Comments on: Marginal Added Value http://tleaves.com/2008/04/14/marginal-added-value/ Creativity x Technology Sat, 17 Mar 2012 05:09:58 +0000 hourly 1 http://wordpress.org/?v=3.3.1 By: Doug http://tleaves.com/2008/04/14/marginal-added-value/comment-page-1/#comment-4825 Doug Wed, 16 Apr 2008 17:58:01 +0000 http://tleaves.com/2008/04/14/marginal-added-value/#comment-4825 I believe Faisal has a typo and meant to say Real Old programmers use FORTRAN. At least at my last job that was the case. Company spent loads of money converting these sims to C. Meanwhile old men refused to learn C and modified the sim in FORTRAN. I believe Faisal has a typo and meant to say Real Old programmers use FORTRAN. At least at my last job that was the case. Company spent loads of money converting these sims to C. Meanwhile old men refused to learn C and modified the sim in FORTRAN.

]]>
By: peterb http://tleaves.com/2008/04/14/marginal-added-value/comment-page-1/#comment-4826 peterb Tue, 15 Apr 2008 23:29:16 +0000 http://tleaves.com/2008/04/14/marginal-added-value/#comment-4826 Real programmers learn every language that appeared as level names in the TRON arcade game. IN ORDER. <small>(RPG, COBOL, BASIC, FORTRAN, SNOBOL, PL1, PASCAL, ALGOL, ASSEMBLY, OS, JCL)</small> Real programmers learn every language that appeared as level names in the TRON arcade game. IN ORDER.

(RPG, COBOL, BASIC, FORTRAN, SNOBOL, PL1, PASCAL, ALGOL, ASSEMBLY, OS, JCL)

]]>
By: Faisal http://tleaves.com/2008/04/14/marginal-added-value/comment-page-1/#comment-4828 Faisal Tue, 15 Apr 2008 17:43:35 +0000 http://tleaves.com/2008/04/14/marginal-added-value/#comment-4828 Real programmers use FORTRAN. Real programmers spell FORTRAN in all caps. Real programmers use FORTRAN. Real programmers spell FORTRAN in all caps.

]]>
By: Doug http://tleaves.com/2008/04/14/marginal-added-value/comment-page-1/#comment-4827 Doug Tue, 15 Apr 2008 16:41:34 +0000 http://tleaves.com/2008/04/14/marginal-added-value/#comment-4827 The reason I use python is because it is very easy to design test programs for embedded software. It has a good serial port interface, and decent math support. And I can run python in a shell, trying different commands until I get what I want. This combination makes it invaluable for me when developing an interface to a dsp. Once the kinks are ironed out then I move to c# or something "real" but that initial development is so much easier in python. Maybe other languages can do that too but I chose python because it's serial port worked the same in windows and in linux. So far the only other language I've found that can say that is c#. The reason I use python is because it is very easy to design test programs for embedded software. It has a good serial port interface, and decent math support. And I can run python in a shell, trying different commands until I get what I want. This combination makes it invaluable for me when developing an interface to a dsp. Once the kinks are ironed out then I move to c# or something “real” but that initial development is so much easier in python. Maybe other languages can do that too but I chose python because it’s serial port worked the same in windows and in linux. So far the only other language I’ve found that can say that is c#.

]]>
By: psu http://tleaves.com/2008/04/14/marginal-added-value/comment-page-1/#comment-4824 psu Tue, 15 Apr 2008 01:11:51 +0000 http://tleaves.com/2008/04/14/marginal-added-value/#comment-4824 FWIW, perl 5 has decent namespace management and a sort of module/object system. It's a little crufty, but so is Python. I've worked on 20-30K lines of code on teams of three or four without much more suffering than with a "real" language like C or Objective C. You just have to be careful about not writing code that is intentionally obscure or too clever. That's about as big as I'd go. But that's also about as big as I'd go with anything in this class and nothing you say can convince me that any of the current script languages will really be any better at that scale. FWIW, perl 5 has decent namespace management and a sort of module/object system. It’s a little crufty, but so is Python. I’ve worked on 20-30K lines of code on teams of three or four without much more suffering than with a “real” language like C or Objective C. You just have to be careful about not writing code that is intentionally obscure or too clever.

That’s about as big as I’d go. But that’s also about as big as I’d go with anything in this class and nothing you say can convince me that any of the current script languages will really be any better at that scale.

]]>
By: Mike Collins http://tleaves.com/2008/04/14/marginal-added-value/comment-page-1/#comment-4823 Mike Collins Tue, 15 Apr 2008 01:05:12 +0000 http://tleaves.com/2008/04/14/marginal-added-value/#comment-4823 So, I'm a python nazi. One of the major reasons that I'm a python nazi is because I expect that the scripts we write in it will not only be used, but read and, in a lot of cases, used to teach people how to use our other tools. I would seriously argue that many of the features that make perl attractive as a quick-bash-em-out language also make it totally unsuitable for developing a nontrivial program (that, as far as I can tell, everything in perl is shoved into one hungus namespace being part of it). If people stopped writing perl code at, say, 300 lines that'd be fine. But they don't. I figure every program over 5 pages becomes mission critical for somebody, and when that happens, we're left with line noise to decipher. So, I’m a python nazi. One of the major reasons that I’m a python nazi is because I expect that the scripts we write in it will not only be used, but read and, in a lot of cases, used to teach people how to use our other tools. I would seriously argue that many of the features that make perl attractive as a quick-bash-em-out language also make it totally unsuitable for developing a nontrivial program (that, as far as I can tell, everything in perl is shoved into one hungus namespace being part of it).

If people stopped writing perl code at, say, 300 lines that’d be fine. But they don’t. I figure every program over 5 pages becomes mission critical for somebody, and when that happens, we’re left with line noise to decipher.

]]>
By: Andrew Plotkin http://tleaves.com/2008/04/14/marginal-added-value/comment-page-1/#comment-4822 Andrew Plotkin Mon, 14 Apr 2008 20:02:24 +0000 http://tleaves.com/2008/04/14/marginal-added-value/#comment-4822 xkcd386. Sorry. xkcd386. Sorry.

]]>
By: Jon F http://tleaves.com/2008/04/14/marginal-added-value/comment-page-1/#comment-4821 Jon F Mon, 14 Apr 2008 19:26:25 +0000 http://tleaves.com/2008/04/14/marginal-added-value/#comment-4821 I actually think that one-key access to hashtables (that key being Shift, to change [] to {}) is even more important than =~ for getting perl scripts onto the road quickly. I love Lisp, but it fails on both of these features. I actually think that one-key access to hashtables (that key being Shift, to change [] to {}) is even more important than =~ for getting perl scripts onto the road quickly. I love Lisp, but it fails on both of these features.

]]>
By: DavidR http://tleaves.com/2008/04/14/marginal-added-value/comment-page-1/#comment-4829 DavidR Mon, 14 Apr 2008 16:53:55 +0000 http://tleaves.com/2008/04/14/marginal-added-value/#comment-4829 I actually want there to be a special team involved here. Picture a movie set and a bunch of stereotypical tough guys of varying stripes. Their commander says "Somewhere out there, a new scripting language is being developed. Your job is to stop it." Why? Because I am lazy and afraid that one of the other ones will be so good that I will have to spend the time to learn it. I actually want there to be a special team involved here. Picture a movie set and a bunch of stereotypical tough guys of varying stripes. Their commander says “Somewhere out there, a new scripting language is being developed. Your job is to stop it.”

Why? Because I am lazy and afraid that one of the other ones will be so good that I will have to spend the time to learn it.

]]>