Re: [hylafax-users] HylaFAX and PCL

[Lee Howard <faxguy@xxxxxxxxxxxxxxxx>]

Philip,

Forgive me for top-posting, but I felt like quoting your entire message 
was appropriate... and my comments didn't really fit well in-line... and 
they weren't really a proper response to your post, but rather my 
comments hopefully will "add to" your understanding of PCL support in 
HylaFAX and get you where you need to go.

A few weeks ago I did a fair amount of work in HylaFAX+ to "fully 
support" PCL both server-side and client-side.  Yes, code work did need 
to occur on the server-side to support PCL.  You'll find that work in 
HylaFAX+ 5.2.1.  I don't use PCL, myself, so the work was done mostly to 
permit other users who do use PCL to provide me with whatever other 
feedback was needed to fine-tune and/or complete the task.  So the PCL 
support in 5.2.1, although more-present, may still be rough on a few or 
more edges.

I notice that you mention that PCL -> PS -> TIFF doesn't seem render 
very well in your testing/usage.  Be wary that in HylaFAX+ 5.2.1 the 
bin/pcl2fax script does PCL -> PS -> TIFF... and thus you may want to 
replace the bin/pcl2fax script with your own.

There are a couple of complications of which I want to make you aware...

First (the minor issue), page counting PCL documents is apparently a bit 
complicated.  Thus, right now if you submit a PCL fax using sendfax and 
use a cover page (generated by faxcover) the coverpage will not have the 
correct page count.  See: http://www.fea.unicamp.br/pclcount/ for some 
example of how involved page-counting can be.

Second (the more significant issue), it appears the there is no 
sure-fire way to identify a PCL file.  You can see from your system's 
"magic" file (often /usr/share/magic) that PCL file detection is a bit 
of a best-guess/probability approach... and it's further-muddied by HP's 
PJL wrappers around both PCL and non-PCL (usually Postscript) data and 
the likelihood that users will want to process PJL-PCL files as if they 
were PCL (because GhostPDL seems to usually support that).  So we need 
to take a look at how to type the PCL files you'll be sending, at 
least.  That typing information will need to be added to typerules (for 
the client-side) as well as hfaxd/FileSystem.c++ (for the server-side).  
Once we can reliably type the PCL files that you'll be sending then this 
issue should not impact you any longer.  (I am told that the <ESC>E<ESC> 
is the most-reliable way to detect non-PJL PCL data, and this typing 
detection is *not* present in 5.2.1.)

So, let's get together and work this out.

Thanks,

Lee.


Philip Gwyn wrote:
> Hello all,
>
> A client needs to send faxes based on PCL.  They currently turn the PCL into PS
> with an old version of GhostPCL, which is scalled down and sent with HylaFAX. 
> The result (because of the scalling) leaves a lot to be desired.
>
> After some tears, I managed to get compile TIFF support into GhostPDL 1.51
> (http://www.awale.qc.ca/ghostpdl/) and learned what they mean by "cathedral".  
>
> I've also modified ps2fax into a working pcl2fax
> (http://www.awale.qc.ca/pcl2fax/).  
>
> I notice that hfaxd has vestigal PCL support.  I hacked Fax::Hylafax::Client to
> allow uploading PCL format files (see submitfax).  However, even if I upload
> the file as PCL, it ends up as "data:0::docq/doc67.pcl.86" in the queue file. 
> The PCL in that filename means it sees and respects the "FORM PCL" but not up
> to the saving file part (should start with "pcl:").
>
> Here is the syslog of the submit :
>
> Jan 18 20:49:41 jimmy-file-server HylaFAX[26516]: Read config file
> /etc/hylafax/hfaxd.conf
> Jan 18 20:49:41 jimmy-file-server HylaFAX[26516]: logfacility = daemon (line
> 57)Jan 18 20:49:41 jimmy-file-server HylaFAX[26516]: servertracing = 0xffffffff
> (line 58)
> Jan 18 20:49:41 jimmy-file-server HylaFAX[26516]: HylaFAX INET connection from
> jimmy-dev.localdomain [10.0.0.44]
> Jan 18 20:49:41 jimmy-file-server HylaFAX[26516]: USER fil
> Jan 18 20:49:41 jimmy-file-server HylaFAX[26516]: FAX LOGIN FROM
> jimmy-dev.localdomain [10.0.0.44], fil
> Jan 18 20:49:41 jimmy-file-server HylaFAX[26516]: TYPE I
> Jan 18 20:49:41 jimmy-file-server HylaFAX[26516]: FORM PCL
> Jan 18 20:49:41 jimmy-file-server HylaFAX[26516]: PORT 10.0.0.44;15746
> Jan 18 20:49:41 jimmy-file-server HylaFAX[26516]: STOT
> Jan 18 20:49:41 jimmy-file-server FaxQueuer[25519]: SUBMIT JOB 86
> Jan 18 20:49:41 jimmy-file-server FaxQueuer[25519]: JOB 86 (suspended dest  pri
> 127 tts 0:00 killtime 2:59:00): CREATE
> Jan 18 20:49:41 jimmy-file-server FaxQueuer[25519]: JOB 86 (ready dest
> +15305786560 pri 127 tts 0:00 killtime 2:59:00): READY
> Jan 18 20:49:41 jimmy-file-server FaxQueuer[25519]: DEST (0x962ca50: 1 ready, 0
> batches, 0 sleeping:) SORT INSERT DONE
> Jan 18 20:49:41 jimmy-file-server HylaFAX[26516]: QUIT
>
> On to my question:
>
> - Why was PCL support abandoned?  I seem to be seeing all the necessary hooks,
>   in the source code, but its hard-coded off.  Ideally, a run-time check for
>   the pcl6 binary would turn the ->supported flag on and off.
> - How do I turn it on?  Beyond HAVE_PCL5, that is.
>
> Looking in the list archives, PCL faxing seems to come up every once in a
> while.  But never fully resolved.
>
> Thank you, keep up the good work.
>
> -Philip
>
>
> ____________________ HylaFAX(tm) Users Mailing List _______________________
>   To subscribe/unsubscribe, click http://lists.hylafax.org/cgi-bin/lsg2.cgi
>  On UNIX: mail -s unsubscribe hylafax-users-request@xxxxxxxxxxx < /dev/null
>   *To learn about commercial HylaFAX(tm) support, mail sales@xxxxxxxxx*
>
>
>   


____________________ HylaFAX(tm) Users Mailing List _______________________
 To subscribe/unsubscribe, click http://lists.hylafax.org/cgi-bin/lsg2.cgi
On UNIX: mail -s unsubscribe hylafax-users-request@xxxxxxxxxxx < /dev/null
 *To learn about commercial HylaFAX(tm) support, mail sales@xxxxxxxxx*