#musicbrainz-devel

ianmcorvidae: ping

pong

I just wanted to catch you up on the developments of today.

I went back to DWNI just as they were winding down and opening wine bottles.

perfect timing.

I asked about more space, since there is one other server in the rack besides us.

I forget the quoted price, but it comes with 3 amps of power.

offsetting the 3 amps we already use, its not much extra. like $75/month or so.

but, the servers are not accessible to me right now, since the owners of the place are out of town for a funeral.

so, this deal includes Tim, the owner of DWNI paying a visit to the owners of the place and expressing his condolences.

and picking our three servers in storage.

he’ll then get his team to rack the servers

so, we will have 3 more dark servers coming in, including tails, would would be an adequate replacement for lenny.

I think its three servers.

I have to check my records.

so, this is effectively splitting the difference. or something like that.

sure

more computing power, more options, not much more money.

I’ll check out what machines we have and worst case I might prep one machine in the trunk and install it at DWNI tomorrow.

I’ll see.

is logging

Users on #musicbrainz-devel: MBJenkins sampsyo njh derwin Muz warp JonnyJD

is there anyone awake that is working on acousticbrainz?

dns53: "Working" how?

well i'm trying to work out how it will work

Then it's probably better to ask that.

so you generate a json file and you are trying to match it against the database

No.

Or maybe. I'm not sure what the question is.

well you seem to be storing the results as text and using python to do the comparison

why not use the json data type in postgres and / or the hstore data type

dns53: Using Python to do... what comparison?

dns53: Also, the data is stored server-side in a pgSQL db. Check https://github.com/metabrainz/acousticbrainz-se... for how it's retrieved.

dns53: I'm still not sure what your question is...

not really a question, just trying to think aloud

i'm waiting for the data dumps, i think there are better ways of storing the data

dns53: The current process is that people run one of the essentia extractors on their music which generates a JSON document. Then a submitter is used to send it off to ABz (currently abzsubmit or acousticbrainz-gui - I'm working on a Picard plug-in, maybe ijabz will make one for his tagger too) where it will be stored in a postgresql db.

nikki: hi

We use the json data type in pg

Dumps will come as a json document per mbid, which is the same as what the extractor returns

Freso: dns53 found the json diff tool in the utils/tools folder of acousticbrainz and thought that was part of the actual workflow

Ah.

502s atm

many 502s

I’ve been kicking search servers.

should be better now.

ianmcorvidae: ping

ruaok: pong

so, tomorrow.

shortly before 6pm at firefly car rental near SAN airport?

I think that was the notion

just confirming.