[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

316.0. "Clarification of "NAME" lengths needed." by NSSG::R_SPENCE (Nets don't fail me now...) Mon Sep 10 1990 18:53

    In the Ethernet Station USE guide on page 1-10 it says that the station
    name must be LESS than 255 characters. How many less? Is 254 ok?
    
    In the Bridge AM USE there isn't any mention of a length for the bridge
    name.
    
    In the DECnet Phase IV AM USE guide on page 3-26 is says that
    "registered name" is UP TO 255 characters. Up to but not including?
    
    For these VERY LONG NAMES, how can DECmcc support them? If a user
    Chooses a name that comes close to the "supported" limit, how can
    the user type a command that uses the name given the VMS limitation of
    a 255 character command buffer?
    
    I can see a case where some very long names could legitimately get in
    to the system using the IMPM and as a result be inaccesable by the FCL.
    
    It would seem that it would be reasonable to give guidelines on
    choosing names and perhaps for the first version or so impose a lower
    limit on the name length in order to reduce support calls?
    
    For the first version there are not going to be very many experts in
    the field that have the DNS experience to choose reasonable directory
    paths either and since the name is actually the full DNS name it would
    be good (in my opinion) to encourage short (20-30 char) names.
    
    I think that conservative numbers would be best at first.
    
    s/rob
T.RTitleUserPersonal
Name		DateLines
316.1is 1024 enough?TOOK::DITMARSPeteMon Sep 10 1990 20:2837
I think your suggestions on reasonable name lengths are important, and I hope
the folks with responsibility for the relevant parts of the documentation will
try to implement your suggestions.  I have no idea which document should include
this kind of information.  Steve Jong must, right Steve?

>    For these VERY LONG NAMES, how can DECmcc support them? If a user
>    Chooses a name that comes close to the "supported" limit, how can
>    the user type a command that uses the name given the VMS limitation of
>    a 255 character command buffer?

Actually, you can enter 512 characters at the FCL command line prompt before it
encounters a "buffer full" condition and forces a CR.  With judicious use of
line continuation characters (the dash "-" character), you can enter up to (and
including) 1024 characters in a single FCL command.  The 1024 character limit on
the command line length is enforced after symbol substitution, so that's not a 
way around the limit.  A line continuation character can be placed anywhere 
(i.e. in the middle of an entity specification, quoted string, etc.) as long as
it is the last character on a line of input.  Line continuation characters
are removed from the command line and any subsequent input is appended directly
to the previously entered text.

Do you think 1024 characters are enough?

In LINE FORMAT of FORMS MODE, the screen I/O package we're using gives us a 
"buffer full" error when we reach the screen boundary, but instead of 
interpreting this as an implicit CR, we pretend the user entered a line 
continuation character.  Kind of like an editor that supports automatic line
wrapping.  This seems much friendlier to me.

I will QAR FCL so that LINE MODE behaves like LINE FORMAT of FORMS MODE when
a "buffer full" error is encountered.  This might make it into V1.1 EFT.  I
don't see it as a major piece of functionality, but it's not a lot of work 
either.

regards,

Pete
316.2Yes, I must 8^)WORDY::JONGSteveMon Sep 10 1990 21:404
    I will take care of it.
    
    Rob, do you prefer statements such as "up to X characters," "less than
    Y characters," " 1-Z characters," or what?
316.3Be real clear on definitions.NSSG::R_SPENCENets don't fail me now...Tue Sep 11 1990 14:1613
    I think a name spec should state the range of both length and content.
    
    Ie; "Name length can be from 1-123 inclusive. It can contain
    alphanumeric characters including underscore though the name must begin
    with an alpha character. Case of the characters may be upper or lower
    or mixed."
    
    My rule of thumb is that if in doubt as to if someone could
    misinterpret the definition, they will, and DEC will get a support
    call.
    
    s/rob