Re: [boost] GSoC 2010 idea

Subject: Re: [boost] GSoC 2010 idea
From: Robert Ramey (ramey_at_[hidden])
Date: 2010-03-23 00:28:01

Joel de Guzman wrote:
> On 3/23/2010 8:02 AM, Andrew Sutton wrote:
>> Hi Robert,
>>
>> Unlike most of the GSOC projects proposed, it would actually
>>> be doable in the time alloted.

> codebase, providing user support, etc. I'd say this is an exemplar
> project. And, make no mistake, the commitment extends far beyond the
> time frame of GSOC for it to be considered as really successful.

This is my point.  Just about all of the proposals are beyond what
one can expect to accomplish in 3 months.  Yet, the proponents
seem to expect that they will start now with an idea and three
months will have something ready to submit to boost.  History,
and my personal experience in software development tell me
that this is very unlikely.  So one has two alternatives:

a) prepare to make a commitment way beyond 3 months

b) or pick a project smaller than a whole new library.

It's certainly not my intention to discourage anyone.  It's just
a fact of software development that finish a project to the
level that boost demands is way underestimated.  Any author
of an accepted library can verify that.  I look at some of the
proposals, the questions being asked, and the experience
of the submitters, I can't help but think that they are taking
on a a task/schedule that will turn out to be undoable.  The
unfortunate thing is there are a lot of things that could be done
which ARE in fact doable - like enhancements to existing
libraries.  Unfortunately, these likely lack the "sex appeal" that
makes for a winning proposal.

In any case, I would suggest that proponents of these project
should consider that there is nothing like the feeling that your
code is being used in 1000's of applications.  And this feeling
you will only get if a task is taken to completion.  A proposal
for a whole new library which ends up unfinished might be
interesting and fun, but it can't provide the satisfaction and
respect from one's peers that the actual completion of a task
can.

Just my 2 cents.

Robert Ramey

>
> Regards,