02:07 <bobthecow> ddfreyne: feature request for deployers — would be awesome if i could see that they were doing something :)

02:07 <number-six> bobthecow: ddfreyne (6 hours ago): Do you think output.diff is useful? I'm thinking of scratching that, and telling people to use git diff (with the output dir being a submodule or in a separate branch)

02:08 <bobthecow> i'm doing a fresh rsync of a fairly large site, and *i* know it's doing something, because i checked the server end of it.

02:08 <bobthecow> but all i see on the cli is...

02:08 <bobthecow> "Loading site data… done"

02:09 <bobthecow> it eventually spit out the entire rsync stdout, but only after it was done.

02:32 <bobthecow> ddfreyne: i do think output.diff is useful.

02:32 <bobthecow> i use it a lot.

02:32 <bobthecow> i don't like checking my output into version control, just my source.

07:12 <ddfreyne> bobthecow: Ah yeah

19:44 <musicmatze> How does nanoc store checksums? In a JSON?

19:44 <musicmatze> it does not look like a JSON, that's why I ask...

19:45 <musicmatze> My question is: Wouldn't it be faster just storing a (sorted) list of checksums? If a checksum cannot be found, recompile the item, if it can be found, don't? That will be my approach, I think...