[Printing-summit] [lsb-discuss] Printer/driver testing
Jens.Stark at seeg.sharp-eu.com
Mon Aug 21 00:58:37 PDT 2006
>> "All LSB certified printer drivers will work on all LSB certified =
>I just want to emphasize how much I support this. I would simply it, =
>though, to: "All LSB Certified printers work on all LSB Certified =
>Please be assured that if you (with Ian's help) can develop the =
>testing infrastructure to make the certification meaningful, we will =
>do everything possible to get the distros and printing vendors onboard.
Offering a private opinion if I may...
This Will Be Difficult.
While it sounds easy to put a small inkjet in a box and ship it to someone =
testing it, this is how the typical private end user imagines printing. How=
ever, if you look ad the mid- to high volume area, you talk about devices s=
hipped on palettes, weighing up to (and sometimes over) half a ton and need=
ing professional installation.
How many testing centers do you expect ? =
Or should every manufacturer be able to do their own testing ? (Now, this m=
akes sense - a printing validation test suite...)
How much is testing expected to cost - and who assures neutrality and confi=
(As in: Who can the printer vendor sue if anything goes wrong?)
What do others do ?
- Microsoft tests printer drivers. Some of the tested ones (not ours, glad =
to say) blue screen a PC on installation. Apart from being a revenue genera=
tor for MS and a constant source of trouble, their driver certification mak=
es sense to MS - not much to the customer.
- SAP ? If you want the very best printing support for your device under SA=
P, they will sell you their services at a reasonable price. You end up with=
something like a device type that has been tested on actual hardware and y=
ou pay for both the device type and the testing. SAP printing is big enough=
to justify this.
- Linux ? There is CUPS (and ESP Print Pro). Again in my personal opinion, =
the only way worth going.
ESP will write drivers for you (if they must) and test it on actual hardwar=
e (again, for a fee), but preferably license and support their DDK. They wi=
ll even host any PPDs that come their way and maybe even have a look at you=
The logistics of hardware testing make printer certification difficult at l=
What DO we need ?
- A printer certification suite for CUPS.
Okay - "it prints the test page" is more or less that, but is it really suf=
- CUPS working out of the box on any non-niche distro (i.e. the typical gen=
eral purpose ones. Not the kind that runs on a smark card, wrist watch or I=
No major changes, no "we know better than Mike Sweet how this is supposed t=
o work", no "oh, let's invent yet another GUI and make things soo much simp=
ler for the user". Again, the ability to run CUPS without having to look at=
the specifics of certain distributions. We have supporters to train and th=
eir job is already pretty difficult. And no matter if you sell a distro wit=
h a toad, pink pith helmets or documented in hungarian portugese, or one th=
at is built and updated exclusiviely via UUCP file transfers - if it runs C=
UPS, we want to administer CUPS. If distro X breaks that, it might end up o=
n a "not supported" list. It requires extra training just for one distro. S=
ee the point ?
IF there must be yet another GUI (the web interface is fine, thank you very=
much), it should be one and the same for everybody.
- Portable printer drivers.
This is hopefully what the whole topic is about. Printer drivers that will =
be buildable (if not only a PPD) and runnable on every system imaginable. I=
t Must Compile and It Must Not Crash.
The same applies to any funky back-ends, special utilites, GUIs etc. pp. =
(Can we do without ?)
[Getting off my private soap box now]
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Printing-summit