[Tolriq]

Because tags are not read for videos. Tags and NFOs are still very much both local information.

Have you only read the whole thread and the PR discussion before jump in and answer that ?

[zag]

Just a few ideas for the future, forgive me if I'm just repeating stuff people have already mentioned. The github topic went a little off topic so I stopped reading.

- Combine the Album and Artist scrapers
- Ask the user when they add a music folder if they want to scrape metadata automatically (remove other options in settings to reduce confusion)
- Let them choose scraper there and then
- Standardize the way audio is added as much as possible with videos so the process is similar.

Just remember most users wont care about these options, and should not have to understand them. By default music should be scanned in with a scraper so the user gets the best experience of artwork and details.

Get these things in, and music will be a lot more used by the general userbase, I hope

[Tolriq]

Well not sure this is repeating, since there was a lot of argument against 1st point that invalidate point 2, and lot's of discussion about why point 4 is false too.

Point 3 is still valid

Anyway I know some do not like me because I may sound aggressive and other bla bla like I say what I think directly without trying to be nice since time is precious and this is not really the way to go after trying for many years.

Anyway I'd just like to repeat again something I have real world precise stats and can have more precise ones about how users use Kodi, and those data are real data, not feeling about this or that like most of the time decision are made from assumptions.

For example over the last month I have stats about media played through Yatse (this is limited to user having a phone on Android and using Yatse ok, but it's still 800 000 monthly active users)

Over the last month for users that have not disabled anonymous stats or using blocker I have data about 12 000 000 items played.



Those are big numbers and should be used by the Team if there was possible discussions.

Unfortunately I no more have the stats about scraped data or not as I no more need them to take decisions, but I can put those back.
Again the number I give are real numbers not fantasy from someone imaginations.

[DaveBlake]

Just a few ideas for the future, forgive me if I'm just repeating stuff people have already mentioned. The github topic went a little off topic so I stopped reading.

- Combine the Album and Artist scrapers

Not good. User may want to only scrape Artists e.g. classical music where scraper can correctly identify artists like "Beethoven" but is weak at identifying recordings by different orchestras and conductors. Or user may want to scrape albums, but because of language may want to read Artist biographies etc. in own language or from separate source.

- Ask the user when they add a music folder if they want to scrape metadata automatically (remove other options in settings to reduce confusion)

Also need to ask which artists, albums or both, and before that do they want to scan (tag data) to library and if not you can't scrape anyway.

For this and previous point, to be honest users either want scraping to be completely automated and invisible, so don't need to even see an option (but might wonder why something is so slow), or they want total control over what is scraped with what and when. We need to cater for both.

- Let them choose scraper there and then

Yes, consensus on that!

- Standardize the way audio is added as much as possible with videos so the process is similar.

More closer yes, but with great care given the fundamental differences between material. This means listening to those of us that see the differences even if it seems all the same to you. Maybe we have a point you have missed.

Shame you didn't read all the PR thread.

Just remember most users wont care about these options, and should not have to understand them.

Yes, most users just want a big "Play my music" button

I think we should be able to produce a system that is easy for novices and still powerful and flexible for the many diverse experienced users Kodi has. I have done so many times in other projects. Power users don't have to be marginalised or disregarded either.

By default music should be scanned in with a scraper so the user gets the best experience of artwork and details.

To be more precise music needs to have its tags scanned so that extra info can then be scraped from external sources. I know you are the audioDB guy, and when it works it can be great. So yes have this automated behaviour as the default for the novice (with a big warning on how slow it can be if you have a lot of music files). But please also make it very easy for things to be manually overridden.

[DaveBlake]

a) yes it should immediately after you add a new source
b) the "the refresh now" should only pop up when you changed a source and not on a new scan ("bug" in video set content)

So I add a new music source and up comes a dialog where I select scanning and scraper options and say OK. Does it then start scanning then scraping? Or does it prompt before actually doing the work?

My thinking is this: sometimes you just want to add a folder as source, not actually populate the library at that time. Say you just need to play music immediately from a network drive, not wait for all the info and artwork or even the tag data to be scanned. Up comes the dialog, all nice scraper settings etc., just what I normally want. How do I avoid it starting the scans? Going thorough turning off the scrapers etc. is slow, and not what I want next time either. A confirmation prompt would allow me to skip scanning and scraping, or have second thoughts on my settings if it is going to be a slow process.

Having a confirmation prompt on video scraping could be a good thing too, so maybe that "bug" in set content for a new scan just needs rewording rather than removing?

[DaveBlake]

Since, from the feedback I have, lot's of users just adds folder and never scan or scrape them, there's things to be done, and if done improperly this will increase the non library usage for music due to data not corresponding to user expectations.

Thanks for your data, this is always nice. But I think in this case, it's unfortunatly very misleading. I think most of these users are simply not aware, that there is a feature like this. And thats mostly our fault, as it is prominent in now way right now.

@Razze
I think you have misunderstood Toliq. He agrees with you something needs to be done to improve things for users, his data shows that users aren't using the library. But he is also saying we need to be careful not to make things worse.

