[Chicago] GitHub & SpamBayes

Brian Curtin brian at python.org
Tue Feb 5 16:44:36 CET 2013

On Tue, Feb 5, 2013 at 9:39 AM, Carl Karsten <carl at personnelware.com> wrote:
> On Tue, Feb 5, 2013 at 9:05 AM, Skip Montanaro <skip at pobox.com> wrote:
>>> I think I see Skips problem.  He thinks (and may be correct, he sure
>>> knows more about it than any of us do) that no one is going to
>>> contribute regardless of how easy it is.
>> ...
>>> What he really needs is for someone else to become an active
>>> maintainer, and they can move it to github.
>>> Skip, how close am I?
>> Close.  I look at it this way.  Moving will require some work to
>> migrate from SF/svn to SF/git or GitHub/git, with no obvious payoff.
>> It's not like someone has told me, "I would be happy to fix your
>> Windows installer problems if SpamBayes was hosted on GitHub. It would
>> be so much easier for me to contribute."  My assumption at this point
>> is that I need to do the work up front (probably not much, but
>> non-zero).  Then I will have to make people aware of the project and
>> its needs.  (More work.) Someone else mentioned that GitHub's issue
>> tracker more-or-less sucks, so maybe I need to set up a tracker at
>> code.google.com.  (A bit more work.)  Still, there is no guarantee
>> anybody will come along and even see my tracker issues, much less work
>> on them.  Then, should any of the existing (dormant) developers want
>> to track progress, they will have to set up clones, maybe learn Git
>> from scratch, etc.
>> Let's assume I have no developers banging down my door to help, and I
>> do all the work above.  How do I make people aware of the project and
>> its needs within the GitHub universe?  That's what I see as the most
>> challenging part, unless GitHub has a button I can press which reads:
>> "Send a note to all GitHub folks with Windows installer and Python
>> experience pleading for help."
>> Then, let's assume I have everything set up and am ready to process
>> pull requests.  I no longer even use SpamBayes, so my own ability to
>> test other peoples' changes will be a bit compromised.
>> I'm really not trying to be dense or difficult here.  I am hesitant to
>> put any effort into this without at least seeing that there might be a
>> path from A to B.
> AH!  Everybody listen up!
> Skips question (now anyway)  is not about how to handle contributions.
> Skips question is: will moving to github get the project new interest?
> If the answer is "No" then it doesn't matter how easy it is for 0
> people to submit 0 contributions.   If the answer is "Yes" then we
> need some hint as to how this works to get some feel for the
> probability  that this new awareness will actually result in interest
> and desire to contribute.

The answer is between yes and no, siding towards no. If no one has
actively been contributing to it for years, Github isn't going to be
the silver bullet that has people jumping out from behind bushes to

Without doing or saying anything as far as PR goes, you might get a
few drive-by contributors that find it easier now to fix their minor
gripes. I would be shocked if anyone said "holy crap SpamBayes is on
Github, I have 30 features ready to go right now." Without trying to
drum up activity it will just be an exercise in changing source
control tools.

More information about the Chicago mailing list