Re: Maintaining packages with complex relationships (Was: Suggesting change in DPT policy)

[Timo Röhling]

Hi,

* Andreas Tille  [2024-02-28 11:51]:

I think it could be useful for the routine-update command to stop 
when such file is hit, in order to raise the importance that the 
package has quirks, and should not be casually updated without 
involved scrutiny.  I wonder whether this can be generalized, 
like if d/README.source file is present?  (Although the latter 
use is codified[2] and I'm not confident it is 100% suitable for 
such purpose: I see many such files on my radar which do not 
necessarily hint for quirks.)



I like all your suggestions.  When reading Timo's suggestion about
debian/README.DPT I also thought about rather using the more generic
debian/README.source.  In any case I agree that routine-update should
respect such debian/README.* (except debian/README.Debian which is
user oriented).


I also thought of README.source at first, and I remained on the 
fence until Étienne brought up the potential use for routine-update, 
which makes me think that a dedicated "quirks" file makes more 
sense. I'm not too attached to the .DPT suffix though, maybe 
something like README.team or even README.quirks signals the 
intention behind the file even better. I'll leave the bikeshedding 
to interested readers for now. :)



Cheers
Timo


--
⢀⣴⠾⠻⢶⣦⠀   ╭╮
⣾⠁⢠⠒⠀⣿⡁   │ Timo Röhling   │
⢿⡄⠘⠷⠚⠋⠀   │ 9B03 EBB9 8300 DF97 C2B1  23BF CC8C 6BDD 1403 F4CA │
⠈⠳⣄   ╰╯


signature.asc
Description: PGP signature

Maintaining packages with complex relationships (Was: Suggesting change in DPT policy)

[Andreas Tille]

Hi Étienne,

Am Wed, Feb 28, 2024 at 09:37:59AM +0100 schrieb Étienne Mollier:
> > > Instead of restricting collaboration, we could let policy encourage
> > > maintainers to state such constraints in debian/README.DPT and ask team
> > > members to check that file before they team-upload.
> > 
> > I think this is a very good idea.  In case MR[1] will be accepted this
> > should be added to the policy as well.  I'm not sure whether the
> > "Maintainership" paragraph is the best place to add this.  I wonder if
> > you (or someone with the same doubts) might want to suggest another MR
> > to add this debian/README.DPT feature.
> 
> Policy changes aside,
(Thus changed subject. ;-) )

> I think it could be useful for the
> routine-update command to stop when such file is hit, in order
> to raise the importance that the package has quirks, and should
> not be casually updated without involved scrutiny.  I wonder
> whether this can be generalized, like if d/README.source file is
> present?  (Although the latter use is codified[2] and I'm not
> confident it is 100% suitable for such purpose: I see many such
> files on my radar which do not necessarily hint for quirks.)
> 
> Of course this could be overred with a --readme-reviewed flag
> once ready to finalize the package with automation for instance.

I like all your suggestions.  When reading Timo's suggestion about
debian/README.DPT I also thought about rather using the more generic
debian/README.source.  In any case I agree that routine-update should
respect such debian/README.* (except debian/README.Debian which is
user oriented).

I admit I like this kind of constructive discussion.

Kind regards
   Andreas.
 
> > [1] 
> > https://salsa.debian.org/python-team/tools/python-modules/-/merge_requests/20
> [2]: 
> https://www.debian.org/doc/debian-policy/ch-source.html#source-package-handling-debian-readme-source

-- 
http://fam-tille.de