Re: [Patch, fortran] PR25084, PR20852, PR25085, PR25086 & PR25416 - assumed character length functions

[Tobias SchlÃter <tobias dot schlueter at physik dot uni-muenchen dot de>]

Paul Thomas wrote:
>>This is ok, modulo two remarks.  Also, please make sure that the ChangeLog
>>lines are not longer than 78 columns (or whatever emacs' fill-paragraph makes
>>them), as that way you avoid wrapping in mail and in the editors of people who
>>don't use windows wider than 80 columns.
> 
> OK - is it 78 columns?  I use SciTe and can set the width to whatever I 
> want. You've made more than two remarks - which ones shall I take?

78 leaves two colums for quoting in e-mail, so that's probably the right
number.  Last time I looked I didn't find a documented number, so I jsut rely
on emacs getting it right.

>>Please add an assert that the callee is indeed an intrinsic. I think the
>>
>>comment is overly verbose, especially with the assert added, but I'm not
>>complaining.
>>
> 
> I tried that - for some reason attr.intrinsic is not set.  I'll work my
> way back to see if I can understand why.

Hm, that's weird.

>  I did not feel that the 
> comment was overly verbose - there are too many corners of gfortran that 
> are absolutely inexplicable for the lack of comments.  I thought that 
> here, where what is being done is not at all evident, a few words would 
> not come amiss.  I will take a fresh look at it.

I think it's sufficient to say that anything but an intrinsic would be an
error that should be caught in resolve.c, and leave the justification in
resolve.c.  But please let the comment in, I'm glad you're adding it, I just
wanted to add another perspective to the issue of commenting.

> This is the way that I had it in the first place.  I decided to reduce 
> the memory tied up in repeated strings!  Explain to me why this helps 
> internationalization, please; I feel an outbreak of ignorance coming on.

All I _know_ is that FX removed exactly these kinds of constructs in his
internationalization patch from 20050830.  I believe this has to do with
different sentence structures in different languages (e.g. the reason might
have to move to the beginnign of the sentence) and with the problem that the
same string may appear in different messages, and might require different
cases, say an adjective might have to be used in the feminine or masculine
case in two different contexts in the French translation.

- Tobi