[Search for users] [Overall Top Noters] [List of all Conferences] [Download this site]

Conference azur::mcc

Title:	DECmcc user notes file. Does not replace IPMT.
Notice:	Use IPMT for problems. Newsletter location in note 6187
Moderator:	TAEC::BEROUD

Created:	Mon Aug 21 1989
Last Modified:	Wed Jun 04 1997
Last Successful Update:	Fri Jun 06 1997
Number of topics:	6497
Total number of notes:	27359

366.0. "PROBLEM: FullName arg via Iconic PM not right" by COOKIE::KITTELL (Richard - Architected Info Mgmt) Sun Sep 30 1990 01:21

A directive argument of type Fullname representing a global entity results
in different ILV showing up in the in_p depending on which PM invokes the
directive.

With FCL, the command is 
MCC>Archive NodeStorage R.OREO VMSFile TEST Archive R.AIM

This results in the directive receiving this ILV in in_p:

[  0 ] (
    [  21 ]     aa 00 04 00 fb 20 80 13 83 84 98 ce 93 00 07 00 01 03 41 49
    4d 00 00
    )

The directive can extract the FullName from the ILV and build an entity spec
that will match the right instance in the MIR.

The default namespace is LAMAMA_NS:, by the way.

With the iconic PM, the Archive operation box has the proper required Archive
field into which I put Archive R.AIM. The directive gets:

[  0 ] (
    [  21 ]     aa 00 04 00 fb 20 80 13 83 84 98 ce 93 00 00 00
    )
 
Looks like something is missing, but I don't how to decode FullName. The
entity spec resulting from this ILV doesn't match a MIR instance.

T.R	Title	User	Personal Name	Date	Lines
366.1	Tried explicit namespace name	COOKIE::KITTELL	Richard - Architected Info Mgmt	`Sun Sep 30 1990 01:26`	7
	It looks like the difference in ILV length is about worth LAMAMA_NS:, so I tried putting that and various other combinations on the Archive field: ARCHIVE LAMAMA_NS:.R.AIM The ILV didn't change a bit.
366.2	INFO - no recording from map yet	GOSTE::CALLANDER		`Mon Oct 01 1990 11:59`	6
	Sorry I can't she light on what is happening here, I will leave that to some one else, but I do remember something in the release notes saying that recording is not yet supported from the iconic interface.
366.3	I grok DNS	MARVIN::COBB	Graham R. Cobb (Wide Area Comms.), REO2-G/H9, 830-3917	`Tue Oct 02 1990 08:47`	29
	Quick tutorial on FullName internal forms. The first 14 bytes are the namespace CTS (a binary identifier for the namespace), the next 2 bytes are the length of the remainder, the rest is the simplenames which comprise the name plus a null simplename (two bytes of 0) at the end. A simplename consists of a type byte followed by a count byte followed by the data. aa 00 04 00 fb 20 80 13 83 84 98 ce 93 00 07 00 01 03 41 49 4d 00 00 <-- NSCTS, presumably for LAMAMA_NS: --> <---> Length of remainder of name <------------> 3 byte simplename AIM <----> Terminating Null Simplename I.e. LAMAMA_NS:.AIM The other name just has the NSCTS and no further data (16-bit count is zero). So it is the null name in the LAMAMA_NS namespace. I.e. LAMAMA_NS:. in external form. I have no idea why you are getting that, mind you! Graham
366.4	recording?	COOKIE::KITTELL	Richard - Architected Info Mgmt	`Mon Oct 15 1990 14:19`	13
	re: .2 > Sorry I can't she light on what is happening here, I will leave > that to some one else, but I do remember something in the release > notes saying that recording is not yet supported from the iconic > interface. Hi Jill. I'm still a little fuzzy from vacation, but I don't follow your reply: the base note doesn't mention recording. The problem is that a FullName argument is translated into different ILV dependinh on which PM it comes from.
366.5	sorry about that	GOSTE::CALLANDER		`Mon Oct 15 1990 16:26`	9
	Sorry about that, I was in a bit of a rush. I pointed to that becase the similarity in the command. I don't know why recording doesn't work, but thought there might be some correlation. In looking at the data in the fullname the "AIM" is what seems to be missing. Have you checked what you have in the dictionary for the data types of these?
366.6	more info please	TOOK::HAO		`Tue Oct 16 1990 15:44`	9
	Richard, Could you supply more info about your command? I'm not familiar with them and I can't tell which is the Entity and which is the Argument itself (i.e. what's the pres name and datatype of your argument). Thanks, Christine
366.7	INFO:	COOKIE::KITTELL	Richard - Architected Info Mgmt	`Wed Oct 17 1990 14:31`	17
	Christine, it is indeed hard to tell "what is what", and we've already discussed changes to make it clearer. Archive is both a directive and a global entity. The Entity hierarchy for the example in .0 is NodeStorage VMSFile The directive is Archive, with one argument, unfortunately also called Archive. The directive says "archive this entity", and the argument specifies the destination. So, to borrow from FCL form mode Directive: Archive Entity: NodeStorage R.OREO VMSFile TEST Arg/Qual: Archive R.AIM
366.8	Nodestorage == Node	CAPN::SYLOR	Architect = Buzzword Generator	`Fri Oct 19 1990 17:12`	2
	Minor note... NodeStorage is not a legal Global Entity class name. It's abbreviation would conflict with Node.
366.9	Yup, NodeStorage won't see the light of day	COOKIE::KITTELL	Richard - Architected Info Mgmt	`Tue Oct 23 1990 17:45`	7
	re: .8 Mark, Agreed that NodeStorage isn't a valid GE name for use in an environment in which Node lives. Handy for limited demos, though.
366.10	you haven't been forgotten	TOOK::HAO		`Wed Oct 24 1990 17:24`	16
	Hi Richard, Verna and I did some quick investigation into this last week. We've seen some truncation where the simplename portion was less than what the 2-byte remainder length said it was supposed to be, but we couldn't find a case where the length itself was zero when it wasn't supposed to be. By the way, the same datatype conversion routines are called from both the FCL and Iconic Map PM. I will look into this further sometime this week. Sorry for the delay in updating you. Christine