Re: [Gcl-devel] mingw32 paths

[Camm Maguire]

Greetings!

Gabriel Dos Reis <address@hidden> writes:

> Camm Maguire <address@hidden> writes:
>
> | Greetings!  I've successfully tested maxima and acl2 under both wine
> | and a native windows machine.
>
> That sounds promising -- I'll get to your other message soon.
>
> | I have a question regarding paths.  From what I can tell, gcl needs
> | *to be run* under a mingw32 shell if it wants to compile anything.
> | Said shell reports paths as /c/dir/foo.c, but GCL's current truename
> | and probe-file expect c:/dir/foo.c.  What is 'supposed to happen'
> | here?  
>
> Please keep GCL's truename as it is.  The msys's UNIX-style /c/dir/foo.c
> is just convenience for UNIX-centric tools (especially build tools).
> However, since we want GCL to be built natively, it should have Windows
> version c:/dir/foo.c, otherwise trouble.
>
> BTW, you can also use "pwd -W" (only in msys) to report the windows
> style pathname for directories.
>
> Also, please, keep the path separator as '/' -- that requires less escaping.
>

OK.  Working on open-axiom now.  A few minor changes to compiler::link
are pending, but appear working.  I'm now stuck at the 'strap' stage
in src/boot.  The lisp.exe error is "Cannot find strap.o", and indeed
it is not there (or anywhere else apparently.)

Take care,

> -- Gaby
>
>
>
>

-- 
Camm Maguire                                        address@hidden
==========================================================================
"The earth is but one country, and mankind its citizens."  --  Baha'u'llah