only thing in its favour.
|
Still no argumnt from me. Neddy, I think we're at cross-purposes here. I'm not arguing that 7.1 should have remained masked.
As I understand it, (correct me if I'm wrong) 7.1 was supposed to be released with blocks to stop portage itstalling it alongside incompatible drivers. The idea being that anyone running unstable but not closed drivers would get the benefit, while the rest of us would have to wait on NVidia and ATI got their act together. Which is fair enough.
The trouble is, it seems, that those blocks (rather than X.Org or ati-drivers) didn't work properly, and in consequence 7.1 installed itself alongside nvidia-kernel and ati-drivers, and a lot of people got unexpetedly non-funcitonal desktops. This raises the question of whether these blocks were properly tested, and if not, why not. And that's why I see this as a QA issue
| NeddySeagoon wrote: |
Releasing binary drivers to preserve IPR is a design decision too. There are ways of preserving IPR and releasing GPL drivers. There are several precedents in the kernel today. |
Personally I think Nvidia and ATI's driver stategey is short sighted and counter productuve, but I have to conceed that this is their prerogative.
I'm not knocking free software
I'm not praising closed source drivers
I just want to know how this slipped past test procedures when it was known in advance that 7.1 had a huge potential for disruption.
_________________
Don't let THEM immanentize the Eschaton! |
|
| Back to top |
|
|
legine
Guru
Joined: 27 May 2004
Posts: 555
Location: Germany
|
| Posted: Mon Jul 31, 2006 8:20 am Post subject: |
|
|
Hmm I dont see any big issues here. Anyone useing ~Arch has been told that any package can break without warning. So Id might to add that this is usertesting case.
So I fully agree with X7.1 moving to unstable.
You can argue on the timing. But the fact remains. Which difference does a week make? You would run into the same problems then you had now.
With X7.1 in unstable Gentoo had made the point that it should get stable. With keeping them masked we would sending them the message that we do not care. So the point is when will the closed source drivers apear in our portage?
I think NVIDIA will releasing drivers too soon.
Maybe they have problems with it?
Dono. The drivers are closed source, and as for closed source drivers I expect tested products...
_________________
quote from Spaceballs:
Dark Helmet:[...] we were told to comb the desert, so we're combing it! [puts down bullhorn] Find anything yet?!
Soldier: Nothing yet, sir. |
|
| Back to top |
|
|
VinzC
Watchman
Joined: 17 Apr 2004
Posts: 5098
Location: Dark side of the mood
|
| Posted: Mon Jul 31, 2006 8:49 am Post subject: |
|
|
After all if XOrg 7.1 remains in the testing branch till binary drivers are compatible with the ABI it'll be a wise decision - humanly. Because it'll reduce the support load on forums, as NeddySeagoon said.
If it is marked stable it'll be also a wise decision - philosophically. Because it is stable though binary drivers are not compatible.
The only thing is if the second decision is made, there should be enough advices, HOWTOs and notifications to prevent from loading the forums with such support requests.
_________________
Gentoo addict: tomorrow I quit, I promise!... Just one more emerge...
1739! |
|
| Back to top |
|
|
slycordinator
Advocate
Joined: 31 Jan 2004
Posts: 3065
Location: Korea
|
| Posted: Mon Jul 31, 2006 8:54 am Post subject: |
|
|
| VinzC wrote: | | The only thing is if the second decision is made, there should be enough advices, HOWTOs and notifications to prevent from loading the forums with such support requests. |
The number of support questions asked that could be answered through using the forum search says otherwise.
Hell, quite a few people make posts that on a topic that's stickied. |
|
| Back to top |
|
|
legine
Guru
Joined: 27 May 2004
Posts: 555
Location: Germany
|
| Posted: Mon Jul 31, 2006 9:19 am Post subject: |
|
|
| Quote: | | Hell, quite a few people make posts that on a topic that's stickied. |
Maybe that comes from the fact that people start with useing the search. If there is no result, they start asking?
_________________
quote from Spaceballs:
Dark Helmet:[...] we were told to comb the desert, so we're combing it! [puts down bullhorn] Find anything yet?!
Soldier: Nothing yet, sir. |
|
| Back to top |
|
|
boroshan
l33t
Joined: 16 Apr 2003
Posts: 730
Location: upside down
|
| Posted: Mon Jul 31, 2006 10:00 am Post subject: |
|
|
| legine wrote: | | You can argue on the timing. But the fact remains. Which difference does a week make? You would run into the same problems then you had now. |
One of the questions that I'm asking is "what level of testing may we, as users, legitimately expect from ~arch?"
There are a lot of things I can't do on a pure x86 system. Develop in mono is one of them, and that's how I make my living. I've always regarded ~arch integration being release candidate quality. But a debacle on this scale would be cause for concern in an alpha release. So it's important for me to know what level of testing I can expect prior to ~x86 release. Because if unstable is to become significantly more unstable then I need to migrate to a new distro.
Nor am I the only one who thinks these issues are important.
| legine wrote: | | The drivers are closed source, and as for closed source drivers I expect tested products... |
Mmm... I don't think the problem lies either with the drivers or with X.Org. The problem as I see it lies in the way they were represented to portage which was supposed to stop the two being installed together.
And that's the difference I'd expect from another week of testing. A bit more testing would have shown that the blocks didn't actually work and given the developers a chance to fix the problem. If that had happened, there'd be at two less great long threads on these forums, and a lot fewer disgruntled gentooites.
_________________
Don't let THEM immanentize the Eschaton! |
|
| Back to top |
|
|
legine
Guru
Joined: 27 May 2004
Posts: 555
Location: Germany
|
| Posted: Mon Jul 31, 2006 10:32 am Post subject: |
|
|
| Quote: | | One of the questions that I'm asking is "what level of testing may we, as users, legitimately expect from ~arch?" |
Hmm, If you ask me. If the Ebuild works, it should go in testing. Thats why we call it unstable.
We have only developer version or Products that have serious Problems in mask section, if I understand the masked section right. ~arch is unstable. There you find everything that could or could not work because it breaks your system.
You should work with stable branch if you actually want to work with it.
I think thats the ideal position of our setup as I remeber it.
| Quote: |
Mmm... I don't think the problem lies either with the drivers or with X.Org. The problem as I see it lies in the way they were represented to portage which was supposed to stop the two being installed together. |
But thats the only correct technical solution. So you trying to hunt fantoms?
I am not realy seeing an issue.
I am ofcourse usually behind the news since I never run and never will run a ~arch system. I am actually trying to run as view ~arch system as I can. Just not to run in issues you are bugging around.
At least thats my opinion.
_________________
quote from Spaceballs:
Dark Helmet:[...] we were told to comb the desert, so we're combing it! [puts down bullhorn] Find anything yet?!
Soldier: Nothing yet, sir. |
|
| Back to top |
|
|
mrsteven
Veteran
Joined: 04 Jul 2003
Posts: 1939
|
| Posted: Mon Jul 31, 2006 10:44 am Post subject: |
|
|
| boroshan wrote: | | There are a lot of things I can't do on a pure x86 system. Develop in mono is one of them, and that's how I make my living. I've always regarded ~arch integration being release candidate quality. But a debacle on this scale would be cause for concern in an alpha release. So it's important for me to know what level of testing I can expect prior to ~x86 release. Because if unstable is to become significantly more unstable then I need to migrate to a new distro. |
Just unmask the packages needed for mono development in /etc/portage/package.keywords. No one forces you to use a pure ~arch system.
As far as I know, a new upstream release goes into ~arch, unless there is a good reason to hardmask it or to mark it stable immediately. Some projects test their new releases very well, others are more worried about releasing new software quickly... So stability of ~arch varies from package to package...
I remember a similar problem with the kernel module from the ati-drivers and some kernel that was just released but broke that module. The kernel went into ~arch and that was just fine, because if you use ~arch things might break if you don't know what you're doing. That's the price to pay if you want to use bleeding edge software. 
This is not a problem about software quality, it is more about how to handle ABI changes.
If you really want to know, I still use X.org 7.0 because 7.1 has other problems. |
|
| Back to top |
|
|
boroshan
l33t
Joined: 16 Apr 2003
Posts: 730
Location: upside down
|
| Posted: Mon Jul 31, 2006 10:55 am Post subject: |
|
|
| legine wrote: | | Hmm, If you ask me. If the Ebuild works, it should go in testing. |
You mean as long as the ebuild is sufficient to compile the software, in it goes? If that's what the developers want, well it's their call. As far as I can see, that would represent a change from established practice, and I'd question the wisdom of such a change, but if that's to be the way it works then fair enough.
I'm still asking the question though.
| Quote: | | Thats why we call it unstable. |
Yes, of course. But instabillity isn't a binary value. Too unstable and ~x86 will never work for anyone because all the packages conflict. Too stable and there's no point to a stable release.
I want to know where the line is being drawn. I want to know how unstable an unstable I should expect.
I think this is a reasonable expectation.
| Quote: |
But thats the only correct technical solution. So you trying to hunt fantoms?
|
I don't think I follow you here. I don't have a problem with approach chosen. I have a problem with the implementation of the solution, in that it didn't actually work. And therefore I want to know how the mistake slipped past testing.
| Quote: | | I am ofcourse usually behind the news since I never run and never will run a ~arch system. I am actually trying to run as view ~arch system as I can. Just not to run in issues you are bugging around. |
Sensible enough. I was running ~x86 because I once needed NPTL support and ~x86 was the only way to get it. In the three years since then, I've not until now had cause to question that decision.
Because of the 7.1 episode I wound up reinstalling from scratch, and when I did that I did so with as stable a system as I could. But my livelihood depends on unstable packages, and in particular upon another politically charged entity - Mono.
So it's not as simple as "just run x86" for me, and it is very important. Because, if nothing else, another session like that one, and I'm going to get told to take Linux off my laptop and use Windows like a normal person.
_________________
Don't let THEM immanentize the Eschaton! |
|
| Back to top |
|
|
boroshan
l33t
Joined: 16 Apr 2003
Posts: 730
Location: upside down
|
| Posted: Mon Jul 31, 2006 11:18 am Post subject: |
|
|
| mrsteven wrote: | Just unmask the packages needed for mono development in /etc/portage/package.keywords. No one forces you to use a pure ~arch system.
|
And as noted elsewhere, this is in fact what I am doing.
| Quote: | | That's the price to pay if you want to use bleeding edge software. |
Sure. I accept that unreservedly. What I find disturbing is that this seems to have been politicised as a free-vs-proprietory issue in the debate prior to 7.1's release. This (along with a certain amount of schadenfreude from some of the more Zealous forum members) raises concerns that political considerations may have influnced technica ones - which would certainly seem to explain how Gentoo's usually exemplary testing processes came to fail so dismally in this instance.
And if so, how long before someone decides to punish Microsoft for .NET and I end up with a non-functional Mono?
| Quote: | | This is not a problem about software quality, it is more about how to handle ABI changes. |
When I say QA - I mean at the integration level. I'm not compaining about 7.1 or nvidia-kernel or ati-drivers. I'm complaining about the glue that sticks the system together and which should have prevented 7.1 installing alongside closed driver sets. The problem here lay with portage and how the metadata for 7.1 and the driver packages was defined, and in the fact that that metadata seems to have received little or no testing.
And that is a software quality issue, at least to my way of thinking.
_________________
Don't let THEM immanentize the Eschaton! |
|
| Back to top |
|
|
legine
Guru
Joined: 27 May 2004
Posts: 555
Location: Germany
|
| Posted: Mon Jul 31, 2006 11:51 am Post subject: |
|
|
| Quote: | | And if so, how long before someone decides to punish Microsoft for .NET and I end up with a non-functional Mono? |
Calm down.
You would have a warning. The Zealots would dismantle wine first
Gentoo says We are about choice. So which choice would we offer if we dump .net?
Besides there are cool Open Source tools based on mono bases. (Check out Tomboy!)
So you see there won't be such a okishtosh thing. 
And the decision for X is not political in the way you put it. 
It is exactly because it has to be that way in order to make things ongoing and to give the community the hands of handling things.
Well May I ask why you desperently want to use the mono version 1.1.16.1 (unstable) instead of 1.1.13.6 (stable)?
I mean you could be fine with 1.1.13.6, or not?
I would guess most people work with older versions on other distros. Of course I am no expert. And dont know anything about monos versioning. 
Hope you come by soon. There is nothing realy to worry about. Gentoo is the  thing on earth and that will stay that way. 
Lets sing!
Don't worry, be Happy!
dadadada
Edit:
I have found Buuugs!!! and eliminated them! 
_________________
quote from Spaceballs:
Dark Helmet:[...] we were told to comb the desert, so we're combing it! [puts down bullhorn] Find anything yet?!
