Paul Smedley | 1 Oct 2010 12:10
Picon

Re: Re: PostScript driver and fonts

Hi Alex,

------- Original message -------
> From: Alex Taylor <mail.me@...>
> To: ecups-dev@...
> Sent: 30.9.'10,  18:45
>
> On Thu, 30 Sep 2010 00:02:14 UTC, Pete Brown
> <losepete@...> wrote:
>
>> > I think it would be immensely valuable -- and quite important, even --
>> > to get Type 42 font support working in the ECUPS driver.
>> >
>> > Is there somebody who's able (and willing) to take a look at this?
>>
>> I think you may need to post the above to a wider audience in the hope
>> that someone can look into this - I would volunteer but my technical
>> skills and knowledge are less than yours  :-)
>
> Yeah, I mainly posted it here so the other group regulars had a chance
> to comment first (for instance, in case they know something I don't).
>
> My next step will probably be to post it to the eCS dev group.

Interesting stuff, and would be great to have this in ecups.drv

not sure if I have the expertise to implement this - in any case, the 
earliest I'd likely be able to look into it would be over Christmas - ie 
starting December 17.

(Continue reading)

Stefan | 9 Oct 2010 08:35
Picon

What is foomatic for?

I’m using cups to print with Printer Canon IP6000 within a network To be 
able to print out of VPC (thanks for it).
Now I read that there is Foomatic. How do I have to install it? Just 
copy all file into the cups folder and that’s all? Or/and do I have to 
create a new Printer object? Will there be any improvement versus my 
currant installation?
The Readme isn’t very helpful.

Regards
Stefan
Paul Smedley | 9 Oct 2010 08:58
Picon

Re: What is foomatic for?

Hi Stefan

------- Original message -------
> From: Stefan <Stefan.Radke@...>
> To: ecups-dev@...
> Sent: 9.10.'10,  17:05
>
> I'm using cups to print with Printer Canon IP6000 within a network To be 
> able to print out of VPC (thanks for it).
> Now I read that there is Foomatic. How do I have to install it? Just copy 
> all file into the cups folder and that's all? Or/and do I have to create a 
> new Printer object? Will there be any improvement versus my currant 
> installation?
> The Readme isn't very helpful.

If your printer is supported by gutenprint, hplip or splix, foomatic is of 
no value.

There are however, man linux printer ppd's around that rely on foomatic, 
and for these, foomatic is required to get them working with eCups.

Eg foo2qpdl which is a backend for Samsung printers and needs foomatic.

Hope this helps,

Paul 
Alex Taylor | 17 Oct 2010 01:52

Re: Re: PostScript driver and fonts

On Fri, 1 Oct 2010 10:10:23 UTC, "Paul Smedley" 
<paul@...> wrote:

> >> > I think it would be immensely valuable -- and quite important, 
> >> > even -- to get Type 42 font support working in the ECUPS driver.
> 
> Interesting stuff, and would be great to have this in ecups.drv
> 
> not sure if I have the expertise to implement this - in any case, the 
> earliest I'd likely be able to look into it would be over Christmas - ie 
> starting December 17.

Well, with some advice from Gregg Young I was able to get the FNTLIB
support compiled in.  Here's what I posted on the EDG mailing list:

======================================================================
It turns out that the FNTLIB stuff, which does indeed seem to include 
Type 42 (TrueType) font support, is actually ONLY used when manually 
downloading fonts to the printer using the "Font Manager" page of the 
printer properties notebook.  Unfortunately, this support has not been 
implemented for the case of embedding fonts in PostScript job files, which 
is what we really need.

As for that manual "downloading" of fonts in the properties dialog... it 
may work with TrueType fonts now, but I have no way to know.  I have no 
access to a real PostScript printer to test with.  Obviously, font 
downloading won't work with a CUPS printer because the printer isn't 
actually PostScript... and so won't know what to do with the fonts it 
receives (it'll generally assume they're print jobs and try to print 
them).  Same goes for printing to file.
(Continue reading)

Alex Taylor | 18 Oct 2010 16:28

Re: PostScript driver and fonts

On Wed, 29 Sep 2010 13:52:29 UTC, "Alex Taylor" 
<mail.me@...> wrote:

> There's a type of font defined by PostScript known as a "Type 42" font 
> (as opposed to a Type 1), which is basically a PostScript wrapper around 
> a TrueType font.  PostScript printers which support Type 42 fonts (as 
> reported in their PPD file) actually include embedded TrueType 
> interpreters.  Whatever generates the PostScript file in this case is 
> supposed to "convert" the requested TrueType fonts into Type 42 by 
> wrapping a PostScript header around them, which then allows them to be 
> embedded into the PostScript file.
> 
> The problem is, PSCRIPT.DRV and ECUPS.DRV... don't do this.  Any text
> which is printed in a TrueType font gets converted to bitmap instead,
> with the problems described above.  Unfortunately, these days MOST
> fonts people use are TrueType fonts.
> ...
> I think it would be immensely valuable -- and quite important, even -- 
> to get Type 42 font support working in the ECUPS driver.  

I spent a few hours looking through the PostScript driver source, and
I _think_ I may see how this could be implemented.

I'll try and work on it this week or next if I have the time.  But in
case I don't, or if someone else wants to jump on board, I'll post my
analysis here.

The device context used for PostScript generation has associated with it 
an array of font structures (typedef FNT) which basically has the font 
name, the internal PS name (usually the same except without spaces), and 
(Continue reading)

Alex Taylor | 19 Oct 2010 09:59

Re: PostScript driver and fonts

On Mon, 18 Oct 2010 14:28:04 UTC, "Alex Taylor" 
<mail.me@...> wrote:

> I spent a few hours looking through the PostScript driver source, and
> I _think_ I may see how this could be implemented.
>
> ...snip...
>
> There are two particular concerns I still need to iron out.  First, is 
> the font data pointed to by the FNT structure actually parsed (with the 
> assumption that it's a OFM/PFB file pair) anywhere else in the code 
> outside the functions I've listed?  

