New editors: a guided, staged, self-training approach - LONG

mmirG · November 4, 2017, 4:29pm

New editor = lots and lots to learn.

too long, didn’t read:
chunk down the process of adding releases.
direct new users to master one chunk before starting on the next.

How about instead of the intuitively obvious, painfully slow approach of struggle to add release, forget most of the guidelines that were found and struggle to add second release, forget most of the guidelines that were understood and struggle to add third release, …, we suggest a different approach to new editors?

The aim is for “wriggling then crawling then walking” rather than a series of jumping in the deep end.

Process:

New editor selectors the fifteen releases 1CD releases they wish to add first. And are directed to a tutorial that shows them how to make sure that a release is not already in the database.
They are then pointed to a tutorial which shows them how to add a release using only the release name and the artist name and language (as language can be necessary for release title) and edit note. Nothing else. A bare minimum release - no dates, no tracklist, no recordings, no script, no barcodes, etc, etc.
The new editor is encouraged to ask questions on the forum.
They create 15 bare minimum releases. And are given assistance they request and feedback by the community during this time.
When this is completed, the new editor is either steered towards creating more bare minimum releases, or more likely, pointed at a tutorial which shows them how to complete the Release information page.
They complete the “Release information” pages for their 15 releases, again asking for needed assistance and being given feedback.
If they demonstrate proficiency in that they are directed to another tutorial that shows them how to add Tracklists.
The New Editor completes 15 Tracklists.
If they have demonstrated proficiency in that they are directed to another tutorial that shows them how to add the Recordings to a release.
They complete adding the Recordings to their 15 releases.
If they have demonstrated proficiency in that they are asked if they have the desire and facilities to enter coverart for their releases. If yes they are directed to another tutorial that shows them how to add the Cover Art to a release.
They complete adding the Cover Art to their 15 releases.
If they have demonstrated proficiency in that they are directed to another tutorial that shows them how to add the DiscIDs and AcouticIDs to a release.
They complete adding the DiscIDs and AcouticIDs to their 15 releases.
They are suggested to select another 15 releases they wish to add and work through the 1. - 14. sequence again.
The New Editor is then directed to a tutorial that shows them how to; create Collections, add their releases to the their Collections and how to subscribe to their Collections. And the benefits and joys of doing so.
They create, add and subscribe if they so desire.
After adding 30 releases the New Editor is asked if they wish to enter Relationships into their 30 releases, and directed to a short article describing the benefits of doing so. If they wish to learn how to enter relationships into their releases they are directed to a tutorial that shows them how.
They are given assistance as requested and feedback.
The New Editor is by now a Beginner Editor who has become somewhat proficient in each stage of adding a release by doing many real-life tasks in each single stage before moving on to the next stage, they know what to read for answers and they know where to ask for help.

After say 5 new users have tried out this approach we look at their proficiency and user experience, make tweaks or scrap the whole dammed thing and start again.

Basically we would be:
Chunking down the very complex skill of adding a release in accordance with the guidelines.
Providing training to the New Editor in what to read to know the guidelines.
Providing training to the New Editor in where to ask questions.
Helping New Editors get more capable quicker at adding releases they wish to tag.
Exposing New Editors to more advanced editing procedures.
Making their adventure on Musicbrainz more fruitful, enjoyable and understandable.
Probably increasing average contributions from New Editors who wish to tag.
Possibly be retaining more Beginner Editors who get enjoyable experiences adding relationships.
Creating a reputation of Musicbrainz being a friendly place where you can succeed and enjoy tagging and editing - and get help if you need it.
Increasing returning Beginning Editors who enjoyed their last sojourn and know exactly how to get back into the game to tag their new CDs - work through steps 1 - 19.

Jim_DeLaHunt · November 4, 2017, 6:06pm

I love a lot about this approach: the chunking, the tutorials, the friendly support and encouragement. But let’s also keep clear what motivates the new editor. In my case, it was (and to a large extent still is) to be able to do and tag a CD I just bought. The gratification of the tagging isn’t fully delivered until all those stages of entering a Release is done.

Oh, and, did you mention scanning cover art? I would make that a stage, and an early one. If a new user creates Release with partial info, and uploads scans of cover art and liner, someone else can help fill in the release info and track list from the cover art and liner.

One further point: I always create new Releases via a Picard CD Lookup. This means that the Release always has the right number of tracks, track times, and a disc ID. Using Picard this way implies a different division into stages.

derobert · November 4, 2017, 8:01pm

You seem to be presuming that editing MusicBrainz is a serious goal people are trying to achieve, that they’ll be willing to work towards despite a bunch of obstacles. (Which, given the amount of work to enter a release, must be somewhat true for everyone here).

So, I’ve had a MusicBrainz account for a long time, but I’ve only recently (since December 2014, I guess) been active. That was because I finally decided to re-rip my entire music collection, which had previously been in Vorbis and hand-tagged. And also to deal with the huge pile of disks which hadn’t been dealt with at all… So I guess this wouldn’t have applied to me (since my account wasn’t new), but I certainly was a newbie editor back then.

