Review Request: kis random generator improvements

-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
http://reviewboard.kde.org/r/2080/
-----------------------------------------------------------

Review request for KOffice and Cyrille Berger.


Summary
-------

Two changes here.

First adapt Cyrille's simplification of the OS-X fix from reverted commit r1045274.

Second, change the neighborhood region in the unit test from [-2,3) to [0,5) (and also the tested region from [2,1000) to [0,1024)). Testing negative numbers from the reference point is redundant, since we would have compared the same two points earlier (except for the first few rows/columns). So this change should result in the same points being compared as before, plus adds additional comparisons without changing the amount of work done.


Diffs
-----

  /trunk/koffice/krita/image/kis_random_generator.cc 1045276
  /trunk/koffice/krita/image/tests/kis_random_generator_test.cpp 1045276

Diff: http://reviewboard.kde.org/r/2080/diff


Testing
-------

builds on Linux, test passes


Thanks,

Matthew

_______________________________________________
koffice-devel mailing list
koffice-devel@...
https://mail.kde.org/mailman/listinfo/koffice-devel

-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
http://reviewboard.kde.org/r/2080/#review2942
-----------------------------------------------------------


It looks good -- but then, I'm a complete amateur in this field, so I will leave it to Cyrille to express an opinion.

- Boudewijn


On 2009-11-05 17:31:20, Matthew Woehlke wrote:

>
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> http://reviewboard.kde.org/r/2080/
> -----------------------------------------------------------
>
> (Updated 2009-11-05 17:31:20)
>
>
> Review request for KOffice and Cyrille Berger.
>
>
> Summary
> -------
>
> Two changes here.
>
> First adapt Cyrille's simplification of the OS-X fix from reverted commit r1045274.
>
> Second, change the neighborhood region in the unit test from [-2,3) to [0,5) (and also the tested region from [2,1000) to [0,1024)). Testing negative numbers from the reference point is redundant, since we would have compared the same two points earlier (except for the first few rows/columns). So this change should result in the same points being compared as before, plus adds additional comparisons without changing the amount of work done.
>
>
> Diffs
> -----
>
>   /trunk/koffice/krita/image/kis_random_generator.cc 1045276
>   /trunk/koffice/krita/image/tests/kis_random_generator_test.cpp 1045276
>
> Diff: http://reviewboard.kde.org/r/2080/diff
>
>
> Testing
> -------
>
> builds on Linux, test passes
>
>
> Thanks,
>
> Matthew
>
>

_______________________________________________
koffice-devel mailing list
koffice-devel@...
https://mail.kde.org/mailman/listinfo/koffice-devel

[Cyrille Berger]

-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
http://reviewboard.kde.org/r/2080/#review2944
-----------------------------------------------------------

Ship it!


sounds good

- Cyrille


On 2009-11-05 17:31:20, Matthew Woehlke wrote:

>
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> http://reviewboard.kde.org/r/2080/
> -----------------------------------------------------------
>
> (Updated 2009-11-05 17:31:20)
>
>
> Review request for KOffice and Cyrille Berger.
>
>
> Summary
> -------
>
> Two changes here.
>
> First adapt Cyrille's simplification of the OS-X fix from reverted commit r1045274.
>
> Second, change the neighborhood region in the unit test from [-2,3) to [0,5) (and also the tested region from [2,1000) to [0,1024)). Testing negative numbers from the reference point is redundant, since we would have compared the same two points earlier (except for the first few rows/columns). So this change should result in the same points being compared as before, plus adds additional comparisons without changing the amount of work done.
>
>
> Diffs
> -----
>
>   /trunk/koffice/krita/image/kis_random_generator.cc 1045276
>   /trunk/koffice/krita/image/tests/kis_random_generator_test.cpp 1045276
>
> Diff: http://reviewboard.kde.org/r/2080/diff
>
>
> Testing
> -------
>
> builds on Linux, test passes
>
>
> Thanks,
>
> Matthew
>
>

_______________________________________________
koffice-devel mailing list
koffice-devel@...
https://mail.kde.org/mailman/listinfo/koffice-devel