Project Proposal: Forum Thread / Edgecase Resolution Documentation

Good morning, everyone.

There is a great deal of combined effort that has gone into building this site in these forums and I'd like to discuss making our work more accessible so that we can all spend less time digging for these discussions or inadvertently retreading over an already resolved topic. This may also hopefully help new s make better contributing decisions if they are able to reference a resolved case.

Using Reference ( https://reference.discogslabs.com/ ), we can create an organized reference page that documents these cases with appropriate links to threads and possibly example submissions/profiles.

I think a good suggested basic framework would be to mirror the layout of the actual guidelines and order edge cases and forum decision links in that manner.

I'd like to hear everyone's thoughts on this. This thread may also be a good staging area to start linking those posts as well.

No experience of using the Wiki but if in effect it's going to be a searchable one-stop shop for this kind of info it sounds great.

A lot of decisions seem to be based around who's logged on with the most up to date saved threads at times.

Diognes_The_Fox
Using Reference ( https://reference.discogslabs.com/ ), we can create an organized reference page that documents these cases with appropriate links to threads and possibly example submissions/profiles.

see https://reference.discogslabs.com/wiki/Undocumented-Guidelines
for an effort in the same direction.

Diognes_The_Fox
making our work more accessible so that we can all spend less time digging for these discussions or inadvertently retreading over an already resolved topic.

This is long overdue and I am very happy that this gets pushed/suggested in an official way.

Diognes_The_Fox
I think a good suggested basic framework would be to mirror the layout of the actual guidelines and order edge cases and forum decision links in that manner.

Are you suggesting something like "the commented version of the guidelines" like https://www.amazon.com/Modern-Criminal-Law-Questions-Casebooks/dp/0314159029

or more a loose collection?

Several pages and subpages or just a single page so it's easy to search even just inside page. (might be more trustworthy then discogs search ;) )

The_Beatles.
A lot of decisions seem to be based around who's logged on with the most up to date saved threads at times.

velove
see https://reference.discogslabs.com/wiki/Undocumented-Guidelines
for an effort in the same direction.

Didn't know about this one. Great starting place.

velove
Are you suggesting something like "the commented version of the guidelines" like https://www.amazon.com/Modern-Criminal-Law-Questions-Casebooks/dp/0314159029

or more a loose collection?

Both? I was thinking pretty much an index. The problem with forum decisions is that you need to know they exist before you can search for them usually. And then you need to know how to search the forums

Hopefully as well, this may give us a chance to at least document (and possibly resolve, just not in this thread please) some potential conflicting resolutions.

Diognes_The_Fox
Didn't know about this one. Great starting place.

https://www.discogs.sie.com/forum/thread/692292
for the thread one year ago. you even participated ;)

Diognes_The_Fox
The problem with forum decisions is that you need to know they exist before you can search for them usually. And then you need to know how to search the forums

how to search the forums isn't good enough as the search is still not good enough to be useful in many situation.

Diognes_The_Fox
Hopefully as well, this may give us a chance to at least document (and possibly resolve, just not in this thread please) some potential conflicting resolutions.

Definitely.
This will also help reduce the conflicts that come with your double role as staff and member. If it's documented in the wiki it's official. Else it's just an opinion. At least that's how I could see one possible way.

But even if forum search was working that still wouldn't help in knowing when decisions/consensus has been achieved and which one is the most up to date one.

May I suggest that Eviltoastman is tempted out of retirement here as he seemed to have the most complete list of these decisions.

so i'm down for helping, but am a bit confused as to what digoness is asking us to do - edit and do the work on the wiki? i have no experience editing the current wiki, and was only mildly acquainted with the old one.

i have several discussions and forum threads saved, typically ones for issues i had encountered which had not yet been covered in the forums, so while they may be old i think in most cases the rulings issued in them remain relevant, and usually delivered by nik. can we post links to them here, or should we wait on that for now?

PabloPlato
but am a bit confused as to what digoness is asking us to do

I think for the moment it's gathering input on what to do, how to structure it etc etc.

And once that is clear enough the actual work of compiling the past decisions can be done.

Maybe a sticky thread for discussing what should go in there could help as well.

Diognes_The_Fox One problem with the wiki is that it's not multi capable. so if person A edits and person B edits at the same time, there is no blocking/locking and the 2nd edit will just overwrite the previous edit. Until this can be fixed, there should be a gatekeeper who'll add the things to the wiki or something similar.

Diognes_The_Fox
I was thinking pretty much an index.

