On Tuesday 02 April 2002 4:35 pm, Jean-Marc Lasgouttes wrote:
Angus == Angus Leeming [EMAIL PROTECTED] writes:
Angus Jean-Marc, I can confirm that I need neither CPP nor CPPFLAGS
Angus when configuring with autoconf 2.52. All is fine and dandy.
Great.
Tell me (again?) what the problem
On Thursday 04 April 2002 12:32 pm, Angus Leeming wrote:
On Tuesday 02 April 2002 4:35 pm, Jean-Marc Lasgouttes wrote:
Angus == Angus Leeming [EMAIL PROTECTED] writes:
Angus Jean-Marc, I can confirm that I need neither CPP nor CPPFLAGS
Angus when configuring with autoconf 2.52. All is
Angus == Angus Leeming [EMAIL PROTECTED] writes:
I've just checked that autoconf 2.13 works with my configure
script. All is fine and dandy here too. deccxx and deccc are NOT
needed at all when using automake = 1.5.
Excellent.
Angus I forgot to mention:
Angus The AC_LANG(C, C++) calls I
On Thursday 04 April 2002 1:53 pm, Jean-Marc Lasgouttes wrote:
Angus == Angus Leeming [EMAIL PROTECTED] writes:
I think it is fine as it is. Currently ac2.13 + am 1.5 works with cxx.
When we switch to autoconf 2.5x, you will be able to use ac2.53 +
am1.6 and all should still work.
All we
Angus == Angus Leeming [EMAIL PROTECTED] writes:
Angus On Thursday 04 April 2002 1:53 pm, Jean-Marc Lasgouttes wrote:
Angus == Angus Leeming [EMAIL PROTECTED] writes: I think
it is fine as it is. Currently ac2.13 + am 1.5 works with cxx. When
we switch to autoconf 2.5x, you will be able to
On Tuesday 02 April 2002 4:35 pm, Jean-Marc Lasgouttes wrote:
> > "Angus" == Angus Leeming <[EMAIL PROTECTED]> writes:
>
> Angus> Jean-Marc, I can confirm that I need neither CPP nor CPPFLAGS
> Angus> when configuring with autoconf 2.52. All is fine and dandy.
>
> Great.
>
> >> Tell me
On Thursday 04 April 2002 12:32 pm, Angus Leeming wrote:
> On Tuesday 02 April 2002 4:35 pm, Jean-Marc Lasgouttes wrote:
> > > "Angus" == Angus Leeming <[EMAIL PROTECTED]> writes:
> >
> > Angus> Jean-Marc, I can confirm that I need neither CPP nor CPPFLAGS
> > Angus> when configuring with
> "Angus" == Angus Leeming <[EMAIL PROTECTED]> writes:
>> I've just checked that autoconf 2.13 works with my configure
>> script. All is fine and dandy here too. deccxx and deccc are NOT
>> needed at all when using automake >= 1.5.
Excellent.
Angus> I forgot to mention:
Angus> The
On Thursday 04 April 2002 1:53 pm, Jean-Marc Lasgouttes wrote:
> > "Angus" == Angus Leeming <[EMAIL PROTECTED]> writes:
> I think it is fine as it is. Currently ac2.13 + am 1.5 works with cxx.
> When we switch to autoconf 2.5x, you will be able to use ac2.53 +
> am1.6 and all should still
> "Angus" == Angus Leeming <[EMAIL PROTECTED]> writes:
Angus> On Thursday 04 April 2002 1:53 pm, Jean-Marc Lasgouttes wrote:
>> > "Angus" == Angus Leeming <[EMAIL PROTECTED]> writes: I think
>> it is fine as it is. Currently ac2.13 + am 1.5 works with cxx. When
>> we switch to autoconf
On Tue, 2 Apr 2002, Angus Leeming wrote:
[...]
static bool build_name (string,const vectorstring,allocatorvoid );
--^
cxx: Error: /usr/include/cxx/locimpl, line 309: #266 string is ambiguous
etc.
Any ideas?
I suppose your compiler is seeing a choice of
On Wednesday 03 April 2002 9:55 am, Allan Rae wrote:
On Tue, 2 Apr 2002, Angus Leeming wrote:
[...]
static bool build_name (string,const vectorstring,allocatorvoid
); --^
cxx: Error: /usr/include/cxx/locimpl, line 309: #266 string is
ambiguous
etc.
Any
On Wed, 3 Apr 2002, Angus Leeming wrote:
On Wednesday 03 April 2002 9:55 am, Allan Rae wrote:
...
In particular, why isn't the #define __STRING__ sufficient to stop
string being loaded by some other header? Dumb STL implementation?
Possibly. It appears that here it's called __STD_STRING.
On Wed, Apr 03, 2002 at 06:55:20PM +1000, Allan Rae wrote:
So the question is why? Does `cc -E` provide any insights?
In particular, why isn't the #define __STRING__ sufficient to stop
string being loaded by some other header? Dumb STL implementation?
Why should it?
