[cups.general] Full speed via socket, but /slow/ via ipp

[Michael Sweet]

On Apr 30, 2010, at 3:02 PM, James Cloos wrote:
> I'm working on the setup for a networked printer.
> 
> The manufacturer suggests ipp://<PRINTER>/ipp/ for non-Mac cups users,
> but that 301-redirects to https://<PRINTER>/ipp/.  Cups seems to ignore
> the 301 and keeps on trying port 631.

I've never seen a printer do this before, but it sounds like a conformance problem to me - the "ipp" scheme and port are REQUIRED by the specs, and we also REQUIRE support for TLS and HTTP Upgrade so that nonsense like redirections to https are unnecessary.  Furthermore, we're going to have to be VERY careful about this since we don't want to redirect to a different address than we started (otherwise you run the risk of sending the data to the wrong place...)

Short term you can do the following as root:

    cd /usr/lib/cups/backend (or /usr/libexec/cups/backend)
    ln -s ipp https

and then use "https://<printer>/ipp/" for the device URI.

What printer are you using?

> Some futher tests showed that http://<PRINTER>/ipp works, but unlike
> when using the socket: protocol, cups seems to send one page at a time,
> pausing until that is finished before sending the next.

Currently we collect all of the print data (saving to a temp file) and then send it all at once to the printer.  This is because a lot of vendors don't support HTTP chunking (which is required for all versions of IPP!)

If you are sending a lot of 1-page jobs I can see this happening, but for a normal multi-page job it should all come out at full speed, just delayed initially while we save the processed print job to disk...

> ...
> And is cups (1.4.2)'s inability to follow a 301 in need of a bug report?
> I don't want to waste everyone's time if the bug has already been fixed.

Go ahead and file a bug report, although I don't think it will be fixed before CUPS 1.5. We certainly should be following HTTP redirects, however the printer MUST support the IPP scheme (without redirects) if they claim conformance to IPP/1.1 or later.

________________________________________________________________________
Michael Sweet, Senior Printing System Engineer, PWG Chair