This project may be useful: https://www.discogs.sie.com/group/6269

PabloPlato
i have several discussions and forum threads saved, typically ones for issues i had encountered which had not yet been covered in the forums, so while they may be old i think in most cases the rulings issued in them remain relevant, and usually delivered by nik. can we post links to them here, or should we wait on that for now?

velove
But even if forum search was working that still wouldn't help in knowing when decisions/consensus has been achieved and which one is the most up to date one.

velove
Diognes_The_Fox One problem with the wiki is that it's not multi capable. so if person A edits and person B edits at the same time, there is no blocking/locking and the 2nd edit will just overwrite the previous edit. Until this can be fixed, there should be a gatekeeper who'll add the things to the wiki or something similar.

This sounds like a staging area is a good first place to start. Here would be a good place to start dumping links. We can create threads to deal with conflicts as they arise.

I don't mind acting as a gatekeeper to prevent overwriting, but once we're past the initial hump, it shouldn't be an issue for casual updating as the need arises.

Diognes_The_Fox
I don't mind acting as a gatekeeper

*giggle*

Diognes_The_Fox
I think a good suggested basic framework would be to mirror the layout of the actual guidelines and order edge cases and forum decision links in that manner.

This is something that I actually have experience with and have suggested privately that an annotated version of the guidelines would be extremely helpful, so I obviously think this is a fantastic (and perhaps long overdue) idea. Moreover, there are many tried and true examples found elsewhere that you can essentially use as a template (PM me if you want some of those).

For online annotated documents, there are really two formats. One is the wiki-style multi-page hierarchical approach, which is vertically restrictive. The second is a vertically unrestricted single page approach (much like the runout groove forum list, i.e., the list keeps growing and stays on one page), which is what you initially suggested. The latter is easier to use, but requires more work on your part to maintain and works best if the annotations are as concise as possible and follow a uniform format. The former provides more freedom as different formats can be used for different pages, and is easier to maintain (and especially by more than one person), but more difficult to use because the pages essentially create visual barriers between sections that are otherwise interconnected, which in turn, can pose a challenge to the curator(s).

If you go with the annotated version, there are two ways to do it. One way is to annotate it inline. The second way is with the online equivalent of footnotes, i.e., hyperlinked placeholders, that link to mini-landing areas either at the bottom of the page or at natural breaks, such as after each major subsection. For inline text to be effective, it must be stylized different from the guideline text. I think the footnotes-like option is the way to go because that's consistent with my reading style, for other people, it's the opposite, so there's no right or wrong option. Either way, for each one, what you will basically need is a citation to the discussion and a brief explanation, e.g., Topic #1234: Stereo Format For CDs? (discusses exceptions to RSG §6.13 for CD format releases) (MM/DD/YYYY). The date is very important to me because it provides context, e.g., is this still a fresh idea? is this established precent and common practice? is this something that was discussed so long ago that it's not really practical anymore and therefore should be revisited in the forums? etc.

Regardless of whether any of the above is useful for you, one thing that I find lacking in the guidelines and think would be perfect for this kind of a document are examples of releases that exemplify some of the rules' operation. For instance, if you found a release with notes that illustrate RSG §11.1.4 in practice, then I'm sure that would go a long way towards eliminating some of the back and forth commenting about whether an edit that removed a '© 19## ABC Name Inc.' was ok. Bad examples are just as helpful, but that would probably require creating dummy releases, which I don't think is necessary, but also good.

in short, this is a great idea, but obviously a pretty big undertaking (just ask sebfact about moderating the living runout groove document — I'm sure that's a huge black time hole).

berothbr
Regardless of whether any of the above is useful for you, one thing that I find lacking in the guidelines and think would be perfect for this kind of a document are examples of releases that exemplify some of the rules' operation. For instance, if you found a release with notes that illustrate RSG §11.1.4 in practice, then I'm sure that would go a long way towards eliminating some of the back and forth commenting about whether an edit that removed a '© 19## ABC Name Inc.' was ok. Bad examples are just as helpful, but that would probably require creating dummy releases, which I don't think is necessary, but also good.

This here.

The_Beatles.
the discussion here.

I actually following that discussion (I don't think I posted).

