| 1 |
>>>>> On Wed, 23 Jun 2021, Fabian Groffen wrote: |
| 2 |
|
| 3 |
> On 23-06-2021 08:47:58 +0200, Ulrich Mueller wrote: |
| 4 |
>> It's somewhat hidden, but it's there: |
| 5 |
>> https://projects.gentoo.org/pms/8/pms.html#x1-950009.1.10 |
| 6 |
>> |
| 7 |
>> 9.1.10 pkg_preinst |
| 8 |
>> ... immediately before merging the package to the live filesystem. ... |
| 9 |
>> |
| 10 |
>> 9.1.11 pkg_postinst |
| 11 |
>> ... immediately after merging the package to the live filesystem. ... |
| 12 |
|
| 13 |
> Aha, so does this mean pkg_prerm and pkg_postrm are run with replacing |
| 14 |
> package in place, e.g. if they refer to scripts installed by the |
| 15 |
> replaced package they may no longer exist or be the same? |
| 16 |
|
| 17 |
PMS has similar wording for pkg_prerm and pkg_postrm, so (at least |
| 18 |
according to the spec) the old package will still be present in |
| 19 |
pkg_prerm and will be gone in pkg_postrm. |
| 20 |
|
| 21 |
I don't think that the ebuild can rely on any particular status of the |
| 22 |
new package in pkg_*rm (of the old package), or the status of the old |
| 23 |
package in pkg_*inst (of the new package). |
| 24 |
|
| 25 |
Ulrich |
Archives