PWG> PWG Process update [comment on file naming scheme]

PWG> PWG Process update [comment on file naming scheme]

PWG> PWG Process update [comment on file naming scheme]

Hastings, Tom N hastings at
Tue Mar 4 18:46:32 EST 2003

Ira filled me in with the discussions that I missed in PSI on the process
stuff.   I may not be able to join the SM WG call this Thursday on the
process stuff, but here are some comments: 

1. File name fields
So the idea is that file names have three fields, so that software can more
easily parse the file names.  It would be good to add this "why" in the
process document.  Then people will be able to remember it much more easily
and people will be more likely to follow it.

2. Depth of tree structure to one level for each WG
Also it is important not to let the tree structure for a WG be deeper than
one level, else we lose the ability to let FTP sort the sub-directories by
date immediately under the WG directory to see the directories that have
most recently been updated with new files.  

So WG should either have a sub-directory for all Working Drafts: /wd/ 
or should have separate sub-directories for each specification.  In the
latter case, the sub-directories for Working Drafts could be: /wd-xxx/, not

3. Keep IPP subdirectories (/new_XXX/) as is
Also it seems better not to change the IPP sub-directories at this late date
in IPP WG, so we'll keep the /new_XXX/ (for example, keep /new_ACT/, instead
of changing to /wd-act/).

So here is the file name I'll use for the next Working Draft for the IPP
Document Object, which I'm editing now:, .pdf, .htm

and the next IPP Document and Pages Overrides (with a new title of IPP Page
Overrides) Working Draft with a maturity level of 0.01 and a target version
of 1.1):, .pdf, .htm

and the next Working Draft of the "IPP Production Printing Attributes - Set
2" which will be entitled "IPP Job Save and other Production Printing
Attributes", with a maturity level of 0.02 and target version number of 1.0:, .pdf, .htm


4. Maturity level concept needs explaining
Also the Maturity level concept needs some words in the Process Document.
That you jump to 0.80 and 0.90 when getting nearer to Last Call, but if you
have more than 10 versions after 0.90, you go from 0.99 to 1.0.0 and start
counting in the third field.

5. Middle field minus the target version number is a good cross reference
Finally, the "ippdoc", "ippover", and "ippsave" would be good symbolic names
for cross references to Working Documents as in "See [ippdoc] section xxx".



-----Original Message-----
From: McDonald, Ira [mailto:imcdonald at]
Sent: Monday, March 03, 2003 14:04
To: 'Hastings, Tom N'; Harry Lewis
Cc: pwg at
Subject: RE: PWG> PWG Process update [comment on file naming scheme]

Hi Tom

We're speaking at cross purposes on filenames for working drafts
(possibly because you were unable to attend any of the telecons
where we discussed this).

The FORMER practice of including the (IETF-like) version of a
working draft in the PWG spec filename has been abandoned.

The new (W3C-like) practice is that the version in the filename
is INVARIANT and represents the TARGET version of the adopted
spec (like 'psi10' for PSI/1.0).  Only the DATE of the filename
indicates successive working draft "versions".

Further, the 'ipp-xxx-Vyy' naming breaks the (desirable)
quality that a working draft filename has EXACTLY three
components (separated by hyphens) to allow programmatic
access by scripts:

(1) 'wd' - to indicate a PWG Working Draft status
(2) 'tag' - the spec acronym suffixed with the target
    major/minor version for the adopted spec
(3) 'date' - the YYYYMMDD format unique document date

The long names of 'ipp-document-object' (for example) are
unacceptable because they become part of the durable 
permanent name of the adopted PWG Standard.  The W3C
practice is to assign a unique SHORT mnemonic to any
given spec, for example 'psi10' for the PSI/1.0 Protocol
Spec and 'psidev10' for PSI/1.0 Developers Guide, or
'ippdoc10' for IPP Document Object/1.0.  Note that EACH
IPP (or other) extension has it's own INDEPENDENT stable
version (such as '10' for '1.0'), which is NOT related
to the version of IPP being extended.

- Ira McDonald
  High North Inc

-----Original Message-----
From: Hastings, Tom N [mailto:hastings at]
Sent: Monday, March 03, 2003 1:37 PM
To: Harry Lewis
Cc: pwg at
Subject: RE: PWG> PWG Process update [comment on file naming scheme]


Its good to get the filenaming scheme into the process document.

ISSUE 01: There is confusion as to what the "XYZ", "xyz", and "/ps/" in the
file name table means.  Which are something that varies with WG and/or
specification and which are "constants"?  So the Table needs some

For example, some have interpreted the XYZ to be the acronym for the WG,
such as SM or PSI.  If so, then the current proposal has a flaw:  It only
works for a WG that is producing a single specification.

The PWG process document URL current pattern is:

ISSUE 02: I assumed that /ps/ was the example for the PSI WG, rather than
being a constant for all WGs, such as "Proposed Standards", right?

Howver, for IPP where we have produced 32 different documents, we need to be
able to put something into the file name to distinguish each document.  Also
I think that a WG should have more freedom to have separate sub-directories
for separate Working Draft specifications, still under the /wd/
sub-directory for all WDs for that WG.

For example, for the IPP Document Object Working Draft, I'd like its URL to
c, .pdf, .htm

rather than:, .pdf, .htm

And perhaps better than /new_ABC/ sub-directories like we currently do, have
all of the sub-directories that hold Working Drafts under /wd/ with further
sub-directories for each individual specification /wd/abc/ (all lower case
and without the "new_" prefix so we'd have for the Document Object and
Actual Attributes specifications:
, .pdf, .htm,
.pdf, .htm

ISSUE 03: Add "-v" before the version number.  I want the version number to
collate ahead of the date but after the rest of the file name.  This is
because editors often make several edits before publishing an updated
version (incremented by 1).  I also like taking up two more position to
include the "-v" to help make it clear that the magic number is a version
number and not something else.

So for the Semantic Model WG and its Semantic Model specification, we have
something like:


 -----Original Message-----
From: Harry Lewis [mailto:harryl at]
Sent: Friday, February 21, 2003 21:42
To: pwg at
Subject: PWG> PWG Process update

An update has been posted. Thanks to Dennis for adding some corrections,
observations and pointing out some issues (yellow). We've also added a
diagram like we had in the old process. Hope it is helpful. 
Harry Lewis 
IBM Printing Systems 

More information about the Pwg mailing list