Anyone mind if i go about fixing some obvious problems in the gnats DB

[Daniel Berlin <dberlin at dberlin dot org>]

Does anyone mind if i fix obvious problems in the gnats db?
I can't do it by the web interface, unfortunately, so nobody would get 
notification, but i can list the bugs i changed and why in an email.

What do i want to change? Corruption and bugs caused by gnats itself.


Fer instance, invalid dates, like on PR 5967:
State-Changed-From-To: analyzed->closed
State-Changed-By: amodra
State-Changed-When: Fri Sep 13 16:37:22
State-Changed-Why:
    Fixed on mainline and 3.2 branch


(This should be Fri Sep 13 16:37:22 2002)


I'm a bit tired of special casing things in the converter, which i'm 
rewriting for official submission to bugzilla (and cleanup and whatnot).

There are other places caused by clear bugs that fixing would improve 
conversion.

For instance, we have reports with something like:

>How-To-Repeat:
begin 644 blah.i.gz
<...., but last 3 lines of uuencoded attachment not here>
>Audit-Trail:
....
>Unformatted:
<last 3 lines of uuencoded attachment>

Which is also clearly wrong, and in fact, causes us to truncate the 
gzip'ed file by a few bytes right now.

I'm only talking about fixing problems in bugs that were obviously 
caused by gnats screwing up. Not changing the categories or anything of 
existing bugs.

Some of them i can't fix of course, since they are just plain old 
missing data, like this cute one (bootstrap/3250)
>Unformatted:
----gnatsweb-attachment----
Content-Type: application/octet-stream; name="toplev-hpux.i.bz2"
Content-Transfer-Encoding: base64
Content-Disposition: attachment; filename="toplev-hpux.i.bz2"


<end of bug>