http://brainstorm.ubuntu.com/item/9219/
There are also many users that are not totally happy with the beta version installed by default and are willing to get the official version as soon as possible, in order to fully use the browser (some web pages and plugins doesn't work with the beta).

[173 votes] Solution #1: Auto-generated solution of idea #9219


]]> en-us Thu, 29 May 2008 08:40:08 +0000 Sat, 31 May 2008 16:16:29 +0000 QAPoll module http://brainstorm.ubuntu.com/idea/9219/ Thu, 29 May 2008 12:17:31 +0000 Guinness is important!]]> Thu, 29 May 2008 13:08:43 +0000 Thu, 29 May 2008 13:20:28 +0000 ]]> Thu, 29 May 2008 13:46:40 +0000 Thu, 29 May 2008 16:17:42 +0000 Thu, 29 May 2008 16:56:17 +0000
I have to mark it -1 because getting it into the repo would do absolutely nothing for the record push.]]> Thu, 29 May 2008 18:35:15 +0000 Thu, 29 May 2008 19:31:16 +0000 Thu, 29 May 2008 20:33:23 +0000 Even if you are downloading from the repositories you are still downloading it.
It would be nice if Ubuntu and Mozilla would come up with a way to count our repo downloads for the world record.
I mean... Why not?]]> Thu, 29 May 2008 22:25:39 +0000 - tested
- stable

I don't think they will get pushed out faster than usual and I really hope staff will be professional about it. Testing new release of main application for only one day is not enough.
-1]]> Thu, 29 May 2008 23:18:11 +0000
Because last I checked:
a) We know 3.0 RC/Betas have bugs. No doubt about it. And, we know that 3.0 has less errors. So what exactly are we testing here? That the bugs in Beta5 have been fixed?

b) If you are saying it takes longer then a day to test the installation package, then its a problem with the packaging process, and highlights why we need to use packages which are more universal. It shouldn't take a day to ensure a package deploys properly.

c) All we really need to test is if it starts up, and test it for 2 or 3 hours, with maybe the flash plugin, and the totem plugin. However, the update could maybe be blocked if the user has any updates that are installed that are not compatible.

d) Canonical SHOULD be testing with the release candidates anyway, even if they aren't officially released. A lot of the time, the final release candidate is exactly the same as the final. That means they may actually have 3 days to test. If they properly test with the RC's, it means on final release, they should be ready to just package it on the spot!

e) If we need to force people to wait more then 24 hours to use any program then good luck in the market-share department. Because my vista and OSX computers which are running RC already, WILL be updated same day, within 15 mins of me waking up. No waiting for Canonical to catch up on packaging. No waiting on anyone else. I could get it within minutes of release (and yes, I can uninstall it as easily in windows/osx as with APT).

So if it cant be done same day, what you are REALLY saying, is that linux isn't ready yet for mainstream. People don't want to have to wait a week after a program is released officially to start using it, especially, when they are already using the beta. Thats a great signal to send to people. People love to be told that the beta, is possibly more stable then the final.


f) How could final possibly be any less stable then BETA5?



g) Just remember, this isn't normal circumstances. Ubuntu normally doesn't ship with beta software installed. So the same rules shouldn't apply.

Any mods here can shed any light on this on how long it should take?

I think 1 day is enough, you could have a test package out within a few hours, shoot it off to everyone in #ubuntu-dev, get them to test it for 2 or 3 hours, and BOOM.]]> Fri, 30 May 2008 02:10:15 +0000 It should be checked on different architectures, on different sites. If there are any ubuntu specific patches, they need to be reviewed before applying of course. Etc.... I can give you a list of things to check that you won't complete in a week.

If new FF satisfies all of those points, then I'd be happy to install it. "linux isn't ready yet for mainstream"? Because it doesn't include packages the same day? Tell that to one of certified mainstream distributions like RH which run and support software older than a year and don't upgrade it, just to make sure everything is working fine.

Noone's stopping you from installing FF from sources on the release day yourself, or 'forcing' you to do anything else. But pushing new version on *every* user's box without proper testing would be just crazy. If they make it the same day, I'm happy, if they won't I won't complain. Take your time - do it right...]]> Fri, 30 May 2008 02:43:24 +0000 I agree with viraptor, because he have something like a "developers big picture view". That's the way the wheels work.

@Auzy : I think you work for M$ ...

Ubuntu should not pay the price of being more bug-prone to be more popular. It's not wise and we have another name for that ...]]> Fri, 30 May 2008 03:50:17 +0000
Just because I compare Linux to Windows or OSX doesn't mean I work for other companies, I am simply stating the facts. So that we can rectify areas we still have problems with.

Thats one thing I noticed about Apple. A lot of people working with Apple haven't used anything else in a long time, so start believing many features are unique. An example is spaces (which is virtual desktops). While every developer thought Steve was an idiot (you could hear it in the background), I seemed to be the only one at my Apple premium reseller who knew that linux/windows had support for ages. So thats why I compare, to ensure that long term linux users are aware of how we compare, so we don't disillusion ourselves. I'd like to see linux be the best/most usable in all areas. Because I believe the best thing for us is to draw a massive userbase, so that more hardware companies and users support us.




But I will admit that both sides have a good point.
On one side, you are right. We should test plugins and all that too.

However, on the other side, we are using a beta, so people upgrading from 2.0 to 3.0 are less likely to have issues with the final. And there aren't hundreds of plugins to test (nor do we have to test every single architecture, as ix86 is the one which most ubuntu users use).

Furthermore, where I'm coming from, is that everything should start to be properly tested every step of the release candidates. So we may have more testing time of the final release then some believe.


I guess by just offering the package for install in APT to everyone, (in synaptic, not software update) till its perfectly tested benefits everyone.

1) My side everyone can still install firefox 3.0 final in one click.
2) Your side don't accidentally install firefox 3 till its ready.

We get the one clicks, you don't get firefox 3 shoved down your throats ;) Then if there are any patches, my side can always install later. ]]> Fri, 30 May 2008 05:05:51 +0000 Peace!]]> Fri, 30 May 2008 06:03:29 +0000 Fri, 30 May 2008 06:18:49 +0000 Fri, 30 May 2008 06:27:20 +0000 Fri, 30 May 2008 06:57:27 +0000
btw 6205, this is different. Its regarding "on the day"]]> Fri, 30 May 2008 07:04:12 +0000 Auzy, your comment written here on 30 May 08 at 05:05 is very reasonable. Giving an option is a better option indeed.]]> Sat, 31 May 2008 16:16:29 +0000