more than one way to do it

[Aaron Swartz]

> The problem is that we still don't have one obvious way to build web
> applications in Python -- and we may never. But the proliferation of
> (often, barely indistinguishable) options makes life harder for the
> Pythoneer.

I'm hoping that this may become the obvious way, by making it the
Pythonic way. Maybe the other frameworks have that as their goal as
well, though. I wish I knew a solution to this problem.

[tyler]

There isn't one.

But you're still doing the right thing. You're giving people options.
"the proliferation of .....options makes life harder or the Pythoneer",
is the most absurdley naive thing I've ever heard from such a
progressive community. It's not "This is how it will be" its "What
works for you?"

I love Ruby, but Rails frightens me to death. There's others. Catapault
is nice, Camping is better (and remarkably similar to web.py) but this
(web.py) just feels right to me. The guy sitting next to me disagrees,
he thinks Rails is beautiful. What, you're going to deprive him of his
pleasure for the sake of "the one true way"? Sounds like a whole lot of
egoistic crap that would inevitably lead you nowhere because people
will build it if it doesn't exist. Sure, you might create "the pythonic
way" but realize the fundamental truth - you can't please everyone.
Plan on it. Build your own army and do what works for you.

Aaron, you're contributing to the solution to this problem. You're
giving people options. What's the wohle solution? The whole solution is
your individual effort, taken on a limit across all the developers of
the world and averaged out. The solution is "whatever works for you" -
if we all have our own framework that we couldn't live without, then
we've solved the problem.

[David Terrell]

On Thu, Feb 23, 2006 at 10:27:19AM -0800, tyler wrote:
> But you're still doing the right thing. You're giving people options.
> "the proliferation of .....options makes life harder or the Pythoneer",
> is the most absurdley naive thing I've ever heard from such a
> progressive community. It's not "This is how it will be" its "What
> works for you?"

I'm trying to understand and explain the complaints. If you reread
what I said, I'm not saying web.py is wrong, just why people keep
saying "do we really need a 19th new templating api?"

As a python community, we're still in the experimental phase of web
frameworks. I hope that at some point there's some sort of shakeout,
where there's at most a couple of commonly used frameworks that serve
clearly different niches. There's so many different O/R mappings,
database APIs, templating languages, frameworks, metaframeworks, session
managers and form handlers. Why? Why is the right answer to "web.py
is almost exactly what I want but there's a and b that I don't like",
to go ahead and create your own that's almost exactly the same?

[tyler]

>> But you're still doing the right thing. You're giving people options.
>> "the proliferation of .....options makes life harder or the Pythoneer",
>> is the most absurdley naive thing I've ever heard from such a
>> progressive community. It's not "This is how it will be" its "What
>> works for you?"

>I'm trying to understand and explain the complaints. If you reread
>what I said, I'm not saying web.py is wrong, just why people keep
>saying "do we really need a 19th new templating api?"

I did read what you said. If you reread what I said, I'm not accusing
you of saying web.py is wrong. I'm just saying I think it's a
bit...unnatural to want to say "No more frameworks/templating!"

I do understand that parenthetical "barely indistinguishable" part, but
what are you going to do? Force people to merge? What if its a
difference thats horridly subtle? Like the use of "+" vs "," ? Some
people get really religious about that stuff and they'll do what they
want because thats..just...the way...people do it. Yea, its frustrating
to navigate those differences, but I'd rather have those subtle
differences then web framework communism.

That's the right answer because everyone works differently and I'll
guarentee it - you push 3 or 7 frameworks on people and someone's bound
to come around and shake it up.

But you said one word that makes me very, very happy: metaframeworks.
Frameworks that build frameworks? There ya go. Then the answer really
is "create your own thats almost exactly the same", and you know why
because you suggested it. ;)

I don't mean to pick - I love how tight-knit everything is around here.
But a lot of it seems counter-productive to me. Let the man make a
19th, 20th and 40th templating API. If the last one gives someone else
a brilliant idea and they create the 41st that's even better then so be
it! What ever happened to the more the merrier? So do we need a 19th
new templating api? If you wrote it and I loved it, I'd give you a
large smile and name my children after you. You'd be my hero and I'd
devote myself to the betterment of you're templating API.

Maybe that's just me. But if other people liked it I'm sure that they'd
show gratidude in similar fashions.

I don't know. I'm not really deep in the python community, but I can't
imagine that developers would be suffering from "option overload" and
people would want to be limiting themselves like this.