Unfortunately in English your reply sounds patronising. I am going to assume that is a language issue, and maybe Tolriq didn't notice anyway.

[Martijn]

Well I agree that data is nice but it tells absolutely nothing.

And his post does not sound patronising at all.

[Martijn]

a) yes it should immediately after you add a new source
b) the "the refresh now" should only pop up when you changed a source and not on a new scan ("bug" in video set content)

So I add a new music source and up comes a dialog where I select scanning and scraper options and say OK. Does it then start scanning then scraping? Or does it prompt before actually doing the work?

My thinking is this: sometimes you just want to add a folder as source, not actually populate the library at that time. Say you just need to play music immediately from a network drive, not wait for all the info and artwork or even the tag data to be scanned. Up comes the dialog, all nice scraper settings etc., just what I normally want. How do I avoid it starting the scans? Going thorough turning off the scrapers etc. is slow, and not what I want next time either. A confirmation prompt would allow me to skip scanning and scraping, or have second thoughts on my settings if it is going to be a slow process.

Having a confirmation prompt on video scraping could be a good thing too, so maybe that "bug" in set content for a new scan just needs rewording rather than removing?

It will immediately start scanning the tags after you hit ok. The big OK button is already confirmation enough. Another confirmation dialog is redundant and not needed. If you have second thoughts just don't hit the OK button. It's simple as that.

If you just want to add a folder with scanning you do exactly the same as in video. Add source, set content to none and hit ok. So not scanning.

About not doing online scanning I already told you that's why there are options and even async scanning of tags and possible scraping of online info. Again if you hit ok you agree with these settings.

You are describing abnormal use cases. You set up a source once and that's it. If you are doing this daily on the same installation your work flow is broken and we are not here to cater for that. You are expecting and extremely powerful settings menu with thousands of buttons to tweak it in every conceivable way. That is not how it's intended.

[jjd-uk]

Just a few ideas for the future, forgive me if I'm just repeating stuff people have already mentioned. The github topic went a little off topic so I stopped reading.

- Combine the Album and Artist scrapers

Not good. User may want to only scrape Artists e.g. classical music where scraper can correctly identify artists like "Beethoven" but is weak at identifying recordings by different orchestras and conductors. Or user may want to scrape albums, but because of language may want to read Artist biographies etc. in own language or from separate source.

I think maybe this is where your lack of understanding for how adding stuff to the video library works, as this case should work perfectly with a combined scraper together with how we wish it to work based on the same type of logic for the video library together with the mock up in Post #18.

Add your Music folder as a source with your desired default options selected, after source is added you'll be asked to refresh source (start scanning/scraping process in other words), however you can select No at this time to continue setting everything up for subfolders that may need something different. So if you have a subfolder of classical music, you simply go to that folder "edit content" on that folder to change to whatever scraper/settings you what for those artists and deselect anything you don't want.

So workflow could be:

Music folder (1. Add as a source and the specified settings will apply to all subfolder then select no when asked to refresh)
-> Rock music subfolder (inherits settings from source)
-> Pop music subfolder (inherits settings from source)
-> Classical music subfolder (2. edit source on this folder and change settings, thus the settings inherited when adding source are overridden for this folder)

3. Select Update library to scan/scrape.

I apologise if I'm getting you wrong, but you seem unaware of the crucial aspect taken from the video library where you can go to any subfolder within the structure of the added source and override the content settings for that folder. So if wanted you could have different content settings for every subfolder so I don't see how that doesn't give you the flexibility to do whatever you want.

[jjd-uk]

By the way I hope that didn't come across as patronising and I certainly didn't mean it to.

[DaveBlake]

No worries JJD we speak the same language. I think stepping through the proposed sequence of things is really useful design wise. The devil of good design is in the detail!

Add your Music folder as a source with your desired default options selected, after source is added you'll be asked to refresh source (start scanning/scraping process in other words)...

Unfortunately no, Martijn says there will not be any confirmation prompt. I specifically asked about it. Video currently does as you say but this is a "bug" according to him. He plans for scanning and scraping to be immediate on you clicking OK to set the options.

I feel this is a mistake for the reasons I gave. The flexible work flow you propose will not be possible.

I apologise if I'm getting you wrong, but you seem unaware of the crucial aspect taken from the video library where you can go to any subfolder within the structure of the added source and override the content settings for that folder. So if wanted you could have different content settings for every subfolder so I don't see how that doesn't give you the flexibility to do whatever you want.

No need for apology, but I do get that aspect and think it is a great improvement.

Obviously my post was confusing! My turn to apologise!! I am not saying don't do it, I am just saying take care with the details.

By points are all about design details, what Martijn calls the abnormal use cases if you like, for me it is the meat that makes a system work well all the time for every one. Good design is all about spotting all the odd cases and handling them as well as the normal one. And no it does not mean an unwieldy UI, that is just silly, but it does take a little bit more thought in design stage. For me praising the good aspects of the change, and how much better it will be, is a waste of time. I take that as read, and anyway there are plenty of people ready to cheer enthusiastically. I am into honing the gritty design detail, making the dream actually work well on release. That is not something you fix at code level, it is what you build in at every level.

