| 1 |
On Sun, 24 Jul 2016 00:04:53 +0200 |
| 2 |
"Andreas K. Huettel" <dilfridge@g.o> wrote: |
| 3 |
|
| 4 |
> -----BEGIN PGP SIGNED MESSAGE----- |
| 5 |
> Hash: SHA512 |
| 6 |
> |
| 7 |
> Am Freitag, 22. Juli 2016, 15:57:36 schrieb Ciaran McCreesh: |
| 8 |
> |
| 9 |
> > > > Wrong. PMS specifically requests you to account for such a |
| 10 |
> > > > possibility. |
| 11 |
> > > |
| 12 |
> > > Common sence must prevail over formal approaches. While PMS is |
| 13 |
> > > great, it is not perfect in all possible aspects, and this one is |
| 14 |
> > > one of them. |
| 15 |
> |
| 16 |
> [snipping irrelevant blather] |
| 17 |
> |
| 18 |
> > Slots are not the only way in which you can end up with multiple |
| 19 |
> > installed versions of the same package. Another way is if there's a |
| 20 |
> > fatal error during certain parts of the upgrade process. |
| 21 |
> |
| 22 |
> 1) If a package only ever had one slot, it cannot ever have two versions |
| 23 |
> installed at the same time. That guarantee (of only ever one slot) can be |
| 24 |
> given for the portage tree (sic). Obviously it doesn't work for overlays, |
| 25 |
> but there are many things we don't care about for overlays. [A] |
| 26 |
> |
| 27 |
> 2) If a package manager leaves two versions of a non-slotted package |
| 28 |
> "installed" somehow, that package manager is fundamentally broken and its |
| 29 |
> author should forever refrain from specifying anything. It's not our job to |
| 30 |
> work around Paludis failure modes. [B] |
| 31 |
> |
| 32 |
> |
| 33 |
> [A] Let's say there are overlays which package StarOffice, Go-OOO and some |
| 34 |
> other random OOO fork. Do I have to block them all because of file collisions |
| 35 |
> then? |
| 36 |
> |
| 37 |
> [B] The package manager could be broken to leave some random files on the |
| 38 |
> system too... maybe we need some more blockers or specific error handling in |
| 39 |
> all ebuilds? |
| 40 |
|
| 41 |
So, to summarize: we should dump PMS and get back to caring only what |
| 42 |
Portage seems to do for a few developers with big mouths, because |
| 43 |
adding a single 'for' loop is so complex you can't handle it? |
| 44 |
|
| 45 |
Instead you prefer wasting hours of time of others to discuss ignoring |
| 46 |
the specification rather than doing things properly. If you don't like |
| 47 |
PMS, start the procedure for changing it. Or dump it altogether. But |
| 48 |
stop this nonsense of 'we use this spec that was written specifically |
| 49 |
for us but random developers can ignore random points because they |
| 50 |
can'. |
| 51 |
|
| 52 |
-- |
| 53 |
Best regards, |
| 54 |
Michał Górny |
| 55 |
<http://dev.gentoo.org/~mgorny/> |
Archives