diff options
| author | neodarz <neodarz@neodarz.net> | 2018-08-11 20:21:34 +0200 |
|---|---|---|
| committer | neodarz <neodarz@neodarz.net> | 2018-08-11 20:21:34 +0200 |
| commit | 0ea5fc66924303d1bf73ba283a383e2aadee02f2 (patch) | |
| tree | 2568e71a7ccc44ec23b8bb3f0ff97fb6bf2ed709 /pipermail/nel/2000-November/000015.html | |
| download | nevrax-website-self-hostable-0ea5fc66924303d1bf73ba283a383e2aadee02f2.tar.xz nevrax-website-self-hostable-0ea5fc66924303d1bf73ba283a383e2aadee02f2.zip | |
Initial commit
Diffstat (limited to 'pipermail/nel/2000-November/000015.html')
| -rw-r--r-- | pipermail/nel/2000-November/000015.html | 105 |
1 files changed, 105 insertions, 0 deletions
diff --git a/pipermail/nel/2000-November/000015.html b/pipermail/nel/2000-November/000015.html new file mode 100644 index 00000000..95f27a17 --- /dev/null +++ b/pipermail/nel/2000-November/000015.html @@ -0,0 +1,105 @@ +<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN"> +<HTML> + <HEAD> + <TITLE> [Nel] matrix.cpp std::min/max problems?</TITLE> + <LINK REL="Index" HREF="index.html" > + <LINK REL="made" HREF="mailto:bryce%40neptune.net"> + <LINK REL="Previous" HREF="000013.html"> + <LINK REL="Next" HREF="000007.html"> + </HEAD> + <BODY BGCOLOR="#ffffff"> + <H1>[Nel] matrix.cpp std::min/max problems?</H1> + <B>Bryce Harrington</B> + <A HREF="mailto:bryce%40neptune.net" + TITLE="[Nel] matrix.cpp std::min/max problems?">bryce@neptune.net</A><BR> + <I>Fri, 10 Nov 2000 03:43:08 -0800 (PST)</I> + <P><UL> + <LI> Previous message: <A HREF="000013.html">[Nel] matrix.cpp std::min/max problems?</A></li> + <LI> Next message: <A HREF="000007.html">[Nel] How do i compile this beast? Empty FAQ...</A></li> + <LI> <B>Messages sorted by:</B> + <a href="date.html#15">[ date ]</a> + <a href="thread.html#15">[ thread ]</a> + <a href="subject.html#15">[ subject ]</a> + <a href="author.html#15">[ author ]</a> + </LI> + </UL> + <HR> +<!--beginarticle--> +<PRE>Heh, well sorry but that's MS for ya... + +As I recall, if you go MFC-free this particular issue will vanish. +However there are other insidious STL bugs with Visual C++. + +If in fact you are using MFC, I would urge you to cease doing so at the +earliest convenience. Not only for portability reasons (MFC *ain't* +portable, especially not backwards-wise) but also because it is a poor +architecture for GUI development, and is not really in the GPL spirit of +things. If you intend to conduct your development openly as a net +project, then crossplatformness is going to be especially treasured by +you. There are many options for crossplatform GUI libs. Let me know if +you'd like to discuss this more. My two preferred libs are wxWindows +(good for OS-matching widgets for editors and similar), or libuta (SDL +compatible and skinnable). Mozilla has another cross platform GUI lib +but I've not used it in its current incarnation so don't know whether +it's usable independently of Mozilla. + +FWIW, I always stick this boilerplate code in any file that uses STL +that I'd like to compile with MSVC 4.2: + +#ifdef _MSC_VER +// Fix broken STL implementation in MSVC 4.2 +// Include files (for MSVC) +#pragma warning(disable: 4786) // Kludge to get around bug in MSVC 4.2 +#pragma warning(disable: 4788) // See above +#endif + +I think you may be stuck including your own max/min template functions. +That's a pretty minor thing compared with some of the patches you may +have to make to work around some of the other issues (especially if you +want to support pre-6.0 MSVC). + +Ultimately, you're going to want to trim down use of STL anyway, so you +can recover the performance loss. We still use STL at WorldForge, for +instance, but some simple tests show that hand coded, non-templated map +and list classes provide some modest performance gains, and some day +when we're down to squeezing out the last bits of performance, we'll +probably be pulling STL out ourselves. + +Bryce + +On Fri, 10 Nov 2000, Vianney Lecroart wrote: + +><i> > The VC++/STL min/max problem is pretty well known, and is a result of a +</I>><i> > very buggy STL implementation in VC++, and some lawsuits. +</I>><i> > +</I>><i> > Simple answer: Define NOMINMAX in your Build Settings +</I>><i> +</I>><i> ok, this define disables the macro min & max, it's same behavior as doing +</I>><i> #undef min #undef max +</I>><i> but I search min in the VC include directory and subdirectory and there s +</I>><i> *no* min +</I>><i> definition in the STL (no template min) so we have to define our own min & +</I>><i> max +</I>><i> template function in our code... no very cool... +</I> +-- +Bryce Harrington +bryce @ neptune.net + + +</pre> + +<!--endarticle--> + <HR> + <P><UL> + <!--threads--> + <LI> Previous message: <A HREF="000013.html">[Nel] matrix.cpp std::min/max problems?</A></li> + <LI> Next message: <A HREF="000007.html">[Nel] How do i compile this beast? Empty FAQ...</A></li> + <LI> <B>Messages sorted by:</B> + <a href="date.html#15">[ date ]</a> + <a href="thread.html#15">[ thread ]</a> + <a href="subject.html#15">[ subject ]</a> + <a href="author.html#15">[ author ]</a> + </LI> + </UL> +</body></html> |