| 1 |
On 2023-02-19 10:24+0100 Matthias Hanft <mh@×××××.de> wrote: |
| 2 |
|
| 3 |
> Hi, |
| 4 |
> |
| 5 |
> according to https://bugs.gentoo.org/718972 (and a posting |
| 6 |
> in the dev list), gsutil (and all needed libs) will be removed |
| 7 |
> from the tree because "it's broken" (which is not true - works |
| 8 |
> like a charm). |
| 9 |
> |
| 10 |
> Since I do need gsutil urgently, I have already copied the needed |
| 11 |
> ebuilds from /usr/portage to a local overlay in order to "emerge |
| 12 |
> -c gsutil::gentoo" and then "emerge gsutil::local". |
| 13 |
> |
| 14 |
> But then came to my mind: Is that necessary at all? As far as I |
| 15 |
> know, I can just leave discontinued packages installed and continue |
| 16 |
> to use them - regardless of whether they are still in portage tree |
| 17 |
> or not. True or false? |
| 18 |
|
| 19 |
Yes, you can just continue to use them. The ebuild will be kept in |
| 20 |
/var/db/pkg/ in case portage needs to reinstall them or so. |
| 21 |
However, it might make sense to copy it to a local overlay to make it |
| 22 |
easier to get an overview of self/un -maintained packages. And since |
| 23 |
gsutil requires Python, you'll have to modify the ebuild at some point |
| 24 |
anyway to add support for newer Python versions. |
| 25 |
|
| 26 |
> Or is it better to get rid of portage/ebuilds at all and do a |
| 27 |
> manual installation, as described at |
| 28 |
> https://cloud.google.com/storage/docs/gsutil_install?hl=en ? |
| 29 |
|
| 30 |
I'd say use the ebuild since it already exists. Makes breakage less |
| 31 |
likely. |
| 32 |
|
| 33 |
Kind regards, tastytea |
Archives