Although that's a great idea (especially for the examples of what not do), I actually think the active/real examples are even more helpful because those are concrete, have images, and because the pages usually have some history with a changelog. This is crucial because anyone looking for guidance can see not only the end result, but also the process that led to it, which is equally (and in some instances even more) informative. Regardless of how those are identified/presented, I strongly believe that, in many instances, examples are just as helpful as the staff 'IMO's in the discussions (and please don't take that the wrong way staff — Discogs /staff is/are fantastic and incredibly helpful/awesome to deal with). This is because sometimes the best way to make something click/figure it out (at least for me) is to see something in action and use it as a model/template to guide a submission/edit.

berothbr
This is because sometimes the best way to make something click/figure it out (at least for me) is to see something in action and use it as a model/template to guide a submission/edit.

Yep same for me as i was trying to explain, poorly perhaps, in that thread.

berothbr
Regardless of whether any of the above is useful for you, one thing that I find lacking in the guidelines and think would be perfect for this kind of a document are examples of releases that exemplify some of the rules' operation.

Isn't the problem the fact that a such a release when "perfect" could then be edited, so in effect invalidating it's status as a good example of anything?

The_Beatles.
This release came about following the discussion here.

Thanks for that link and some back history, I think I was only dipping into the forums on a patchy basis at that time and missed that. Good came of it I think, especially now.

berothbr
see something in action and use it as a model/template to guide a submission/edit.

Perhaps, building on the original idea and taking the above point about visuals being vital and imperfections being corrected (gist), I would suggest -

Create an example (basic) submission for each guideline, add full images. The submission may have one or two errors. Correct those errors with full reasons why. Add a few more details, add a few more and more until C&C staus is reached. That submission should be marked (as in Dave's example) "do not edit", with the reason why.

Getting back to the thrust of this thread, whenever a guideline or practice is changed, edit that example submission and post a link in edit notes to the revised guideline that caused the change and the resolved threat that prompted the guideline change.
Again, when a forum thread is resolved and the example submission is brought up to date based on new guidelines, conclude the thread with a link to the example submission. Ideally, the initiator of the thread could do the edit required.

That idea may need more input but it came to me from the most recent comments, it also doesn't cover the points raised at the outset of this thread, but it would work for me.

I'm sure there's more to follow on the thread theme so I'll just leave that thought out there for now.

The_Beatles.
Isn't the problem the fact that a such a release when "perfect" could then be edited, so in effect invalidating it's status as a good example of anything?

That is a reasonable concern, however, this is basically what's done in the runout groove thread and seems to work well. By comparison, the inverse is also true, i.e., if a guideline/forum interpretation changes, then the 'fake' examples would also have to be edited. Unless this is going to be open and not curated by someone like diognes, then changing/adding one small annotation, which would otherwise take a few minutes to update, would become much more of a time intensive task, which is lousy for the curator. In theory, I of course think mcr1's addition to that idea would be great and helpful (at least to those of us who like to use this sort of thing), but, in practice, unless this is going to be open, we should be realistic about how much the voluntary curator (probably diognes) can really be asked to do in of maintaining the resource (would probably delay their inevitable achievement of 1 million rank points)

berothbr
forum interpretation changes, then the 'fake' examples would also have to be edited.

Thanks for considering the above, I thought about the practicalities too. I suggested that each resolved thread where guidelines were changed, or practices updated, could have the originator of the thread (or a nominee agreed in the thread) actually do the updating of the example, with links all round.
In practice though, there aren't that many guideline changes or threads where guidelines are requested to be changed and I could see it working.

Here's an example already set up for guideline Lemonheads* - Create Your Friends, but it's a terrible example as it doesn't have images!
It's also in need of a vote.

mcr1
there aren't that many guideline changes or threads where guidelines are requested to be changed and I could see it working.

I kind of disagree with this — there might not be a lot of threads each month that are like 'proposal to change RSG §#.#.#' (which I usually don't find particularly informative in improving my submissions), but there constantly are threads about figuring out something specific. For example, I just participated in Diognes_The_Fox!), but I think there's a lot more that we're unaware of just because, for example, I try to stay walled off from classical releases, so that's a whole side of things that I don't pay attention to.

One thing I do like about the runout groove thread is that even though the primary page is essentially locked in, s can still post comments and discuss at the same time. This is nice because, for example, if I add an image of a weird runout etching or if some else finds a broken link, etc., (for better or worst) it's really easy to communicate that there's something to update. I think the same word work well for this, e.g., someone could post 'we just had a really good discussion about XYZ topic' that might be helpful to add to §#.#.#'

berothbr
there constantly are threads about figuring out something specific

Yes, that's true. In my scenario (it is a scenario for discussion, based on ideas started earlier) those resolved threads would have the conclusion linked to an example too.
But yes, there's the "runout" thread scenario, well worth discussing I agree. However, 3 years down the line I've just started following that thread and I suspect (from other forum comments) that participation in that and use of the forums is only for a minority of s. A correct example on each issue, linked to in the guidelines would be there for all s as a reference, from experienced s needing confirmation of updates at a glance to newbies looking to follow an "approved" format from the outset.
All are ideas though at this stage, it's a big topic and I'm sure once other s get to look in we'll read more.

