#metabrainz

Gentlecat:

LordSputnik: I am afraid that it doesn't, sorry

are you trying to wake me up or something?

just post your questions here or on GitHub

Gentlecat: nvm. It can wait. Sorry for waking you

I'll respond as soon as I can, or somebody else will

LordSputnik: I've forgot to add it.

(to add id=1, id=2 and I thought that it is working, because I've seen no errors)

Strangely, after deleting all this languages adding from fixture.py, it worked. I might also say, that printing language_id shown numbers such as 1008 and 1009, and it haven't cleared in the next execution of my program.

LordSputnik: ?

stanislas: it will show 1008 and 1009 if you haven't set IDs, because it's using the next value of the language_id sequence

stanislas: It might actually be better to put the TRUNCATE in fixtures.py, right before you add the languages

opatel99: wait for what?

Gentlecat: Wait until you have woken up

after sleep^

may I ask why?

"are you trying to wake me up or something?" > I thought that implied you were sleeping

do you have a question?

I wanted to discuss http://stackoverflow.com/questions/3412275/load... . I was trying to find a way to render Jinja in the js after it is compiled

the only way jinja and js should be connected is by jinja calling functions

stanislas: Let me know if you're having trouble with the language stuff or are ready to do the commit squashing

scripts (bundles) should be imported from html, as usual

Then another concern. I reference a script, but when I call the function from the load more results button, I get not defined. If I place a console.log in the bundle, the log does output to browser

probably because definition should be before the call

I tried that

I don't remember exactly how it is implemented

you might need to replace `onclick` attribute with a listener

trying to reference a function defined in a bundle?

opatel99: wait for document.ready before adding the listeners

It's either that the DOM isn't loaded when you try to assign the handler or that the handler JS isn't loaded at the point when you bind it to the event

(most likely anyway, I haven't been following too closely)

bitmap: yep https://github.com/metabrainz/critiquebrainz/bl...

all files are wrapped in functions, so the load_more() function only exists in the file which it's defined in

LordSputnik: I've not seen that I've lost the connection, sorry.

if you need to reference it from the template, add load_more to the window or an object defined on the window

Bitmap: zas is busy today and I'm about to go flying.

Can you please keep an eye on things?

LordSputnik: it doesn't work

I can do

Note: Dora is out and will stay out.

LordSputnik: even if I put this line before lan1 = Lan...

okay

I think having two servers sometimes causes issues.

Index rotation is off.

stanislas: OK, I'll have a look

LordSputnik: but now different errore

LordSputnik: It says something about alias table

bitmap: I will try that.

HINT says that I should cut at the same time alias or use TRUNCATE ... CASCADE.

stanislas: add CASCADE at the end :)

opatel99: the better option is to just attach an event listener in the file where load_more is defined

you can either do what LordSputnik said or use event delegation, which doesn't need the elements to exist

LordSputnik: doesn't working too

bitmap: Like this document.getElementById("more-button").addEventListener("click", load_more);

LordSputnik: the same error as before

LordSputnik: id=1 exists

opatel99: $(document).on('click', '#more-button', load_more)

event will bubble up to the document, second arg filters on #more-button selector

ruaok: any advice if search freaks out?

Follow the instructions in syswiki.

If that doesn't work, take both search servers offline for a minute or two.

If that doesn't work, take everything offline.

sudo crm_mon -1r

To find out which server is running which services.

LordSputnik : ?

ruaok: got it, thanks

Good luck. :-)

stanislas: run db.session.execute(...) when truncating, that should fix it

when ?

What do you mean by ... ?

And this previous TRUNCATE should be in db.engine.execute or in db.session.execute ?

db.session.execute

LordSputnik: changing only this previous line worked :)

commited and pushed

LordSputnik: Do you want me to clean up these commits ?

stanislas: yes please - do you remember how to do it? If not, check that chatlogs, the instructions should be pretty much the same :)

And I've been wondering if changing fixture.py to more iterative style would be appropriate (it is mainly copy-paste, but I left it since I've encountered it thay way)

It would be less readable that way, but I don't like how it's done there. What do you think LordSputnik ?

LordSputnik: I will look in bash_history

stanislas: You mean with loops to create larger numbers of entities? I think so too

We can probably randomise most of the entity content, I think

These are all real examples now :)

We could have a set of functions to generate random values for the scalar types, and use those in loops

I've done it for fun really

I've pasted some real names to google translate and selected some weird language

bitmap: crashed again.

the dora out trick didn't work. :(

But I see you point, doing it automatically would make some better tests.

everything is back and I now dropped the global rate limit to 1500.

bitmap: zas: I changed the following socket related items back, in hopes they may have made a difference:

applied on bert and ernie

LordSputnik: How can I move myself to another branch, I've some old data now

Did you add them to gateway cookbook ?

Safer if we re run cookbooks ....

oh hai.

no not added.

stanislas: Are there commits in the current branch you don't want to keep?

LordSputnik: I am afraid, I can't do that myself. There are branches now, and I have no idea how to do that.

anyway, the global rate limit is 1500 now.

see if we can be stable with a lower limit.

LordSputnik: No

stanislas: What branch are you in locally?

I think that it is master, but i don't really know. I've used PyCharm git client before.

stanislas: OK, go to the directory your source code is in, and run "git status" to find out the local branch

Mineo: can you please help zas understand what is required to deploy the solr search stuff?

I think we may just want to bite the bullet and deploy it.

any bugs that need fixing are likely less troublesome than the shit we're dealing with now.

creator_list_test

and I am sick and tired of wasting massive amounts of energy into fixing code we're going to throw away.

so, I would like to start the conversations about how to deploy the search servers onto our infrastructure.

stanislas: OK, isn't that what you want to be on? Why do you want to move branch?

and then we can print out the search server code and have ourselves an office style PC LOAD LETTER moment.

ok, enough for now.

LordSputnik: I don't want to move it

stanislas: but you just said you wanted to move to another branch?

LordSputnik: But when I am looking on the files the new ones aren't there, It might be usable that I've done git reset --hard HEAD~1 21 times

LordSputnik: I don't remember saying that.

Maybe we didn't understand each other, when I was saying that.

Ok hold on, could you paste the commands you've done so far in a gist?

ruaok: ok

zas: ping

bash_history is deleting the duplicates, I have copied the whole directory somewhere else on my computer, we can use that copy and forget this local thing.

LordSputnik: What do you think ?

OK, let's start over with your up to date work then

First, you'll want to do "git reset HEAD~20" (not --hard!)

Actually

let's make it really easy

"git reset --soft 178269e1cb5a2d99ef6957f03bead7241b59fa7f"

Then go through and make the commits you want

(with git add and git add --patch)

ok, done

LordSputnik: ^

now commit and push

?

stanislas: are you only going to put it in 1 commit?

yes

Generally it's best to have 1 commit correspond to a complete feature at a particular point

LordSputnik: so now commmit and push :) ?

So having everything in one commit isn't great - ideally you would have created the EntityTestCase class, committed, added some working complete functions, committed, and then kept going like that, writing useful commit messages as you went

Now, I think it would be best to separate your work into one commit for the test class and one commit for each entity

(but please try to do more useful commit messages in future so we don't have to go back and rewrite history in this way)

LordSputnik: Since there is an avatar on the creator display page eg. https://bookbrainz.org/creator/4e8f8704-7c34-4a... ; does that mean it is an option to add a picture to a creator? Like MB?

gcilou: there will be, but not yet unfortunately

we need to sort out stuff with the CAA, and additionally implement linking to wikimedia commons images