[asterisk-dev] Mantis bugs seen as Feature request, when it may only be an issue.

Olle E. Johansson oej at edvina.net
Wed Nov 18 02:41:46 CST 2009


18 nov 2009 kl. 09.17 skrev Alec Davis:

> Discussion at Leif Madsen's request:
>  
> Using the following bug as an example which was closed as was seen as a Feature request, with no patch attached.
> https://issues.asterisk.org/view.php?id=16248
>  
> Reading the bug guidlines, I believe the issue reported meets the bug-guidelines of 'unexpected or detrimental results in the Asterisk system'.
> The issue is not detrimental to asterisk, but it's not the expected behaviour of a globally deployed PBX solution.
There are many things in the standards that we do not support today. If you want us to support new features in those standards, I would agree that it's a feature request. Supporting a new variant of caller ID signalling is certainly a new feature.

Also remember that closing an issue is not the same as deleting it. It will always be available for searching and re-opening. I search among all issues, open and closed, quite often to find information about various topics.
> 
>  
> The question:
> Should mantis bugtracker be the forum for requests for specific implementations of existing features?

Please remember that the issue tracker is a tool for all of us working with the process and we have decided earlier that filling it with feature requests will block the view from the work we have to do. Experience tells us that if no one puts money or manpower behind it, which sadly doesn't happen in most cases, the request will just hang there and won't help any process at all.

I tried setting up www.asteriskideas.org as a long-term repository for ideas and feature requests, a site which hasn't been successful so far, but at least was a try at one solution. If you have other ideas and energy to move them forward you are more than welcome.

I do think having a repository for ideas and feature requests is a good thing, it gives inspiration and also connects people with similar ideas together. Having it in the same tool as the one we use for processing new code and resolving bugs is wrong, since the feature requests are a very different set of issues and they follow a very different process, if any at all. As soon as that process leads to code submissions, the issues should move to the issue tracker.

The main issue for me personally is how do we get funding for handling those feature requests? We can't really expect Digium to fund all work with Asterisk in the long run. There are many companies who get a large part of their revenue out of Asterisk and seldom contribute back, if at all. If they could fund a feature request now and then, I would be very happy.

Sorry for hijacking your issue and moving it on to a boring an lonely sidetrack, but, well... 

Regards,
/Olle





More information about the asterisk-dev mailing list