sjh - mountain biking running linux vegan geek spice - mtb / vegan / running / linux / canberra / cycling / etc

Steven Hanley hackergotchi picture Steven




Other online diaries:

Aaron Broughton,
Andrew Pollock,
Anthony Towns,
Chris Yeoh,
Jeremy Kerr,
Martijn van Oosterhout,
Michael Carden,
Michael Davies,
Michael Still,
Tim Potter,
Tony Breeds,


Linux Weekly News,
Girl Genius,
Planet Linux Australia,

Canberra Weather: forecast, radar.

Subscribe: rss, rss2.0, atom

Mon Tue Wed Thu Fri Sat Sun



Archive by month:

Wed, 02 Aug 2006

The LA Emperor^WPresident's new wardrobe - 17:27
I notice a mention of anti rfid wallet sleeves or similar on Boing Boing to protect your wallet or RFID cards from being scanned. Now thinking of sleeves, Jon needs a new wardrobe! (I wonder if he wants some long sleeved hello kitty jumpers). Just think if the sleeves on all his clothing were Anti RFID, no one would go around stealing his identity.

[/various] link

Do they really make fake crappy network cards? - 16:24
This is interesting, as Bob mentioned on the CLUG list we had some problems with some network cards that appeared to be Realtek 8139 based recently. As suggested in a few (linux netdev posts with mention of the pci id 1904:8139 there) places, there may (a copy of the code here from the INTEX Zip file mentioned in the previous post) be a 2.4 driver (of somewhat questionable licence, quality and capability)

It was interesting to see, as Bob pointed out, the driver supplied with the cards will load on windows, and appear to say it is a 8139 card, yet it was not recognised as a 8139 by the default windows 8139 driver, nor does this driver work with other 8139 cards.

I kind of wonder what details can be extracted from the 2.4 driver file, as suggested in the netdev posts it may be weird, however if we are allowed to use those register details and such it should be possible to get a working 2.6 driver and maybe even make a driver that does not suck. Of course I do wonder why you would want to fake a 8139 rather than badge it as if it were a much better network card.

[/comp/hardware] link

Getting a Kyocera FS 820 Laser Printer working under Linux - 14:42
I needed to buy a laser printer that would be used on a POS system and make it work under Linux. I looked at the Linux printing suggested printers page and found they recommended a Kyocera FS 820 as it is fully supported and provides a very low cost per page. I purchased it at a local computer shop for AUD $185 which is pretty damn cheap.

Kyocera provide a lot of ppd's that can be used with Linux, however they do not provide one for this printer. The Linux Printing page for the printer is not much help, however I find the PPD for the F-820 drives it successfully at 300x300 and also at 600x600 if I manually add the line

*Resolution 600x600dpi/600 DPI: "<</HWResolution[600 600]>>setpagedevice"
to the ppd file. I also found the margins needed a bit of adjustment, at the moment the best I can get (losing a few mm off the top of the page currently) is with
*Margins Custom/Custom : "<</.HWMargins[0 20 30 0] /Margins[0 0]>>setpagedevice"
as a Custom margin in the ppd file.

To get the printer working I needed the usblp driver in the 2.6 kernel and to install a few packages (cups and related), at the moment I have installed "cupsys foomatic-filters-ppds cupsys-driver-gutenprint cupsys-driver-gimpprint foomatic-filters python-foomatic python-ipy printconf hpijs hplip pconf-detect" (on debian) however I suspect I do not really need all of them.

One problem I had with it for a while is it did not seem to be able to find the usb printer most of the time. It took me a while to realise (by running lpinfo -v a few times that it seemed the printer utilities no longer were able to see the usb printer device after the first time they looked for it after it was plugged in. I hope this is due to some bugs in the kernel version or cups version I am using (notably I compiled usblp with gcc 4.0.3 and the kernel I inserted it into was compiled using gcc 4.0.2) and this is on a machine running sid. When I get the production machine set up it is running a debian kernel image and also a sarge system everywhere. For now I can put up with testing from my laptop by re plugging the cable every time I need to print.

Interestingly the device shows up as usb://Kyocera/FS-820 (also you can speak to /dev/usb/lp0) which as they say means it can be plugged in with other printers and not be dependent on plug in order (though if you plug in multiple FS-820's that may now work <g>) Oh and I wonder if the above mentioned need to re plug the usb interface all the time is a cups utils bug due to the fact /dev/usb/lp0 is there all the time an if I do "echo text > /dev/usb/lp0" at any time it prints a page with that plain text on it quite happily.

[/comp/hardware] link

home, email, rss, rss2.0, atom