Where is written that
Allan == Allan Rae [EMAIL PROTECTED] writes:
Allan On Tue, 2 Apr 2002, Angus Leeming wrote: [...]
static bool build_name (string,const vectorstring,allocatorvoid
); --^ cxx: Error:
/usr/include/cxx/locimpl, line 309: #266 string is ambiguous
etc.
Any ideas?
On Wednesday 03 April 2002 10:19 am, Jean-Marc Lasgouttes wrote:
Allan == Allan Rae [EMAIL PROTECTED] writes:
Allan On Tue, 2 Apr 2002, Angus Leeming wrote: [...]
static bool build_name (string,const vectorstring,allocatorvoid
); --^ cxx: Error:
On Wed, Apr 03, 2002 at 09:30:39AM +, Angus Leeming wrote:
What I don't get is why formala.C is #including locale info anyway.
Not to worry, I'm not going to waste much time on this.
Everything std:string related carries around locale baggage.
I've just removed two unneeded #include's
On Tue, 2 Apr 2002, Angus Leeming wrote:
[...]
> static bool build_name (string&,const vector&);
> --^
> cxx: Error: /usr/include/cxx/locimpl, line 309: #266 "string" is ambiguous
>
> etc.
>
> Any ideas?
I suppose your compiler is seeing a choice of
On Wednesday 03 April 2002 9:55 am, Allan Rae wrote:
> On Tue, 2 Apr 2002, Angus Leeming wrote:
> [...]
>
> > static bool build_name (string&,const vector > >&); --^
> > cxx: Error: /usr/include/cxx/locimpl, line 309: #266 "string" is
> > ambiguous
> >
On Wed, 3 Apr 2002, Angus Leeming wrote:
> On Wednesday 03 April 2002 9:55 am, Allan Rae wrote:
...
> > In particular, why isn't the "#define __STRING__" sufficient to stop
> > being loaded by some other header? Dumb STL implementation?
>
> Possibly. It appears that here it's called
On Wed, Apr 03, 2002 at 06:55:20PM +1000, Allan Rae wrote:
> So the question is why? Does `cc -E` provide any insights?
> In particular, why isn't the "#define __STRING__" sufficient to stop
> being loaded by some other header? Dumb STL implementation?
Why should it?
Where is written that
> "Allan" == Allan Rae <[EMAIL PROTECTED]> writes:
Allan> On Tue, 2 Apr 2002, Angus Leeming wrote: [...]
>> static bool build_name (string&,const vector> >&); --^ cxx: Error:
>> /usr/include/cxx/locimpl, line 309: #266 "string" is ambiguous
>>
>>
On Wednesday 03 April 2002 10:19 am, Jean-Marc Lasgouttes wrote:
> > "Allan" == Allan Rae <[EMAIL PROTECTED]> writes:
>
> Allan> On Tue, 2 Apr 2002, Angus Leeming wrote: [...]
>
> >> static bool build_name (string&,const vector >>
> >> >&); --^ cxx:
On Wed, Apr 03, 2002 at 09:30:39AM +, Angus Leeming wrote:
> What I don't get is why formala.C is #including locale info anyway.
> Not to worry, I'm not going to waste much time on this.
Everything std:string related carries around locale baggage.
I've just removed two unneeded #include's
On Thursday 28 March 2002 2:48 pm, Jean-Marc Lasgouttes wrote:
Angus == Angus Leeming [EMAIL PROTECTED] writes:
Angus I think that compiling on Tru64 is now sufficiently
Angus straightforward that users should be able to do it with the
Angus help of a Readme only. Nonetheless, it's still
Angus == Angus Leeming [EMAIL PROTECTED] writes:
Angus Jean-Marc, I can confirm that I need neither CPP nor CPPFLAGS
Angus when configuring with autoconf 2.52. All is fine and dandy.
Great.
Tell me (again?) what the problem was with autoconf 2.13.
Angus I'm willing to test autoconf 2.13
On Tuesday 02 April 2002 4:35 pm, Jean-Marc Lasgouttes wrote:
Angus == Angus Leeming [EMAIL PROTECTED] writes:
Angus Jean-Marc, I can confirm that I need neither CPP nor CPPFLAGS
Angus when configuring with autoconf 2.52. All is fine and dandy.
Great.
Incidentally, when running
Angus Leeming [EMAIL PROTECTED] writes:
| static bool parse_name (vectorstring,allocatorvoid ,const char*);
| -^
| cxx: Error: /usr/include/cxx/locimpl, line 309: #266 string is ambiguous
| static bool build_name (string,const vectorstring,allocatorvoid );
|
On Thursday 28 March 2002 2:48 pm, Jean-Marc Lasgouttes wrote:
> > "Angus" == Angus Leeming <[EMAIL PROTECTED]> writes:
>
> Angus> I think that compiling on Tru64 is now sufficiently
> Angus> straightforward that users should be able to do it with the
> Angus> help of a Readme only.
> "Angus" == Angus Leeming <[EMAIL PROTECTED]> writes:
Angus> Jean-Marc, I can confirm that I need neither CPP nor CPPFLAGS
Angus> when configuring with autoconf 2.52. All is fine and dandy.
Great.
>> Tell me (again?) what the problem was with autoconf 2.13.
