I’ve been wondering about this for years, but now it’s come to a head. Discogs will often have multiple versions of the same release with the only thing separating them being things like matrix/runout inscriptions. Recently I got ahold of a repressing of this album and it looks visually identical to the original 1997 version, only that the inner matrix on each disc has different inscriptions. Is this enough to qualify for a new release? I’ve already gone ahead and attached Disc IDs; if I have to make a new release I’ll move them.
4 Likes
I agree with what @kepstin said in #musicbrainz,
on a CD? probably not, a release in continuing production (or even mass production) will probably have several pressings made
Of course, it’d be great to record this data somewhere. Yay for Annotations ?
5 Likes
If there’s a difference, and you have the time and inclination to create a separate release, with proper annotations/ disambigs, why not!
Some people might think it unnecessary for their useage scenario, but that’s not really a good enough reason not to have it imo. Maybe if the DB had limited space for releases, but I don’t think it does! 
3 Likes
The issue is not the space in the database, it’s one of maintainability. If there are thirty different pressings of some major-artist release, differing only in the ring on the CD, who will enter all these, and who will fix them all if there was a typo in the tracklist?
In fact, individual release copies aren’t 100% identical, either. There may be tiny fluctuations in the print colouring, and it’s very likely the orientation of the CD groove compared to the label print will never be exactly the same between two copies (though you won’t notice without a microscope and precise measurements). So, according to Wikipedia, Adele sold more than 100 million records to date; add 100 million releases for her, then?
7 Likes
[quote=“chirlu, post:4, topic:79829”]
add 100 million releases
[/quote]That’s clearly not what I said or implied, but if we’re going for hyperbolic arguments, then no, I don’t agree with you wanting to merge everything in the DB into 1 release, for ease of maintenance.
But seriously -
How do we decide if we split one vinyl release, that has 10 different vinyl colours, into separate releases? If someone finds enough value in it to do it properly, then they do it. Otherwise they don’t.
If a fluctuation in a print colouring is considered important enough to someone for them to add it to the database, with the appropriate information and hopefully scans, then that seems appropriate to me.
In @HibiscusKazeneko 's case in particular, if this re-press represents something of significance to her (eg collectors try to collect both, one is worth more etc), then why not, we often do this with limited edition vinyl re-presses, even if the content and packaging is pretty much exactly the same.
That’s not to say that I think it automatically qualified/ requires a new release, that would be daft…
6 Likes
I think it’s totally appropriate to say—as a community or through a style dictator—“these are the properties that we use to determine if two items are considered to be the same release; these other properties are not taken into consideration (but by all means, note them in the annotation!)”
I don’t base my opinion on my own useage scenario, but on what I consider to be a useful, maintainable data model. By definition, that’s going to involve some trade-offs.
11 Likes
Would it make sense to have separate releases if they are intended to be different? For example, remastered, different cover art, extended/limited editions, etc. Multiple CD pressings wouldn’t qualify because they are just making more copies. Small variations in vinyl color wouldn’t qualify, but a major color shift would (e.g., black vs red).
If the minor variations are important to someone, collecting the information in the annotation would be a great place for it.
3 Likes
[quote=“CallerNo6, post:6, topic:79829”]
I don’t base my opinion on my own useage scenario, but on what I consider to be a useful, maintainable data model[/quote]
Do you think one more release addition is maintainable?
I think I might not have made myself that clear - I’m not saying any negligible difference automatically requires a new release. I personally would not ad a new release in this case.
If we encounter problems this becoming unmanageable/ someone adds 100 million Adele releases, then I would very much welcome a style guideline that tries to roughly box in what the majority thinks is ‘useful’ 
3 Likes
I really disagree here, and it likely would get merged together quickly. This is one difference between MB and Discogs where I strongly believe MB is superior. Discogs is a mess here. Not arguing against having the info somewhere, but separating identical releases because of this minor manufacturing detail does not make much sense. MB is very strong in avoiding data duplication.
2 Likes
Even different cover art isn’t always a surefire reason to have separate release. Last year there was an album release (by Hot Chip), where, via a combination of different colors and a line pattern, every copy of the album had a unique cover (well, not every copy, but there were over 130K variations).
8 Likes
One man’s minor manufacturing detail is another man’s important distinction.
Compare https://musicbrainz.org/release/40e0e68b-8141-444d-88e0-12224f8d7c2b and Release “Ring ring (bara du slog en signal) / Åh, vilka tider” by Björn & Benny / Agnetha & Anni-Frid - MusicBrainz – should they be the same release or different?
2 Likes
I think there is no question, these examples have different printed credit, they are different enough for no one to wonder. 
2 Likes
I agree, but I don’t think it can be said that the difference is intended. No one at the label sat down and went, “I think for this release we’ll credit them a little differently from the last issue”. So in that sense it is merely a minor manufacturing difference. You sometimes see similar things with a slightly different position or size of font on a record label or the like, my point is that there’s many degrees of difference possible
2 Likes
It’ll always be a trade-off, won’t it? What’s the benefit to creating a new release for each matrix number (compared to adding multiple matrix numbers to a single release)? Does it outweigh the cost in complexity (entering, maintaining, and using the data)?
To me, it would be like creating a different release for each Disc ID. I guess we could, but the resulting thing would no longer feel like a “release” to me. It’d feel like some lower-level thingy. If we need a lower-level thingy, then fine, let’s add one to the schema 
4 Likes
When and where did I ever advocate for this??
Isn’t that what we’re talking about in this thread? Maybe I’m confused.
Sorry, I don’t mean to misrepresent anybody.
No worries, I’m just feeling a bit misunderstood, because I feel like I’ve already tried to clear this up a couple of times in this thread #feelings 
[quote=“aerozol, post:8, topic:79829”]
I’m not saying any negligible difference automatically requires a new release.[/quote]
I might be repeating myself but -
I don’t think a difference in matrix numbers means that somebody should by default enter a new release - but that if someone finds it important, in a specific case, for instance it makes a large difference to the value of the release for collectors or represents other interesting data (eg someone’s trying to track what a specific pressing plant has produced, dates for repressings, or it’s a different © holder that’s not reflected on the packaging etc), then our database is equipped to have another release added.
But I might have been misconstruing @HibiscusKazeneko’s question as well -
If the question is “the matrix number being different is important to me (for whatever reason) and I would like to add a new release, can I?”, I would say, sure, if the differences are made clear.
If the question is “should I enter another release just because this release has a different matrix number?”, I would say, no.
8 Likes
The releases that have multiple TOC are for the almost unanimity pre‐NGS releases (all similar tracklist editions were grouped as one release), IMO.
I usually cleanup my editions by removing such TOC, I remove all pre‐NGS Disc IDs from my edition when they are still present on at least one other NGS migrated edition and when nothing in the edit history says what edition it is.
You usually see about one TOC per split release.
3 Likes
Well I still think having a medium as an entity would help. They’d be the ideal site for mastering credits, they’d allow for box sets to reuse entire discs.
Then it would be entirely logical to have separate mediums per discid and/or runout.
Releases would then have one or more medium sets associated with them.
But then it would probably become hard to decide between adding a new medium set to an existing release, or a new release to a release group.
3 Likes
As much as I hate to necro-bump a thread, I have to bring this back up. I understand the issue of maintainability for differing runout inscriptions, but there’s another issue: transparent disc labels. I have a later pressing of this release in my possession and its runout inscription is different from that present in the uploaded scans; since the disc label is mostly transparent, the inscription is visible from the non-playing side. Should I create a new release based on this?