PMP Mail Archive: Re: PMP> Revised Test Plan

PMP Mail Archive: Re: PMP> Revised Test Plan

Re: PMP> Revised Test Plan

Gail Songer (
Fri, 17 Jan 1997 10:26:35 -0800

On Jan 16, 11:56pm, Chris Wellens wrote:
> Subject: PMP> Revised Test Plan

>For all the items listed in #6 below, InterWorking
>Labs has tests. We are modifying the output of the
>tests so that only the data containing the name of the
>object and its current value is included in an output
>file. That way we get only the test results reflecting
>a list of all objects and their values for each printer.
Exactly what do the tests do? How does it differ from the mib walk?

> 4. Do a MIB walk on each product and count the returned objects.
I assume that you are saving the responces?

> 5. Generate controlled (in a pre-defined script) printer specific
> transactions, including jobs in various formats and various sizes
> (known page counts), postscript files, raw ascii, and specific applications
> (e.g. Word, Navigator, Quicken).
What is the purpose of printing lots of jobs? The only thing I can think of is
to make sure that the page count increments appropriately, and that can be done
with one or two jobs.

> 6. Start the test software application for the following areas. After
> completing the test for each line item, check to make sure the values of
> the individual objects are contained in the specified output file.

> Trap Test: coverOpen
> Trap Test: noPaper
> Trap Test: noToner
> Trap Test: paperjam
Is the trap table going to be read? What about the sub-unit status and the 3
magic host resourses objects?

> 6. Compare the values of the objects in each area returned from each
> printer. Make a list of all variations. Decide on how the spec should
> be modified.
Deciding on the spec should be modified will be done at some later date, right?

>-- End of excerpt from Chris Wellens

It is my opinion that you will find lots of differences in very key areas.
Most developers can figure out the size of media path and what is on the front
panel. The problem is that the sub unit status variables and the 3 magic host
resources variables are ill-defined and as such there has been lots of

For example, the definition of off-line varies between manufacturers. What
should prtInputStatus say when there is load paper for the tray? What should
it say when there is no paper in the tray but there is not a load paper
condition for this tray? What should hrPrinterDetectedErrorState say(this
object defines noPaper as "down" for hrDeviceStatus, but the printer is
definately not in that state)?

There needs to be some focus on these types of conditions since this is where
the spec is the weakest.



Gail Songer Electronics For Imaging 2855 Campus Drive (415) 268-7235 San Mateo, CA 94403