So most of the discs, I think (did not keep stats—and do not feel like it’s worth the work to gather them), were already in MB. Some were trivial to find (discid), others take some searching (and then adding disc id). So selecting 15 releases to add would be annoying, but not insurmountable.

Now, I’d enter in 15 releases and… still not useful. Can’t tag with only an artist and name (and this is presuming the tutorial understands these are classical releases). And they’re not useful to MusicBrainz either — that is no where near enough information to identify a release. At least not a classical one. (E.g., because the title is often just the single symphony that is on the release, and especially for popular symphonies a given orchestra and conductor will play and record it many times). It’d take until step 6 until those were useful entries in the catalog, as they’d at least be (hopefully) identifiable. And hopefully I’d kept them separate on the shelf, else I might not be able to find the same ones again…

However, despite entering 15 discs, and waiting several days for forum replies, I still haven’t been able to rip a disc yet. So the goal has been thwarted so far.

I’m not sure how you enter track lists (7) without recordings (9). I thought the schema didn’t allow that. But adding the track list is, depending on release, a bunch more hours for 15 releases. Same with cover art. Still haven’t gotten to rip a disc yet, though, many days & hours into this process. After adding the discid (14), finally morituri will allow ripping the disc (which is good, otherwise it’d be hard to add AcoustID data).

MB still isn’t useful, though, after probably a full day’s work spread out over two weeks (remember, goal was to re-rip music collection). Could just come up with a local tagging solution again, seems like much less work.

Anyway, and then step 15, do it again.

The goal of ripping and tagging my frist release can’t actually be accomplished until step 18 because until then, I’m not able to enter the works (and this composers, lyricists, arrangers, etc.) or the performers, recording location, soloists, etc. (the recording ARs). That’s at least two or three weeks of elapsed time and two full days of work (and probably a similar amount of work for existing users, who have a lot of stuff to review).

A tutorial would be useful, but I think it’d probably be better if it guides you through adding one disc, and at the end of it you have something useful for tagging (and also for the database). And you can stop early (say after the tracklist/recordings, before ARs) and still have something useful. And you’re welcome to go through the tutorial again for the 2nd disc, or third (but probably won’t need it again). Each time though you do a reasonable amount of work and actually accomplish the goal you started with.

mmirG · November 5, 2017, 12:10am

The failings of the wriggle, crawl, walk approach for New Editors with only a few CDs that need to be added are real and well described. I suspect too that more failings remain to be identified.

The wriggle, crawl, walk approach is not suitable for all New Editors. Some way of streaming only those likely to benefits from it would need to be developed.

However for editors with +10 releases they need to add, most of the failings you describe seem likely to become minor in comparison to the benefits of proficiency in each step of adding a release.
The total time taken to add 10 releases using the chunked approach would be less than the time taken by using the intuitive “series of 10 jumps in the deep end” if familiarity with applying the guidelines speeds things up and that proficiency in adding a release is, like many other skills, better learnt by mastering small steps.

There is a fine tutorial that goes through the process of adding release. Let’s if I can link to it using my phone … here we go
https://musicbrainz.org/doc/How_to_Add_a_Release
The problem I find with it is that there is so much information in it that little of it sticks. Even with repetition. And that each skill, is applied only to the release I am trying to add. Which acts against getting a broader grasp of the skill and how to apply it quickly and accurately when the next release needs something slightly different.

I don’t assume that a New Editor has any other goals than adding their albums so they can tag their digital copies. What I am pointing at is that a New Editor with many releases to add can likely save time, reduce stress, be more accurate, do a better job and enjoy themselves more if they take advantage of the freedom to chunk together tasks requiring one skill. And then, with that chunk completed and some mastery gained they can move onto the next chunk and master the skill needed there.

As you point out this approach won’t suit all. Just getting past the fear inducing idea of temporarily leaving releases incomplete in the Musicbrainz encyclopedia is an obstacle to the wriggle, crawl, walk approach. I know cause it was intuitively obvious to me that that would be a bad thing.
But it ain’t.

mmirG · November 5, 2017, 12:19am

Looking back at the short period in which I tried to tag my collection I see that I tried to add missing releases to the database whenever I came across albums I couldn’t tag.

The wriggle, crawl, walk approach would recommend New Picard Users finish tagging what they can tag from releases already in the database. And then progress to adding releases to the Musicbrainz database.

Doing things in that manner would let a New Picard User concentrate on developing their skills in using Picard to tag - and gain the benefits of having a simpler, less complex task to master.

mmirG · November 5, 2017, 12:38am

Yes - building skills from there does make good sense. Having just inherited the family’s CD and record collection I can soon begin to experiment on myself by adding batches of rips, batches of CDs and batches of LPs.
And observe what chunking does for my workflow.

lightsussex · November 5, 2017, 1:32am

Everyone is different. They learn in different ways, they have different things they think are important, they have different motivations, levels of commitment, etc etc etc.

