| 1 |
On Wed, Jan 21, 2015 at 9:00 AM, Sam Bishop <sam@××××××.email> wrote: |
| 2 |
> |
| 3 |
> I don't see why it can't be all the combinations, the issue is |
| 4 |
> storage, and the storage costs could be a lot lower than expected |
| 5 |
> given how hard it is to guess. |
| 6 |
|
| 7 |
I don't believe that binpkg filenames contain the use flag settings, |
| 8 |
and I'm not sure that given multiple copies of a binpkg with different |
| 9 |
filenames portage goes through them and figures out which ones are |
| 10 |
which. This isn't an area I have looked into seriously. However, it |
| 11 |
obviously would be a blocker for getting what you propose to work, |
| 12 |
even theoretically. |
| 13 |
|
| 14 |
I don't really see the value in having EVERY combination of use flags |
| 15 |
on call though. |
| 16 |
|
| 17 |
Practically speaking I doubt this could be done. You're talking about |
| 18 |
a LOT of combinations. |
| 19 |
|
| 20 |
However, I think it would be very useful to have a binpkg repository |
| 21 |
all the same. Perhaps have one for each of a few common profiles with |
| 22 |
the default flags. That alone would be a significant undertaking. |
| 23 |
|
| 24 |
Just about everybody who has talked about running Gentoo in a |
| 25 |
datacenter has set up a binpkg repository. They may very well deviate |
| 26 |
from the default USE flags, but for the most part they try to keep |
| 27 |
their systems identical. They would build updates as binpkg, install |
| 28 |
to a test system, and after testing deploy them to production and that |
| 29 |
would of course go quickly. |
| 30 |
|
| 31 |
I have a script I use to build binpkg nightly for the day's updates. |
| 32 |
That lets me review updates and deploy them quickly. Any rebuilds/etc |
| 33 |
still take time, but the bulk of my updates are very fast this way |
| 34 |
with minimal time spent staring at the screen. This would be another |
| 35 |
route to take if your really did need highly varied deployments. |
| 36 |
|
| 37 |
-- |
| 38 |
Rich |
Archives