Re: LDAP Comments

[[e-mail]]

> From: Alan Shepherd <[e-mail]>
> 
> this issue has come up before and it may well be the case that the
> quipu implementation is lacking (side-issue and I don't want to
> discuss it here), but I'm fairly sure that the X.500 standard says
> that DSAs should be able to cope with the sort of search that PAP
> doesn't like.  It doesn't at the moment and I expect that every
> implementation has its own set of problems, but sooner or later, I
> think that you are going to have to stop dissuading every from doing
> searches on pizarro DSAs and make them work better !
> 
> Alan


Sorry Alan, 

I regret you seem to be hurt by my message, but the case is not
to defend or attack a product
  but to help/inform users of X.500 services, about some important issues
  AND 
  help developers (including NeXOR) getting caught into 	
	developing solutions with uncertain future, for their
	own sake

I understand that you may feel unhappy at getting conscious that 
QUIPU will have to undergo a rather drastic evolution, but frankly
being aware of that before there is any real market pressure
is certainly an advantage and not something you should complain
about.

here are a few points:

1. even if I am from INRIA and PIZARRO has it roots within INRIA,
I am not at all making a case PIZARRO against QUIPU.
Moreover I am not directly involved with PIZARRO or derived products evolution,
as PIZARRO comes from Christian Huitema team.
If you have somtehing to say to Pizarro people, please contact them directly.


2. We (my team) are being funded within PARADISE by CEC DGXIII to take
care of interoperability in an heterogeneous environement.
That is taking into account plenty od implementations such as
Siemens Dir.X, Marben and many more.
  You should have noted that I have been careful at not limiting to
  pizarro, even avoiding as much as possible using the name, and 
  giving a generic view.
Thus that type of activity and concern for other than QUIPU is for
a commitment we are being paid for.

3. We consider as our duty to notify interested parties as soon as possible,
of potential difficult interop/interworking issues.

4. If I am taking usualy the french example
   . it is for one point because it seems to be the only country where
	the master is non-QUIPU, thus where we are really faced to these 
	interworking problem. And where probably there exist a handful
	of people with concrete experience of these issues
   . it was also to avoid criticizing any other third party, prefering
	to see all the critics come to us, that pointing out any
	vendor (for which in some case I KNOW, that the problem would
	be very similar). Please not that very often I am even pointing out
	our own deficiencies, and not taking a partisan approach.
   . if others (eg. people involved in interworking within Surfnet)
	want to contribute, they are free to do so, but I am not going
	to disseminate information related with a particular implemntation
	(not PIZARRO) if they decide not to do so themselves.

5. Fundamentaly I disagree with you about the search approach, and though 
	convinced that QUIPU and QUIPU oriented tools will have to eveolve 
	I am not going to enter comparative debate, nor even present what
	I consider to be the right approach. If you want to debate this
	with Pizarro people, and they are willing to present their plans
	and future design it is up to them and you. But I will not let you
	get me invaolved in any fight. I will limit to advice you to
	talk within the IC with some key people who perfectly know
	what is the issue about, and perceive things with a larger
	spirit of mind


best regards,

-- PAP