On further inspection, I think the answer is "no".  So that's good.

> Second, this needs to be conditional on the current printer device 
> reporting (via its imported PPD) that it actually supports "Type 42" 
> fonts.  

A bit of bad news here: it turns out PIN doesn't actually import that
information into the driver.

So to do this properly, I'd need to not only modify PIN, but also 
change (slightly) the actual format of the *.PAK file entries to
include an extra field with this info.

I think I know how to make the change (it's a pretty minor one), but 
I'm not sure if I should just dive in and do that.  For one thing, it
would irrevocably obsolete all previously-existing versions of PIN.

(Continue reading)

Paul Smedley | 19 Oct 2010 10:49
Picon

Re: Re: PostScript driver and fonts

Hi Alex,

On 19/10/10 18:29, Alex Taylor wrote:
> On Mon, 18 Oct 2010 14:28:04 UTC, "Alex Taylor"
> <mail.me@...>  wrote:
> I think I know how to make the change (it's a pretty minor one), but
> I'm not sure if I should just dive in and do that.  For one thing, it
> would irrevocably obsolete all previously-existing versions of PIN.
>
> What do people think?  Is it worth doing?
Yes I think so, I wonder if there's some way to make the updated pin NOT 
work with pscript.drv, but only with ecups.drv somehow - just so 
somebody doesn't try use it with an odler IBM pscript.drv

> The alternative is to just assume that ALL printers that report
> PostScript level 3 support include TrueType rasterizers.  I'm really
> not sure how safe an assumption that is.
I wonder what the Postscript v3 spec says - the Ghostscript guys may be 
able to point to the docs or comment on whether TTF is part of the spec 
for a Postscript v3 printer...

Cheers,

Paul
Pete Brown | 19 Oct 2010 16:06

Re: PostScript driver and fonts

Hi Paul

Paul Smedley wrote:
> Hi Alex,
>
> On 19/10/10 18:29, Alex Taylor wrote:
>> On Mon, 18 Oct 2010 14:28:04 UTC, "Alex Taylor"
>> <mail.me@...> wrote:
>> I think I know how to make the change (it's a pretty minor one), but
>> I'm not sure if I should just dive in and do that. For one thing, it
>> would irrevocably obsolete all previously-existing versions of PIN.
>>
>> What do people think? Is it worth doing?
> Yes I think so, I wonder if there's some way to make the updated pin NOT
> work with pscript.drv, but only with ecups.drv somehow - just so
> somebody doesn't try use it with an odler IBM pscript.drv
>

As pin.exe does not have a bldlevel I suspect the way to go about this 
would be to include an identifier into new builds of pin.exe - or, even 
simpler, name the new pin.exe to something else eg ecupspin.exe or 
newpin.exe

That would avoid confusion with the IBM pin.exe

Regards

Pete

>> The alternative is to just assume that ALL printers that report
(Continue reading)

Pete Brown | 19 Oct 2010 21:15

Re: PostScript driver and fonts

Hi

Pete Brown wrote:
> Hi Paul
>
> Paul Smedley wrote:
>> Hi Alex,
>>
>> On 19/10/10 18:29, Alex Taylor wrote:
>>> On Mon, 18 Oct 2010 14:28:04 UTC, "Alex Taylor"
>>> <mail.me@...> wrote:
>>> I think I know how to make the change (it's a pretty minor one), but
>>> I'm not sure if I should just dive in and do that. For one thing, it
>>> would irrevocably obsolete all previously-existing versions of PIN.
>>>
>>> What do people think? Is it worth doing?
>> Yes I think so, I wonder if there's some way to make the updated pin NOT
>> work with pscript.drv, but only with ecups.drv somehow - just so
>> somebody doesn't try use it with an odler IBM pscript.drv
>>
>
>
> As pin.exe does not have a bldlevel I suspect the way to go about this
> would be to include an identifier into new builds of pin.exe - or, even
> simpler, name the new pin.exe to something else eg ecupspin.exe or
> newpin.exe
>
> That would avoid confusion with the IBM pin.exe
>
> Regards
(Continue reading)

Alex Taylor | 22 Oct 2010 16:28

Re: Re: PostScript driver and fonts

On Tue, 19 Oct 2010 08:49:17 UTC, Paul Smedley 
<paul@...> wrote:

> > I think I know how to make the change (it's a pretty minor one), but
> > I'm not sure if I should just dive in and do that.  For one thing, it
> > would irrevocably obsolete all previously-existing versions of PIN.
> >
> > What do people think?  Is it worth doing?
> Yes I think so, I wonder if there's some way to make the updated pin NOT 
> work with pscript.drv, but only with ecups.drv somehow - just so 
> somebody doesn't try use it with an odler IBM pscript.drv

Well, before I do this I'll make sure I can actually get the Type42
support working in general.

Still poking away at it... I've already made the changes I described
in my earlier post, but it doesn't seem to be enough.  Obviously I'm
going to have to comb through the code some more.

There's debug logging facility built into the source, but I haven't
managed to figure out how to get it working just yet...

> > The alternative is to just assume that ALL printers that report
> > PostScript level 3 support include TrueType rasterizers.  I'm really
> > not sure how safe an assumption that is.
> I wonder what the Postscript v3 spec says - the Ghostscript guys may be 
> able to point to the docs or comment on whether TTF is part of the spec 
> for a Postscript v3 printer...

The Type 42 font spec from Adobe says:
(Continue reading)


Gmane