I’m a noob, and I’ve mostly been ok with figuring out what to do (especially with your and other members’ help I’ve had on here) but the main thing is the volume of information. There’s no way of making that information less, because it’s the nature of a huge database to have all those fields and “stuff”.

Personally, I wouldn’t go through any sort of “training process” where I’d be doing little bits of something. I’m happy to contribute where I can just “do it” but if it’s too much effort, I won’t. I just want as much clear information as I need to “just do it”. I’d prefer less information, not more. What I found/find hard to get my head around is what is actually happening and what the terminology means (yes, despite the guide - it is still a bit daunting when you think you might be “messing something up”). Even if I did go through all that, I’d probably still forget how when I got around to entering another of my ancient cassettes into my computer/your system.
The only thing I wish I’d come across earlier is that “add cluster as release” button. If not for that, I never would have proceeded to add the titles that I have. There’s no way I’d have typed out all those titles again (I’m doing it when I’m converting my records to put into the computer). The other thing I’d have liked was a glossary in one spot (is there one and I’ve missed it?). Read, read, come to a term, link to another page to explain it… I’d have liked one spot (or those popup hovering things) to have it tell me basically what’s being talked about.

For somebody more motivated to be a regular contributor, your approach might be useful for them. It’s horses for courses. I’m the run-down nag, the thoroughbreds might go for it.

Jim_DeLaHunt · November 5, 2017, 2:11am

Here’s another way to look at the staged approach: it will work better if it matches a reasonable workflow for actually getting useful work done with MusicBrainz.

To that end, I took some time to summarise my workflow, in _How do _you_ add a Release? What's your workflow? . Maybe others will contribute their workflows to that thread.

A guided, staged, self-training approach that fit with that workflow would be what would work best for me as a new editor. But everyone is different.

Yes, agreed.

mmirG · November 5, 2017, 7:13am

The search box in the top right of the Musicbrainz page has included in the menu (both) “Editor” and “Documentation”.
https://musicbrainz.org/doc/Terminology
The other place which has searchable things is the wiki.
https://wiki.musicbrainz.org/Main_Page

mmirG · November 5, 2017, 7:32am

My very rough estimate of the percentage of a modern western educated adult pre-geriatric population who can add a release is based on my experience learning languages and dances and the parallels I draw between learning those activities and learning Musicbrainz protocols.

My estimate is that with the current approach to learning, 2% of that population could add a release.

I’d like to get that to 8%.

mmirG · March 2, 2018, 4:02am

After many weeks of understanding the basis of the Hyperion clean up in terms of availability of cover art and Hyperion metadata the penny finally dropped.

Such a collection of cover art and accurate metadata provides the ideal raw material for new editors of MB who want a structured training in adding (classical) Releases.

Are there other troves like the Hyperion website?

Are there new editors who want a structured training?

thwaller · March 3, 2018, 5:32am

I am no longer a new editor, but I say100% that looking back, I would have wanted something like this. I had some editors who helped out and it was more struggle than it needed to be. Help came mostly in response to edits that were either wrong or could have been better. To me that was a bad start to a good thing. When one starts out by voting no to your edit and you are a new user, that is, or can be, discouraging. While I look back and appreciate the editors that struggled with me, today I have put that aside as well as the disagreements that may still exist and look at them for guidance, especially since many of them have a area of specialty.

EDIT: I wanted to add on the no votes… there is an auto-editor who explained a better way to handle such things. Mainly to vote yes and make the correction to it if it is not a lot of hassle. Then you avoid voting no and still get the data correct.

Llama_lover · March 7, 2018, 1:00pm

Speaking as a Grandpasaurus, there are many great suggestions above. The last “edit” above hits home with me. Senax has been (and still is) my mentor. He has made numerous corrections for me and has taken the time to show me what I have done wrong or right.

Personally I would like to see a means to be able to practice submissions and editing without formally entering anything into Picard. It is a little daunting moving forward in my quest to be able to edit things such as relationships, etc… The detail I see in the work others have provided is indeed awe inspiring. I have yet to become comfortable adding to this detail for fear of screwing something up. I do not wish to add to the burden of established editors by entering things erroneously thereby adding to their work load in correcting my errors.

I look forward to the future of Picard!! Your work is well respected.

mmirG · May 9, 2019, 9:02am

Seems The_King has created a body of Releases with Cover Art that would allow any self-directed new editor who wants to develop strong skills around editing relationships to do so.

The_King’s post: If you’re bored

If any new editors take up this opportunity, please post about your experiences, what worked well, what could be better.
And if you have questions or need assistance then ask on this forum.

rvb · May 29, 2019, 9:28pm

I have 6 degrees in computer science and 40 years experience in filing, but this is so complicated, I need at least 6 degrees more and 20 years deep knowledge about filing a medium that has sound on it.

rvb · May 29, 2019, 9:41pm

In general all software is way too complicated. In the old days there were often light editions, for people who do not need all the stuff. I suggested this to other software systems, but nowadays they are not so interested in making something that is simple and useful for the majority of the people .