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

Re: finding out which version of a package a bug affects



On Wed, 21 Apr 2010, Robert Lemmen wrote:
> i am trying to write something similar to bts.turmzimmer.net, and
> somewhere in these i need to find out which versions of a package a
> given bug affects. i have read the information on the wiki, and had a
> quick chat with sesse, but there are still some things i do not quite
> understand:
> 
> - how can i get the version graph as used by debbugs? is there a way
>   on the soap to query it? will there be one in the future? is that
>   what getversions() is for? i have of course found the raw data on
>   merkel, but is that accessible without a login to merkel (http or
>   so)? and how do i parse it?

The raw data is rsyncable, but you should be using the soap interface
for everything. If there is something that the soap interface doesn't
do that you want to do, you should file a bug against debbugs
requesting a specific feature to be added.

There currently isn't a way to query the version graph using soap, but
in almost all cases, you actually don't want to do that. You should be
using the BTS's own routines to calculate whether a bug is buggy or
not instead of writing your own.
 
> - i don't quite understand what the "source/version" format in found
>   and fixed is good for. under which circumstances can this be
>   different from the source package the bug is against?

There are cases where a bug exists in multiple different source
packages or different source packages build the same binary package on
different architectures.

> - if a bug gets reassigned, what happens to the found/fixed
>   information?

It's cleared.

> - is it correct that if i have only a "found" and the bug is not
>   "done", all versions in the tree after the "found" are affected.

Yes.

> - in a similar manner, if i have only the "fixed" information, all
>   versions before that have to be considered affected?

No.

> - if the version tree is only one long chain (for the sake of argument),
>   will there be only one affected interval, or can there be multiple
>   (e.g. found: 1;4 fixed 2;6)? can they be redundant (e.g. found: 1; 2 
>   fixed: 4; 5)

Yes, but none of this is important. The BTS handles all of it for you.


Don Armstrong

-- 
He quite enjoyed the time by himself in the mornings. The day was too
early to have started going really wrong.
  -- Terry Pratchet _Only You Can Save Mankind_ p133

http://www.donarmstrong.com              http://rzlab.ucr.edu


Reply to: