FirstSearch matters

[Thomas Dowling]

Let me combine a couple different responses regarding FirstSearch into one.

First, for people who are trying to provide Web access to FirstSearch using
OCLC's own scripts, there is a mailing list that carries, among other
things, announcements from OCLC regarding service changes.  Send "subscribe
webscript-l Firstname Lastname" to listserv at oclc.org.  BTW, the natives on
that list are getting restless; the consensus is not that OCLC provides
top-notch service for a well planned, well maintained operation.

Mack Lundy asks confusedly what version of WebScript we're supposed to be
running.  I believe everyone needed to upgrade by 9/24; I'm using version
1.91 of WebScript and 1.93 of the FirstSearch login script.  It works for
all but one of our databases, and this morning's job is to figure out
what's different about that one.  The WebScript scripting language and the
FirstSearch and EJO scripts are available at ftp.rsch.oclc.org.

Doug Cornwell asks more generally what's being discussed here.  I haven't
tried Bill Drew's script, but OCLC provides a scripting language,
WebScript, with which you can create CGI programs to automate server <-->
server web connections and do autologins, etc.  Kinda sorta like a Tcl
"expect" for the Web, except that expect is much better documented.  Our
impression here is that, for some reason, OCLC is so wedded to its login
procedure that they preferred to create this kludge rather than provide
IP-based user authentication.  Of course, you'll probably still have to
hide your WebScript programs behind some form of IP checking.

Jian Liu asks plaintively if anyone would care to comment on FirstSearch
via the Web.  I'll leave out some of the more colorful adjectives that come
to mind (except perhaps "ugly" and "unhelpful"), but I certainly don't
think the user interface is anything to write home about and definitely not
worth the maintenance hassle they put on web managers.  They seem to have
taken some challenge to see how many Netscape-specific tags they could fit
on a given page, and while I have yet to see a vendor display any regard
for validating HTML, some FirstSearch pages would be candidates for the
most gnarled source I've seen in a long time. 


Thomas Dowling / tdowling at ohiolink.edu / http://www.ohiolink.edu/
Ohio Library and Information Network. Please do not perpetuate the GOOD
TIMES virus hoax, started in 11/94 and circulating again in 9/96.  Refer
to: http://ciac.llnl.gov/ciac/notes/Notes04c.shtml