"Archiving" e-journals

[Bill Pardue]

We're starting to look into the possibility of "archiving" scholarly 
and technical  e-journals, rather than just linking to them.  The 
essential concern is that if the remote web site goes down or we 
somehow lose access to the files there, we don't "have" any of the 
sources (unlike paper journals--if the publisher goes under, you've 
still got the back issues).  The main problem is that e-journals, 
especially those in HTML format, are highly-structured documents.  
You can't just hit the "save" option and get the images, related 
documents from other links, etc.  Even if you take the time to 
recreate the e-journal's directory structure and download all the 
relevant files, you may still have problems of file name 
incompatabilities (unix names that don't stick to the 8.3 naming 
format...etc.).  I've found that even a single short article can take 
over an hour of re-editing just to get it to display properly.  We're 
also aware that there are significant copyright issues, but my 
concerns right now are strictly technical ones.

Is anyone else addressing this issue at their library?  Have you
developed any effective strategies that don't resemble using slave
labor to re-write web pages?  I don't have any experience with
"mirroring" any web sites.  Is this a possible automated solution to
the problem?  Are there any broader initiatives looking into the
problem of archiving and preserving e-journals locally? 

Thanks in advance for your replies!

--Bill Pardue

----------------------------------------------------------
    Bill Pardue--Electronic Resources Librarian
    Galvin Library, Illinois Institute of Tech.
                Chicago, IL  60616
312-567-3615/312-567-5318 (fax) pardue at charlie.acc.iit.edu
----------------------------------------------------------