Thanks - i agree with all your points.
Shoehorning presets into Resources would always be a bit strange.
Smith had (verbally) hoped to perhaps use Resources for finer-grained
access control (i.e., "this is a preset Resource for my user group"), but
even so, there's no getting around that End Users shouldn't be doing
the Printer configuration (by whatever operations), so pushing Client
side presets to Printers from End Users just doesn't work.
I agree with your explicit precedence of attributes for Create-Printer.
Pete's original thought about Job Template resources was that they
formed a neat alternative to a lot of pre-configured Printers (color,
double-sided, etc.). I'll give this a bit more thought.
Ira McDonald (Musician / Software Architect)
Co-Chair - TCG Trusted Mobility Solutions WG
Chair - Linux Foundation Open Printing WG
Secretary - IEEE-ISTO Printer Working Group
Co-Chair - IEEE-ISTO PWG Internet Printing Protocol WG
IETF Designated Expert - IPP & Printer MIB
Blue Roof Music / High North Inc
mailto: blueroofmusic at gmail.com
Jan-April: 579 Park Place Saline, MI 48176 734-944-0094
May-Dec: PO Box 221 Grand Marais, MI 49839 906-494-2434
On Sat, Aug 26, 2017 at 2:45 PM, Michael Sweet <msweet at apple.com> wrote:
>> > On Aug 25, 2017, at 6:32 PM, Ira McDonald <blueroofmusic at gmail.com>
> > Hi Mike,
> > For Create-Printer it's really nice to have a template-printer Resource
> > for each supported service type.
>> Agreed, and for a sometimes-used admin operation with likely vendor
> extension attributes, I think it is appropriate.
>> I also think that we should define the semantics of template resources in
> general, i.e., the order of precedence used in creation operations:
>> 1. Attributes in request (highest)
> 2. Attributes in resource
> 3. Default attributes (lowest)
>> That makes it clear how a template resource is applied during a create
>> > I agree about the template-job (too heavyweight to configure). But
> > we already plan that the HTTP PUT of a font to a well-known URL
> > will quietly create a corresponding Resource object for that font.
> > Why not do that simple HTTP PUT for job templates?
>> Well, while we do have HTTP PUT for creating and allocating Printer
> resources (as defined in IPP INFRA), that interface is also quite limited
> compared to using the Create-Resource operation since you cannot specify a
> "resource-name" different from the filename in the PUT. So either the
> Printer has to pull the "resource-name" value from the file's metadata,
> e.g., the typeface name, and/or you use the "base name" from the URL, e.g.,
> "http://printer.example.com/ipp/print/resources/Foo%20Bold.otf" has a
> "resource-name" value of 'Foo Bold'.
>> For IPP messages we'd be limited to using the base name approach, which
> means UTF-8 with percent escapes in the URL (ugh).
>> > Pete's retired from PWG work or he'd be speaking up for Job Templates
> > (they're part (in XML at present) of PWG Job Ticket spec). I think the
> > MIME type of a Job Template Resource should allow application/ipp
> > or text/xml.
>> *If* an implementation chose to support a PWG PJT (with whichever schema)
> we need to use a specific MIME media type for it, not just "text/xml".
> Also, I believe IETF currently wants us to use application-scoped media
> types for this kind of thing, so that would probably mean registering
>> There is also the issue of schema versions - do we make that part of the
> media type (a parameter) that can be included in
> "resource-formats-supported" values?
>> > Is even the simple HTTP PUT support too much for IPP native job
> > templates (i.e., wire-encoded IPP attributes, not PAPI or XML)?
>> The issue is how hard will it be for the Client to support them? Doing N
> requests (Get-User/Printer-Resources + a bunch of HTTP GET requests) to get
> presets (which is all Job Template resources are) is inefficient and
> complicated compared to looking at a single collection attribute that is
> returned by the (common) Get-Printer-Attributes request that Clients
> already do.
>> I can also say we (Apple) are more likely to support the "job-presets" and
> "job-triggers" attributes than we are to support Job Template resources -
> philosophically they are better aligned with how presets and our internal
> trigger implementations work on macOS, and adding support for them will be
> easier (since we already have the interface for querying Printer
> attributes) than if we have to add additional code to query and pull
> resources. And because there are fewer requests, the time to show the
> Print UI will remain low.
>> Many Printers also already support Set-Printer-Attributes, so I think
> adding support for a couple attributes will be easier than adding all of
> the resource infrastructure.
> Michael Sweet, Senior Printing System Engineer
>>-------------- next part --------------
An HTML attachment was scrubbed...