[Date Prev][Date Next] [Chronological] [Thread] [Top]

Re: xmLDAP?

>Did anyone go to IBM's xmLDAP talk at XML 98?
>There's an infoworld article:
>I can't find any online specs, however.

I did and I owed the list a summary of the talk, but I've been so busy
cleaning up end of year stuff that I haven't had a chance.

I should've done this sooner since the summary is quite easy. About
all you need to do is take the productions for the protocol and turn
them into XML element declarations and you have the language. The xmLDAP
client communicates with an LDAP server via a proxy that accepts the
XML messages and passes them onto the server in proper LDAP messages.

Since I had been one advocating the replacement of ASN.1 with XML, I
asked Doug for a rationale. He said that IBM felt that this would
make LDAP more easily accessible since you could deal with it directly
through the browser. Having thought about it for a while, I think that
this is enough to make the change worthwhile. This would reduce the
API to some very primitive ops and reduce or even eliminate the need
for gateway apps.

At first glance, it didn't seem like that big of a deal. xmLDAP doesn't
do anything to simplify things for the user as I feel I was led to
believe from the abstract in the XML conference brochure. However,
I think it could do a lot to simplify client implementations and 
probably even some server-side issues. One of the nicest, immeditate
benefits that I can see would be the normalization of character sets to

I haven't looked at the code for the front end of the server, but I
bet it'd take less than a week to hook in an XML parser and eliminate
the need for a proxy with openLDAP. If I had some familiarity, I'd
probably do it just for fun, but it'd take me a day or two just to
sort through that code and I have other stuff that I'm already signed
up for that I have to finish first.

Doug, correct me if I've misrepresented your proposal here anywhere.
Do you think you could post your paper to the list? I don't think
there's any need to worry about copyrights, the GCA hasn't bothered
to put any copyrights on papers at previous conferences, so I doubt
they started now.