#metabrainz

ishaanshah[m]: no, I'll try to look today, but no promises.

pristine__: Okay.

Ok, no problem

[sir] yvanzo opened pull request #111 (master…sqlalchemy13): Upgrade SQLAlchemy from 1.0 to 1.3 https://github.com/metabrainz/sir/pull/111

ruaok, iliekcomputers, anyone python/sqlalchemy-savvy: please take a look at these PRs when you have time ^^

I think it might me time to rename this channel to #lookatmyprprettyplease . Good problems to have.

And buying a new keyboard.

I know you don't have much time at the moment, it is more an update about the work-in-progress.

yvanzo: around for whatever is needed :)

Project bookbrainz-site build #2565: passed in 2 min 11 sec: https://travis-ci.org/bookbrainz/bookbrainz-sit...

Project bookbrainz-site build #2566: passed in 2 min 16 sec: https://travis-ci.org/bookbrainz/bookbrainz-sit...

pristine__: ping

Mr_Monkey are you busy with the gci for now?

ruaok: pong!

I have created a PR

hey.

Hey!

When free do have a look as I am not 100% if the logic is correct or not.

*100% sure

pristine__: hey.

so, I was pondering your function naming conundrum.

your function names are sometimes not very descriptive and you've expressed frustration with naming functions.

:(

but then you give a short, succinct description of what the function does.

which is often times a very good function name.

as I've already written in the review, why not name the function something intentionally wrong, then write the quip that describes the function and then go back and update the function name to reflect the quip?

you clearly know how to do it, but I think you feel some pressure about naming functions since they are more permanent than comments are. I think.

loi, this is why all my tickets have shait names too :D

ruaok: thank you for the review :) if you could write comments to point out a bad function name (since there are many), it will help me better. Other wise I can refactor the whole PR.

:)

I do that already, but I am convinced you can do this yourself -- you've demonstrated this several times. I'm just wondering if I can guide to do this yourself...

When I think of it, I have always tried to give nice function names. At this moment I cannot think of any bad function name but I if you and iliekcomputers feel that way, I am convinced that it should be improved.

> def get_spark_dump_path(self, directory, dump_id, ftp_dump_dir, dump_id_pos):

> """ Download dump and get local (spark) path.

[bookbrainz-site] MonkeyDo merged pull request #346 (master…ORM-update): ORM update requires some changes https://github.com/bookbrainz/bookbrainz-site/p...

that is a good example.

download_dump_and_get_path() is a totally good name.

download_spark_dump_and_get_path() is even better, I suppose.

do you see what I am saying?

Project bookbrainz-site build #2568: passed in 2 min 9 sec: https://travis-ci.org/bookbrainz/bookbrainz-sit...

abhishekpanwar: Looking at your PR now

ruaok: cool. Thanks:)

Apologies for making you read bad code.

its not bad, don't get me wrong. the code is doing just fine.

and you know how to describe functions, I'm just trying to nudge you to use your very good descriptions for names.

pristine__: https://github.com/metabrainz/listenbrainz-serv...

here you test adding an invalid path, but I dont see how that test result is being evaluated.

pristine__: finding good function/variables names is an hard task, for everyone, but on the long term it helps all contributors to understand code, without going to each function description and code. Don't worry, every dev went through this, and experienced devs still struggle with it ;)

ohhh, wait. I think I get it now.

Mr_Monkey: Okay

> here you test adding an invalid path, but I dont see how that test result is being evaluated.

That will be evaluated later.

a question to people from India: does it make sense to provide Hindi translation for our apps? https://www.transifex.com/musicbrainz/musicbrai... is pretty empty

k, I just realized that.

zas: while I am not the target audience for your question, the answer should be yes.

zas: Mr_Monkey thanks :)

there are plenty of people in india who do not speak english. and some regions think that english is their official language.

we should not think of india as one monolithic cultural entity, but like many other large countries as an amalgamation of many different cultures. and of these groups needs picard in hindi.

