Adam R. B. Jack wrote:
Stefano
wait, this does not compute: Y affects 2 (self + Z). The fact that Z was already affected should not change the situation.
So look at 'dependees' and stop looking at 'affected' -- and make the dynamic webapp sort by either, or by your choice. Sure I understand your argument against 'affected' (it is obvious, even to me ;-) but I felt it might have some real-world value, so added it. I still feel it adds value, maybe not theoretically, but in the real world.
You point to 'affected' for why Cocoon has not build. Clearly wrong! [Ok, clearly a Stefano exuberant-pointed-exaggeration also, but indulge me...] Cocoon didn't build 'cos (1) we'd not got enough folks caring about gump successes (2) we didn't build from repository (and cocoon has lots of dependencies, so it's chances of a build were unlikely. FWIIW: Communities (many), along with myself (looking at 'affected') and Stefan (just being the natural Gumpmeister) put in the time and effort in to picked candidate failures, fix them, and slowly edged up the Gump wave front to meet Cocoon. So, in a sense, this number resulted in Cocoon getting built. ;-)
all right, it is clearly not your fault that cocoon was never built, but this for sure didn't help, IMHO.
Still, I'd like to see 'good math' to occur at this level. I agree that the links between nodes (projects) is where the best Gump information resides. I've done a few 'reports', but it'd be great to see these done completely.
Depth (deepest sub-tree) and Total Depth (accumulation of all branches) :
http://brutus.apache.org/gump/public/gump_stats/project_depdepth.html http://brutus.apache.org/gump/public/gump_stats/project_totdepdepth.html
Historical Odds of a build (aka FOG)
http://brutus.apache.org/gump/public/gump_stats/project_fogfactor.html
Maybe 'reports' like this ought not be dynamic (although maybe Cocoon copes with that, I don't know) but I feel we really want to be able to express these numbers in some graphical form (or comparisons, or something) in order to get the most from them. I'd enjoy being led here, if your experience/expertise can guide...
working on it.
Ok, I'll stop talking and let action speak.
-- Stefano.
smime.p7s
Description: S/MIME Cryptographic Signature
