#metabrainz

And I'm talking about MusicBottle. :p

fixing that doesn't require a schema change, since the mbids are already in the db

UmkaDK_: It won't be.

:(

And as kepstin said, it doesn't require a schema change.

I don't think it'd be a terribly hard thing to implement, so if you bribe ruaok to tell bitmap to get to it soonishly after the schema change, it might not be too far off.

So once the schema upgrade goes live, we'll have the data to list multiple MBIDs for an artist (for example), but we'll have no way of getting that data out of the database...

Ok, I can do that :)

let's try that one again...

UmkaDK_: We *already* have the data to list multiple MBIDs for an artist.

We just haven't got an endpoint displaying this data.

We have the data for years.

I thought you are leaving? ;)

(MBS-3946 was created in 2011, so we have had the data since before that.)

I am!

Yee, I know the data is there. But there is no way of using it without it being exposed via the API

yep, but that api could be written any time. you could get out your time machine and bribe someone to add it 5 years ago, and it would work

the data's already there, and you can already get it if you're accessing the db directly.

:)

(it's *sort of* exposed - if you access a page via an old mbid, you'll get redirected to the new one on the website, and if you make an api call with an old mbid, it'll return the current one in the response)

Unfortunately, the people that want to use the data are accessing it via the API

also sometimes, the url you're on will be another one, but the entity you're on will be the redirected one, source: happened ot me a few times

usually by old links in edit notes (not edit entities)

actual text lniks in the notes.. the edit itself this never happens

Freso: I have now :p

(I missed it originally)

Freso: re that DuckDuckGo mail, can't they use the same JSONLD stuff Google does?

asterix starts to bug again :(

duckduckgo \o/

heh, that's pretty cool

reosarevok: my impression was that duckduckgo doesn't actually run their own crawlers, they outsource that? all the fancy search modes are implemented by hitting external search apis

kepstin: oh meh. Guess not then :(

http://www.last.fm/dashboard "all time scrobbles"

hey official dev peoples: in the ws/2 docs, the wording about user-agent strings varies in strength.

sometimes one 'should' have one, sometimes one 'needs to use one'.

How strongly should I word it?

it's required

in fact, many standard library user agents (python, java, etc) are outright blocked

I guess 'must' is better, if we're following RFC strengths

required as in "requests that don't provide a user-agent string are blocked"?

I'm not sure how true that is - and if it differs between "no header User-Agent" and "standard header for whatever library you're using"

you'll have to confirm with zas or ruaok or rvedotrc

for context, I'm doing a Q/A format.

A. No, but you [should|must] provide a meaningful user agent string.

I vote must

Different topic, I don't know if this is something chrisskye tracks, but it's nice to be name-checked. http://www.news.cornell.edu/stories/2016/04/gra...

cool. I don't recognise any of the names from the academic world

but we should promote these uses as much as we do companies on metabrainz.org

alastairp: will you have time to get to AB pull requests any time soon?

it's on my "this week" list

so hopefully this week or next :)

next month then? :D

(as opposed to the last few weeks where it hasn't been on any list)

Thanks CallerNo6!

It is nice to be recognized...now I’m going to reach out to that project team...and see if they can support us...

alastairp: I CCd you on that email about Dan's project

Gentlecat: I saw. good recommendations

chrisskye: confirm with rob, but typically we don't ask for anything from universities

I agree with alastairp about how great it is to have academic users.

did you have some other plans with messybrainz?

Gentlecat: no plans yet

alright

just a vague idea that it might be useful to use to generate ids for items in these datasets

also, dimi has been working on some stuff to match audio in datasets that we have to actual mbids

The thing is...we are a foundation. We don’t ask other non-profits for donations, because they too need support. But when you have for profit academic institutions or groups who have funding at that level...it is appropriate to ask!

They got a $1million grant for their data project!

if MeB doesn't want money from academia, maybe look at who's making the grants?

I’m going to remind everyone again - we are not asking for $$$ for service. We don’t sell anything. We are asking for funding from sources who can afford to help.

chrisskye, CallerNo6: we have a section for those https://metabrainz.org/supporters/tiers/8 :)

yes! But, that doesn’t mean organizations can’t donate to us!!!!

right

An institution, non-profit, academic or otherwise...who has millions of dollars can help us!

It is about sharing!!!!

Inter-institution support.

We don’t have customers. We don’t have clients. We have a community.

chrisskye: I've been involved in writing academic grants

And, we will take funding, donations and financial support equally from those with the goodwill to do so.

they normally don't have a line item "$xx to give to some non-profit"

no they do not.

and if they don't, it can be pretty difficult to justify it in the accounts at the end of the year

But, Harvard et.al. has resources...

not saying they shouldn't

alastairp: I have 15+ years of university...trust me I know...

I was a legal academic...and you know what...firms pay to have their name posted on things...

Not all things are equal in academia.

I’ve worked on government funded research where they were begging me to spend $5,000 in less than 3 weeks...to “justify” the books...

"MetaBrainz: Where You Can Sink Your Extra Budget!" :)

anyhow... giving to us is always voluntary

^

alastairp: I'll get back to hacking on my project (and perhaps an API) after I finish that assignment

...now if only we were registered as a charity in another country...where there were tax incentives...like Canada.

...people give more when they can claim 40% of it back in tax refunds...

maybe will come to uni again and discuss things

I want $$$ for a serious server upgrade...so we can serve people who want to hit the database hard and fast...

And, funding for the Task Rabbit project...

Leftmost: ‘extra budget sink’ is a good idea...but maybe I need a list of old important people in the tech world...so we can inherit...

I like it.

alastairp, CallerNo6: no UA string or "-" are both blocked IIRC.

or severely limited.

Or hire a team of elite burglars to change people's wills to inherit all the assets to MeB >:D

elite burgers? NOM!

ruaok, so that's another vote for 'must'?

wait what was the question? I've been talking MB all day and trying to catch up on calories right now...

<CallerNo6> Q. Do I need an API key?

<CallerNo6> A. No, but you [should|must] provide a meaningful user agent string.

(for the ws/2 docs)

ruaok: we had to take asterix out of rotation, bitmap and i found a serious issue, which concerns all web servers but asterix doesn't resist as well as others (perhaps because it is using reiserfs). Basically in certain cases we get huge logs, eating disk resource, causing uber slow down. Bitmap is currently trying to find a way to prevent the main issue, a

complete fix will need more time :(

should we reinstall the OS on asterix?

the main trigger is that the catalyst logger includes all function arguments in timeout stack traces, which can be several megabytes long for certain releases...

so I'm just going to add a sane limit there for now

bitmap: if it is enough to prevent the slow down, good for now

ruaok: reinstalling to get rid of reiserfs ? well, since we should move elsewhere soon, not sure it is worth.

k

we really need to accelerate the schedule of moving.

we're running out of time.

bitmap: thank you for the awesome email about all the banking details. for once someone sent too much banking info. :)

haha. well, no problem

ruaok: yes, i hope we'll be able to proceed during June or early July, not later.

zas: perfect we'll make a detailed plan when you and bitmap are here.

I hope so.

and, consider this. we don't need to move everthing all at once.

discourse, metabrainz, critiquebrainz and some others can come later.

Yes, we need to focus on what is eating the most bandwidth/power

And it is the db + ws

We need to ensure we can install everything fast enough, gateways are now fully in chef, web servers too, search servers too (but not the hacky autorestart yet)

So basically we can do it under a week, if all work together, and we have servers + dist ready