RE: [boost] uBLAS formal review

From: Paul A. Bristow (boost_at_[hidden])
Date: 2002-07-08 01:14:45

I have also seem premature EOF reported by the compiler.

Perhaps the mixture of terminations comes from editing the file provided to
carry out more test,
thus producing a mixture?

My first compile failures came from deleting some of the excessive (for my
taste) blank lines in Jorge's code to try to get my head round how it worked
(I am still in partial disambiguation mode!)

As I recollect (but I could be wrong), this was on code received via a zip
not CVS.

Whatever the cause (and it might be me), I believe a release should be free
of these problems.

Paul

> -----Original Message-----
> From: boost-bounces_at_[hidden]
> [mailto:boost-bounces_at_[hidden]]On Behalf Of Michael Stevens
> Sent: Saturday, July 06, 2002 1:41 AM
> To: boost_at_[hidden]
> Subject: RE: [boost] uBLAS formal review
>
>
>
>
> >From: "Paul A. Bristow" <boost_at_[hidden]>
> >
> >The reason is that many files contain CR, CR (0x0D, 0x0D)
> >rather than CR, LF ( 0x0D, 0x0A) in quite a few places.
> >These are invisible to the MS IDE editor and Textpad,
> >except in binary display mode. Linux/Microsoft difference?
> >
> >After much scratching of my head,
> >and eventually removal of tuffs of hair,
> >I conclude this produces terminal confusion to the compiler,
> >and to me!
> >
> >
> This is most odd. I can certainly confirm that although the VC7 editor
> has no problem with CR only line termination the compile seems sometimes
> to treat them as an EOF.
> The odd thing is I have never seen CR only (the CR CR is infact 2 lines)
> line termination before,. LF only is the usual UNIX format, CRLF the old
> DOS format. The compilers work fine with either of these but sequences
> of CR only are bad news.
>
> The other odd thing is that I have no problems with the uBLAS files I
> pulled from boost CVS. Infact they seem to be CRLF terminated. Maybe
> this is due to some kind of CVS configuration problem.
>
> Michael
>
>
> _______________________________________________
> Unsubscribe & other changes:
> http://lists.boost.org/mailman/listinfo.cgi/boost
>
>