[vtkusers] Re: [vtk-developers] VTK 5.2 Release Planning

[Sean Ziegeler]

My experience in general is that a patch release shouldn't break any 
external interfaces (API's, client/server protocols, etc.).  But AFAIK, 
VTK uses a major.minor.patch version number scheme, so the change in 
minor version number is somewhat accurate.  The usual question is, is it 
enough of a change to warrant a major version number increment.

Normally, I wouldn't have much of an opinion on that for VTK, but David 
Cole raises an important point about the book.  I guess I'm wondering, 
on behalf of my users, how much does the 5.2 release break the 
documented classes, methods, and examples in the VTK 5 book?  If the VTK 
5 book can't reliably be used for a 5.x release, some people would find 
it frustrating, whereas if the new release were 6.x then they might be 
prepared for some incompatibility.

-Sean

Mike Jackson wrote:
> Probably depends on how much API breakage there is. Usually patch
> releases should not break APIs, or at least this should be kept to a
> minimum. Just something to think about.
> 
> Mike
> 
> On Jan 29, 2008 2:26 PM, David Cole <david.cole at kitware.com> wrote:
>> Bill,
>>
>> This revision of VTK/CMakeLists.txt was the first commit to the VTK-5-0
>> branch:
>>   revision 1.341.2.1
>>   date: 2005-09-01 16:49:44 +0000;  author: david.cole;  state: Exp;  lines:
>> +2 -2
>>   ENH: Update version numbers in the VTK-5-0 branch to 5.0.
>>
>> Sept. 1, 2005 was the first VTK-5-0 commit activity after the branch
>> creation.
>>
>> For the purpose of running your script on VTK, I would say changes since
>> then in CVS HEAD are "new in VTK 5.2" -- feel free to run your script and
>> post the results to the Wiki -- I would *love* that! Don't be surprised if
>> it takes a while...
>>
>> As to why we aren't calling it VTK 6.0, I haven't really personally given it
>> much thought. The automatic rollover to the next even minor number seems
>> good to me. Proclaiming it "6.0" sounds like a reasonable idea, although for
>> people buying the books, it might be more "warm/fuzzy" to have a VTK 5 book
>> and a 5.2 software install vs. a 5 book and a 6 software install...
>>
>> Anybody else have a strong opinion either way on the 5.2/6.0 thought?
>>
>>
>>
>>
>>
>> On 1/29/08, Bill Lorensen <bill.lorensen at gmail.com> wrote:
>>> David,
>>>
>>> I have a script that we run on itk to generate a wiki-compatible list of
>> new classes, tests etc.I also have a script that generates a wiki-compatible
>> list of changes. Both take a start and end date as an argument. They are
>> unix shell scripts with two small gawk programs,
>>> The itk list looks like this: http://www.itk.org/Wiki/ITK_Release_3.4
>>>
>>> If you want, I can add links to the vtk equivalents on the VTK release
>> page on the vtk wiki.
>>> What date do you consider VTK5.2 started? Looking at the repository I
>> think September 8, 2005 may be the date just after VTK5.0.
>>> One last comment. Is there a reason that this release is not called 6.0?
>> It seems that with over 300 new classes and 40 removed classes that this
>> release is more than an upgrade from VTK5.0.
>>> Bill
>>>
> _______________________________________________
> This is the private VTK discussion list. 
> Please keep messages on-topic. Check the FAQ at: http://www.vtk.org/Wiki/VTK_FAQ
> Follow this link to subscribe/unsubscribe:
> http://www.vtk.org/mailman/listinfo/vtkusers
>