Thanks, Earnie, I'll keep working on it. I'm trying to put together a
science/math "exporation" tool where I can do some electromagnetics and
stuff... bessel fcns and fem codes, etc. Just getting really sucked in to
Python right now, and on the c front I'm thinking I might be able to get
away with scipy's so kindly provided gcc3.2.3 ATLAS_LAPACK precompiled XP P4
dot-a's! - should be pretty good match for my machine ? (sorry for the
aside, but is there a way to convert the .a's to "shared"
library/dll?-should I be using a shared library on XP box?).
I really do want to learn to use the MSYS/MinGW to compile from source,
maybe somehow the line ends are getting messed up along the way, I'll look
for it....but I'm still confused because the error line
sed: file ./confstat2636-30293/defines.sed line 24: Unterminated `s'
seems to be sed complaining about a defines.sed file having an error it on
line 24, but I know of no confstat2636-30293 directory (supposedly in ./ -
the current directory) in the first place! Is not ./ the directory I am
in/running configure from? There is no such dirctory I know of and I think
I was wrong - it probably never is ever created...huh.
Anyway, Thanks Again and Best Rgds, Rob
----- Original Message -----
From: "Earnie Boyd" <earnie@...>
Sent: Tuesday, March 08, 2005 6:28 AM
Subject: Re: [Mingw-msys] the unterminated s-command sed error!
> <quote who="Robert Wenzel">
>> Earnie and Greg,
>> Earnie, are you for sure convinced I have a end-of-line problem? I had
>> originally run any dos<>unix on it at first, only in trying to fix the
>> problem. And I tried 7-zip to see if that made a difference but it
> I know that the MSYS sed has issues with sed command files ending with
>> Would it be true that during run of configure a temp file with the
>> command is created then deleted?
> If the sed command file is generated, how is it generated? If it uses a
> program that is win32 native the the possibility exists that \r\n line
> endings are in fact stored in the file. You mention configure deleting
> the temp file, you can modify the configure file to remove the ``rm -f''.
>> It seems so and if so then I couldn't
>> you fix on the non-existent file? Might I need to modify the
>> file to fix it or something?
> You need to determine how the sed command file is created. Once we know
> that I can help more.
>> I had similar problems trying to compile
>> and someone told be I HAVE to get cygwin but that was more intense... so
>> it is seeming very hard to use configure scripts, I can't get anything to
>> configure - maybe I should just hack the makefiles but it seems some
>> important system compile flags won't be set then... maybe I'm too newbie
>> still! Maybe the old meschach configure.in file is bad? I did download
>> source code for sed (ironically) just to try configure and it
>> made the sed makefile(s) (although I think it complained a little), so
>> again, maybe it's the old configure.in... might I need an older autoconf
>> that is compatible with it? - has the configure.in syntax changed since 7
> Yes, the syntax is different in the new versions of autotools and
> consideration for old syntax is not given.
>> Thanks Again for any more hints, I really want to use the minimalist
>> approach.. Do I really need cygwin to compile most source packages as
>> guy told me? -
> I sure hope not! Besides, MSYS is really an older version of Cygwin just
> tweaked to my liking for the purposes of MinGW.
> ? Are my expectations just off - is it alot of modification
>> tweaks and hacking to do the porting with MSYS with my little experience
>> this area? - Perhaps I keep scraping around for binary distributions or
>> do I graduate into the "compiled it myself" category?
> No the expectations are not off. You complete your mission one step at a
> SF email is sponsored by - The IT Product Guide
> Read honest & candid reviews on hundreds of IT Products from real users.
> Discover which products truly live up to the hype. Start reading now.
> Mingw-msys mailing list