Re: Blame caching (was Re: Case study: Mono switches to Subversion)

> a.) The actual effort involved in a lazy calculation may be
> significantly higher (since the fulltexts would need to be
> reconstructed, again).

Hmm, I am not sure whether I understand this point. The price of
calculating the current result must be paid whatever solution is
choosen, only in case of pre-calculating it is paid by the
commiter and in case of lazy caching by the first user of blame
on given file. The latter seems a bit more reasonable.

I can also imagine merging lazy caching with some precomputing
activities (say some low priority daemon which processess recent
commits list and actualizes cached data - maybe for all, maybe
for most frequently used files). But I would not consider such
optimizations for the first shot of implementation.

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Thu Nov 18 17:12:49 2004