| 1 |
On 01/26/2010 12:40 AM, Volker Armin Hemmann wrote: |
| 2 |
> On Montag 25 Januar 2010, Nikos Chantziaras wrote: |
| 3 |
>> On 01/26/2010 12:06 AM, Volker Armin Hemmann wrote: |
| 4 |
>>> On Montag 25 Januar 2010, Alan McKinnon wrote: |
| 5 |
>>>> On Monday 25 January 2010 23:11:18 Nikos Chantziaras wrote: |
| 6 |
>>>>>> Today it is finally released. I don't know why it took so long. :) |
| 7 |
>>>>> |
| 8 |
>>>>> It's a bit sad though that an open source project keeps secrets... I |
| 9 |
>>>>> found out that the tarballs were available for days now but only for |
| 10 |
>>>>> maintainers of KDE packages of various distros. For example Arch Linux |
| 11 |
>>>>> had RC2 binaries available in its [kde-unstable] repo since Jan 21, but |
| 12 |
>>>>> no one would give a link to the tarballs used to build the sources |
| 13 |
>>>>> because they are "secret". |
| 14 |
>>>>> |
| 15 |
>>>>> Am I the only one to find this frustrating (to say the least)? |
| 16 |
>>>> |
| 17 |
>>>> There's nothing secret here. Nothing to see, move along folks. |
| 18 |
>>>> |
| 19 |
>>>> The code is always available in svn, if you wanted it that bad you could |
| 20 |
>>>> have checked it out. It's common release engineering practice to build |
| 21 |
>>>> a tarball from the repo and provide limited access to *those* files to |
| 22 |
>>>> distros and mirrors to give them time to sync and build their binaries |
| 23 |
>>>> before making them available to the public. I mirror Ubuntu and Fedora. |
| 24 |
>>>> The deal is that I get the rpms and debs two days before the public |
| 25 |
>>>> does. This is so that every official mirror on the planet is ready to |
| 26 |
>>>> simultaneously publish the release *all*at*the*same*time*. |
| 27 |
>>>> |
| 28 |
>>>> This is a very excellent way of working. You have no idea how much chaos |
| 29 |
>>>> thousands of eager beaver users cause pounding away at my inbox if I |
| 30 |
>>>> don't have the latest Ubuntu for them the day it is released because |
| 31 |
>>>> sharks bit the undersea cable to London and the inbound bandwidth slows |
| 32 |
>>>> to a crawl. I'm not making this up - stuff like this happens. |
| 33 |
>>>> |
| 34 |
>>>> Gentoo erred in publishing ebuilds for tarballs that were not generally |
| 35 |
>>>> available yet. |
| 36 |
>>> |
| 37 |
>>> no, the user erred in stupid unmasking. |
| 38 |
>> |
| 39 |
>> The user used the provided unmask file |
| 40 |
>> (Documentation/package.unmask/kde-4.4). |
| 41 |
>> |
| 42 |
> |
| 43 |
> before release. |
| 44 |
> |
| 45 |
> Gentoo has a proud tradition of providing the ebuilds before the software is |
| 46 |
> released so as soon as the tarballs hit the mirrors you can start compiling. |
| 47 |
> |
| 48 |
> If you unmask earlier and get errors you are the only person to blame. |
| 49 |
|
| 50 |
And besides, I unmasked for KDE 4.3.90 (RC1) which was released, not |
| 51 |
4.3.95. But the provided unmask doesn't distinguish between them; it |
| 52 |
only specifies the :4.4 slot, not -4.3.90 version numbers. |
| 53 |
|
| 54 |
Perhaps you should first check that you know what you're talking about. |
Archives