| 1 |
On 29 August 2013 07:09, Michael Weber <xmw@g.o> wrote: |
| 2 |
> On 08/28/2013 01:15 PM, Markos Chandras wrote: |
| 3 |
>> Hi, |
| 4 |
>> |
| 5 |
>> I'd like to ask the council to vote on the following topics regarding the |
| 6 |
>> 'minor arches' based on the feedback I received on the respective |
| 7 |
>> thread in the gentoo-dev mailing list |
| 8 |
>> |
| 9 |
>> http://marc.info/?l=gentoo-dev&m=137708312817671&w=1 |
| 10 |
>> |
| 11 |
>> Drop the following arches to ~arch |
| 12 |
>> |
| 13 |
>> - s390 |
| 14 |
>> - sh |
| 15 |
>> - ia64 |
| 16 |
>> - alpha |
| 17 |
>> - m68k |
| 18 |
>> - sparc |
| 19 |
>> -(maybe ppc and ppc64?) |
| 20 |
> make that x86 to be consequent. |
| 21 |
|
| 22 |
I don't think being sarcastic adds anything valuable to the conversation |
| 23 |
|
| 24 |
>> |
| 25 |
>> The feedback on the original question was mostly positive. |
| 26 |
>> Most people agree that the long stabilization queues for these |
| 27 |
>> architectures create problems |
| 28 |
>> for maintainers wishing to drop old versions. |
| 29 |
> Is this the only motivation? Drop all the effort that has been put into |
| 30 |
> stabilization work on minor arches just for some impatient maintainers? |
| 31 |
|
| 32 |
Others reasons are explained to the thread in -dev ML |
| 33 |
|
| 34 |
> |
| 35 |
> Keywording/Stabilization is a process we all agreed on joining, so live |
| 36 |
> with it. |
| 37 |
> |
| 38 |
>> The council should also take into consideration that the stabilization process |
| 39 |
>> for these arches is mostly a one-man job (Agostino). |
| 40 |
> It's the same one man show for amd64 and x86. |
| 41 |
|
| 42 |
It is not. |
| 43 |
|
| 44 |
> Minor arches tend to have less cpu/io performance than this fancy |
| 45 |
> show-off amd64 dev machines. |
| 46 |
> Running the @world\@system on bleeding edge might be a never ending |
| 47 |
> compile job. |
| 48 |
|
| 49 |
This is one of the reasons they are so slow. And combining it with the |
| 50 |
one-man show argument, |
| 51 |
you see why we have problems. Moreover, as you can see, nobody from |
| 52 |
these arch teams |
| 53 |
ever replied (apart from Matt) which kinda proves my point. |
| 54 |
|
| 55 |
> |
| 56 |
> Please give it thorough consideration before throwing all the work out |
| 57 |
> of the window. |
| 58 |
|
| 59 |
The council members are supposed to read the original thread in |
| 60 |
gentoo-dev before they make a decision. |
| 61 |
I believe adding more feedback to this thread might end up in |
| 62 |
/dev/null so if you want to contribute to the |
| 63 |
conversation please do so in the original thread. It would be easier |
| 64 |
for them to have everything collected |
| 65 |
in a single thread when they discuss this. |
| 66 |
|
| 67 |
-- |
| 68 |
Regards, |
| 69 |
Markos Chandras - Gentoo Linux Developer |
| 70 |
http://dev.gentoo.org/~hwoarang |
Archives