[Martijn]

The current dialog you get after adding a source is wrong as it forces a complete rescan. That particular dialog only need to be shown after editing a source.
If the scan should or shouldn't start right after the set content is another discussion.

[Tolriq]

Well I agree that data is nice but it tells absolutely nothing.

And his post does not sound patronising at all.

So you again miss the important part, I did say I have no more the details about in library or not, this is a very very small excerpt of the possible available data. (And we can debate hundred of hours about patronising, not being english native (my case) and the tone of your posts on the PR, or majority of your posts but it's not the important point here)
(Maybe I sound rude in my posts, because I improved my English on reading this forum )

I can add back easily the data about music to see if they have client id or not and prove the number I had before and used to build Yatse.
I can add needed data for the Team also, I can give real world data for tons of things that could be useful to you, but there's no way to work correctly with non Team members (see my 2013 post about PRs and time loss).

For example when there was a discussion about transcoding and flac or not. I did take time to check real world data, for the numbers I have (8.2M songs) it's 40% mp3, 10% flac, rest is other codecs or addons. So 10% is less than what I thought, but it's the second major codec and not a niche at all, and 8 200 000 songs is a big enough number to be quite representative in my world.

Data centric dev and corresponding to users needs make the ratings of Yatse what it is, not luck and random guesses.

Unfortunately your answer shows that no one care, and the discussion on this PR and the vanish when talking about organisation for PR are just more proof.

I was hoping since the problem I report since 5 years was a new time public with someone else that speak better than me would be a new start to have normal discussion about this, it seems not and even worse for some Team members.

So I'll continue to do the tests when asked, and post on PR that will break things or have impact not thought due to my little knowledge, but have finally really understood that it's a lost cause.

Too bad that there's no way to work correctly and intelligently so that I could bring my knowledge and the quality I like to put in what I do to Kodi code base to fix very long lasting bugs and add the missing stuff at least to remote parts. (And again if some have doubt, not only for Yatse benefits but for Kore and all other and in the end the users because I only reflect their needs)


Just remember that having a Favorites API that only support list / add and no delete / edit just because no one can get to a consensus on how it should be is not normal.
And that when there's a recursive flag on 95% of the API and the 5% remaining are still on the TODO list of the main JSON dev it's not normal that a random guy comes and say no to a PR without valid points and let an unfinished API. And it's even worse that this feature stays in the TODO list, so someone else could come and take the feature and get kicked out at the end if the same guy shows up again ....

And this kind of messages : http://forum.kodi.tv/showthread.php?tid=...pid1277609 should not happens (my answer with details : http://forum.kodi.tv/showthread.php?tid=...pid1279821)
When something is asked for 3 years multiple times, is publicly refused for some reasons (I will not debate about them here) a Team member can not answer condescendingly such things.
This drop and the refusal of the new API made me fix in urgency EventServer to support Unicode characters https://github.com/xbmc/xbmc/pull/1612 with hours of lobbying to get it accepted in time.

[Tolriq]

I apologise if I'm getting you wrong, but you seem unaware of the crucial aspect taken from the video library where you can go to any subfolder within the structure of the added source and override the content settings for that folder. So if wanted you could have different content settings for every subfolder so I don't see how that doesn't give you the flexibility to do whatever you want.

Well well well, this is something that I think not a lot of people knows about.
I had made support for thousand of users to help them build their library and linked them to the Kodi wiki that I've read a few times and never noticed that. (I've read it again and have now noticed it http://kodi.wiki/view/Set_content_and_scan).

This should be more advertised as effectively this solve a lot of things. And maybe the Wiki should make it more clear, from experience with Yatse wiki for support, when putting some text at the start of a large page and there's a second part with images, users eyes jump to the first images then left to read the text and most of the time never get back to up to read the text they missed, as they think it's still part of the introduction or content links.

[evilhamster]

Not sure if they are on-topic or not but there are some things that I would like to highlight.

When scraping artists and/or albums, scraping is done based on information in the db and not the files themselves. So there is no guarantee that an artist and/or album only exist in one location, which could cause issues if the different locations are in places that have different scrapers set. Especially with musicbrainz tagged music, artists can be all over the place (since they are by default not placed in the title of the song but rather as separate artists in the artist field).

Currently, we don't have any breakdown of song information from the scrapers but when discussing the future of scrapers for the music section we should probably consider artist / album / song. There are some use-cases listed in this thread where different scrapers are being used, but I think we could solve those use cases while still merging the scrapers. If for example the settings for a scraper was not global but set on each place it was supposed to be used, a scraper could have settings for which data source to use for artist / album / song.

When adding a source do we really need to present the users with a lot of options they should preferably not need to change? What do you think about only being able to select music as a type when adding a source and to do more advanced settings you would need to do a set content? Or maybe only show the more advancedsettings if settings level has been set to advanced or expert under settings?