Greetings all,
Sorry I was unable to be present for the discussion! (I'm on "vacation" helping my elderly father pack his apartment.)
I think we have a bit of a "chicken and egg" issue here. It may be that many clients have backed away from using "printer-name" because of the variability of the values printers provide. We can either deprecate that attribute / discourage its use, or we can try to push it towards some consistency. These others are much more specifically prescriptive but not always pertinent:
- printer-make-and-model
- printer-dns-sd-name
- printer-device-id
The scenarios a client might use these and "printer-name" include constructing the "default name" it might give a persistent "old-school" print queue (e.g., Linux, macOS, Windows). I personally have a habit of setting the DNS-SD Instance Name to something like "_Smitty ChickenLips_" and then I also set its .local domain name to a downsampled version of that: "smitty-chickenlips" so that I can identify it across client platforms regardless of how the queue gets set up.
The queue might be set up using a discovery protocol (DNS-SD etc.), or via a URI / hostname / raw IPv4 address / raw IPv6 address. One thing I'm concerned about is that, if the queue / printer is set up using one of the latter non-DNS-SD methods, how would the user know that that is the device's name? For instance, if I set up the queue with "192.168.1.37", and the user read that off a config page or the printer's control panel, they have acquired the IPv4 address but may never have seen "_Smitty ChickenLips_" or "smitty-chickenlips" or "smitty-chickenlips.local" before. So how will they connect the name to the device?
(I'd like to find a pathway to allow us to begin discouraging the use of raw IPv4 addresses, because DHCP leases can cause IPv4 address rotation, causing "printer offline" issues / stale queues, but in many semi-managed SMBs their networks' DNS infrastructure and multicast propagation is just not reliable enough so they fall back to IPv4 addresses with fixed leases, I guess.
So what I might suggest for IPP Implementor's Guide v2.1:
- For printers set up using DNS-SD, the default queue name ought to be "<DNS-SD Instance Name / printer-dns-sd-name>"
- For printers set up by the user providing a hostname, the default queue name ought to be "<hostname>"
- For printers set up by the user providing a raw IP address, the default queue name ought to be "<IP address>" (gross for IPv6, I know...but that's what the user provided...)
- Regardless of how the printer / print queue is set up, it seems like a model name should be presented as well. How this gets presented to the user is a UX issue that is out of scope.
- All printers ought to present their DNS-SD Instance Name and their hostname and/or their current IPv4 address on their control panel if they have one.
Thoughts?
Smith
/**
Smith Kennedy
HP Inc.
*/
On Jun 20, 2024, at 4:52 PM, Michael Sweet via ipp <ipp at pwg.org> wrote:
CAUTION: External Email
Summary of discussion at June 20, 2024 IPP workgroup conference call:
- Chris: Are we going to try to standardize this?
- A: No, more curiousity
- AirPrint/IPP Everywhere only require specific values for printer-make-and-model and printer-dns-sd-name
- Bill: We might suggest a default? And recommend that the value be settable?
- Uli: If there was a recommendation that would more strongly influence firmware design decisions
- Mike: Not used by CUPS clients
- Uli: Windows seems to default to using the printer-name value
- Michael: I'll check
- Ira: IIG2.0 says nothing about printer-name
- Maybe file errata for IIG 2.0?
- Mike: IIG can talk about client recommendations, IPP Everywhere 2.0 can talk about printer recommendations
- Bill: What's the point of requiring printer-name if nobody uses it?
- Mike: It has been required since IPP/1.0, and if we take a longer view maybe we could have something useful in the future?
- Mike: Maybe IPP Everywhere 2.0 can make some recommendations/requirements
- Bill: Suggest defaults
- Mike: Also require it to be settable (EWS, Console, and/or Set-Printer-Attributes), not blank and not the resource path, etc. from the URI
> On Jun 20, 2024, at 10:08 AM, wamwagner--- via ipp <ipp at pwg.org<mailto:ipp at pwg.org>> wrote:
>> Hi Smith et al,
>> For KonicaMinolta MFPs, the default value of "printer-name" is left blank. But the User Guide, under "Configuring basic settings for the IPP Printing" instructs administrator to "Enter the printer name of this machine (using up to 127 characters)"
>> Thanks,
> Bill WagnerFrom: ipp <ipp-bounces at pwg.org> on behalf of Charles Armstrong via ipp <ipp at pwg.org>
> Sent: Thursday, June 13, 2024 12:40 PM
> To: ISTO-PWG Internet Printing Protocol workgroup discussion forum <ipp at pwg.org>
> Cc: Charles Armstrong <carmstrong at cusa.canon.com>
> Subject: Re: [IPP] What value do IPP Printers provide for "printer-name"?
