#metabrainz

we'll see.

ok, everyone, thanks for your time. and please keep patient when using mb.org, ok?

no shooting rampages, ok? had enough of those to last a lifetime.

Freso: am I late for the meeting again?

no, go!

I was about to close,but we've got plenty of time

The PR seemed good to the mentors but needs some modifications.

This week I plan to start writing the client which will make request to the server for getting all the pending jobs for the user

fin :)

\ø/

ok, thanks everyone!

</BANG>

this is weird.

this was the 1st time i was waiting for the Freso's queue as a call for the start of meeting.

and sadly it did not happen. :'(

gentlecat: I’m working a little bit more on the client side, but I already implemented the API endpoint to receive the datasets, still missing the evaluation part

armalcolite: you can just set a reminder. it's happening every week at the same hour

i set a reminder for 2 hrs before meeting so i can get ready

then i engage in other stuff. (today was very critical one)

gentlecat: I think that the two PR I opened are becoming very big and I would like to discuss with you how to divide the future work in smaller chunks to permit faster merge

server PR is fine

I'd just prefer not to merge functions that are unused

you can split up meaningful parts into commits

I don't see the API endpoint btw

oh yes I wanted to add the dataset evaluation and then commit!

between tonight and tomorrow :)

what's in dataset evaluation?

I would suggest to focus on one thing at a time

it would be easier to review and merge

otherwise you are writing parts that depend on each other at the same time

if one needs changes, others probably too

yes, this is why I had the feeling that the PR where becoming big

that's not a problem

the problem is that we can't merge any of them because they are unfinished

oh yes I agree, so I’ll try to reach a consistent point in next days

I’m very good in making things more complex that needed :P

here's my suggestion: finish server changes first

without them you can't do anything

yeah, it’s the second time you tell me that! :( I will focus on that untill I finish!!!

then implement support for these features in the client

if we are doing it backwards, then it will never get done

merging everything at the end of gsoc is not going to work

…yes… I agree, by the way now I can focus on server and still have a ‘beta’ of the client to test the API endpoints!

can you push your endpoint updates?

oh yes if you want, but it’s really basic, check if the input is a dictionary and returns a response to the client!

if you want feedback on what you already have

I definitely recommend to get what you've done uploaded as soon as possible

especially for feedback

not sure if if you've been following the discussions we've been having here

but I said the same to kartikgupta0909 today

yes I read the discussion today, that’s why I was thinking also to the size of the PR …

we should try and get things merged in as soon as they're finished. especially many of the things that you are working on are very distinct from the existing site, so there will be no problem to release often

pushed the endpoint!

open a pull request too

it helps us keep all of our comments in the same place

oh, is this the same branch that you had?

Yes it is in the same branch!

our network future is in the hands of the USPS.

how sturdily built is the switch? :D

well, at least our switch is purported to be shipped.

that is another aspect, yes.

I'm more worried by the bizarre time-warp issues that seem to follow the USPS.

Project acousticbrainz-server build #92: SUCCESS in 5 min 30 sec: https://ci.metabrainz.org/job/acousticbrainz-se...

tracking entered into the system 3 days after delivery.

or package arrives 3 days after the site said it was delivered.

QuoraUK: ping

LordSputnik: Pong

QuoraUK: how has it gone today? :)

Good I think, made some decent progress

got a good start on the profile display for achievements

Hello from opatel99 :)

started on the Creator Creator achievement, just trying to figure out how to get the number of Creators created out of the db, should have it done tommorow

once I have it, i basically have a tonne of new ones for free

hmm, you'd want to get creator revisions which have no parent

QuoraUK: you started doing the achievement page with Jade, but if you want to be able to select achievements on that page to show on the profile, you'll probably need to use React and JS

ahh, ok

I change that tommorow :P

Prototyping the HTML and CSS in Jade is probably fine though

yeah, I at lease have an idea of how it will look now anyway

But start learning React in the background so that you're ready for moving it over later in the week

sure, can do

good work on getting the refactoring done too

wasn't as bad as I thought tbh

And I think you're right on the VARCHAR vs. TEXT thing, having a limit helps ensure nothing crazy can happen with display

https://www.irccloud.com/pastebin/AYOxjiWF/ was chrisskye's update for the meeting.

yeah, I dont think they should be too long anyway, trying it on my own username, it looked pretty silly after 20 or so

QuoraUK: is 2083 the standard maximum length for a URL?

From what I found yeah

at least most browsers wont handle a url longer than that

thats why I chose 2047 anyway

thats why I chose 2087 :P

yeah, thats what i found

OK, that makes sense

Let's make that 2000, to be on the safe side, and to make it easier to remember

ok sure

Then I'll merge the PR

I was thinking, how will we store the chosen achievements?

will we need another table? or something to add to another table

Good point

perhaps modify unlocks

Yes, there are two ways to do that - add a "chosen" flag to the unlock table, or add a separate table to store the chosen ones

In the past I would've said do the latter, but practically, the former is gonna lead to simpler code, so it's probably better

I think I prefer that practically to be honest

could perhaps rank them as well

And if each person unlocks 50 achievements, and we have a million users, that's still only 50 MB even if PG is storing booleans as bytes

Yeah, might as well rank them too

ok, that sounds good :p

so you could have a null value for not being displayed, and then a small integer giving the display order for those that are

cool, Ill add that in with the 2000char limit and I think that should be it for sql

OK good :) What's your plan for tomorrow?

I'd like to work on the x_revisions achievements a little bit and see if I can figure it out

have another look into ui

do that sql stuff

move over to react for the profile achievments

actually, I still need to do the svg images, I could do that instead of ui perhaps

OK

That sounds good :)

I have plenty to get on with anyway :P

Just wanted to add as a side note: It's always best to be careful when deciding whether or not to allow NULL, because it doesn't have any built-in meaning and complicates logic (see http://stackoverflow.com/questions/163434/are-n...)

But here, it's OK because the alternative would probably complicate the code even more

And to be honest, some of the tables we currently have for this reason would be better off being merged (eg. suspended_users)

alright fair enough, Ill keep that in mind :P

Just so you know if you need it in future ;)

Anyway, think we're done, anything you want to talk about?

I think im good for today :D

OK, great :) Speak tomorrow!

alright, till tommorow :D

oh well, first switch purchase went bad. it will arrive on wednesday.

purchased another one from a more reputable place.

and paid the astronomical shipping fee. sigh.