Soldier: Nothing yet, sir. |
|
| Back to top |
|
|
boroshan
l33t
Joined: 16 Apr 2003
Posts: 730
Location: upside down
|
| Posted: Mon Jul 31, 2006 1:03 pm Post subject: |
|
|
| legine wrote: | | Quote: | | And if so, how long before someone decides to punish Microsoft for .NET and I end up with a non-functional Mono? |
Calm down. |
I guess that came across a bit strident, huh? It isn't that I can't see the funny side - I just don't want to have to explain the joke to my boss.
| Quote: | | Gentoo says We are about choice. So which choice would we offer if we dump .net? |
Or dump closed source drivers. Only we didn't need to dump them to put a lot of people to a lot of avoidable trouble.
| Quote: | | And the decision for X is not political in the way you put it. |
I don't think the decision was. But I think the discussion beforehand undoubtedly was. The problem is that, quite possibly by co-incidence, there was a silly mistake in the release process that would have normally been caught by testing but which slipped through on this occasion, and which just happened to have the effect of punishing those running closed source graphics drivers.
I don't want to accuse anyone of anything in any way improper. But I think I'm justified in asking a few questions.
| Quote: | Well May I ask why you desperently want to use the mono version 1.1.16.1 (unstable) instead of 1.1.13.6 (stable)?
I mean you could be fine with 1.1.13.6, or not? |
Actually, it's Boo that only has an unstable ebuild. Something in Boo's dependency chain must have depended on unstable Mono, since I didn't include any unstable packages if I could avoid it. Or it could have been MonoDevelop that needed it - I can't remember exactly.
| Quote: | | Hope you come by soon. There is nothing realy to worry about. Gentoo is the thing on earth and that will stay that way. |
I agree. I think gentoo is a fabulous system. And I too want it to stay that way.
Besides, where's the harm in asking a few questions?
_________________
Don't let THEM immanentize the Eschaton! |
|
| Back to top |
|
|
batistuta
Veteran
Joined: 29 Jul 2005
Posts: 1384
Location: Aachen
|
| Posted: Mon Jul 31, 2006 1:49 pm Post subject: |
|
|
I'm so sick about this discussion that I don't care anymore.
But I just wanted to suggest you boroshan, that if your work depends on a stable system and on lots of ~x86 tools, that you might want to consider checking out freeBSD. Their "testing" branch" (notice testing rather than "unstable") is in my opinion something between unstable and stable in Gentoo, but it tends to brake much less frequently.
I'm not suggesting you to switch, since they are different things and like everything, is a matter of needs. But if Gentoo's unstable branch is really giving you such a hard time then consider it as an option. |
|
| Back to top |
|
|
boroshan
l33t
Joined: 16 Apr 2003
Posts: 730
Location: upside down
|
| Posted: Mon Jul 31, 2006 2:15 pm Post subject: |
|
|
| batistuta wrote: | | I'm so sick about this discussion that I don't care anymore. |
meh. maybe you have a point. This is certainly getting repetitive.
_________________
Don't let THEM immanentize the Eschaton! |
|
| Back to top |
|
|
legine
Guru
Joined: 27 May 2004
Posts: 555
Location: Germany
|
| Posted: Mon Jul 31, 2006 2:25 pm Post subject: |
|
|
hehe.
Well what matters I try to answer your question as good as I can. Wont request anyone stopping to ask.
But what I want to add is. That the X Problem is from my point of view a bit unique. so bear with our devs, and their decision. You had no Problem with your system in 3 years?
Well be carefull and you wont have Problem with it in the next 3.
I don't think that Gentoo is getting more unstable then it used to be, and it is surely not as unstable as most people think it is.
The work of Distribution is a bit overrated if you ask me. ( It is not that they do nothing but I think people dont differ between distribution and the Project itself. )
Maybe we end this at this point
_________________
quote from Spaceballs:
Dark Helmet:[...] we were told to comb the desert, so we're combing it! [puts down bullhorn] Find anything yet?!
Soldier: Nothing yet, sir. |
|
| Back to top |
|
|
slycordinator
Advocate
Joined: 31 Jan 2004
Posts: 3065
Location: Korea
|
| Posted: Mon Jul 31, 2006 6:14 pm Post subject: |
|
|
| legine wrote: | | Quote: | | Hell, quite a few people make posts that on a topic that's stickied. |
Maybe that comes from the fact that people start with useing the search. If there is no result, they start asking? |
You completely misunderstood me:
Lots of people make NEW THREADS about something ALREADY STICKIED.
edit:
He was saying that the number of howtos and notifications in the forums would be enough. The fact that people continually ask stuff that's already answered 100 times over proves otherwise.
