Monday, August 6, 2007

Pylons - first impressions

If you didn't hear there's a new python framework called pylons.

It's aimed to be a competitor for django or ruby on rails. Last week I played with it a little bit - read some docs, did some tutorials.

My first impression was really positive - pylons tries to utilize already existing products in python world, they don't invent new things on their own (as django does). With clean design and python as a backend I thought it could be useful for me.

But it seems that pylons it's still in early stages of development - there are some documents that address basic stuff. But then I tried to add something crucial for me - authentication and authorization to tutorial application.

I was surprised that it didn't work :-) I found authkit, read about. I downloaded some still-in-development-version that was supposed to be usable (the old version was to different compared to documentation to try with it). So I managed to get a login page (after a lot of time struggling with documentation and code). But I didn't manage it to work - I could authenticate but I was redirected to wrong page.

I tried to issue a bug report but issue tracker wanted me to login. I managed to found credentials that allowed anonymous users to issue bugs but when I tried to use them the issue tracker responded with error. Just great! I mailed administrator but got no response for few days.

That gets me thinking if all those frameworks are usable at all. Maybe it's better idea to stick to some known, maintained (and maybe not so great and easy to use) Java framework.

What do you think? Is there anything stable and usable on the market that's not on Java or .NET? By usable I mean - easy object persistence, MVC, web support (dynamic flows, AJAX, web authentication and authorization with pluggable security providers), easy to use and learn, with good deployment model (production ready web servers), web services support (SOAP, REST)? Anyone? Point me into the right direction and I'll follow :-)


  1. Pawel,

    Here is my long comment, but don't count on my help.

    I understand your concerns, ... but:
    - there was some clever person who wrote Ruby and now JRuby is "trendy"...
    - there was some clever person(s) who started Spring and now JEE in fact is Spring...
    - there was some clever person who invented Java and now it is supposedly the most popular general-purpose language (at least outside Microsoft :) )...
    - there was a man called Larry who long time ago wrote Perl. I dare say my colleague admins and Linux hackers would kill me if I said that Perl is stupid. They cannot live without it.

    So what I am trying to say: technologies, languages, frameworks change over time. They appear, majority falls into oblivion, some are embraced and become very popular but finally all (or almost all - see embedded word) vanish.

    If you just want to keep up with this pace you have to learn and use new technologies. Even more: if you want to be recognized or just make a lot of money as a developer you have to be brave and "descent into waters" unknown by others are rarely visited. I think that if you have lot of patience, lot of knowledge and lot of lack - then you may become another Mr. Gosling, Wall, Matsumoto or Johnson.
    Just by following the trends and exploring already explored terrains you will be always maximum in the second rank.

    So: if you feel that some framework (maybe Pylons) makes sense and there is a lot of things to do to make it robust - just spend your time (it will be only more difficult once you are married or have children...) and contribute...

    Easier said...


  2. What do you think of the Python TurboGears web application framework? It looks like a reasonable competitor to Rails and my first impression is that it feels much more agile to work with than ASP/.NET for doing the same work.

    And I'm really sorry to hear about the layoffs. That really sucks.

  3. Well, I haven't used it (neither heard of it). I'll try it in my free time and I should have a lot of free time now ;-)))

    Yeah, that's a bummer but we've expected this.

  4. True, the whole efficiency thing and layoffs were predictable and unfortunate. I'm surprised your whole team got it, though, I'll admit I didn't think *that* was going to happen. Best of luck!