Re: Comments from Christian H. on LDAP

[Steve Hardcastle-Kille <[e-mail]>]

 >From:  Christian Huitema <[e-mail]>
 >To:    Steve Hardcastle-Kille <[e-mail]>
 >Subject: Re: Comments from Christian H. on LDAP
 >Date:  Thu, 14 Jan 93 10:56:52 +0100

 >Steve,
 >
 >We obviously disagree with the importance of keeping the name service
 >compatible with OSI-X.500, which you describe as "the whole world". After

I think that this is very important.  I'm not quite clear on your
take here.

 >playing in this field for quite a long time, I believe that OSI is now doomed
 >and is just collapsing under its own weight.

I think that this really depends on definitions.  If you mean by OSI,
the original all encompassing vision of a total solution to all
communications problems, I will definitely agree with you.   On the
other had I believe that other components of OSI are doing very nicely
indeed.

 >
 >If one buy's that line, the next think to do is to try to salvage a few OSI
 >jewels out of the wreck. Such jewels, IMHO, include ASN.1 and also X.500 --
 >although X.500 has to be fixed in at least one aspect, i.e. by "removing" the
 >correlation between navigation and naming hierarchy. 

Well, we at least agree that X.500-derived technology has a future.  I
think that the hierarchy is a fundamental strength of X.500, not a
weakness.  There has been lots of research and all sorts of
great-sounding research projects on non-hierarchical naming.  I belive
that it will continue in this vein forever - a holy grail that will
generate many research grants, but which will never have much impact on
operational networking.  If you want to play with non-hierarchical
naming, this is fine.  If you want to use X.500 as a base, even
better.  However, lets not tamper with the core X.500 service as a
part of this research.

 >Thus, the next think to
 >do would be to make a light weight version of the DSP, and possibly unify it
 >with LDAP in order to obtain a reasonable "Internet white page service" that
 >would reuse large chunks of the X.500 technology -- and working codes. Note

I think that a lightweight DSP could be useful.   However, unifying it
with LDAP is not.   DSP of any form has issues which can be avodied in DAP.
This is not the highest priority issue.

 >that there is an alternative, i.e. to observe the development of an entirely
 >different competing technology like "whois++", which may result in a situation
 >where X.500 would have about as much future 

I'm happy to compete!   I think that X.500-based technology has much
benefit, and will show well in a competition.

 >as X.400 (who would deploy X.400
 >now that MIME is available?).

Just wait - X.400 has an excellent future ahead.   

 >
 >Obviously, *you* do not believe that OSI is doomed. Otherwise, you would not
 >have founded a company to sell OSI products! 

I think that some parts of OSI (certainly X.400 and X.500) and
technology derived from them have excellent prospects.   For a company
based on "doomed technology" we are not doing at all badly.

 >This will not be the first time
 >that we differ.

Indeed.   You are one of my favourite people to disagree with.

 >
 >Christian Huitema


Steve