|
From: | Robert T. Short |
Subject: | Re: Removing broadcasting from Octave |
Date: | Wed, 14 Dec 2011 14:50:47 -0800 |
User-agent: | Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.9.1.4) Gecko/20091017 SeaMonkey/2.0 |
Juan Pablo Carbajal wrote:
Well, you have the option of not using things that don't translate automatically...On Wed, Dec 14, 2011 at 10:58 PM, Robert T. Short <address@hidden> wrote:Przemek Klosowski wrote:On 12/14/2011 03:55 PM, Juan Pablo Carbajal wrote:The idea of having an easily deactivated warning when about bsx is perform is not a bad idea...of course standard code gets verbose output :(MATLAB_ME_HARDER? (in a way of explanation, Richard Stallman, who didn't like POSIX much, proposed an environment variable POSIX_ME_HARDER that made several GNU utilities behave in a POSIX-mandated way that Stallman didn't like (reporting disk space in 521B units instead of 1kB, etc). People found it a little too 'geek/offensive' so it's called POSIXLY_CORRECT now :)Are we worrying too much about translation to MATLAB? Easy translation is absolutely critical for me, but I think about it as I write code. If the manual has a warning then I consider myself warned. BobRobert has a point...I guess I will have to go through my functions at the time of collaboration and check if they are all matlabotomized. Guess an entry in the manual should be enough for humans (and we say good bye to automatic translation (basically search and replace) :( ).
If a customer really wants MATLAB (and most do) I work in octave and deliver MATLAB. That means I have to be cool about what features I use. It is my library functions that I really struggle over, but there is always a solution.
These features that really break compatibility are worth this kind of discussion and soul-searching, but this one is a powerful extension with lots of potentially dangerous side effects. It seems to me that the community is pretty split on the whether it should be permanent or not, and therefore the argument against is no stronger than the argument for. It seems to me that without an overwhelming argument against, we should include it along with a warning in the manual.
Bob
[Prev in Thread] | Current Thread | [Next in Thread] |