Angus> I'm willing to test
On Tuesday 02 April 2002 4:35 pm, Jean-Marc Lasgouttes wrote:
> > "Angus" == Angus Leeming <[EMAIL PROTECTED]> writes:
>
> Angus> Jean-Marc, I can confirm that I need neither CPP nor CPPFLAGS
> Angus> when configuring with autoconf 2.52. All is fine and dandy.
>
> Great.
Incidentally, when
Angus Leeming <[EMAIL PROTECTED]> writes:
| static bool parse_name (vector&,const char*);
| -^
| cxx: Error: /usr/include/cxx/locimpl, line 309: #266 "string" is ambiguous
| static bool build_name (string&,const vector&);
Angus == Angus Leeming [EMAIL PROTECTED] writes:
Angus I think that compiling on Tru64 is now sufficiently
Angus straightforward that users should be able to do it with the
Angus help of a Readme only. Nonetheless, it's still sufficiently
Angus convoluted that this info is probably best placed
On Thursday 28 March 2002 2:48 pm, Jean-Marc Lasgouttes wrote:
Angus == Angus Leeming [EMAIL PROTECTED] writes:
Angus I think that compiling on Tru64 is now sufficiently
Angus straightforward that users should be able to do it with the
Angus help of a Readme only. Nonetheless, it's still
Angus == Angus Leeming [EMAIL PROTECTED] writes:
Angus On Thursday 28 March 2002 2:48 pm, Jean-Marc Lasgouttes wrote:
Angus == Angus Leeming [EMAIL PROTECTED] writes:
Angus I think that compiling on Tru64 is now sufficiently
Angus straightforward that users should be able to do it with the
On Thursday 28 March 2002 3:11 pm, Jean-Marc Lasgouttes wrote:
Why do you need to set CPP?
[snip...]
as I said, I'll check.
Is the problem with std::count still relevant now that we have
lyx::count?
Angus I believe that we check only for the existence of std::count,
Angus not whether
Angus == Angus Leeming [EMAIL PROTECTED] writes:
Angus I believe that we check only for the existence of std::count,
Angus not whether it returns a sane value.
Indeed. That's just stupid.
Angus It's stupid to check for a sane value? Fair enough; then the
Angus comment should stay.
No, I was
> "Angus" == Angus Leeming <[EMAIL PROTECTED]> writes:
Angus> I think that compiling on Tru64 is now sufficiently
Angus> straightforward that users should be able to do it with the
Angus> help of a Readme only. Nonetheless, it's still sufficiently
Angus> convoluted that this info is probably
On Thursday 28 March 2002 2:48 pm, Jean-Marc Lasgouttes wrote:
> > "Angus" == Angus Leeming <[EMAIL PROTECTED]> writes:
>
> Angus> I think that compiling on Tru64 is now sufficiently
> Angus> straightforward that users should be able to do it with the
> Angus> help of a Readme only.
> "Angus" == Angus Leeming <[EMAIL PROTECTED]> writes:
Angus> On Thursday 28 March 2002 2:48 pm, Jean-Marc Lasgouttes wrote:
>> > "Angus" == Angus Leeming <[EMAIL PROTECTED]> writes:
>>
Angus> I think that compiling on Tru64 is now sufficiently
Angus> straightforward that users should
On Thursday 28 March 2002 3:11 pm, Jean-Marc Lasgouttes wrote:
> >> Why do you need to set CPP?
[snip...]
as I said, I'll check.
> >> Is the problem with std::count still relevant now that we have
> >> lyx::count?
>
> Angus> I believe that we check only for the existence of std::count,
>
> "Angus" == Angus Leeming <[EMAIL PROTECTED]> writes:
Angus> I believe that we check only for the existence of std::count,
Angus> not whether it returns a sane value.
>> Indeed. That's just stupid.
Angus> It's stupid to check for a sane value? Fair enough; then the
Angus> comment should
o On Tru64 Unix with cxx, you may have a compilation error in
lyx_main.C if you have GNU gettext installed. This is due to a bug
in gettext. To solve this, you can either (1) configure with
--with-included-gettext or (2) add -D__STDC__ to cxx flags.
Jean-Marc, is there any
Angus == Angus Leeming [EMAIL PROTECTED] writes:
Angus o On Tru64 Unix with cxx, you may have a compilation error in
Angus lyx_main.C if you have GNU gettext installed. This is due to a
Angus bug in gettext. To solve this, you can either (1) configure
Angus with --with-included-gettext or (2)
o On Tru64 Unix with cxx, you may have a compilation error in
lyx_main.C if you have GNU gettext installed. This is due to a bug
in gettext. To solve this, you can either (1) configure with
--with-included-gettext or (2) add -D__STDC__ to cxx flags.
Jean-Marc, is there any
> "Angus" == Angus Leeming <[EMAIL PROTECTED]> writes:
Angus> o On Tru64 Unix with cxx, you may have a compilation error in
Angus> lyx_main.C if you have GNU gettext installed. This is due to a
Angus> bug in gettext. To solve this, you can either (1) configure
Angus> with
46 matches
Mail list logo