I would prefer us to find someone to fill up the hindi translations.

ruaok: I totally agree, we're about to blog post about new Picard beta, with a call to translators, I expect someone would pick Hindi translation, with all nice Indian people we have around. If interested, head to https://www.transifex.com/musicbrainz/musicbrai... (it requires a free transifex account)

current status ~2% translated...

I bet we could also get people interested in contributing to picard for gsoc to write the translations as a first community task.

zas: maybe add that to the gsoc ideas page as a first task for getting aqcuainted with the community?

good idea

hi, I'd love to help with this if possible! (even if it is a gsoc task?)

antara: please do, no need to help. see the link above to jump in.

no need to wait. sigh.

okay, found it - thanks!

zas: yes

(to the i18n question)

For that there's a need to help...

reosarevok: call the lost luggage dept at Swiss?

or go buy chocolate?

antara: if you need help regarding i18n and Picard, ask me

we plan a beta on Thursday, which means string freeze until final release (prolly in few weeks), it'd be amazing to have Hindi translation by this time ;)

just curious, how many string are there in picard?

okay, thank you :)

ruaok: 1177, 3824 words

I do have school, so I'm not sure how active I can be, but I'll try to do as much as possible

and there are few more in picard_appstream

11 lakhs worth, zas?

ahah, 1177 strings ;)

ahh, I see.

lots.

yup, huge app ;)

I'm getting started on countries now :D

thanks!

note: MusicBrainz website and Picard are sharing some translations, like countries

oh okay! overlap might be helpful

I would love to see Picard translation getting improved in the next release cycle. I'm not happy with a few things

I don't think so, at least, never heard about using this alternative. Something we may evaluate, what are pros/cons compared to transifex?

I see we can self-host it (for free/full featured) or just use free version (for open source projects) but it has few limitations compared to paid versions

It has pretty great integration with git, is free software, can be self hosted but the developer also provides free hosting for open source and the developer is a nice guy

I see one advantage: it is open source, https://github.com/WeblateOrg/weblate/

the thing is moving current translator community from transifex to weblate may not be very good, as we may lose few people during the process

Also nice is priorization of strings (so you can mark what is really important), adding screenshots and MT integration

pontoon has been suggested as a GSoC project last year.

Don't know that one

transifex has git integration too btw.

What I really don't like about Transifex is that it does not really make it transparent who has translated what. But it does not make sense for a sine MB project to use something different

ok, perhaps time to list alternatives and evaluate if it is worth moving to one. For sure, moving to an open source one is a strong argument for me... and if it helps to have a wider translator community while providing similar or better features than our current solution...

Transifex gained a proper GitHub integration, that was lacking for a long time. I never used it, but we should try if it works for us with Picard

Imho, Github integration shouldn't be top requirement, we still don't know where github will go in future years ... (reminder: it is M$ property, and open source community should be careful). But if it has also gitlab (or other alternatives to gtihub) integration. GitHub integration is a plus for sure, at least until we are forced to move our projects away from it....

I think I evaluated weblate for GSoC ideas two years ago, pontoon was pretty bleeding-edge was featuring in-place localization (for websites), but I don’t recall what dismissed weblate.

zas: it is not just github, it is git: https://www.transifex.com/git-extensions/git-ex...

oops

wrong link but statement is the same.

Transifex just feels pretty limited in most regards, e.g. glossary is very limited, nearly no quality checks, TM export I think is not possible in the free version

outsidecontext: I tend to agree (about transifex)

no gitlab integration :( I just imagined it

+1 glossary is very limited.

Also while the translation editor itself is fine, the rest of the UI is confusing. Also requiring user's to register at a commercial nthird party is not so great

Weblate actually would allow oauth 2 based single sign on using MB account

Maybe we should start a comparison of different tools and MB requirements. I can start something to get into a discussion

zas: not to ping you too on this much, but any chance you looked into the code search livegrep thing?

Mr_Monkey: So for each of the relationship in the relationship_set of the deleted work, we would have to go to the target of these relationships and then delete the relationships from these target entity's relationship_set where the deleted work is the source.

iliekcomputers: I didn't had time today, but it is planned ;) I think about setting up a VM for it

Thank you! It'd be amazing!