SUCCESS!! RE: [UMN_MAPSERVER-USERS] MapServer incompatibility wit h RHEL4/CE NTOS4.2

[Doyon, Jean-Francois]

Phew! Took 5 hours, but I figured it out :)

The problem seems to be with FreeType compiled on CentOS 4.2 ...

Using LD_LIBRARY_PATH and a mapserver binary from RH7.3, I got a working
solution that involved moving ALL the library dependencies from 7.3 to
CentOS.

Once that was done, I removed one by one the dependencies, switching to the
OS provided ones instead, check the maps at every step.

Once I managed to eliminate ALL remaining libraries, I was left with
freetype.

The freetype binary lib from 7.3 moved to CentOS works fine.

The freetype provided by CentOS does not.

If you compile freetype 2.0.9 (the version on 7.3) on CentOS however, the
problem appears!

For some reason, compiling freetype on CentOS results in the symptoms I was
showing.

I also saw these problems appear on my development platform mind you which
is still 7.3.  There I was using freetype 2.1.10 for a while, and it was
causing problems ... Once I backtracked to the RPM one (2.0.9) things came
back to normal.

I'm thinking RedHat was doing something to the code for their RPM's and
stopped doing it, like maybe that hitting thing that's in the library but
disabled by default?

J.F.

-----Original Message-----
From: Lowell.Filak [mailto:lfilak at medinaco.org] 
Sent: March 3, 2006 1:46 PM
To: Doyon, Jean-Francois
Cc: MAPSERVER-USERS at LISTS.UMN.EDU
Subject: Re: [UMN_MAPSERVER-USERS] MapServer incompatibility with RHEL4/CE
NTOS4.2

Doyon, Jean-Francois writes: 

> As mentionned in the initial post, tried a wide variety of 
> combinations of libraries and mapserver :)
> 
> At the root though, a given combination of mapserver/libs works on RH 
> 7.3, the exact same combination on CentOS doesn't.
> 
> (GD 2.0.33 with FreeType 2.0.9 and MapServer 4.8.1). 
> 
> I suppose it is possible GD or FreeType have problems on CentOS, but 
> I'm hoping/guessing google would've told me about it by now, I've had 
> no luck trying to find out about similar problems, they're hard to 
> describe in a search engine :)
> 
> I'm going ot try taking the mapserver from RH7.3 and "chrooting" it on 
> CentOS, and see how that works, I'll let you know :)

J.F., 

I did see you mention that in your original post, but considering the amount
of work it is to do that I guess I was hoping that you hadn't.
Either way you/this-thread do seem to have narrowed down how to reproduce
the problem without requiring specific hardware. Of course without any
solution as of yet ;-) 

Lowell