And if the notifications were enough we wouldn't have had the huge problem a while back with gentoo changing the way apache is configured. Tons of users were left with useless apache servers. If notifications had been enough, that wouldn't have happened. |
|
| Back to top |
|
|
mrsteven
Veteran
Joined: 04 Jul 2003
Posts: 1939
|
| Posted: Mon Jul 31, 2006 6:49 pm Post subject: |
|
|
| boroshan wrote: | | mrsteven wrote: | Just unmask the packages needed for mono development in /etc/portage/package.keywords. No one forces you to use a pure ~arch system.
|
And as noted elsewhere, this is in fact what I am doing. |
Then I got you wrong. Sorry!
| Quote: | When I say QA - I mean at the integration level. I'm not compaining about 7.1 or nvidia-kernel or ati-drivers. I'm complaining about the glue that sticks the system together and which should have prevented 7.1 installing alongside closed driver sets. The problem here lay with portage and how the metadata for 7.1 and the driver packages was defined, and in the fact that that metadata seems to have received little or no testing.
And that is a software quality issue, at least to my way of thinking. |
Ok, I agree with you here, portage should not allow to install both packages together on the same system. This is a problem, yes.
I thought you were complaining about the ABI change itself. These changes have to happen from time to time in order to improve the X server (or the kernel). You also can't use a driver written for Windows 98 on Windows XP, even Microsoft changes things sometimes.
In this case it was actually a developer from Nvidia who suggested the ABI change, but as they have not released a driver for the new version, they are to blame for this situation. |
|
| Back to top |
|
|
Kensai
Guru
Joined: 09 Mar 2004
Posts: 570
Location: Puerto Rico
|
| Posted: Mon Jul 31, 2006 7:54 pm Post subject: |
|
|
| mrsteven wrote: | | In this case it was actually a developer from Nvidia who suggested the ABI change, but as they have not released a driver for the new version, they are to blame for this situation. |
They are woking on it, suggesting an ABI change doesn't mean that at the second the change was made the driver will be available too you know? They are on testing periods and I believe the results will show that is worth the waiting, resulting in a better tested and working integration of Nvidia and Xorg. So nobody is to blame as you said this changes has to happen.
_________________
Gentoo: Gigabyte: nFORCE 2: nVIDIA GeForce 6600: AMD Athlon XP 3200+
Leaving the above specs to immortalize the first system I Installed Gentoo on! |
|
| Back to top |
|
|
VinzC
Watchman
Joined: 17 Apr 2004
Posts: 5098
Location: Dark side of the mood
|
| Posted: Mon Jul 31, 2006 9:08 pm Post subject: |
|
|
| slycordinator wrote: | edit:
He was saying that the number of howtos and notifications in the forums would be enough. |
No. I was saying that to make XOrg stable there ought to be a certain number of warnings, notifications and HOWTOs explaining what to do. I was not saying that the HOWTOs and forums were already enough in their current state.
| VinzC wrote: | | The only thing is if the second decision is made, there should be enough advices, HOWTOs and notifications to prevent from loading the forums with such support requests. |
I used the term "should" in the meaning "there has to be" but not in the meaning "there is probably". You picked up the wrong meaning.
_________________
Gentoo addict: tomorrow I quit, I promise!... Just one more emerge...
1739! |
|
| Back to top |
|
|
slycordinator
Advocate
Joined: 31 Jan 2004
Posts: 3065
Location: Korea
|
| Posted: Mon Jul 31, 2006 9:49 pm Post subject: |
|
|
| VinzC wrote: | | slycordinator wrote: | edit:
He was saying that the number of howtos and notifications in the forums would be enough. |
No. I was saying that to make XOrg stable there ought to be a certain number of warnings, notifications and HOWTOs explaining what to do. I was not saying that the HOWTOs and forums were already enough in their current state. |
As mentioned, there were all sorts of warnings and notifications about the changes that happened with apache yet TONS of people had problems with it.
The devs made announcements in the forums, bug reports, newsletters, warnings in the ebuilds,...
All I'm trying to say is that putting warnings everywhere doesn't do a thing with a lot of people. It's like at my job at a movie theater: we'll put up 3 signs saying "[insert movie here] is sold out" and people walk up asking for that movie or during slow times they'd come in and say "Why is no one selling tickets out there?" ignoring the neon sign that says "Purchase tickets inside the lobby."
Notifications fail. You can't count on them. |
|
| Back to top |
|
|
VinzC
Watchman
Joined: 17 Apr 2004
Posts: 5098
Location: Dark side of the mood
|
| Posted: Mon Jul 31, 2006 10:13 pm Post subject: |
|
|
| slycordinator wrote: | As mentioned, there were all sorts of warnings and notifications about the changes that happened with apache yet TONS of people had problems with it.
