Oh Yeah! Where's Your Patches Zed!?
I admire Guido quite a lot, he’s definitely much more successful than myself at software project management and probably just about everything he does with his life. Yet when I read comments like this:
You seriously think I’d complain about something without actually having any experience working with it or attempting to replace it?
Alright Guido, you want patches, I’ll give you some patches. Try my current project for starters.
I’ve already got gear for nearly everything I complained about, and the more I work the more I’ll eventually replace all of the crap I have to use from Python while maintaining the good stuff. Give me six months and I’ll have the following:
- Totally new, very well written, complete SMTP server replacing the smtpd lib entirely. It’s gonna die because I’ll run into problems with it.
- A complete rewrite of the horrible optparse crap everyone uses. Why the hell does a dynamic language like Python have a C style option parser?
- A complete and easy to use email generation and processing system to replace the bizarro design decisions in email and friends.
- A standard that templating libraries will follow so that guys like me can have the “WSGI of Templating”.
- Libraries and possibly python patches for doing efficient state machines that will work with generators.
- A complete replacement for the scanner library.
I will make a long bet right now, that no matter how nice I am, no matter how many people use my software, no matter how much better it is than anything Python has that’s similar, Guido will not accept them as changes.
This is the history of the Python project. This is fine, all open source projects are like that. I can’t fault them for it since everyone does it, even me to a certain extent.
What is dumb about Python though is they demand higher quality input from critics than they do from contributors in their club. They say someone like me should have patches (totally ignoring all my work on email of late), go through the PEP process, but their buddy creates a crap ass install tool and it’s accepted with open arms even though it is totally and completely broken for an obvious reason.
So Guido, with much respect and in the spirit of debate, I will submit patches and work from my Lamson project on two conditions:
- You apply the same demands of quality on all contributors by rejecting any solution that does not have consistency and symmetry.
- You acknowledge that the email systems in Python need an overhaul and that my work will seriously be the basis of the effort.
Otherwise, why should I bother when I’m held to a higher standard then regular contributors who create obviously broken software, and when there’s a high probability all of my work will be for nothing?
