> "Arnd" == Arnd Hanses <[EMAIL PROTECTED]> writes:
Arnd> Indeed, bracing
That's what I thought, but I prefered to be sure.
>> 2/ Why do you comment out the code when need_shell is false?
Arnd> Was it me, who commented this out??? ;)
Arnd> In fact, I can't remember doing this. What does t
On Thu, 8 Jul 1999 14:17:02 +0200 (MET DST), Jean-Marc Lasgouttes
wrote:
>I began to apply it and have two questions:
>
>1/ The indentation of the following code is strange; doesn't it need
>some braces?
>
>+if (sh.empty())
>+// COMSPEC is s
On Thu, 8 Jul 1999 14:17:02 +0200 (MET DST), Jean-Marc Lasgouttes
wrote:
>I began to apply it and have two questions:
>
>1/ The indentation of the following code is strange; doesn't it need
>some braces?
>
>+if (sh.empty())
>+// COMSPEC is s
> "Arnd" == Arnd Hanses <[EMAIL PROTECTED]> writes:
Arnd> Can't you simply detete that? Or cut and paste by hand? The
Arnd> patch should be simple to read and understand. In fact there are
Arnd> only five or six lines of code changed.
I began to apply it and have two questions:
1/ The inden
On Thu, 8 Jul 1999 11:44:26 +0200 (MET DST), Jean-Marc Lasgouttes
wrote:
>Two things:
>
>- do you want me to apply the patch, or is it still for discussion?
Well, it should be safe...
>
>- even if you want me to apply the patch, I won't ;) The problem is
>that it removes a lot of blank lines in
> "Arnd" == Arnd Hanses <[EMAIL PROTECTED]> writes:
Arnd> wrt to different shells with different syntax on OS/2:
Arnd> Here's a gzipped patch.
Hello,
Two things:
- do you want me to apply the patch, or is it still for discussion?
- even if you want me to apply the patch, I won't ;) The p
Hello,
wrt to different shells with different syntax on OS/2:
Here's a gzipped patch.
os2-shells.patch.gz
On 15 Jun 1999 06:44:07 +0900, [EMAIL PROTECTED] wrote:
>>As far as I remember, the last report of yours suggested it was really
>>a bug in the temporary path handling in LyX. Will you please send
>>the output of
>> lyx -dbg 163
>> ls -lR /tmp (while lyx is running)
>OK.:
Setze Debug-Level
No, better is:
cmd = "start/min/ " + cmd;
On Mon, 28 Jun 1999 13:46:18 +0100, Arnd Hanses wrote:
>... and therefore needs a command interpreter. So this is a bug,
>correctly it should read:
>
>cmd =+ "start/min ";
Ahem,
no,
>cmd = "start/min " + cmd;
Regards,
On Sat, 26 Jun 1999 14:54:11 +0100, Arnd Hanses wrote:
>OK, finally my paranoia version for lyx_cb.C
>(worked fine on my box):
Ahem...
Murphy's law:
XDVI likes to start Rexx-scripts to render masses of missing fonts,
etc. ...
>
>#ifndef __EMX__
> if (!wait)
>
On 24 Jun 1999 06:49:06 +0900, [EMAIL PROTECTED] wrote:
>"Arnd Hanses" <[EMAIL PROTECTED]> wrote:
>
>> Is there a reason *not* to use it? (Because it's plain C and emx
>> specific?)
>
>Mixing C char* and C++ string is not a good programming practice.
>Everytime a conversion from one to the other
"Asger Alstrup Nielsen" <[EMAIL PROTECTED]> wrote:
> When it comes to executable size, you save a lot more in LyX 1.0.x by adding
> the flags
>
> -fno-exceptions -fno-rtti
>
> This works for all g++-platforms that support the flags (at least egcs 1.0.x
> and above.) LyX v1.0.x does not use
"Arnd Hanses" <[EMAIL PROTECTED]> wrote:
> Well, it may even now get fooled by certain *evil* directory names. I
> vaguely remember EM using a kind of fancy test for /c or -c usage
> somewhere in emx sources. I will have to dig them out and grep for it.
If you want to make it completely NLS safe
On Tue, 15 Jun 1999 12:59:28 +0200, Asger Alstrup Nielsen wrote:
> LString sh = "";
>
>is sub-optimal. Just do
>
> LString sh;
>
>LString is default constructible, and defaults to the empty string.
AAhh, yes:
I see, C++ is so much more intelligent than C. Unfortunately I've no
ide
"Asger Alstrup Nielsen" <[EMAIL PROTECTED]> writes:
| On the other hand, exceptions are not easy to emulate, so I think we should
| avoid using those.
Well, I don't agree. (I agree as for now).
But we need to be very concious about using eceptions, it is not
something you just throw in.
Much
> Yes, -malign-jumps=4 -malign-functions=4 is what any i486 really needs.
>
> I like the 20% smaller executable; a performance change isn't so
> obvious on my old box (classic pentium). Your mileage may differ.
When it comes to executable size, you save a lot more in LyX 1.0.x by adding
the flag
On 15 Jun 1999 06:44:07 +0900, [EMAIL PROTECTED] wrote:
>> limited testing. Its fast. Only the old inline image bug seems to
>> remain here.
>
>As far as I remember, the last report of yours suggested it was really
>a bug in the temporary path handling in LyX. Will you please send
>the output of
On 15 Jun 1999 06:44:28 +0900, [EMAIL PROTECTED] wrote:
>> No, the problem is upper/lower case difference in:
>> if (!sh.contains("sh") && !sh.contains("SH"))
>
>Thanks for finding the bug. But you have to use lowercase() here.
>Consider "Cmd", "Sh" etc.
Well, it may even
> // OS/2 cmd.exe has another use for '&'
> if (!wait) {
> LString sh = "";
> sh = getenv("EMXSHELL");
> if (sh.empty())
> sh = getenv("COMSPEC");
>
"Arnd Hanses" <[EMAIL PROTECTED]> wrote:
> >sh.empty() appears not to work if emxshell is unset:
>
> No, the problem is upper/lower case difference in:
>
> if (sh.contains("cmd.exe")
>
>
> I propose:
>
> if (!sh.contains("sh") && !sh.contains("SH"))
> or
>
On Fri, 11 Jun 1999 16:27:45 +0100, Arnd Hanses wrote:
>sh.empty() appears not to work if emxshell is unset:
No, the problem is upper/lower case difference in:
if (sh.contains("cmd.exe")
I propose:
if (!sh.contains("sh") && !sh.contains("SH"))
or
On Sat, 05 Jun 1999 23:57:01 +0900, Shigeru Miyata wrote:
>It seems LyX 1.0.3 works fine here. If you don't find
>any problems, could you possibly release the binary?
sh.empty() appears not to work if emxshell is unset:
SET EMXSHELL
EMXSHELL=(null)
LYX_CB.C:599
// OS/2 cmd.ex
23 matches
Mail list logo
