Why Python?

Benjamin Smedberg benjamin at smedbergs.us
Wed, 09 Nov 2005 17:11:48 -0500

Brendan Eich wrote:

> If someone thinks doing embedding for key clients other than Firefox, 
> Thunderbird, Minimo, Camino, and the other apps we tinderbox or 
> otherwise care about (Eclipse included), or doing embedding as a 
> general-purpose SDK and runtime _a la_ XULRunner or the FlashPlayer, *as 
> an end in itself* has similar such strategic advantage in bringing 
> high-quality large volumes of new blood quickly to our world, please 
> make a case.

It's hard for me to judge what "new blood" means here, but I certainly do 
see XULRunner already prompting significant discussion from developers on 
npd.xul and other forums who I haven't really heard of before. These are 
mostly extension-developer types are graduating up to writing larger 
standalone apps using the roughly-familiar XUL technologies; I may be 
misinformed or pessimistic about python, but I believe that this community 
has at least as much potential as Python. (creating a union of these 
communities would be ever better)!

What I'm concerned about is that the roadmap should be a document saying 
"this is what we are actually planning to do", rather than "this is what 
we'd like to do, and we'll cut stuff later". This doesn't mean that we're 
cutting python support, but it does mean that I don't think we should be 
promising it (and we can generate excitement within the Python community 
without making promises). Building a developer community around Mozilla 
(Firefox) requires that companies/people who are planning business models 
around FIrefox+python or xulrunner or whatever, know whether when Mozilla 
say "we're doing Python support" that means that they will actually be able 
to *deploy* something in Q1 07 which reliably uses XUL+Python (deployment is 
an extremely important part of the battle that I think we have historically 
tended to neglect). A generic "we can solve this with stub installers and 
dynamic downloading" response isn't good enough!