Jim Kleckner wrote:
Jim Kleckner wrote:
The attached patch for gnuplot speeds it up on Cygwin
by around 2.5x due to unbuffered reads for all file types.
I will submit it upstream as well.
See this link for the original discussion:
http://tinyurl.com/6c4sse
http://thread.gmane.org
Jim Kleckner wrote:
The attached patch for gnuplot speeds it up on Cygwin
by around 2.5x due to unbuffered reads for all file types.
I will submit it upstream as well.
See this link for the original discussion:
http://tinyurl.com/6c4sse
http://thread.gmane.org
The attached patch for gnuplot speeds it up on Cygwin
by around 2.5x due to unbuffered reads for all file types.
I will submit it upstream as well.
See this link for the original discussion:
http://tinyurl.com/6c4sse
http://thread.gmane.org/gmane.comp.graphics.gnuplot.devel/2189/focus=2207
d
I have XP, links is not the problem...
I have mingw installed, this is not the problem...
I included such a simple example, yes, I broke down and subscribed to
[EMAIL PROTECTED]'s mailing list, wonder if they'll be any more willing to
listen
- Original Message -
From: "Jim" <[EMAIL PROTECTED]>
To: <[EMAIL PROTECTED]>
Sent: Sunday, December 15, 2002 11:15 AM
Subject: Gcc 3.2 -mno-cygwin
> Someone broke GCC somewhere
>
> echo 'int main( void ) { return 1; }' >test.c
&g
back to 2.95.3 or whatever...
Jim
What's worse - is after it makes the directory, for some reason, the
existing tmp.lnk doesn't work anymore, so I have to delete it, and then
remake it... then I can resume my common practice of cp mtmp.lnk tmp.lnk
cp ztmp.lnk tmp.lnk ... etc... and those work...
>
> His request was to _not_delete_a_symlink_he_made_himself and replace it
> with a directory named /tmp.
>
It doesn't delete the link, it just makes a directory
> I assume you mean as a link to a windows temp directory somewhere.
However,
> what if that gets changed by someone? Suddenly a lot of stuff starts
> breaking, and people start posting here asking why application foo isn't
> working. The majority of the people are fine with a hardwired /tmp
> di
Would be really really nice if /tmp existed as a
link that a directory /tmp weren't created after running
setup...
y
resembles the message queues... not even constructing atoms and using
postthreadmessage... since atoms are limited to 255 ascii characters
(terminating '\0' and all)
Jim
ystem
without a great deal of effort... the cygipc library would have been this,
but I found that it's implementation of SYSV message queues to be somewhat
flawed...
Jim
> Hi,
>
> > >>/etc/hosts -> ${SYSTEMROOT}/system32/drivers/etc/hosts
> >
> > So, like all great ideas, all that it needs is an actual patch to make
> > it happen.
>
Uhh - what about just /windows/hosts (for those that are non NT based) ?
> Ok, I will have a go at this. As it involves mods to the
Ignore prior messages - was using old version of make/cygwin1.dll from tools
directory and then sh.exe from the cygwin install...
Apologetically.
Jim
ke: *** [all] Error 1
>
>
> makefile content
>
> all:
> (tab) #don't do anything
>
>
> so - which update broke my make files? make, or sh?
>
> Jim
>
M:\tmp>make
#don't do anything
E:\cygwin\bin\sh.exe: *** MapViewOfFileEx 'shared'(0x610A60E8), Win32 error
6.
Terminating.
make: *** [all] Error 1
makefile content
all:
(tab) #don't do anything
so - which update broke my make files? make, or sh?
Jim
- Original Message -
From: "Mumit Khan" <[EMAIL PROTECTED]>
To: "Jim" <[EMAIL PROTECTED]>
Sent: Thursday, June 13, 2002 3:18 PM
Subject: Re: Question about GCC predefined symbols
> On Thu, 13 Jun 2002, Jim wrote:
>
> > Was wondering -
>
> You may dream all you like, but "portable" means "works on all the
> systems we support". We cannot control the systems. We can control
> our sources. Portability precludes //-style comments in binutils.
Well - personally I thought you had gcc to support all your platforms, which
for ages
>
> > + // RH: prevent generating reimported functions
>
> Do not use C++ style comments in C code. It is non-portable.
Don't you mean C99 style comments?
Would be portable if the world adopted a standard now 2 years old...
Subject: Midnight Command 4.5.55 on Cygwin
Ya know... Midnight commander's good and all - but FAR which appears to have
had a common anscestor has much more functionatlity, kinda wish someone
could port FAR back to linux for all the various plugins, the process list
works great, etc etc ( www.ra
- Original Message -
From: "Markus Hoenicka" <[EMAIL PROTECTED]>
To: <[EMAIL PROTECTED]>
Sent: Monday, April 08, 2002 1:54 PM
Subject: Re: SGML/XML packages available for testing
> Hi Jim,
>
> you mean a package dependency on Jade? This would be o
> > finally Jon Foster and I got something that seems to work. The
> > following packages are available for testing:
>
> > docbook-dsssl
> > docbook-sgml
> > docbook-xml
> > docbook-xsl
> > openjade
Ahh was just playing with postgresql and it wanted JADE - suppose it'll be
weeks before this is a
>Subject: more and base
>Can we remove more from base?
More is what? 3k? I'd love to have had it in the base install when I
installed my cygwin - I want the basic development environment too - since
what use is the environment without gcc?
>
> This version of NASM has been modified by SciTech Software such that it
I'm all for going to the heart of the issue - why are you going to a
modified version? When they're still making improvements to the base? I
mean sure it's modified, but that locks it into an old version path
There
Do they have cygwin versions? Are they interested in having it available
via setup.exe?
Rob
*shrug* they have a forum - the version compiles with cygwin without a
hitch... most recent is 0.98.22 (I'm pretty sure)
I kinda thought the people at nasm.2y.net were doing a good job of
maintaining it
> On Sat, Mar 16, 2002 at 11:42:48AM +0100, Gerrit P. Haase wrote:
> > Hallo,
> >
> > Is someone willing to maintain NASM, the netwide assembler?
>
> You?
>
> Corinna
>
> --
27 matches
Mail list logo