6 ms·

The disqus comments on the page are entirely correct: it should come standard with python (or something like it). I've used it a few times and it's just elegant

by drats 15y ago

The disqus comments on the page are entirely correct: it should come standard with python (or something like it). I've used it a few times and it's just elegant and would match "batteries included" perfectly.

briancurtin 15y ago
A full web framework is not something you want to come standard with Python. Putting Bottle, Flask, Django, {your_favorite} inside the standard library would pretty much kill it. Development would screech almost to a halt since Python generally releases less often than any of these frameworks. For example, we released 3.2 in Spring '11. There will be no new features released until 3.3, which is slated to come out in Fall '12 - about 1.5 years. In comparison, the popular Flask framework is about 1.5 years old and has had ~8 point releases, and it can add features whenever it wants at whatever rate it wants. If Armin wants to make a new feature today, he can do that and release it today if he wants. On top of that, web frameworks are like a religion that we would have to choose. There's currently a separation of church and state that has worked well in the community and allowed different frameworks to do different things at different paces. I think we should keep that.
- andybak 15y ago
  I agree on you regarding a full stack web framework with it's own community such as Django but a lightweight, minimal web framework would be a useful addition to the standard library, surely?
  - briancurtin 15y ago
    Sure, it would be great, but whatever we consider would have to be very complete and very mature. There's not much room for evolution in standard library APIs, and with lightweight frameworks being relatively new to the scene, the framework developers themselves probably do not want to give up their current freedom to grow in whatever direction they please. If we look back at this in a few years, maybe it's worth considering as usage grows, APIs are more locked down, and a clear winner shines through. Right now, any inclusion would be months of bike shedding, assuming any of these framework developers are willing to give up external maintenance of their products and go 100% into the standard library.
  - rednaught 15y ago
    And end up with something like Perl's CGI? Request/Response favorite flavors seems to change every 2-3 years. I wouldn't want Python to suffer from that bloat.
  - sitkack 15y ago
    Putting something in standard lib effectively kills it. It doesn't mature, it turns into an adult child actor. Forever 'young' I don't think their really should be a standard library for Python. There should be a base, and a much more proactively revved base library, but not a standard library. Part of installing python would be installing the base.
    - wisty 15y ago
      I think the standard library is great. It's good to know that there is a "batteries complete" library that you can really on, that won't change its API for years. Being standard, it's well documented, well vetted, and lots of people know it. There needs to be a wider repo, but pip / easy_install does that OK. What does need to change - there needs to be more "library2" modules. urllib wasn't good enough, so they made "urllib2". Likewise, there's lots of old, outdated modules that could be supplanted.
      - sitkack 15y ago
        Brian, you are completely wrong. The x2 modules need to go away. They are confusing, the "standard lib" is mostly shit. And it is fixed in time. If it were writing I would describe it as a, "hack, rambling, incoherent, amateur" piece of work. The usability of pip is abysmal. You don't even get usage by just typing "pip" with no arguments. Look at brew for direction. Python makes me sad.
  - cageface 15y ago
    Better to have pip included as standard and make the framework of your choice a one-line install.
  - thristian 15y ago
    wsgiref is already in the standard library and is as about as minimal as you can get. Why add more?

by drats 15y ago

briancurtin 15y ago

A full web framework is not something you want to come standard with Python. Putting Bottle, Flask, Django, {your_favorite} inside the standard library would pretty much kill it. Development would screech almost to a halt since Python generally releases less often than any of these frameworks. For example, we released 3.2 in Spring '11. There will be no new features released until 3.3, which is slated to come out in Fall '12 - about 1.5 years. In comparison, the popular Flask framework is about 1.5 years old and has had ~8 point releases, and it can add features whenever it wants at whatever rate it wants. If Armin wants to make a new feature today, he can do that and release it today if he wants. On top of that, web frameworks are like a religion that we would have to choose. There's currently a separation of church and state that has worked well in the community and allowed different frameworks to do different things at different paces. I think we should keep that.

andybak 15y ago
I agree on you regarding a full stack web framework with it's own community such as Django but a lightweight, minimal web framework would be a useful addition to the standard library, surely?
- briancurtin 15y ago
  Sure, it would be great, but whatever we consider would have to be very complete and very mature. There's not much room for evolution in standard library APIs, and with lightweight frameworks being relatively new to the scene, the framework developers themselves probably do not want to give up their current freedom to grow in whatever direction they please. If we look back at this in a few years, maybe it's worth considering as usage grows, APIs are more locked down, and a clear winner shines through. Right now, any inclusion would be months of bike shedding, assuming any of these framework developers are willing to give up external maintenance of their products and go 100% into the standard library.
- rednaught 15y ago
  And end up with something like Perl's CGI? Request/Response favorite flavors seems to change every 2-3 years. I wouldn't want Python to suffer from that bloat.
- sitkack 15y ago
  Putting something in standard lib effectively kills it. It doesn't mature, it turns into an adult child actor. Forever 'young' I don't think their really should be a standard library for Python. There should be a base, and a much more proactively revved base library, but not a standard library. Part of installing python would be installing the base.
  - wisty 15y ago
    I think the standard library is great. It's good to know that there is a "batteries complete" library that you can really on, that won't change its API for years. Being standard, it's well documented, well vetted, and lots of people know it. There needs to be a wider repo, but pip / easy_install does that OK. What does need to change - there needs to be more "library2" modules. urllib wasn't good enough, so they made "urllib2". Likewise, there's lots of old, outdated modules that could be supplanted.
    - sitkack 15y ago
      Brian, you are completely wrong. The x2 modules need to go away. They are confusing, the "standard lib" is mostly shit. And it is fixed in time. If it were writing I would describe it as a, "hack, rambling, incoherent, amateur" piece of work. The usability of pip is abysmal. You don't even get usage by just typing "pip" with no arguments. Look at brew for direction. Python makes me sad.
- cageface 15y ago
  Better to have pip included as standard and make the framework of your choice a one-line install.
- thristian 15y ago
  wsgiref is already in the standard library and is as about as minimal as you can get. Why add more?