<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
<HTML>
<HEAD>
<META HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=iso-8859-1">
<META NAME="Generator" CONTENT="MS Exchange Server version 6.5.7653.38">
<TITLE>RE: [geos-devel] 3.0.3 To Do</TITLE>
</HEAD>
<BODY>
<!-- Converted from text/plain format -->
<P><FONT SIZE=2>Sorry it took so long to test on VC9 (VS 2008 Professional on Vista). Before Sanak's patch and after upgrading solution to 2008 doesn't compile, but after applying Sanak's patch and upgrading solution to 2008 - it compiles.<BR>
<BR>
The geos_test compiles and after running against the CAPI/test.wkt and doing a diff compare between the test.out and what I got, I see no difference.<BR>
<BR>
HOWEVER: Why does it say - GEOS version 3.0.0rc4-CAPI-1.3.3 -- shouldn't we change that version. To I don't know something like 3.0.3<BR>
<BR>
The geos_unit doesn't give any errors either.<BR>
<BR>
One thing I did notice is a lot of warnings about how Wp64 is deprecated and will be removed in a future release. I don't recall getting that when compiling under VC8. Get same when compiling against trunk.<BR>
<BR>
also just a lot of somewhat scary looking warnings - which I'm sure everyone on the GEOS team is aware of (I wasn't paying too much attention when compiling under VC8, so they were probably there too).<BR>
<BR>
So for 3.0 branch<BR>
geos_lib - 0 error(s), 371 warning(s)<BR>
<BR>
For Trunk<BR>
1>geos_lib - 0 error(s), 453 warning(s)<BR>
<BR>
3.0 Warnings:<BR>
lots of "assignment operator could not be generated"<BR>
lots of unreferenced parameters (no big deal)<BR>
lot of conversion from 'size_t' to 'int', possible loss of data - assume no big deal<BR>
There was a warning about int64 being cast to long - can't remember whether that was in trunk or branch.<BR>
<BR>
<BR>
Trunk warnings:<BR>
Things like - lots of "assignment operator could not be generated"<BR>
"'geos::io::WKTReader' : type name first seen using 'class' now seen using 'struct'"<BR>
monotonechainbuilder.cpp(103) : warning C4018: '>=' : signed/unsigned mismatch<BR>
nodedsegmentstring.h(88) : warning C4018: '<' : signed/unsigned mismatch<BR>
conversion from 'size_t' to 'int', possible loss of data (assume no big deal)<BR>
<BR>
The geos_test is not in trunk for vc8, do we want it there or do we not even care to have this test at all for VC?<BR>
<BR>
Anyrate I guess the moral is for 3.1 we should probably cleanup some of these warnings since its hard to tell which warnings are important and which are not. But Paul already stated the obvious before, I'm just restating it.<BR>
<BR>
-----Original Message-----<BR>
From: geos-devel-bounces@lists.osgeo.org on behalf of Sanak<BR>
Sent: Thu 10/23/2008 1:52 PM<BR>
To: GEOS Development List<BR>
Subject: Re: [geos-devel] 3.0.3 To Do<BR>
<BR>
Ok, I had added geostest project to msvc80 solution file.<BR>
Please see the attach<BR>
file(geos-branches-3.0-build-msvc80_add_geostest.patch).<BR>
geostest.vcproj will generate capi/test.out.<BR>
<BR>
I think that the cause of difference between generated test.out and<BR>
test.expected is precision of floating point, but I don't know details.<BR>
<BR>
And another Visual C++ specific problem may be capi version.(capi/<BR>
geos_c.h.in)<BR>
---<BR>
Index: geos_c.h.in<BR>
===================================================================<BR>
--- geos_c.h.in (revision 2208)<BR>
+++ geos_c.h.in (working copy)<BR>
@@ -51,9 +51,9 @@<BR>
#if defined(_MSC_VER)<BR>
#include <geos/version.h><BR>
#define GEOS_CAPI_VERSION_MAJOR 1<BR>
-#define GEOS_CAPI_VERSION_MINOR 3<BR>
+#define GEOS_CAPI_VERSION_MINOR 4<BR>
#define GEOS_CAPI_VERSION_PATCH 3<BR>
-#define GEOS_CAPI_VERSION "3.0.0rc4-CAPI-1.3.3"<BR>
+#define GEOS_CAPI_VERSION "3.0.3-CAPI-1.4.3"<BR>
#else<BR>
#define GEOS_VERSION_MAJOR @VERSION_MAJOR@<BR>
#define GEOS_VERSION_MINOR @VERSION_MINOR@<BR>
---<BR>
<BR>
Regards,<BR>
<BR>
2008/10/24 Frank Warmerdam <warmerdam@pobox.com><BR>
<BR>
> Obe, Regina wrote:<BR>
><BR>
>> I know Frank already tested on Visual C++ Express (which is basically<BR>
>> Visual 2008 Express) and got some test fail errors. Do we have a solution<BR>
>> file for that (or would people just use the msvc80 solution file for that if<BR>
>> they wanted to use a solution) .<BR>
>><BR>
> ...<BR>
><BR>
>> Running regression tests using geos_unitd.exe...<BR>
>> ===============================<BR>
>> GEOS Test Suite Application<BR>
>> ===============================<BR>
>> capi::GEOSCoordSeq: ..... 5<BR>
>> capi::GEOSSimplify: . 1<BR>
>><BR>
><BR>
> Folks,<BR>
><BR>
> I would note that I wasn't running geos_unitd.exe, I ran geostest.exe<BR>
> against capi/test.wkt and compared to capi/test.expected. If I get<BR>
> back to my windows build I'll look into the geos_unitd.exe which I wasn't<BR>
> aware of.<BR>
><BR>
> Best regards,<BR>
> --<BR>
><BR>
> ---------------------------------------+--------------------------------------<BR>
> I set the clouds in motion - turn up | Frank Warmerdam,<BR>
> warmerdam@pobox.com<BR>
> light and sound - activate the windows | <A HREF="http://pobox.com/~warmerdam">http://pobox.com/~warmerdam</A><<A HREF="http://pobox.com/%7Ewarmerdam">http://pobox.com/%7Ewarmerdam</A>><BR>
> and watch the world go round - Rush | Geospatial Programmer for Rent<BR>
><BR>
> _______________________________________________<BR>
> geos-devel mailing list<BR>
> geos-devel@lists.osgeo.org<BR>
> <A HREF="http://lists.osgeo.org/mailman/listinfo/geos-devel">http://lists.osgeo.org/mailman/listinfo/geos-devel</A><BR>
><BR>
<BR>
<BR>
<BR>
<BR>
<BR>
<BR>
<BR>
<BR>
<BR>
<BR>
<BR>
<BR>
<BR>
</FONT>
</P>
</BODY>
</HTML>
<HTML><BODY><P><hr size=1></P>
<P><STRONG>
The substance of this message, including any attachments, may be confidential, legally privileged and/or exempt from disclosure pursuant to Massachusetts law. It is intended solely for the addressee. If you received this in error, please contact the sender and delete the material from any computer.
</STRONG></P></BODY></HTML>
<P><hr size=1></P>
<P><STRONG><font size="2" color="339900"> Help make the earth a greener place. If at all possible resist printing this email and join us in saving paper. </p> <p> </font></STRONG></P>