Python as replacement for PHP?

Paul Rubin http
Thu Mar 4 11:52:26 EST 2004


paul at boddie.net (Paul Boddie) writes:
> As a PHP replacement, I'd argue that a number of existing frameworks
> are good and obvious enough - cue developer adverts for Spyce,
> CherryPy, etc. ;-) 

While those are nice pieces of code, they're not in the official
distribution, which means they're a separate point of failure for
maintenance, security auditing, and so forth.  I.e., if you're trying
to pitch Python for some project and convince management that it's
secure and adequately maintained, you now have to separately make the
exact same pitch for Spyce.  And you then have to pick out a database
interface layer and make another separate pitch for THAT.  With PHP,
pitch once, get a yes or no, and you're done.

And no, that's NOT a pointy-haired boss problem, those are real
issues, not bureaucrats being stupid.

> With the PSP support in mod_python, I'd argue that
> the bar has been lowered further.

I'm not familiar with this and will have to check into it further.

> But in the short to medium term, expect better application
> portability between frameworks - the need to run various popular
> applications in different environments is making itself increasingly
> evident.

You mean the problems of having a "multitude" of environments are
making themselves evident.  If Python is to have anything like PHP's
wide deployment for web apps, the maintainers have to pick one of
those environments (probably Zope, I guess), include it in the
official Python distribution and declare it to be under official
maintenance, and do the same for a database interface.  To first
approximation, if something isn't in the official distribution, it
doesn't exist.  

> > > We could bring the PHP ADODB, or many of the other numerous PHP APIs
> > > for DB connectivity into the picture, but which of these are the
> > > obvious choice, and officially endorsed?
> > 
> > That's precisely the situation Python is in now!
> 
> Really? At the fundamental level, the DB-API is pretty well
> standardised, and there's a collection of well-written modules that
> mostly adhere to it.

Nope.  Look in the Python distribution, there's nothing like that
there, and similarly in the Python reference manual.  Yeah, there's
some random third party modules and docs, but see above, those don't
exist ;-).



More information about the Python-list mailing list