Re: Super MARC to code RDA?

Posting to RDA-L

On 29/09/2011 21:33, Jonathan Rochkind wrote:

I am suspicious of your claim that either catalogers or programmers have any special facility with binary MARC. Certainly catalogers, but even most programmers working with MARC use pre-built libraries for reading/writing binary marc, and have no facility with it themselves. (With gratitude, because it is a _bear_ to work with).

To make sure we’re talking about the same thing, here is a literal binary MARC record. Is this really something you think catalogers have any familiarity or skill to transfer? If you are looking at anything other than something that looks like what I paste below, you are not looking at the actual MARC format, but a representation of it that could be created from anything that uses the same schema/vocabulary as marc (for instance MarcXML or a marc-in-json).

This is what an actual binary marc record looks like, in the format you are talking about expanding to take for instance three-digit tags. This is not a format that anyone can write or read by hand, it is not something catalogers actually look at. This is not an error in copy and paste into this email, this is really what binary marc is (yes, it’s all one line — it’s possible my email client or yours will try to seperate it into multiple lines of reasonable length. If that happens, it’s no longer valid binary marc, this should be all one line).

[an ISO2709 MARC record]

Yes, and it is also important to keep in mind that in modern catalogs, the only time this form of the record exists is when it is transferred (or communicated) from one library catalog into another. That is why they call it a “communication format”. Once it is in your own catalog, it is reworked in all kinds of ways, normally into a type of relational database format. In Koha, one cell of the relational database contains the entire MARC record in MARCXML. From there, in Koha, the information is reworked into a series of indexes by a system called Zebra, which allows for the headings you see in the left column after you do a search,

When someone else wants the record from your catalog, the computer recompiles it into ISO2709 format and sends it on. Someone mentioned that in one catalog, the ISO2709 format is retained. But I don’t know why.

So today, this form of MARC exists for just a few milliseconds while it is compiled, transferred, and decompiled. The only software that uses it is software that is designed for libraries, not for World Wide Web  transfer of information. Browsers can’t use it, spreadsheets can’t use it, word processors can’t use it, citation management software can’t use it, that is, until it is converted into another format, such as a type of XML. MARC served its purpose and should now be laid aside.

I agree that the numbers of the fields and subfields transcend languages, but I fought that fight and lost. Developers refuse to work with the numbers and insist on the names–even in English. OK. Let them have it–they want our records and we want to share them (I hope). Let’s save our strength for other battles and move forward.