[Liblas-devel] RE: reading/writing las files windows c++
Howard Butler
hobu.inc at gmail.com
Tue Aug 25 19:59:04 EDT 2009
On Aug 25, 2009, at 5:47 PM, Gary Robinson wrote:
> Just more evidence of memory allocation problems.
Please provide patches and/or specific links to problems in the source
code. Anything else is just unsubstantiated conjecture. I won't
claim there aren't issues, but you have to show me, not tell me.
>
> Ok I’ve tried a bunch more stuff, now it looks like I can have
> success if I run in release mode, but not in debug. Is there a
> package made that I can use to run with debug mode that just isn’t
> posted on the website? I suppose I can recompile everything to get
> debug versions of the libraries if I really need them, but I’d
> rather not if it already exists.
Yes, you'll need to compile your own debug versions if you need that
> Am I the only one trying c++ on windows, or am I missing something
> obvious that everybody else is doing correctly?
Plenty are using libLAS on windows, but I think most are using the C
API rather than the C++ API. If you are writing code in C++, you
should look to the C API for hints on how things should be expected to
operate. The test suite also provides great examples of how to
operate with libLAS in C++ instead of C or Python.
> I am using the 1.2.0 version linked from the home page of
> liblas.org. I also tried installing the OSGeo4W linked from the home
> page, but it made no difference.
>
> The writer crashes claiming a buffer overrun error. The call stack
> goes from my function to LASHeader::Init(), to
> LASSpatialReference::SetVLRs(), and then to _report_gsfailure().
I would suspect this is evidence that some sort of exception is being
thrown that is not caught. We need to see your code...
> Thanks for any help you may provide. If I can’t get past this
> problem I will have to try another solution for .las files.
I would be very interested if you find any with libLAS' price point ;)
More information about the Liblas-devel
mailing list