> Hi Smith,
> Here's the response from Canon:
>> • printer-name : Model name. The value is fixed (e.g. XK500 series, iR-ADV C3520)
> • printer-make-and-model : Specific printer manufacturer and model name, and is fixed. (e.g. Canon XK500series, Canon iR-ADV C3520)
> • printer-dns-sd-name : Bonjour Service Name
> Canon doesn't provide the hostname for the "printer-name" attribute.
> Best Regards,
> CharlesFrom: ipp <ipp-bounces at pwg.org> on behalf of Anton Thomasson via ipp <ipp at pwg.org>
> Sent: Wednesday, June 12, 2024 11:39 AM
> To: ISTO-PWG Internet Printing Protocol workgroup discussion forum <ipp at pwg.org>
> Cc: Anton Thomasson <antonthomasson at gmail.com>
> Subject: Re: [IPP] What value do IPP Printers provide for "printer-name"?
> Email was not sent from a Canon Network!
> Hi
>> I too expect something human-useful here.
> Looking through my user-supplied attribute dumps, i see most printers doing something decent; mostly model-based.
>> Some interesting outliers:
> •
> Xerox WorkCentre 6515: "192.168.1.74"
> • Various Brother printers: "" (empty string)
> • Brother VC-500W: "print" (from /ipp/print presumably)
> • EPSON SC-P900: "ipp/print"
> Br,
> Anton
>> Den ons 12 juni 2024 kl 12:38 skrev Arun Holla via ipp <ipp at pwg.org>:
> Hi Smith,
> For the Fiery Digital Front End servers, here are the IPP attribute values being populated:
>> • printer-name à ‘hostname’
> • printer-make-and-model à Specific printer manufacturer and model name, and is fixed.
> • printer-dns-sd-name à Bonjour Service Name concatenated with the last 3 octets of the Mac address (similar to Xerox). Default is the printer manufacturer and model name concatenated with the 3 octets of Mac address. This helps to keep it unique as it is used during the printer discovery. This value changes when the Bonjour service name is changed by the user.
> Thanks,
> -Arun-
> From: ipp <ipp-bounces at pwg.org> On Behalf Of Rizzo, Christopher via ipp
> Sent: Wednesday, June 12, 2024 4:44 AM
> To: Kennedy, Smith (Wireless & IPP Standards) <smith.kennedy at hp.com>
> Cc: Rizzo, Christopher <christopher.rizzo at xerox.com>; PWG IPP WG Reflector <ipp at pwg.org>
> Subject: Re: [IPP] What value do IPP Printers provide for "printer-name"?
> [EXTERNAL] Of course - I just thought of what happens when there is a name conflict with another device on the network, and tested it.
> The conflict caused the registered name via MDNS to have "(2)" appended to it, but neither printer-name nor printer-dns-sd-name were updated to match.
> So now I'm wondering if that is a bug. Is that a bug? I think that is as I read PWG 5100.13 - printer-dns-sd-name at a minumum should have changed to match the MDNS (updated) registered name.
> Chris
> Christopher Rizzo
> Engineer II, Software Engineering
> Design & Development Engineering
><image001.png>
> Xerox Corporation
> Virtual Office Employee
> 26600 SW Parkway Ave
> Wilsonville, OR 97070
><image002.png>
><image003.png>
><image004.png>
><image005.png>
><image006.png>
> From: Christopher Rizzo <christopher.rizzo at xerox.com>
> Date: Tuesday, June 11, 2024 at 4:03 PM
> To: Kennedy Smith <smith.kennedy at hp.com>
> Cc: PWG Workgroup <ipp at pwg.org>, Michael Sweet <msweet at msweet.org>
> Subject: Re: [IPP] What value do IPP Printers provide for "printer-name"?