[...]
Notifications fail. You can't count on them. |
I think the problem is, no offense, you misunderstood what I meant. My reasonning is if you did once with me you probably did with other people. Another fact is you seem to read only the words that are present in the posts. You based your whole point on a wrong idea that I didn't express.
For instance I didn't say there should be HOWTOs, notifications, forums and nothing else. No. I meant (but didn't say) it's up to those who make the decision to evaluate what information means best fit such a situation so as not to reproduce what happened with Xorg, Apache...
Maybe you have been implied too much (or too heartily) in this thread to read behind the lines and catch the real intents.
Back to your idea now. Of course there are people who ignore warnings, people who don't read notifications, people who forget almost immediately... Well, so many people who don't do what they're told, I know. My point is (but I didn't express it here) first care for those who carefully read the instructions. Care for the rest afterwards. If you care at first for those who don't read, who forget, who don't act as expected, well, you might risk a heart attack or you'll end up dumbing things down .
But never forget: you won't satisfy every single individual at once. I'd rather be criticized by people who read the instructions than by those who ignore warnings and notes and HOWTOs and...
_________________
Gentoo addict: tomorrow I quit, I promise!... Just one more emerge...
1739! |
|
| Back to top |
|
|
slycordinator
Advocate
Joined: 31 Jan 2004
Posts: 3065
Location: Korea
|
| Posted: Mon Jul 31, 2006 10:59 pm Post subject: |
|
|
| VinzC wrote: | I think the problem is, no offense, you misunderstood what I meant. My reasonning is if you did once with me you probably did with other people. Another fact is you seem to read only the words that are present in the posts. You based your whole point on a wrong idea that I didn't express.
For instance I didn't say there should be HOWTOs, notifications, forums and nothing else. |
Well, what else is there? I already said that they have tried howtos, forums, newsletters, ebuild warnings, and general notifications.
What else is there, exactly?
| Quote: | | No. I meant (but didn't say) it's up to those who make the decision to evaluate what information means best fit such a situation so as not to reproduce what happened with Xorg, Apache... |
What I'm saying is that no matter what they do to get the information out there, a good chunk of people will still miss the message.
Like I mentioned, with Apache, every method that the devs could think of were used. They put out the info in newsletters, forums, ebuild warnings, etc. And it still failed. About the only thing left they could've tried was sending out individual emails.
| Quote: | | Back to your idea now. Of course there are people who ignore warnings, people who don't read notifications, people who forget almost immediately... Well, so many people who don't do what they're told, I know. My point is (but I didn't express it here) first care for those who carefully read the instructions. Care for the rest afterwards. If you care at first for those who don't read, who forget, who don't act as expected, well, you might risk a heart attack or you'll end up dumbing things down . |
If the update process were more automated, most of this would be avoided to begin with.
Also, with the number of places that one must look to try to get a piece of information, many people likely just give up. Plus, many of those pieces of info aren't official, as some users decide to put up a piece about some new portage feature that's undocumented. Some of us feel a lot more secure with official info rather than stuff written up by some guy (not to belittle the work some of those guys go through to give us that info).
But by now I'm beating a dead horse. I think you get what I'm saying. |
|
| Back to top |
|
|
SnEptUne
l33t
Joined: 23 Aug 2004
Posts: 656
|
| Posted: Mon Jul 31, 2006 11:18 pm Post subject: |
|
|
No, because I don't use binary drivers
However, support wise, it maybe better off marking it ~x86/~amd64. However, it should be marked stable on ppc since the binary driver wouldn't work there anyway, right?
_________________
"There will be more joy in heaven over the tear-bathed face of a repentant sinner than over the white robes of a hundred just men." (LM, 114) |
|
| Back to top |
|
|
slycordinator
Advocate
Joined: 31 Jan 2004
Posts: 3065
Location: Korea
|
| Posted: Mon Jul 31, 2006 11:22 pm Post subject: |
|
|
| SnEptUne wrote: | No, because I don't use binary drivers
However, support wise, it maybe better off marking it ~x86/~amd64. However, it should be marked stable on ppc since the binary driver wouldn't work there anyway, right? |
I think the general rule is that something has to be in testing for 30 days w/o any unresolved bugs before it's made stable.
So if there's an open bug with 7.1 or hasn't been 30 days yet since it last had one, it'll wait. |
|
| Back to top |
|
|
|
|
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
|
|
Gentoo Chat
