Description of problem: While configuring the first printer queue, after the old one got deleted, I got this: # Traceback (most recent call last): File "/usr/share/system-config-printer/system-config-printer.py", line 2322, in on_tvNPMakes_cursor_changed self.NPMake = model.get(iter, 0)[0] TypeError: iter must be a GtkTreeIter Traceback (most recent call last): File "/usr/share/system-config-printer/system-config-printer.py", line 2322, in on_tvNPMakes_cursor_changed self.NPMake = model.get(iter, 0)[0] TypeError: iter must be a GtkTreeIter /usr/lib/cups/filter/foomatic-rip: found Yes, this indeed showed up twice although I am not sure in which moment. That printer is a network located printer using built-in JetDirect server, and Samsung ML-2250 pxlmono driver. Version-Release number of selected component (if applicable): system-config-printer-0.7.50-1.fc7 How reproducible: Looks consistent
All described above "attractions" apply to system-config-printer-0.7.50-1.fc6 from updates-testing as well. Only this time I was watching closely what happens the tracebacks like the above show up when during configuration one is trying to type names in order to get the right printer. Also with Samsung ML-2250 highlighed I tried to pickup a correct driver and got: ERROR: foomatic-ppdfile: Printer 'Samsung-CLP-300' and driver 'splix' are not compatible ERROR: foomatic-ppdfile: Driver 'splix' not in database! Going "Back" and "Forward" indeed showed up that system-config-printer highlights now CLP-300 with 'splix' as the only driver. How it got there I have not the slightest idea. On the second attempt to chose ML-2550 with pxlmono it got it right.
I tried again to start that printer configuration from scrach using system-config-printer-0.7.50-1.fc6 and while on "Select printer from database screen" I clearly had picked up and highligthd "Samsung" a list of models was clearly showing various Brother models. Going "Back" and "Forward" two or three times eventually brought to a screen a list of various Samsungs. OTOH it looks like that whatever is written in /etc/cups/ is mostly immaterial becuse if I will start with an empty /etc/cups/printers.conf it stays empty while a printer queue does show up. If this writes some semi-random stuff in secret locations then no wonder that I cannot get repeatable results in various configuration aspects.
Please file one bug per report. So far in this one I see: * traceback --> That's what *this* bug report is about then. Appears harmless but should be fixed. * complaint about splix --> just needs packaging I think -- want to volunteer? * manufacturer confusion --> please try to isolate what causes it, and file a separate bug report * complaint about how s-c-printer works --> s-c-printer uses the CUPS API -- no hidden config
Hmm, we're getting None for get_selected() from the tree view's selection, which is a bit unfair when the row is clearly selected. :-/
Fixed in CVS (had to use get_cursor instead).
Fixed in 0.7.51-1.fc7 for devel.
FC6 test update: 0.7.52-1.fc6.
Fixed in update: system-config-printer-0.7.52-1.fc6