I agree with a lot of what you wrote mcr1. In my mind the ideal form would be:

§#.#.1. Blah Blah Blah This is a rule.
n.1: normal example and/or bad example
n.2: discussion URL that explains an exception to this rule
n.3: discussion URL about some issue about this rule
n.4: example of an exception
n.5: a useful tip/anecdote/extra explanation from a veteran
n.6: insightful/helpful/good quote from a forum (does not have to be staff — could be post that is just super)

§#.#.2. Blah Blah Blah This is the second rule.
n.1. + n.2 + n.3, etc.

mcr1
there's the "runout" thread scenario

The thing I like about it is that there's one person who has control over the first post, but there's a place for everyone to offer their input directly or ask a quick question that's right there in one place. The downside is that it could inevitably just end up as a single repository for guideline gripes, which be entirely unsuccessful.

Thanks Opdiner. I do have a lot of discussions saved and I have all my requests backed up on my hard drive.

berothbr
This is something that I actually have experience with and have suggested privately that an annotated version of the guidelines would be extremely helpful, so I obviously think this is a fantastic (and perhaps long overdue) idea. Moreover, there are many tried and true examples found elsewhere that you can essentially use as a template (PM me if you want some of those).

Good suggestions here! Thank you for the outline.

berothbr
I strongly believe that, in many instances, examples are just as helpful as the staff 'IMO's in the discussions

Yes, very true. And also by forum resolutions, I meant community lead ones as well. This is part of a larger goal I have of renovating how we (the community) are able to develop our database with more autonomy. Having a better system of documenting these cases will help reduce the need for people to act as "database elders" as well as having to wait for staff to become involved in order to make progress.

As time goes on, staff involvement in most database threads has had diminished impact. As the database grows and develops, situations are becoming more esoteric and affect a smaller amount of submissions. The amount of time required to properly devote to researching, providing insight on and decisions to these forums can take several hours at a time and only affect at most a few hundred submissions, where as roughly 3300 submissions a day are being contributed daily, and that number is growing.

Ultimately, I'd like a system where these types of decisions can be made by vote and the whole process automated to the effect of something like we have for merging releases: issue arises, s can vote and discuss the issue until a percentage is attained or the request times out. Obviously, that's the a very simplified version of the idea, but that's where I'd like to take it.

With that system, I'd actually be able to voice opinions on how the site should be run. I don't like being in a position where anything I say may accidentally be interpreted as a decision or canon. Sometimes my ideas aren't great. I don't want things to be a way in the DB just because "staff said so". Plus, anything I can do to eliminate any potential conflict of interest with my own submissions/practices, the better.

Anyways, as I'm sure many of you have noticed over time, I'm a far better contributor than a dictator.

berothbr
The downside is that it could inevitably just end up as a single repository for guideline gripes, which be entirely unsuccessful.

Yes, this is something I'd like to avoid. Things will definitely have varying levels of "resolved" which could be determined. It'd also potentially give us room to discuss if those resolutions should or should not be applied to other cases. Some cases might just apply to only specific subs and not

I have no idea how to organize anything in the meantime though. Which is also why I'm not rushing either. There's a good amount of ideas to discuss and let soak in.

Eviltoastman
Thanks Opdiner. I do have a lot of discussions saved and I have all my requests backed up on my hard drive.

Great! This should be most helpful.

Thanks for the thoughtful post Diognes_The_Fox.

Diognes_The_Fox
As the database grows and develops, situations are becoming more esoteric and affect a smaller amount of submissions

Unless overall activity and forum participation is decreasing, then that's probably a good thing.

Diognes_The_Fox
I'd like a system where these types of decisions can be made by vote

Not to go too far off topic, but I've always wondered why releases can only be merged and/or removed by vote, but labels/companies can be obliterated by one person. Sometimes a label/company profile can be created by mistake, such as a typo, and quickly rectified. However, Delicious Vinyl Inc. is an example of where a simple merge request would be very effective.

