[Web4lib] RE: library automation software

C.S. Durfee csdurfee at gmail.com
Tue Jul 19 18:27:13 EDT 2005


Okay, I took what you said in the wrong light.  I've just encountered
a lot of people (and this is by no means limited to the library world)
who seem to have a mental picture that software companies have
unlimited engineering resources that can be summoned if only they yell
and scream loud enough (or even ask really nicely).  It just doesn't
work that way, which as a programmer you've probably experienced
firsthand.  The resources are always very limited and fixing problem x
always means delaying work on something else, which is why it is
important to make sure you know what your priorities are and to know
how to push for what is important in the most successful manner.

As far as complexity... without going into too much detail, some
things that in my experience make library automation software complex
is: implementing byzantine standards like Z39.50 (and interfacing with
other peoples' mis-implementations), I18N issues (esp. when dealing
with a legacy backend + complex diacritics), all the standard
practices and workflows of the library world, data
migration/corruption issues, fairly complicated business rules for
circulation, acquisitions/cash management and indexing, user security
config, etc.

Actually, the hardest part is probably designing an easily
configurable, robust, intuitive administrative interface to do all
this stuff, and to make the whole thing useable for relatively
non-technical people.  That's something that a company like Amazon
doesn't have to worry about at all.  I can't buy my own copy of the
Amazon back end software and then, without knowing any HTML or
programming or anything, make the toolbar purple, replace Jeff Bezos'
picture with my own and make the search box search a subject browse
index instead of a general keyword one...

Certainly none of it is as difficult as doing realtime 3D graphics,
designing algorithms or building realistic AI like you'd have to
handle when programming games, but add all those factors together and
you've got something pretty tricky.  It's not so much that any one
piece is very complex, but that there are so many outside factors that
have to be dealt with, and so many points of failure.


More information about the Web4lib mailing list