1 |
Hi Peter, |
2 |
|
3 |
On 2021/02/09 19:51, Peter Stuge wrote: |
4 |
|
5 |
> Joonas Niilola wrote: |
6 |
>> There's a script by jkroon in data/api.git |
7 |
>> (https://gitweb.gentoo.org/data/api.git/) that prints the next available |
8 |
>> UID/GID pair for new acct-* packages. |
9 |
> This is super nice! Thanks a lot jkroon! |
10 |
You're most welcome. |
11 |
>> It is not forbidden to mix and mash UID/GID between different packages, |
12 |
>> but I'd still suggest to find a new "pair" even if you push just an UID |
13 |
>> or GID. Since we don't seem to be in danger of running out any time soon. |
14 |
> Mh - so the obvious first feature request for the script is to also output |
15 |
> Free UID+GID pairs. Counting them manually in your screenshot I get 36. |
16 |
|
17 |
https://github.com/gentoo/api-gentoo-org/pull/367 |
18 |
|
19 |
> |
20 |
> That's not a whole lot; just 7% of 500. |
21 |
Recommended GID only: 460 |
22 |
Recommended UID only: 458 |
23 |
Recommended UID+GID both: 326 |
24 |
Free UIDs: 200 |
25 |
Free GIDs: 177 |
26 |
Free UID+GID pairs: 160 |
27 |
|
28 |
Full count of 32% looks a tad better than 7%. |
29 |
|
30 |
Re always picking a pair (raised in original email), my recommedation |
31 |
would be: |
32 |
|
33 |
Pick a pair if there is any chance that you might want the other |
34 |
eventually too, else pick free uid/gid. |
35 |
|
36 |
The problem with that is, if you don't take BOTH right now, someone else |
37 |
might take the other, so perhaps the strategy of always picking a pair |
38 |
is better ... |
39 |
|
40 |
Then again, there are packages that need one UID and multiple GIDs, and |
41 |
probably the other way around too, so not always picking pairs in such |
42 |
cases makes a lot of sense too. |
43 |
|
44 |
Kind Regards, |
45 |
Jaco |