Quote:
That is now - Finland won at eurovision

General problem is with buggy parallel.device. March found the problem and promised to write replace for this parallel.device, when he find free time...

[ Edited by Kaczus on 2006/5/22 13:15 ]

Another solution would be to report what he found so the device could be fixed without rewriting a clone from scratch.

I heard, March had been ignored...

What ever happened with this? Good printing support is one of the things I want most from MorphOS. If it only takes some minor rework of TP to get it, then why havent they released it? I'd pay $100 for a working, fully native and modern Turbo Print even.

Acill
Another option for now is to get a fully compatible printer.. I bought an Epson 740 usb printer for this reason. Works nicely with TP and mos and is pretty full featured..

magnetic

Well i have a nice color lazer I want to use, but I cant since its not on the list. It is in the linux verson though. Its a shame. I dont have a good working word processor. I paid all that cash for papyrus and its not updated to get the bugs out yet.

> I paid all that cash for papyrus and its not updated to get the bugs out yet.

Have you tried to contact Titan about that? I would like to know if they still work on the Morphos Version.

Quote:

That's quite a spectacular feat since the first we heard of it was today, from Morgoth (and the stuff he sent us is only dated 2 days ago) .. who/what/where did he try to contact?

Anyway, AFAICT what we received of info from Morgoth seems to be wrong, it implies a certain bit is not set, however it is according to the sources, so we need more info, preferably through a more reliable channel than whichever he tried the last time .. how about the MorphOS ML?


- CISC

Quote:
I first heard about this exactly 55 minutes ago, from Morgoth.

I tried to contact Emm - the parallel.device author directly. I sent him a privmsg on #morphos few times and tried to get in touch with him by email. However without any success, so I gave up - I was a little bit busy and didn't have time to prepare whole bug report. That was my fault - next time I will simply sent a mail to morphos ml or here on MorphZone.

The bug in parallel.device concerns reset bit in control register. Now I got a few hours of spare time so I prepared more detailed report.

To prove that the bug is there it I wrote a simple tool - lpt_reset, which shows, sets and clears reset line directly by polling the hardware registers. It is available (with source) here: http://march.w.staszic.waw.pl/lpt_reset.lha

I checked my lpt_reset tool with StarLC100 printer and doing "lpt_reset set" then "lpt_reset clear" in fact causes the printer reset (printer flashes its leds and moves the carriage to the init possition) so the tool works.

Next lines are the output from a shell window:
8.Work:Dev/ParFix/lpt_reset> lpt_reset
LPT status register: 0x0d - reset line inactive
8.Work:Dev/ParFix/lpt_reset> echo "test" >par:
8.Work:Dev/ParFix/lpt_reset> lpt_reset
LPT status register: 0x00 - reset line active

As you can see, the reset line after the system start (first line) is inactive. However, sending a few characters directly to printer makes this line active, which shouldn't happen! This causes all the problems with printing. Printer gets reset signal before it even manages to print any character.

Quote:

Well, Emm is hard to get in touch with these days, even for us, he does indeed come to IRC, but he idles alot. ;)

Quote:

Well, that is rather strange, as the bit you call CTRL_RESET (which is actually the ENDINIT bit) is always set (as your example does) in parallel.device, so the problem must be elsewhere (some timing issue with the printer?)...

Ok, mystery solved .. the bit was cleared when device is closed, this could race and cause complete or partial loss of printout...

A new parallel.device that fixes this issue, and blockage when no printer is connected (now just returns an error) has been uploaded to the MorphOS FTP .. it should also be available at MorphOS Download when our resident webslave wakes up... ;)


- CISC

[ Edited by CISC on 2006/6/3 15:41 ]