I imagine it would be pretty hard to do business if I owned a small brewery but insisted on selling my beer in spherical beer bottles year round.

That would probably create all kinds of problems in terms of packaging, delivery, and shelf space in a retail setting. Sure, I could do it if I insisted but I imagine the financial consequences for my small brewery would be too great versus just complying with the bottling norm.

Having said that, I sometimes I get a little sick and tired of talks of "standards" in Library Land with respect to digital information. It's as if some people still have a physical-world mental barrier clouding their thinking.

My take is that, standards aside, it's pointless to re-invent the wheel in any walk of life so, yeah, give standards a chance but if the wheel doesn't do what you want then one shouldn't hesitate to "roll their own". Moreover, I'd argue that a great many "standards" are in fact born out of that very kind of dissatisfaction with the status-quo. Reminds me of this great quote by George Bernard Shaw:

The reasonable man adapts himself to the world; the unreasonable one persists in trying to adapt the world to himself. Therefore all progress depends on the unreasonable man.

And isn't that the very freedom that digital information provides? Isn't that one of the points of the web – that everyone can contribute with very minimal pre-requisites?

But it seems like some people think metadata formats and ways of doing things always have to be determined by committees.

I just don't get that especially since the device anyone is reading this on exists in large part due to private entrepreneurship and not a bunch of people sitting around a table talking about how great the world would be if we all did things their way. Actually, that's exactly what happened, but it wasn't just a bunch of academics without the resources to transform theory into practice.

The difference in rolling one's own is, of course, that you have to make your data comply to existing standards or practices if you want to reap the benefits of doings things the same way as others like linked data and what not.

Big deal. If the data exists at a granular level it can be "shipped" in spherical AND regular bottles simultaneously.

And if you're a big enough player you can make people go along with your vision and indeed ship a product in round bottles after all …

Too bad those aren't filled with a nice IPA.

If anyone out there knows anything about this, please post a comment or drop me a line.

It just seems to me that a couple of things are happening that create major problems:

1. so called "best practices" often seem to be born more out of individualistic grant-receiving concerns than they are of truly getting several institutions on-board with effective, affordable, and shared approaches to workflows

and

2. an industry so concerned with standards and organization seems to have such little concern for organizing a do-able approach to ensure certain standards – and those "best" practices – get met in a relatively easy and affordable way.

So if there isn't already, why isn't there major activity to develop a Linux distro specifically for digital library functions?

If that were the case, couldn't some of these "best" practices actually get built into the very OS people use?

All the scanning, audio-video, metadata, and delivery applications could be integrated within the OS, ensuring compatibility for whatever institutions use it.

Lists of compatible scanners and A/V equipment could get published, ensuring that people will buy equipment that will run on the OS.

Software updates could me made and get implemented in real-time across institutions.

Using all open-source software could standardize technical metadata outputs and save institutions lots of time/money – allowing them to pay for better and more talented employees.

Etc, etc, etc.

You could even call it LibOS* and use an image of Tux making a libary-shushing pose!

Maybe it's time to stop paying lip service to collaboration and really get things moving.

"It's the operating system, Stupid!"

*BTW: MIT appears to have had a late 90's project that used the term LibOS, but it's something totally different.