> Yes, both are updated together.
> And they are independent of the hostname
> Chris
> Christopher Rizzo
> Engineer II, Software Engineering
> Design & Development Engineering
><image007.png>
> Xerox Corporation
> Virtual Office Employee
> 26600 SW Parkway Ave
> Wilsonville, OR 97070
><image008.png>
><image009.png>
><image010.png>
><image011.png>
><image012.png>
> From: Kennedy Smith <smith.kennedy at hp.com>
> Date: Tuesday, June 11, 2024 at 3:49 PM
> To: Christopher Rizzo <Christopher.Rizzo at xerox.com>
> Cc: PWG Workgroup <ipp at pwg.org>, Michael Sweet <msweet at msweet.org>
> Subject: Re: [IPP] What value do IPP Printers provide for "printer-name"?
> CAUTION: This email originated from outside the organization. Do not click links or open attachments unless you recognize the sender and know the content is safe.
> Thanks Chris! HP printers do something similar for the default DNS-SD Instance Name. If Xerox printers provide the same value for "printer-name" as "printer-dns-sd-name", if one changes then so does the other?
> Smith
> On Jun 11, 2024, at 3:21 PM, Rizzo, Christopher <Christopher.Rizzo at xerox.com> wrote:
> CAUTION: External Email
> The Xerox devices that I work on use the Bonjour advertisement name for IPP printer-name. I cannot speak for all Xerox products.
>> The factory default for this is "Xerox <model> (<last 3 bytes of eth mac addr>)".
>> Reason for the mac addr append is for customers that purchase more than one device they can be identified while avoiding the need for initial conflicts in name resolution.
>> Chris
>> Christopher Rizzo
> Engineer II, Software Engineering
> Design & Development Engineering
>> <http://www.xerox.com/>
>> Xerox Corporation
> Virtual Office Employee
> 26600 SW Parkway Ave
> Wilsonville, OR 97070
>> <https://www.linkedin.com/company/xerox/> <https://www.youtube.com/user/XeroxCorp> <https://twitter.com/Xerox> <https://www.instagram.com/xerox/> <https://www.facebook.com/XeroxCorp>
>>>>>> On 6/11/24, 12:46 PM, "ipp on behalf of Michael Sweet via ipp" <ipp-bounces at pwg.org <mailto:ipp-bounces at pwg.org> on behalf ofipp at pwg.org <mailto:ipp at pwg.org>> wrote:
>>> Smith,
>>> > On Jun 11, 2024, at 3:22 PM, Kennedy, Smith (Wireless & IPP Standards) <smith.kennedy at hp.com <mailto:smith.kennedy at hp.com>> wrote:
> > ...
> > So the attributes involved in a discussion were:
> >
> > printer-name (RFC 2911 / 8011 - required)
> > printer-make-and-model (RFC 2911 / 8011 - recommended)
> > printer-dns-sd-name (PWG 5100.13 - required)
> >
> > The wording of "printer-name" makes it sound like it should be a unique name, not a model name...
>>> A printer by any other name would smell as sweet?
>>> From RFC 8011, "printer-name" is *not* a unique name, just a more friendly one:
>>> 5.4.4. printer-name (name(127))
>>> This REQUIRED Printer attribute contains the name of the Printer. It
> is a name that is more End User friendly than a URI. An
> Administrator determines a Printer's name and sets this attribute to
> that name. This name can be the last part of the Printer's URI, or
> it can be unrelated. In non-US-English locales, a name can contain
> characters that are not allowed in a URI.
>>> ________________________
> Michael Sweet
>>> _______________________________________________
> ipp mailing list
>ipp at pwg.org>https://www.pwg.org/mailman/listinfo/ipp> _______________________________________________
> ipp mailing list
>ipp at pwg.org>https://www.pwg.org/mailman/listinfo/ipp
________________________
Michael Sweet
<signature.asc>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.pwg.org/pipermail/ipp/attachments/20240621/15e94983/attachment.html>