[sldev] sljirastats.com Linden Metrics Report

[Rob Lanphier]

On 4/16/08 2:45 PM, Mike Monkowski wrote:
> Rob Lanphier wrote:
>> On 4/16/08 12:02 PM, Jason Giglio wrote:
>>> "Open" bugs have been on a steady climb.  It started at 1547 when
>>> sljirastats.com began collecting data, and is now at 2717.   This means
>>> we are averaging 8.6 additional unresolved bugs per day.  This is after
>>> subtracting fixed and resolved issues.  This number is a critical
>>> metric; Linden Lab is falling more and more behind every day.
>>
>> How does this compare to other open source and public projects?  My 
>> suspicion is that almost every active open source project experiences 
>> net positive growth in number of open bugs.  I'd like to make sure 
>> we're not being measured against an unreasonably high standard.
>
> I don't think being open source would *add* to the number of 
> unresolved bugs.  Why do you think it would?

Because anyone can file issues, and you can't get the people who 
disagree about whether an issue is really an issue into the same room 
and lock them there until they agree on a resolution.  With proprietary 
software you can.

It's also more a question of observability.  I'm not interested in 
abstract discussions about issue trackers we don't have access to.  I'm 
interested in having a discussion about something with a public audit 
trail that we're all on a level playing field when we discuss pros/cons.

It's also about what I've observed:

Linux kernel.  1298 open issues out of 10460 since the tracker was set 
up in late 2002:
http://bugzilla.kernel.org/

Firefox:  Query timed out before I could get a result:
https://bugzilla.mozilla.org/

MediaWiki:  2869 open issues out of 13768 created since the tracker was 
created in mid 2004.  Issues I've personally been watching have sat open 
for years:
https://bugzilla.wikimedia.org/

I think all of these are fine pieces of software that I use on a daily 
basis, and hope we can have the same success as these projects.  Yet 
they are also not reaching the platonic ideal of zero net growth in open 
issues as near as I can tell.  If there are active projects of 
comparable size/complexity which do achieve the zero-open growth 
standard that we can all observe the public bug tracker for, I'd be very 
interested in seeing it.  While it'd be cool to be a pioneer, I'm much 
more interested in achieving something that is sensible to achieve given 
community norms.

Rob


-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 249 bytes
Desc: OpenPGP digital signature
Url : http://lists.secondlife.com/pipermail/sldev/attachments/20080416/2f4a185b/signature.pgp