[mythtv-commits] Ticket #1945: DVB-S/diseqc patch

MythTV mythtv at cvs.mythtv.org
Wed Jun 14 17:04:38 UTC 2006

#1945: DVB-S/diseqc patch
 Reporter:  yeasah at schwide.net  |        Owner:  danielk
     Type:  enhancement         |       Status:  new    
 Priority:  minor               |    Milestone:  0.20   
Component:  mythtv              |      Version:  head   
 Severity:  medium              |   Resolution:         
Comment (by anonymous):

 Replying to [comment:5 danielk]:
 > Yesah, can you break this up into smaller pieces.
 > Preferably, the size of each patch should be something that I can read
 through an understand in about 10 minutes. Also, each change in MythTV
 behaviour should have it's own patch.

 Yeah, I know it's kind of annoyingly monolithic at this point. There's
 definitely stuff that can be split off -- basically anything that isn't
 actually new code, including the potentially controversial stuff like the
 retuning or FEC_AUTO, so I can split that off for sure. But most of the
 complexity is in 2 totally new files (dvbdevtree.cpp/h, which contains all
 the new diseqc tree code, and dvbdevtree_cfg.cpp/h, which contains the GUI
 for the same), and I don't know how I would split up code that is all new
 -- it's a cohesive whole.

 > Some parts need to be discussed, for instance:
 >  * Always tune with FEC_AUTO if card can do it.
 > This is much slower than explicit tuning parameters for some cards, if
 we really need this feature, it needs to be optional.

 On both of my cards FEC_AUTO is just as fast, so I thought it might be
 nice, but if it's going to reduce tuning performance I agree it should
 probably not be included, certainly not by default.

 The reason I thought it might be nice is because there's at least one
 instance of FEC being wrong in transmitted NITs. Another solution would be
 to attempt to tune at scan time with FEC_AUTO, and take note of the actual
 FEC, but some cards that are FEC_AUTO capable just report back FEC_AUTO as
 the fec type after tuning, so that isn't 100% either.

 I'll split it up as much as I can and post the patches -- it will likely
 take the form of the new dvbdevtree files as a single patch (without any
 changes to the existing code), and then a series of other patches that
 contain the various other DVB-S fixes and also the patch that integrates
 the new dvbdevtree stuff with everything else.

 Sound good?

Ticket URL: <http://svn.mythtv.org/trac/ticket/1945>
MythTV <http://www.mythtv.org/>

More information about the mythtv-commits mailing list