[Insight-developers] Problem with new version of gdcm

[Mathieu Malaterre]

Ok,

	So I went over the diffs in between gdcm CVS and gdcm 1.2. I backported 
all the bug fixes and patches I could see. Hopefully I did not miss any. 
Anyway this is not merged into ITK right now, this will wait until I 
check all our dashboards are passing all tests.
	I did check the output of gdcm against D. Clunie tools (dciodvfy), they 
all seems ok. But one thing I cannot check is if GE workstation accept 
those files. Bill once this is merged could you try to load some files 
again.

Mathieu
Ps: I finally got a copy of eFilm 1.5.3 to check gdcm output...

Mathieu Malaterre wrote:
> Bill,
> 
>     Ok thanks a lot for tracking it down. This is indeed the proper fix, 
> as this was removed from the gdcm CVS HEAD, but I guess I missed this 
> patch. I'll try to go over -again- difference in between gdcm CVS HEAD 
> and 1.2 branch.
> 
> Sorry for troubles,
> Mathieu
> Ref:
> http://cvs.creatis.insa-lyon.fr/viewcvs/viewcvs.cgi/gdcm/src/gdcmFileHelper.cxx?annotate=1.91 
> 
> 
> Lorensen, William E (GE, Research) wrote:
> 
>> Mathieu,
>>
>> The dicom series I create with the new gdcm no longer are acceptable 
>> to the GE Advantage Windows Workstations. Apparently, the new gdcm is 
>> not putting the (0020|0052) Frame of Reference UID
>> in the headers. I tracked down these statements in gdcmFileHelper.cxx 
>> that are causing the problem:
>>
>>    ValEntry *e_0028_0008 = FileInternal->GetValEntry(0x0028, 0x0008);
>>    if ( !e_0028_0008 )
>>    {
>>       Archive->Push(0x0020, 0X0052);
>>    }
>>
>> If I remove them, then the files I create are OK with the A/W and in 
>> fact are the same as files I created in the prior version of gdcm.
>>
>> I would like to remove these lines and check in the changes. This is a 
>> real show stopper for us. What is the reasoning behind including these 
>> statements?
>>
>> Bill
>> _______________________________________________
>> Insight-developers mailing list
>> Insight-developers at itk.org
>> http://www.itk.org/mailman/listinfo/insight-developers
>>
> 
> _______________________________________________
> Insight-developers mailing list
> Insight-developers at itk.org
> http://www.itk.org/mailman/listinfo/insight-developers
>