From: "Doug Bissett" <virtualized_ecs_users@2rosenthals.com>
Message-ID: <000.a04a0b007fbef84b.015@telus.net>
To: "Virtualized eCS Users Mailing List" <virtualized_ecs_users@2rosenthals.com>
Date: Sat, 22 May 2010 23:34:55 -0600 (MDT)
In-Reply-To: <list-2432243@2rosenthals.com>
References: <list-2432243@2rosenthals.com>
Priority: Normal
User-Agent: PMMail/3.07 (os/2; U; Warp 4.5; en-CA; i386; ver 3.07.32.1531)
MIME-Version: 1.0
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: 7bit
Subject: Re: [Virtualized eCS] Printer support

On 2010-05-23, at 11:54:47, Ed Durrant wrote:
>Doug Bissett wrote:
>> On 2010-05-23, at 08:10:38, Ed Durrant wrote:
>>   
>>> I've been trying to get a sort of Wintendo print server in virtualbox 
>>> working for sometime - it would of course only work over network 
>>> connectivity - I take it you are "riding" on the eCS Wifi support to get 
>>> to the wifi enabled printer. 
>>>     
>>
>> No, I am using the  Virtual Switch to access the router, which uses the
>> WiFi to access the printer. I suppose it would also work if I used the
>> eCS WiFi to access the router (not sure how, or if, the printer could
>> be made to work using the eCS WiFi, without the router). 
>
>So you are saying you have been able to get the Windoze WiFi drivers to 
>see the WiFi hardware in the host system via the virtual switch?

Actually it should be possible to use the WiFi link, BUT windows guest
only sees the virtual wired link supplied by the Virtual Switch
software. It knows nothing about the WiFi that eCS can use. WiFi is
nothing more than a cable replacement between a NIC and something else
(usually a wireless router). Think of it as a completely separate
machine, running windows, and connected to the network by a cable, not
WiFi. If that network has a router, that has wireless capability, then
other wireless devices can be seen as part of the network that windows
is connected to. The printer is part of that network, and is connected
to the router by a WiFi link. Windows itself doesn't need to know
anything about the WiFi link, all it needs to do is communicate with
the router, and the router communicates with the printer.

> If 
>that's possible it opens up other options to support currently 
>non-supported by eCS WiFi cards - i.e. VLAN into the virtualised windows 
>session and route out through the WiFi link to the outside world.

No, that doesn't work. It *may* be possible to do that, if the virtual
machine software could be made to expose the actual hardware to the
guest software, but it currently doesn't do that. All that the guest
sees is the virtual cabled NIC. At the other end of that "cable", the
"physical" link to the router can be a WiFi link, but the guest knows
nothing about that.

>> As far as the
>> guest XP system is concerned, it just sees the printer on the network,
>> and has no idea that it is running in a virtual machine. The host
>> machine knows nothing about it, except the part that causes the data to
>> be transferred  to/from the router.
>>
>>   
>>> I have never found a way to have the data 
>>> "captured" by a Win32 program and auto sent to the printer without 
>>> hitting the printer driver problem however - I need something to convert 
>>> incoming standard (i.e. PS or PCL) data into native printer language 
>>> within the Wintendo Virtual session. Haven't found anything that works 
>>> so far. If I could do this I could load a stripped down (and locked 
>>> down) Windoze image in Virtualbox (or VPC/2) and have it running to take 
>>> care of printing to any Windoze compatible networked printer.
>>>     
>>
>> I think the WiFi setup would work with any windows machine sharing the
>> printer, but for eCS (or Linux etc.) to use the printer, they would
>> need a compatible printer driver so they could use the shared printer.
>> The trick seems to be that the windows in the virtual machine sees the
>> printer as network connected, and it is not dependent on anything else.
>> At least the guest windows can use the printer as if it was a real
>> windows using a network connected printer (and scanner).
>>
>>   
>
>> Your problem would seem to be that you want to capture the data, then
>> convert it to something that the windows driver can send to the
>> machine. The only thing, that I can think of, would be to print to a
>> PDF file, in eCS (or Linux etc.), and send that to windows. Then, you
>> would need to, somehow, get the windows PDF support to print it
>> (possibly as simple as opening the PDF file in Acrobat Reader, then
>> printing it). I am not sure, but I think that GhostScript has a windows
>> version that might do that, with a little help from a script of some
>> sort.
>>
>>   
>That's exactly what I have been trying to achieve. There are some 
>Windows Daemons that claim to be able to do this but I've had no success 
>with them so far.

Not very surprising. You may have better luck trying to get that part
to work with a real windows machine first. Not that it will work any
better, but it would remove a lot of uncertainty about what is
happening.

>>> By the way I guess you know the latest versions of CUPS/2 work via USB 
>>> as well as parallel and network?
>>>     
>>
>> I never tried parallel, but I have got USB to work, up to a point. My
>> brother has the Canon Pixima MP620, and a Canon i960. I also have an
>> i960, but never got it to work  with CUPS (haven't tried for over a
>> year). I did get my brother's i960 to work, but there is something odd
>> about the colors. 
>Yes I've seen colour change using CUPS as well - strangely when using 
>Adobe reader vs Lucide, Lucide tends to give better colours through the 
>same drivers as Abobe but neither are 100% correct on the colours.

What I see with the i960, using the CUPS test print, is that all of the
black parts seem to be correct. Part of that is a circle, that has
segments that are supposed to be filled with various colors. The colors
appear to be correct, but the circle that they make is about twice the
width of the black circle that they should fit into.

>> I can get the MP620 to work with CUPS too, but it
>> puts a line of blank dots, between each line of dots, causing only the
>> top half of the page to be printed on a whole page. CUPS is still not
>> usable with either machine, but it is close. When (if) CUPS ever
>> supports these machines properly, it should be possible to use a shared
>> printer to do the printing, but I think it should also be possible to
>> use CUPS to print directly to the network attached printer (not sure
>> how though, I didn't have the time to try to figure that out). The
>> really interesting part is that the scanner also works with the windows
>> guest.
>>   
>
>I currently have two printers connected via the network and printable to 
>from any machine on my network (eCS and Windoze - haven't got any Linsux 
>systems currently) - one is the HP OfficeJet 6210 all-in-one MFC with a 
>built in (cabled) network connection and the other is a Canon PiXMA 
>IP4500 connected from it's USB cable to a Linksys  WPS54G Wireless/wired 
>print server. The HP is also USB connected to my main eCS system and can 
>print using CUPS either via USB or over the LAN.

The 6210 (cabled) should be very much like having a WiFi connection.
WiFi simply replaces the cable with a radio.

It seems that you are doing better than what I was able to achieve with
my brother's printers. I haven't had time to set up CUPS for my own
printer yet. I have only been home for a couple of days, and I had to
put a new power supply into my main machine to get it to come back to
life (don't know how a power supply can go bad when it is turned off,
but...). That, and catching up, have taken all of my spare time, so
far.
-- 
****************************
From the eComStation of Doug Bissett
dougb007 at telus.net
****************************
... If at first you don't succeed, destroy all evidence that you tried.