[mythtv] diseqc
Yeasah Pell
yeasah at schwide.com
Mon Jun 5 03:31:50 UTC 2006
Ok, I finally found some time this weekend to do the GUI part of the
hierarchical diseqc stuff, and it's basically feature complete and seems
functional enough -- I've been running it for the last week or so.
Except for any bugs I may have missed, there's only a couple things left:
1) GUI finishing work -- mainly adding help text in the GUI, possibly
other minor stuff of that nature.
2) Database upgrade -- the queries to create the new tables are
straightforward and will be trivial to add to dbcheck.cpp, but
converting old diseqc setups to new setups will be more challenging,
especially because LNB information is treated in a fundamentally
different way now.
I'm assuming that auto-upgrade of old diseqc setups is a must-have feature?
The patch as it stands is pretty much just a wholesale replacement of
all the diseqc code scattered throughout various files with new code
that lives in a couple new source files (one for the device objects
themselves, and another for the UI), but there's one behavioral change
that may be more controversial.
Previously, DVB cards with diseqc configurations had a fixed number of
card inputs based on the maximum number of combinations of diseqc inputs
(or some approximation thereof) -- which was sort of ok when the number
of possible configurations was small, but obviously can't work for a
system capable of representing arbitrarily complex diseqc setups (even
the 50 inputs for the rotor was really annoying to work with in the UI)
Instead, I've changed it so that DVB cards with non-trivial diseqc
setups (configured in the card setup section) show up with an extra "add
new input" input at the bottom of the input list, which allows you to
always add one more input.
It certainly looks much cleaner to not have any unused inputs littering
up the screen, but I'm not really crazy about the fact that you can't
delete unused inputs (without deleting the card) -- of course it isn't
really any different than the way it was before (it just showed a whole
bunch of inputs that didn't really exist yet, and if you populated one
there wasn't really any way to get it unpopulated), and adding a delete
function to the card input list would be objectionble because only
certain kinds of inputs would be subject to deletion (plus it would make
coming up with new input names challenging, since you'd end up with gaps
in the input numbering sequence)
Anyway, if anybody is interested in doing some testing on more diverse
setups at this point, I'll open a ticket and stick the patch up there.
More information about the mythtv-dev
mailing list