[postgis-users] Mac OS X regression failures - 13/35 (was no "coveredby" in liblwgeom.1.3.so")

[William Kyngesburye]

Webb,

I finally got off my butt and worked thru those tests on my Mac.  All  
OK on my end.  I'm using the release 1.3.1 source, but I didn't see  
any major changes in the latest SVN that might have such effects.

So, I'm wondering how you built your PostGIS and Postgres. (and GEOS,  
PROJ...)

- do you build GEOS and PROJ yourself or use binaries?

They're both pretty straightforward to compile on OSX.

- do you build your own Postgres or use binaries?

Also pretty easy to compile on OSX.

- what configure flags are you using for PostGIS?

Also simple.  But I do one patch (in two places).  PostGIS is  
hardwired to include debug symbols.  For my binary release, I remove  
the -g flag from lwgeom/makefile:CSTAR_FLAGS and loader/makefile/CFLAGS.


- kindof goes along with the questions about using prebuilt binaries  
above - do you use Fink or Macports?  Even if you don't use either  
for PostGIS or any of its dependencies.  Their versions of some of  
the basic tools used in compiling may get used instead and interfere  
with a smooth and successful build (even tho it may seem to have  
sompleted successfully).

I can't remember exactly what it was, but someone once had trouble  
with a basic tool like sed or ls.  Something about a GNU version in  
Fink vs. a BSD version in OSX.  messed up the expected build process.

Fink/Macports add to your PATH, and may also fiddle with  
DYLD_LIBRARY_PATH.

They are good tools for Mac users who want to *use* common unix  
software, but can create trouble when casually compiling other  
software (whether or not you intend to use stuff from Fink/Macports  
in said compilation), and sometimes even when *using* other unix  
software.


On Aug 22, 2007, at 12:35 PM, Webb Sprague wrote:

> Here is my log file and a rerun of the regression.  Cheers, and let me
> know if there is anything else.
>
> Here is your request (seems to me that these should fail, but the
> regression test isn't checking for that correctly - maybe a mac thing
> or a funny version of sed or something?).  I don't see any server
> crashes at all:
>

-----
William Kyngesburye <kyngchaos*at*kyngchaos*dot*com>
http://www.kyngchaos.com/

"Those people who most want to rule people are, ipso-facto, those  
least suited to do it."

- A rule of the universe, from the HitchHiker's Guide to the Galaxy