| 1 |
PPC Stages and GRP testing is in progress. |
| 2 |
|
| 3 |
On 07 Jan 2004, at 21:46, John Davis wrote: |
| 4 |
|
| 5 |
> Additionally, if you want livecds for your arch, *please mail Daniel |
| 6 |
> and |
| 7 |
> CC the list your kernel .configs and bootloader info ASAP.* |
| 8 |
|
| 9 |
I've cc'ed trance, he's our most qualified person to provide info about |
| 10 |
these things. |
| 11 |
|
| 12 |
> There is no guarantee that you will get a livecd if you do not do so - |
| 13 |
> you have been |
| 14 |
> warned. |
| 15 |
|
| 16 |
Could you elaborate on this? What do you mean by 'getting' a livecd? |
| 17 |
|
| 18 |
Is your request intended for people building/integrating/testing livecd |
| 19 |
support for architectures other than intel/amd? Additionally, has |
| 20 |
anyone successfully build a livecd for these architectures? Has the |
| 21 |
catalyst/genkernel code been frozen now? I would like to reiterate that |
| 22 |
I stated a few weeks ago that the 01/01/2004 deadline for stable ppc |
| 23 |
livecd building was unrealistic, here's why (again): |
| 24 |
|
| 25 |
Unlike for x86, gentoo ppc developers need to 'guess' what minimal set |
| 26 |
of kernel .configs and minimal set of livecd configurations will work |
| 27 |
to provide the same level of support as distributors who have access to |
| 28 |
the machines, and thus can guarantee support. PPC depends heavily on |
| 29 |
community feedback for that reason. It's like the chicken and egg |
| 30 |
problem: you can't test ppc .configs for the range of machines they |
| 31 |
work on without a livecd and you can't test the livecd without a good |
| 32 |
.config. This process is very frustrating and time consuming. Also, |
| 33 |
there is no such thing as a virtual machine on PPC, that is able to |
| 34 |
boot an iso without burning it. So: you have to build, burn and test, |
| 35 |
rebuild, reburn, retest until you get to a fixpoint (where changing |
| 36 |
something doesn't improve things). This takes time (at least a few |
| 37 |
days) and that time needs to be made available in your schedules. |
| 38 |
Please keep that in mind for scheduling future deadlines. You had been |
| 39 |
warned before ;-) |
| 40 |
|
| 41 |
Frankly, to me your last sentence reads like a threat that you or |
| 42 |
somebody else intends to be very hard on deadlines and start finger |
| 43 |
pointing, even though I stated before they were unrealistic at least |
| 44 |
for the domain I'm managing. Any fingers pointed at me will be rerouted |
| 45 |
to what I've said before. I don't think any ppc developer (volunteers, |
| 46 |
including myself) deserves to be punished for releng scheduling issues. |
| 47 |
This isn't a ppc is holding back gentoo issue, because as everyone |
| 48 |
knows ppc would be the first to applaud multi-architecture initiatives. |
| 49 |
This is just a matter of correct scheduling. |
| 50 |
|
| 51 |
Pieter |
| 52 |
|
| 53 |
|
| 54 |
-- |
| 55 |
gentoo-releng@g.o mailing list |
Archives