he is currently in san francisco, he was on irc a few hours ago.
2012-10-18 29253, 2012
ruaok
I got that much, ya.
2012-10-18 29209, 2012
ruaok
he'll soon realize that he needs a ride to the south bay. :)
2012-10-18 29238, 2012
ruaok
ocharles: take bart to SFO and be at SFO at… 1800
2012-10-18 29247, 2012
ruaok
that should work out ok.
2012-10-18 29200, 2012
warp
well, he can always take the bus
2012-10-18 29210, 2012
warp
;)
2012-10-18 29212, 2012
ruaok
you mean three buses. :)
2012-10-18 29212, 2012
RootWyrm_
Does BART even run that late?
2012-10-18 29214, 2012
ruaok
or 5
2012-10-18 29224, 2012
warp
ruaok: it was three for me last year, yes
2012-10-18 29225, 2012
ruaok
at 6pm?
2012-10-18 29227, 2012
ruaok
yes.
2012-10-18 29235, 2012
ruaok
from SFO to sunnyvale.
2012-10-18 29238, 2012
RootWyrm_
Oh 1800 gods, I'm tired.
2012-10-18 29203, 2012
ruaok
I guess he can bart to milbrae and catch a bus from here
2012-10-18 29205, 2012
RootWyrm_
OOW10 I got caught out because I was a train and a bus away from my hotel.
2012-10-18 29253, 2012
RootWyrm_
"It's only freaking 1AM and the trains stopped?!"
2012-10-18 29257, 2012
ruaok
k bed time for me.
2012-10-18 29258, 2012
warp
ianmcorvidae: alright, shall we look at that edit stuff before I go all around the city preparing for imminent travel?
2012-10-18 29259, 2012
ruaok
nn
2012-10-18 29209, 2012
warp
goodnight ruaok
2012-10-18 29220, 2012
ianmcorvidae
warp: yes, lets, if you have a bit of time
2012-10-18 29257, 2012
warp
ianmcorvidae: have you looked at the code and/or the output?
2012-10-18 29212, 2012
ianmcorvidae
I've looked at the code, not yet the output
2012-10-18 29230, 2012
ijabz joined the channel
2012-10-18 29200, 2012
ianmcorvidae
I'm wondering what cross-checking you've done -- how closely have you compared the sets it produces with known problem sets (e.g. what modbot complains about)?
2012-10-18 29220, 2012
ianmcorvidae
it looks largely obvious and I think we can probably go forward with it, but just for basic checking :)
2012-10-18 29259, 2012
ianmcorvidae runs a copy of it to see what I get
2012-10-18 29206, 2012
warp
the query only grabs edits which have a comment":null in the data field, so that should only grab edits which are actually affected.
2012-10-18 29243, 2012
RootWyrm_
ianmcorvidae: BTW, I forgot to thank you. That agg exposed a bug I missed.
2012-10-18 29251, 2012
warp
we have a comment field on many entities, but as far as I can tell the edits are all either $editdata->{comment} or $editdata->{new}->{comment}.
2012-10-18 29247, 2012
warp double checks that.
2012-10-18 29219, 2012
ianmcorvidae
RootWyrm_: hah; well, glad to be of assistance :P
2012-10-18 29242, 2012
ianmcorvidae
warp: I was just checking that we always double-quote "comment" and that null never has a space before it, looks like we're good on that count though
2012-10-18 29209, 2012
RootWyrm_
ianmcorvidae: heh, I don't know if I'd go -that- far.. I don't know that I can FIX the bug. ;(
2012-10-18 29215, 2012
ianmcorvidae
heh
2012-10-18 29236, 2012
warp
ianmcorvidae: ah, yes, I assumed encode_json() to be consistent there.
2012-10-18 29259, 2012
warp
or atleast has been for the past X weeks, because we're only looking at open edits.
2012-10-18 29258, 2012
ianmcorvidae
yeah, I think we're fine, I just wanted to double-check
2012-10-18 29205, 2012
warp nods.
2012-10-18 29214, 2012
ianmcorvidae
you're happy re: data->comment and data->new->comment being the only targets?
2012-10-18 29224, 2012
warp
it touches 3523 edits on my local dev machine.
2012-10-18 29230, 2012
ianmcorvidae
3798 here
2012-10-18 29236, 2012
warp
ianmcorvidae: yes, I checked all the edit types with comment in them ( git grep comment | sed 's/:.*//' | sort | uniq | xargs )
