if you look at the bar codes in Search Results - MusicBrainz
only one entry has 00 at the beginning. As strings, the others shouldn’t match. They have only one 0, but they do. Are they stored as numbers? It doesnt make sense to me
why arent they all displayed as padded with two 00s?
nice app though!
I don’t want to wake the Barcode\EAN debate up, but I have a quick question from your example @outsidecontext
Is it the bold we use as the label’s catalogue number? Or the whole number?
From what I understand you have said, this 1997 release has two catalogue numbers: CDNODATA 02 and 55229 2
The complete one, for sure.
“9981592” is printed on the spines of the release @outsidecontext mentioned.
EAN or part of it may or may not match the catalog # or part of the catalog #, there are no rules there, different record companies use different methods.
I think I’ve seen somewhere theories why part of the EAN text is printed bold on some label’s releases but can’t recall anything right now, sorry.
Okay. That makes sense. Whole number is on the spine, then whole number is catalogue.
If only 55229 2 had appeared on the spine, then the shorter number would have been the catalogue number.
The examples @outsidecontext linked didn’t have any artwork attached so it was hard to be sure. I thought he was pickout out the bold text as relevant.
Thanks, clarity is good and I had been using the full number most of the time anyway. 
I didn’t want to imply that a part of the barcode is always also to be used as catalogue number 
. That depends a lot on the specific case. I just marked parts of the barcode as bold to visualize which part appears also in the actual catalogue number
I think the numbers in bold are indeed the “internal catalog number” since the last digit is the check digit and the leading digits usually are identical across many (or even all) releases of the same company. In case of bigger companies they are even exclusively using certain EAN ranges for their products.
But if the full number is printed on the release it makes sense to enter all the digits to make sure that a search for the full number will also return the release, not only a search for the bold digits.
I was not reading the Bar Code. I was reading the number ABOVE the barcode box, and the number on the spine. I normally put the full number, but as I had spotted that there is usually a bold bit in the middle of the full number I had wondered if this was the official cat no. An Internal number as @kellnerd picks out. Reading your example I got a little confused. Hence the question.
I know some releases I have seen both printed out on the back. In those cases I’d write both as the label’s cat no. In the example I have shown I have not used the short number is it is not shown separately. Cat Nos on that release are now 7243 8 5529 2 5 and CDNODATA 02 and I go back to ignoring the bold bit.
I guess I am a visual person who needs pictures to understand things 
I think the entire catalogue number vs. barcode story goes a bit like this:
As soon record companies became a thing they needed a way to properly identify the products they produced and sold for their stock keeping and accounting, and also to allow people to order things without ambiguity. So they came up with a scheme for catalogue numbers for all of their products. (Who here actually remembers having ordered things from a paper catalogue by filling out an actual paper form with a list of catalogue numbers, putting this form into an envelope with a stamp on it and putting this into the nearest post box? I guess the younger ones cannot imagine that this actually worked 
).
Anyway, in the 1970 people thought some identifier to uniquely identify any product would be a great thing for international sale, and they came up with the UPC and then EAN number system and made it machine readable with the use of barcodes. Companies started adopting this, but that meant now they actually had two unique identifiers for their product: One internal, the catalogue number, and one universal for the outside world.
Now they could just get rid of the catalogue numbers and use the EAN. But the catalogue numbers usually were much shorter and often contained additional information. E.g. “SHVL 804” is the catalogue number of the 1973 Vinyl release of Pink Floyd’s “Dark Side of the Moon”, “Q4SHVL 804” is the quadrophonic version and “TC-SHVL 804” is the tape cassette. That is something a label employee can easily remember, making the catalogue number more useful.
The barcode “077774600125” is for sure much harder to directly associate with the 1985 US CD release of the same record. But luckily the actual number the company needs to care about internally is shorter. Originally for UPC the first part “077774” is assigned to the company*, and the last digit “5” is a checksum. But they can decide freely on the rest “60012”. So in this case they came up with a catalogue number of “CDP 7 46001 2”, which is at least shorter and easier to read then the full barcode and the “CDP” presumably adds some more information (with “CD” likely indicating a CD release). But it is also possible to derive the full barcode from the catalogue number.
So some labels still kept separate catalogue numbers. But to make it easier to them many use the same number as the product code for the barcode. Or a part of it. But they sometimes add some letters and stuff to add additional information.
I think in your Ok Computer example they highlight in bold the for the label relevant part of the barcode, the actual product code. Whether they also consider this the “catalogue number” or if they have a distinct catalogue number is a different question.
–
*) EAN uses a variable length manufacturer code
I think part of the barcode is as Kellnerd notices - allocated in blocks to companies. The first digits are country related. (Haha - just learnt that the inventor of the barcode also worked on the Manhatten project)
I remember Kimble tags: Kimball tag - Wikipedia