[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Bug#793331: RFS: postsrsd/1.2-1 [ITP]



On 22/08/15 15:06, Oxan van Leeuwen wrote:
> Hi,
>
> On 21-08-15 00:19, Tomasz Buchert wrote:
> >I think that we either:
> >
> >   * need hard dependency on postfix
> >   * need to have a debconf dialog that goes more or less like that:
> >     - if postconf exists, the domain is taken from there
> >     - if not, the current hostname is taken
> >     - then a debconf dialog is shown prefilled with these defaults
> >     - the obtained domain name is used in init scripts
> >
> >Waht do you think? Tell me if you need help with the second.
> The second option seems sensible, I've pushed an implementation. Let me know
> what you think.

Nicely done, but there is one issue. If you modify conffiles in postinst
it marks them as changed, and this is not something you want to do.

In my branch "tomasz-changes" I followed ADVANCED PROGRAMMING WITH
DEBCONF section in "man debconf-devel". The outcome is that the config
file is not tracked by conffiles machinery and hence the problem does
not exist. Please review.

>
> >I think that a sensible thing to do would be to provide POSTSRSD_OPTS
> >as 'Environment=...' and then pull a config file with
> >'EnvironmentFile=-...' (which may contain various configs in comments
> >too).  It seems to be sort-of standard. And, right, if you also use
> >debconf, then you probably need to pull a file created created in
> >postinst.
> I'd prefer to keep separate variables for the most common options. A single
> options variable will be quite cryptic (as postsrsd doesn't support long
> options), and that will make it harder to change a setting without having to
> pull the manpage. Duplicating a few defaults is a small price to pay for
> ease of configuration in my opinion. We could add a options variable to add
> additional command-line options, though I doubt it will be used much.

Ok.

>
> Cheers,
> Oxan
>

Cheers,
Tomasz

Attachment: signature.asc
Description: Digital signature


Reply to: