incubator-general mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
Subject Re: Blondie's Parallel Lines... numerically ...
Date Fri, 03 Jun 2011 11:49:04 GMT
Greg Stein <> wrote on 06/03/2011 02:13:43 AM:

> I don't see any of this discussion about numbers being helpful, only
> divisive. "My numbers are right." "No, they're not. See?" "But those 
> are too small."

I agree, especially if the numbers are not relevant to the question at 

It was brought up, and I think it is a legitimate point, whether the 
proposed project will have enough developers to sustain it.  Numbers were 
tossed around about the hundreds of developers that LibreOffice claims to 
have.  But there is certainly a difference between:

Dev 1:  10,000 units of code
Dev 2:  1 unit of code
Dev 2:  1 unit of code
Dev 100: 1 unit of code


Dev 1:  2,501 units of code
Dev 2:  2,501 units of code
Dev 3:  2,500 units of code
Dev 4:  2,500 units of code

Ignore for the moment that any given metric unit here can be manipulated. 
Think of this in the abstract.  Different distributions follow from 
different community structures, based on depth of talent, the difficulty 
of the code base, the number of corporate sponsored developers, etc.

To have an estimate for the number of developers necessary to create an 
Apache OpenOffice community, knowing the total number of developers on 
another project is irrelevant.  If there is a long tail of 1-time patch 
submitters, that may be useful for community development, encouraging new 
stronger developers who might impact the project more down the road, etc. 
But it doesn't really answer the question of what is *necessary* to get 
the project running.

That is why my estimate looked at how many developers were responsible for 
90% of the code contributions in the last 6 months.  That is a more robust 
estimator, less skewed by the distribution.  It is a better indication, I 
think, for this purpose.

So if our LibreOffice colleagues have an easy way of deriving this number, 
I think it would be very useful to have.  I'd certainly be appreciative if 
they could calc this, if not too much trouble.



To unsubscribe, e-mail:
For additional commands, e-mail:

View raw message