Breaking up the image

Daniel Vainsencher danielv at netvision.net.il
Tue Dec 10 20:36:38 UTC 2002


Andreas Raab <andreas.raab at gmx.de> wrote:
> Hi Daniel,
[snip]

> About finding references to collaborators: How about adopting the DVS
> scheme?! E.g., it might be a more understandable measure to analyze the
> dependencies based on the categories the classes are in. This might
> (note: "might") help in identifying the internal collaborators. 
Oh, SpT does this (a couple of lines), though at the category level -
what categories collaborate with this category. You're right it might be
useful to find "what classes in what categories does this class
collaborate with". Ok, the category graph for the whole image takes a
while to calculate, but I'll start working on it.

> E.g.,
> I'd be interested to see for example how a "category dependency" looks
> for something like the WonderlandCameraMorph - I would expect it to be
> pretty much Kernel-* (some fundamental suff is used everywhere)
> Balloon3D-* and Morphic-* (due to the camera being the integration
> part). At this point you might be able to spot any "strange looking"
> categories. This of course assumes that classes are in "good" categories
> to begin with.
You're right it might be useful to find "what classes in what categories
does this class collaborate with". Ok, the category graph for the whole
image takes a while to calculate, but I'll start working on it.

> 
> Cheers,
>   - Andreas
>



More information about the Squeak-dev mailing list