Comment 98 for bug 35638

Revision history for this message
dracon (picky2k) wrote :

George, thanks.

Since telling Windoze where the printer is was a one time deal, that isn't a problem on that side. Wish that could be done for Ubuntu.
Parallel cables don't solve my problem of easily switching the printer off to my laptop either which has no Centronics port.

Now to cups. I don't think this is really a cups problem but I don't yet really understand well the relationship between hal and cups. I'm under the impression that hal runs before cups. It seems that when hal has found and identified the devices' that exist (even if the identification is "unknown usb device") then cups is capable 1) of allowing driver to be assigned to that device/hal port and 2) follow that device if it is subsequently moved to another usb port between reboots.

It's only when hal only finds the printer cable (Centronics parallel to usb cable device) and does not find a printer on the end of the cable (even a "unknown" printer) that cups does not present that printer for configuration in the "add Printer" options. That leads me to conclude that cups is ok but hal is not.

Now I don't know what hal uses to get the hardware info but from what I've read hal is pretty low level stuff looking for signals on specific pins with certain timings...etc. Since hal has found the Cable device I expect that his timing is PDQ and he doesn't wait a long time for the end point (printer) of the cable to post a signal....Just guessing here but if hal moves on to the next port before whatever device is on the end of that cable sends a "i'm here!", he will not even find a "unknown" device and assumes that no device exists. At that point cups can't find a printer to configure since hal tells him there is no device on the cable.

Net: I believe this could be a simple timing fix in hal for a coder to cause hal to wait a millisecond or so for the device at the cable end to post if he finds a Parallel2usb "Printer" cable device on the port (he can tell the difference), (but not me! unless I get sufficently bugged to go back to school -lol- not likely)

 I don't think this is a cups problem...(is problem for me though) If it's hal this problem will be distribution independent across Linuxes and while the older HP laserjets may be "older" there are a bunch out there, and this has more to do with how hal is handling the Parallel Printer Cable device regardless of manufacturer.. (I lose my HP 890c inkjet sometimes as well the same exact way...just not as often probably because it is quicker at responding to it's interface...and I've seen the problem expressed with Brother, Epson, and other printers of both technologies in several fora, not just this one. (most end with a null answer as this one seems to have))

Any idea how to get a hal type to take a look at this? I guess Ubuntu team could champion it but the interest seems low even from the cups side. Still if Ubuntu is trying to establish itself in the laptop market where parallel ports are appearing less frequently maybe they should take a look at this.