[OSGeo-Discuss] Open Source development metrics

P Kishor punk.kish at gmail.com
Wed May 28 03:01:00 PDT 2008


On 5/28/08, Bruce.Bannerman at dpi.vic.gov.au
<Bruce.Bannerman at dpi.vic.gov.au> wrote:
>
> IMO:
>
>
> Thanks for the comments Puneet,
>
>
>  >
>  > Actually, a variation on the above may be the best metric -- "create
>  > feature X that we need in our organization and that works for us."
>  > That would allow your organization to determine what is meaningful for
>  > your organization first and for open source second. In other words,
>  > you would treat open source development no different from non-open
>  > source development. Open source would simply become a "normal"
>  > activity.
>  >
>  > Once feature X works for you, you could consider "donating" it to the
>  > open source community by whatever process that particular open source
>  > project has.
>  >
>  > Other metrics such as SLOC (source lines of code) or "feature in SVN
>  > trunk" are not only subject to abuse, they are also mostly
>  > meaningless.
>  >
>
>
> We want to avoid anything that could result in an isolated fork or branch.
>
> This is a danger with this approach.


Quite the contrary -- don't fixate on "forking as a bad thing." Do
what is good for your own organization. If it is of sufficient general
interest, and if enough folks find it of use to them, it will get
absorbed in the main branch anyway. Else, it will either die off (in
which case it still served your own organization's purpose) or develop
as a specialized branch. Very Darwinian.

Forking is not a bad thing. I have no idea why it is viewed as such.
Forking is one of the beauties of open source, brings diversity in the
code base, and even ensures longevity. The ability to fork is what
ensures that in open source there will not be any lock-in.

The beauty of this approach is the open source is not treated as
something special. It becomes as normal as non-open source software.


>
>
> However, it has merit in that it would allow a developer to meet their
> performance
> criteria and concentrate on getting the customisation into the trunk in
> their
> own time.
>
>
> Bruce
>
>
>
>
>
>
..
-- 
Puneet Kishor http://punkish.eidesis.org/
Nelson Institute for Environmental Studies http://www.nelson.wisc.edu/
Open Source Geospatial Foundation (OSGeo) http://www.osgeo.org/



More information about the Discuss mailing list