#11 Improves canon bjc-3000 page margin settings.

[David Keegan]

The gimp-print-4.2.5 driver fails to print within 15mm
of the bottom of an a4 page, yet according to the Canon
manual the printable area extends to 5mm from the
bottom of the page.

The Windows 95 Canon driver emits an ESC(p command to
set the page margins so my patch enables that
print-canon.c. It also tweaks the width and length
calculation to allow for rounding and ensure the
command parameters are closer to the values emitted by
the Windows driver and changes the printable area
margins to the values specified in the bjc-3000 manual.

[David Keegan]

Patch file for print-canon.c (gimp-print-4.2.5)

[Robert Krawitz]

Logged In: YES
user_id=5436

Peter, can you take a look at this (at least for 4.3, and
your judgment about 4.2)?

[Peter Missel]

Logged In: YES
user_id=345146

I'll update my stuff and have a look at it mid-month, when
I will (finally!) be on a mentionable vacation.

[Peter Missel]

Logged In: YES
user_id=345146

Review done ... I have no objections against the
modification in the 3000's model capabilities data.

But exactly what is the altered math in
canon_init_setPageMargins2() about? Rounding? Data format
mismatch? Please explain. We need to be careful here, so
as not to break things on all the other printers.

[David Keegan]

Logged In: YES
user_id=851644

I arrived at the altered math in canon_init_setPageMargins2
after comparing the output of the gimp-print driver with
that of the WIN95 driver. The escape sequence parameters
emitted by gimp-print were for a slightly smaller printable
area than those emitted by W95, with the result that
gimp-print was failing to print one or two pixel rows at the
bottom of the page.

I guessed that this might be due to rounding error given
that the width and length parameters are integers and are
involved in some arithmetic in the same function.

I understand the need to be careful here. My patch works for
me on my printer, but I don't understand the driver code
sufficiently to decide for others.

[David Keegan]

Logged In: YES
user_id=851644

I arrived at the altered math in canon_init_setPageMargins2
after comparing the output of the gimp-print driver with
that of the WIN95 driver. The escape sequence parameters
emitted by gimp-print were for a slightly smaller printable
area than those emitted by W95, with the result that
gimp-print was failing to print one or two pixel rows at the
bottom of the page.

I guessed that this might be due to rounding error given
that the width and length parameters are integers and are
involved in some arithmetic in the same function.

I understand the need to be careful here. My patch works for
me on my printer, but I don't understand the driver code
sufficiently to decide for others.

[alan johnson]

Logged In: YES
user_id=943591

maybe why problem prevails

[Robert Krawitz]

Logged In: YES
user_id=5436

Should have reassigned this a while ago.

[Maximilian Engelhardt]

Logged In: YES
user_id=667171
Originator: NO

The patch does work with gutenprint 5.0.0 (Debian etch) and my BJC-3000.
Even if the printer is quite old it would be nice to have this patch included in gutenprint.

[Sascha Sommer]

Logged In: YES
user_id=1185323
Originator: NO

Thanks.
I commited the patch to cvs.

[Sascha Sommer]

Logged In: YES
user_id=1185323
Originator: NO

This should be fixed in 5.1.4.

[Maximilian Engelhardt]

Logged In: YES
user_id=667171
Originator: NO

Thanks, I just did install 5.1.4 and it does work fine with my printer.