[Proj] Geo Transfoms and calculations
hamish_b at yahoo.com
Tue Jan 13 22:51:28 EST 2009
--- On Mon, 1/5/09, strebe <strebe at aol.com> wrote:
> Not to start a war, but
.... but some will try anyway ...
:) I will bite with my tongue-in-cheek,
> GCC is not quite "world
> class". It is usable, but the code generation from MS
> Visual Studio C++ is much better. Intel also supplies their
> own compiler so that developers have access to world-class
> code generation; I have read it's even better than
> Microsoft's but have not used it myself, though I am
> considering it for Macintosh.
> How do I know this? The disinterested benchmarks say so.
> Not content to take others' word for it, my own software
> Geocart runs nearly twice as fast on VS C++ -generated code
> than on GCC-generated code in compute-bound situations.
> That's... important, when I have massive raster
> reprojections that take minutes or even hours.
> GCC's conformance to C++ standards used to be MUCH
> better than VS C++, but that advantage has dwindled since VS
> C++ 7 and pretty much disappeared in VSC++ 8, which I have
> on occasion found to be even more conformant than GCC. VS
> C++ 6 was a standards disaster.
> <<< shudder >>>
> So, I would agree GCC is world-class for standards
> conformance, which is important. The same cannot be said for
> code generation, which is also important. Since the best
> performers in that regard come out of proprietary vendors, I
> have to view "the world-class status of both Linux and
> GCC is a result, largely, of their GPL status" to be
> more ideological than a factual, particularly since there
> are not many metrics the market cares about by which Linux
> leads the world. Still, I am grateful to have the GPL, and I
> agree it makes no sense to fault the GPL community for the
> license or its implications. Fair's fair, and it's
> not like developers would be better off if GPL did not
it is a bit simplistic to judge only based on runtime speed.
The goal of GCC besides standards conformance you should consider is
the very tricky problem of portability across hardware architectures.
Intel & VSC only have one target, while gcc has 14+ to consider*.
Optimization will be least third down on gcc's list of priorities behind
these two. actually, probably 4th behind code readability.
compile-speed is another competing variable to consider, as is
"Freedom" and cost.
[*] debian; probably there are more
and FWIW, millions and millions of professional users and successful
deployments worldwide make these things world-class, not some benchmark
or our opinions/ideologies.
as with most things, be it a tool, license, projection, or a spouse,
the best solution for you depends on your specific needs, and it is
silly to argue which is "best" or unfit. but we try anyway :)
More information about the Proj