[DOC] clearer description of -std= (try 2)

[Manuel López-Ibáñez]

Given the comments on the previous attempt [*], I think we agree the
description of -std= is confusing. So this is a second try. I tried
to:

* Clarify that -ansi is not more than an alias of -std=c89 or
-std=c++98. Thus, the explanations of -std= apply to -ansi (and, to
some extent,  the other way around).

* Make evident that -std=c89 does not disable all gnu extensions.

* Make evident that -std=gnu89 may reject or change the meaning of
valid ISO C90 programs.

* Explain why we have -std=gnu89 and -std=gnu99.

* Move explanations to where people will look, that is, the paragraph
after -std=.

I have avoided words like "subset", "extended". I didn't avoid "GNU
extensions" because that is a name we use pervasively (although I
tried to use "features") as much as possible.

Please, let me know if I got something wrong, or something is worse
than before. It can be done better but I honestly think this is an
improvement over the current text.

I would like this to go in GCC 4.3.

Thanks.

008-01-17  Manuel Lopez-Ibanez  <manu@gcc.gnu.org>

  PR c/28368
  * doc/invoke.texi (-std): Clarify description of -std= and -ansi.

[*] http://gcc.gnu.org/ml/gcc-patches/2008-01/msg00745.html
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: pr28368.diff
URL: <http://gcc.gnu.org/pipermail/gcc-patches/attachments/20080118/dbdf1914/attachment.ksh>