In of your annotated guidelines idea, if you could find a technical solution for ensuring that the actual guideline text is essentially locked and can only be edited by a gatekeeper, e.g., Discogs staff, I do think it could be open to everyone and still be successful, so long as, for example, dozens of capitalization examples aren't posted to RSG §1.2.

One thing I have noticed, and maybe it's just me, but I do not see the same level of editing to the unofficial style wiki descriptions as I do with the database. This is interesting because there appears to be continuous discussions about adding new styles to the list. One possible explanation for this is that reference.discogslabs.com is effectively walled off from the database, forums, and even the pages. In my experience, there are often times where I want to look something up really quickly and, if it were a guideline, I would because those are only 1-3 clicks away at all times regardless of whether I'm posting to the forums, editing a release, perusing the marketplace, etc. However, to access the style wiki, unless I'm mistaken, the main way is to go to RSG §9.1.1 and click hyperlink in the paragraph (naturally, I always click the forum link thread by mistake). The reason for this longwinded description is that I believe that an open annotated guidelines (along with the open areas of the Reference Labs in general) would be more successful if it were more visible/less buried. Whether that's a Reference Labs link below the Help Is Here heading at the bottom of the page, more links in the pages, etc., it's not really my place to say plus I don't really know. But as a I do know that if it wexre, then I would access and, in turn, participate in the Reference Labs wiki-area much more frequently. If other s share similar thoughts on that, then I think that would go a long way to ensuring your term as an annotated guidelines 'dictator' was as short as possible.

berothbr
Not to go too far off topic, but I've always wondered why releases can only be merged and/or removed by vote, but labels/companies can be obliterated by one person. Sometimes a label/company profile can be created by mistake, such as a typo, and quickly rectified. However, Delicious Vinyl, Inc. and Delicious Vinyl Inc. is an example of where a simple merge request would be very effective.

Ancient discogs history: label (and I think artist) merging used to be a thing way many years ago when I was new. The feature is somewhat before my time, so I do not know the specifics of why it was discontinued.

One of the things i've been doing in the background is doing an analysis on July's DB forum posts for some data insight into usage. So far, artist & label page merges are both the most common and most resolved thread. Maybe this is something that could be looked more into.

berothbr
One thing I have noticed, and maybe it's just me, but I do not see the same level of editing to the unofficial style wiki descriptions as I do with the database.

Same. Theoretically everyone should be able to create new style pages in the wiki and there's even a subform for it somewhere. Diving into that is heavy on my to-do list. It's somewhat of a hidden feature. I have it on my agenda to try and write some documentation to make it more accessible in the near future.

More to think about, either way!

Diognes_The_Fox
everyone should be able to create new style pages in the wiki

To clarify, I think I miscommunicated. I just meant that one reason I often overlook or forget to use the Discogs Reference Labs site/sub-site/IDK? is because it's somewhat buried/less visible and therefore more tedious to access. This is not a gripe — maybe it's intentional just because it's a 'lab', but I know that I would personally access it more frequently if I could do so more quickly/in fewer steps. So I think my point is that visibility/placement is something to consider as I believe affects usefulness and overall use.

berothbr
To clarify, I think I miscommunicated. I just meant that one reason I often overlook or forget to use the Discogs Reference Labs site/sub-site/IDK? is because it's somewhat buried/less visible and therefore more tedious to access. This is not a gripe — maybe it's intentional just because it's a 'lab', but I know that I would personally access it more frequently if I could do so more quickly/in fewer steps. So I think my point is that visibility/placement is something to consider as I believe affects usefulness and overall use.

I agree (and bump!)

This is still on my radar and might get some action this quarter.

berothbr
So I think my point is that visibility/placement is something to consider as I believe affects usefulness and overall use.

Exactly. The RSG is an equal pain to browse. There isn't even a navigation bar.

jweijde
The RSG is an equal pain to browse. There isn't even a navigation bar.

so maybe that could be a quick win. Improve the RSG, add navigation to it. etc.
even though I am afraid it might not be such a quick win. Not sure it's a structured document at the backend. or just a collection of pages.

jweijde
Exactly. The RSG is an equal pain to browse. There isn't even a navigation bar.

I was referring to whether it would be located on Discogs.com or on reference.discogslabs.com
By comparison, I actually like how the Guidelines are presently structured and presented. However, an annotated version likely could not function in that form without a commitment from Discogs (which probably really means Diognes_The_Fox's continuously cruising the forums for new annotations. Conversely, the latter offers a more organic less time intensive approach and IMHO works really well in the runout groove forum.