Re: Allowing anonymous users to submit crashers via bug-buddy: Yes/No?

["Christian Kirbach" <Christian Kirbach student uni-siegen de>]

On Thu, 09 Feb 2006 13:19:29 +0100, Olav Vitters <olav bkor dhs org> wrote:

First of all, thanks to fer for improving bug-buddy. the sendmail stuff  
really
sucks and even prevents me from submitting.

> I'm against allowing anonymous users to submit crasher bugs. Mainly

I agree partially. Most bug-buddy submissions are usually useless -  
incomplete
stack traces, lacking debugging symbols. Especially the ones from 'normal'  
users.
I believe painless anonymous xml-rpc bb submissions would flood bugzilla
with new, but mainly useless reports. The only benefit I see is not to open
a bugzilla report, but instead to have the server analyse the trace and
create some statistics so that we can see what bytes users most and focus
on it.

Perhaps bb should allow anonymous submissions (that do not open reports)  
while
kindly asking for and allowing non-anonymous submissions. The latter ones  
should
open a new report since the submitters seems to be willing to help. In  
case of
email provided we should check whether the user already has a bugzilla  
account
and use it. Otherwise perhaps automatically create a new one after asking  
to do so.

(!) bb definitely should give the user the opportunity to add more  
information, and
point to the wiki page 'how to obtain good stack traces' after the user  
has clicked
on some 'I want to help' button.

[Should open a bug-buddy report about this]


> 1. Submit stracktrace automatically (to some function that will not
> create a bug)
agreed.

> 2. Determine on server if it has been fixed yet
> 3a. If fixed: Tell user, exit

Brilliant idea. Do this for the most frequently reported. and ask the user  
to
upgrade.

> Above requires a system to reject known fixed crashers. That would
> system would be manually maintained by the developers of a project
> using the 5 functions (as found by simple-dup-finder).
Sounds good to me, can be implemented at any time.

(!) also we should try to filter for traces that actually have debugging  
symbols
and then open a report (using some bugzilla account?)


> So if you look at lots of (Bug-Buddy) bugs.. would you appreciate
> anonymous posts? Would 'steps to reproduce' be required?
should be optional IMHO. in perhaps 1/3 of all cases users simply
have no idea how they triggered it. that ratio is perhaps 2/3 for
'normal' users.


--
Using Opera's revolutionary e-mail client: http://www.opera.com/mail/