I've work with names and other demographic data darn near my entire career. I've also had name issues due to the length of one of my names.
Things are better and more accommodating than twenty years ago when every byte was precious. But some cultures and individuals get ridiculous. I'm not allocating 400 characters of space in a standard name field just because you want to include your family tree, PGP key, DNA sequence and homeland geology in your name. (I will put this in another table, and link it relationally so the software and the people can access your true name.)
Yes, who cares about those ridiculous cultures. Why can't they just use normal names that can be represented by 10 or fewer ASCII characters like the rest of us? They're just being difficult.
Software can be made to be very accommodating. Most systems I work on allocate a 50 character nvarchar for a full name, various fields for prefix, first, last, multiple middles, suffix, maiden, titles, informal salutation, aliases, nick names, gender change issues, etc. And various flags and other fields/tables for exceptions, notes and warnings, etc...
But you wouldn't believe the bad data and crap people put in name fields unless you've worked with a huge number of sources over the years. I've seen every bad variation possible. And that's not including address fields. Or combined name/address fields with multiple names and multiple addresses in the same record with a complexity that would vex a dyslexic Talmudic scholar.
> But some cultures and individuals get ridiculous. I'm not allocating 400 characters of space in a standard name field just because you want to include your family tree, PGP key, DNA sequence and homeland geology in your name. (I will put this in another table, and link it relationally so the software and the people can access your true name.)
This attitude boils down to "I write software that is easy for me to write, not easy for you to use." Why should any user want to use the software a developer that is hostile to their existence?
The key is obviously to make some trade-off between usability and difficulty during creation. If the software is never finished because running test cases with 10e10 chars-long names takes too long, users won’t even be given the choice of not using it.
That said, I’d love to have my PGP fingerprint on my ID :)
I've a more detailed comment below, but the issue isn't new databases per se, but the huge number of legacy systems out in the world. I have no control of those, I end up building software that has to work with them and help the users and their clients.
You deal with questionable quality data that comes from hundreds of different systems running software that may be decades old with demographics from around the world that may have no common conventions. And you fight for the extra disc space and processors to handle an extra hundred characters per record, because with a huge number of records, even that may cause issue with storage, backups and time to process.
And yes, I have to write code to deal with all of the "Little Bobby Tables", too.
Things are better and more accommodating than twenty years ago when every byte was precious. But some cultures and individuals get ridiculous. I'm not allocating 400 characters of space in a standard name field just because you want to include your family tree, PGP key, DNA sequence and homeland geology in your name. (I will put this in another table, and link it relationally so the software and the people can access your true name.)