Find your local library with a text message!

Today I am very proud to announce that I’ve created the service I wished for ten years ago.

Text LIBRARY 12345 to 41411, assuming that your ZIP code is 12345 (and if it is, how’s the weather in Schenectady?). You will get a text message with the name, address, and phone number of the nearest library.

Here’s how it works:

• You send the text message “LIBRARY 12345″ to 41411 (TextMarks).
• TextMarks uses the word LIBRARY to assign the rest of the message (“12345″) to my TextMarks application.
• My TextMarks application sends the rest of the message (in this case, your ZIP code) to a server at freeshell.org. Running PHP scripts on freeshell.org isn’t free, but for a one-time $36 donation, you get that privilege.
• freeshell.org searches NCES IMLS for branch and central libraries, then finds the first result and identifies the name, address, and phone number in its HTML code.
• freeshell.org sends back a very simple result with newlines where appropriate (view the source to see how it’s really formatted).
• TextMarks sends you the literal answer from freeshell.org, plus a brief advertisement with which TextMarks pays the bills. (I don’t see a penny from this except when y’all phone up potential employers and tell them they’d do well to hire me for 2010.)

Things I’d like to add eventually:

• Support for other location forms like SCHENECTADY[,] NY or 1060 w addison st chicago il, probably with a call to a Google Maps API
• Support for libraries outside the US (though I caution you, if you give me a location on the south end of Grosse Ile, Michigan, my script won’t know that the Essex County Library in Amherstburg, Ontario is nearer than Bacon Memorial Library in Wyandotte, Michigan — though in that particular case you might not care unless you have a boat)
• A minor efficiency tweak: search for libraries within 1 mile, and if none are found, try again with 5 miles, 10 miles, and 50 miles before giving up.
• If no results are found, say so apologetically.

Please tell everyone about it, especially truckers and other business travelers, snowbirds, people who are moving to a new home, etc.

And please let me know what you think, and how to improve it! Thanks especially to Susan Hansen for suggesting IMLS as a more recent source of data.

OCLC should share WorldCat data for commercial and noncommercial use

OCLC’s Review Board of Shared Data Creation and Stewardship asked for comments on OCLC’s policy on sharing bibliographic information with partners who make a profit.  Here’s my comment, which I’ve also submitted to the board:

OCLC supports “WorldCat data sharing that encourages innovation and benefits libraries, museums and archives while protecting OCLC’s members’ investment in WorldCat.”

Sharing bibliographic data freely, without worrying about whether it advances profitable causes, achieves all these criteria.  It certainly encourages innovation; it certainly will benefit libraries, museums, and archives; and the wealth of innovative services based on freeing the records will handsomely repay members’ investment in WorldCat while benefiting the world at large as well.  Let us not hoard metadata like a jealous dragon.

As librarians, we don’t ask our patrons whether their in-house use of business reference books is intended to help a for-profit entity. We don’t have separate usage policies for entrepreneurs and non-entrepreneurs. We trust that the use of our information will benefit society at large in many ways, some obvious and some subtle. Why should OCLC behave differently?

The thing about grant-funded positions is…

Grant-funded positions are liable to run out of funding, and that’s exactly what has happened with my most recent large-scale consulting project.  I’m still available as a consultant, and I’m also looking for work in Atlanta.  Thanks to those who have already wished me luck—I’m sure you’ll be nearly as happy as I am when I have good news about employment!

Where can I submit abstracts in a way that will help researchers?

There are loads of articles out there with no abstracts available, leading to NAA bias (‘No Abstract Available’—i.e. the tendency to overlook highly relevant articles because the title fails to convey the relevance and there’s no other good metadata).

So let’s say I’ve read a scholarly article and I believe there’s no abstract available. I’d like to write an abstract and make it easy for researchers to associate it with that article. How should I do that? Is there an Abstracts Wiki somewhere? Should I use a COinS reference as a trackback of some sort?

My new job: Horizon vs. Evergreen cage match

I’ve been hired as a consultant for the University of Utah for a one-year gig. (Actually, it’ll be a little less than a year, since the LSTA money has to be spent by September 1, 2008.)

They’ve been using SirsiDynix’s Horizon ILS for a while now, and they’ve asked me to conduct a feasibility study on using Evergreen. I’m going to be figuring out how well it works with their data, their workflows, and their business needs, and identifying any potential trouble spots.

I’m thrilled that Evergreen is being seriously considered by large university libraries. I’m not a zealot, though; for example, my preference for a desktop operating system is “whatever seems to work pretty well in the given situation”. I won’t hesitate to burninate Evergreen’s faults, whatever they may be. After all, my report is only valuable to the extent that it corresponds with reality.

In that spirit, I welcome propaganda from all sides. If you’re a SirsiDynix fan (or employee), and you want to point out Evergreen’s most serious flaws, please do. I’ll be glad to verify your comments in the course of forming a well-informed opinion. Similarly, if you’re an Evergreen fan and you want to point out its features or defend it against defamation, I’d love to hear tips about how to verify the truth.

It’s… ALIVE!

My email address, that is.

Color me sadder-but-wiser. For the first month or two of my job hunt in Atlanta, ben@benostrowsky.com was hosted on a site that was sending some very peculiar, non-SMTP messages to the MTAs that tried to contact me:

MAIL FROM: ostrowb@tblc.org
250 2.1.0 ostrowb@tblc.org... Sender ok
RCPT TO: ben@benostrowsky.com
alias([username redacted—Ben])
aliaswait(implicit:/etc/aliases)
250 2.1.5 ben@benostrowsky.com... Recipient ok

Some MTAs had no problem with this. So when I tested the account, I was able to receive mail. Unfortunately, most people just got a cryptic bounce message.

I’m guessing that this has led to several potential employers deciding not to bother phoning me, especially since my cell phone still has a Tampa area code. (I’ve solved that problem by getting a local number from GrandCentral and having it ring on my cell phone—but all my circulating résumés have the Tampa number.)

Two nights ago I asked the readers of my personal blog to email me and post bounce messages as comments. Very few got through. Last night I repeated the experiment; lots of people got through and nobody has reported a bounce. So now I’m confident enough to email my potential-employers-thus-far and say “I’ve switched email hosts because of technical problems. Have I missed anything?”.

And this will be a good story for the interviews, when I’m asked to talk about lessons learned from mistakes.

Break my code—I’m going to Atlanta for a week

I’ll be in Atlanta for the next week, looking for a place to live while my wife goes to graduate school.

Meanwhile, my reworked HIP-based catalog, SunCat 2.0, is at a fairly good place for a first look.  Please check it out and tell me what you think; it’s an ambitious project and I know there’s still quite a lot to do.  The comments link at the bottom of each page will send me your browser/referer info, so please use it.

And if you’re in Atlanta and would like to meet while we’re in town, email me!