Re: Attempting (and failing) to compile natbib branch

[Lars Gullik Bjønnes]

Allan Rae <[EMAIL PROTECTED]> writes:

| Dear me-too's,
| 
| You all use RH7.0 or RH7.1.  Mike in his email stated he was using Mdk8.0
| just as I currently am.  The gcc-2.96-0.48 included therein ICEs on a
| number of files in the LyX source tree and as (I think) John pointed out
| they seem to be using a something even older than RH's 2.96-69.

So Mdk has done an even worse job than RH, since thay havent even
updated the compiler...
 
| So not only can we not easily identify minor numbers for versions (on RH
| systems where there seems to be at least 4 different official updates) but
| we also would need to identify distribution.  This whole nightmare -- call
| it the "gcc-2.96 nightmare" -- just gets worse and worse.

We could enlarge the test a bit, so that instead of (for 2.96 only)
just looking at output from --version we look at output from -v as
well.

| Hence my suggested configure warning message.   John,  such a message
| would be output at the end of the configure run, just like the
| xforms-not-found or Xpm-not-found messages are.

I guess we might need it.

-- 
        Lgb