Comments on: Scherprechter neemt poolshoogte http://4.flowsnake.org/archives/914 A Pythoneer's adventures with Scheme, etc. Mon, 18 Jul 2011 14:37:45 +0000 hourly 1 http://wordpress.org/?v=3.3 By: Juho Vepsäläinen http://4.flowsnake.org/archives/914/comment-page-1#comment-820 Juho Vepsäläinen Wed, 03 Feb 2010 16:32:41 +0000 http://4.flowsnake.org/?p=914#comment-820 No language is perfect. I think Python manages to come up with a decent compromise. It's almost like pseudocode. :) I think the deal with encapsulation is that some people are really insistent on their getters and setters (perhaps this has something to do with Java?). I find it much cleaner to just use properties and a naming scheme ("_" = protected, visible to subclasses, "__" = private, hidden). I know you can easily bypass a naming scheme but that's not the point. It's about communicating the interface. It's possible to get rid of some "warts" by writing a little precompiler, by doing some bytecode hacking (search or "selfless python" for instance) or just by metaprogramming. There are also interesting projects like EasyExtend to make certain modifications easier. No language is perfect. I think Python manages to come up with a decent compromise. It’s almost like pseudocode. :)

I think the deal with encapsulation is that some people are really insistent on their getters and setters (perhaps this has something to do with Java?). I find it much cleaner to just use properties and a naming scheme (“_” = protected, visible to subclasses, “__” = private, hidden). I know you can easily bypass a naming scheme but that’s not the point. It’s about communicating the interface.

It’s possible to get rid of some “warts” by writing a little precompiler, by doing some bytecode hacking (search or “selfless python” for instance) or just by metaprogramming. There are also interesting projects like EasyExtend to make certain modifications easier.

]]> By: mark http://4.flowsnake.org/archives/914/comment-page-1#comment-819 mark Wed, 03 Feb 2010 11:50:23 +0000 http://4.flowsnake.org/?p=914#comment-819 Couple of minor points: The httplib time out thing *has* caught me - but like only as far as "oh this needs a timeout. How do I do it?" I think you just need to set a timeout on socket before you start - no big deal. What have you got against "join" - such a useful function for a task I seem to do all the time. What intrigued me is what you mean by "ultra-dynamic" languages - can you name one or two to give me an example. (or write the blog post implied by "(but that is a different story…)" :-) Couple of minor points:

The httplib time out thing *has* caught me – but like only as far as “oh this needs a timeout. How do I do it?” I think you just need to set a timeout on socket before you start – no big deal.

What have you got against “join” – such a useful function for a task I seem to do all the time.

What intrigued me is what you mean by “ultra-dynamic” languages – can you name one or two to give me an example. (or write the blog post implied by “(but that is a different story…)” :-)

]]> By: Charles McCreary http://4.flowsnake.org/archives/914/comment-page-1#comment-815 Charles McCreary Wed, 03 Feb 2010 04:37:41 +0000 http://4.flowsnake.org/?p=914#comment-815 Hans, The guy obviously doesn't understand much about dynamic languages and the benefits thereof. Python is not the perfect solution for every problem, but unless I'm coding high performance FEA routines and their ilk or Linux device drivers/modules, Python is the first tool I reach for. Hans,
The guy obviously doesn’t understand much about dynamic languages and the benefits thereof. Python is not the perfect solution for every problem, but unless I’m coding high performance FEA routines and their ilk or Linux device drivers/modules, Python is the first tool I reach for.

]]>