Re: [TuxKart-devel] fixes

[James G. <j....@vi...>]

On Wed, 2004-10-20 at 21:57, Red Zebra wrote:
> Hi,
> 
> Not sure this is the right channel but I just compiled the latest CVS.
> I had to made some small changes for compatiblility in order to
> compile with 2.95.3. (running gentoo)  None of these should break 
> anything and are very straight forward
> 
> If someone would be so nice to add them or let me doit that would be
> nice before the next official release.
> 
> So here I go
> 
> --------------------------------------------------------------
> /cvsroot/tuxkart/tuxkart/src/KartManager.cxx
> /cvsroot/tuxkart/tuxkart/src/RaceManager.cxx
> /cvsroot/tuxkart/tuxkart/src/StringUtils.cxx
> /cvsroot/tuxkart/tuxkart/src/World.cxx
> 
> had to add :
>          #include <algorithm>

Committed...

> ------------------------------------------------------------
> /cvsroot/tuxkart/tuxkart/src/start_tuxkart.cxx
> replace: and by &&
> replace: or  by ||

Committed...

> --------------------------------------------------------------
> /cvsroot/tuxkart/tuxkart/src/StringUtils.cxx
> 
> <     return lhs.compare(lhs.length() - rhs.length(), rhs.length(), rhs) ==
>  >     return lhs.compare(rhs, lhs.length() - rhs.length(), rhs.length()) ==
> 
> none of the compare methods seems std++ ?

Not committed (at least by me, someone else might commit it) because
according to "The C++ Programming Language" compare can take:

const basic_string&
const Ch*
size_type, size_type, const basic_string&
size_type, size_typ[e, const basic_string&, size_type, size_type
size_type, size_type, const Ch*, size_type

Currently it uses the 3rd one, your one doesn't seem to be there, so I'm
not quite sure if that would fix more problems than it would create?

> --------------------------------------------------------------
> /cvsroot/tuxkart/tuxkart/src/lisp/Lexer.cxx
> 
> std::streamsize n = stream.readsome(buffer, BUFFER_SIZE);
> 
> This is a bit different. readsome is not availabel in older
> versions so I replaced it with
>          stream.reads(buffer, BUFFER_SIZE);
>          std::streamsize n = stream.gcount();
> Which should be exactly the same. However if that's realy the case then
> there might be an error in the 2 places a bit further where there
> is an check fo stream.eof. because then stream.eof would be true
> before c is at the end of the read buffer. (stream is already read
> and reached the end but there's still data there.) So I
> had to put the stream.eof check to ( while(1) and if (0) )

I'll leave someone else who understands that code to deal with that.

Thanks,
	
	James