116 replies to this topic

[littlesat]

Did you get all these plugins from our feeds???

[BuGless]

Is that using the PLi-HD skin? Because again, I do not see the problem there, it works the same with and without the patches. More specific instructions on how to reproduce this, or where to look, would be appreciated.

The argument 'it works in the PLI-HD skin' isn't valid, if the change breaks all other skins available from the 3-rd party PLI feed.

Nobody is using the argument 'it works in the PLI-HD skin'. As you might notice, I'm merely reporting which skin I'm testing it with (since noone else did), because I full well realise that it might be skin dependent.

[adri]

Nobody is using the argument 'it works in the PLI-HD skin'. As you might notice, I'm merely reporting which skin I'm testing it with (since noone else did), because I full well realise that it might be skin dependent.

If it is skin dependent, why would an image op 19-01 work properly with the same skin?
All I did was do a 'software update' from the setup menu, so the skin wasn't changed.

[BuGless]

If it is skin dependent, why would an image op 19-01 work properly with the same skin?
All I did was do a 'software update' from the setup menu, so the skin wasn't changed.

Because some skins exercise parts of the code that others don't. Meaning that if something works with one skin, it might break with another. In this case a plugin got more functionality, but apparently this extra functionality results in unexpected side effects in the presence of some other plugins and/or skins.

[adri]

If it is skin dependent, why would an image op 19-01 work properly with the same skin?
All I did was do a 'software update' from the setup menu, so the skin wasn't changed.

Because some skins exercise parts of the code that others don't. Meaning that if something works with one skin, it might break with another. In this case a plugin got more functionality, but apparently this extra functionality results in unexpected side effects in the presence of some other plugins and/or skins.

Or it might be caused by the plugin removing or renaming functionality, like was done in the EPGSelection.py.
'Service' got renamed to 'ServiceEvent' with NO backwards compatability!
Showing the name on the top left op the 'Single EPG' window is not some functionality which is used by a particular skin.
It is used in 99% of all available skins!

But lets stop this discussion, since the changes have been reverted, and I can check if the problem has disappeared.

[BuGless]

Or it might be caused by the plugin removing or renaming functionality, like was done in the EPGSelection.py.
'Service' got renamed to 'ServiceEvent' with NO backwards compatability!

That was done after checking that the there were no apparent skins referring to this element. The element has a different name in all other screens (ServiceEvent, to be exact). Putting in backwards compatibility from the get-go would not result in cleaner code; backwards compatibility should be avoided if nothing depends on it.

Showing the name on the top left op the 'Single EPG' window is not some functionality which is used by a particular skin.
It is used in 99% of all available skins!

Which did not get affected, in theory. But apparently practice is different. Which is why I asked which skin you miss it in now.

But lets stop this discussion, since the changes have been reverted, and I can check if the problem has disappeared.

It was not really a discussion; I was merely answering your questions. But, by all means, check if things are back to normal and let me know.

[littlesat]

At least the one getting the issue is working with the basic skin. BuGless is also testing with the same skin... BuGless cannot verify the issue...

So again... where are the plugins destroying the commits comming from?
What buttons must be pressed to trigger the issue?

[BuGless]

So again... where are the plugins destroying the commits comming from?

Well, I installed the special TMDB plugin now, I'll test with that and see if I can notice anything going awry.

What buttons must be pressed to trigger the issue?

That question is still open for the EPGSearch plugin.

[littlesat]

So first we need these answers....

[Dimitrij]

1)Install TMDB plugin->'Add \"Lookup in TMBD\" Info Button to single-EPG'=yes
2)Install EPGSearch plugin->'Add \"Search\" Button to EPG'=yes
Run single EPG and press blue button(Search)

Can close the topic?

[BuGless]

1)Install TMDB plugin->'Add \"Lookup in TMBD\" Info Button to single-EPG'=yes
2)Install EPGSearch plugin->'Add \"Search\" Button to EPG'=yes
Run single EPG and press blue button(Search)\

Thanks. Will try that.

Can close the topic?

Of course not. I'll test first, as soon as I can reproduce it, it can be closed.

[BuGless]

Seems like I found a problem.
I cannot seem to reproduce the GMEG up/down navigation problem, however, I *am* able to spot the Single EPG missing channel name.

It appears that the ServiceEvent in GMEG does not cause a problem. However the renamed Service -> ServiceEvent member in EPGSelection is causing the missing channel name.
So, if you could please reapply the ServiceEvent addition to GMEG which appears to be harmless, but apply the following patch to EPGSelection instead, and can verify that nothing breaks (I tested it here with different skins and the TMDB plugin, and it appears to work), I'd be much obliged.

diff --git a/lib/python/Screens/EpgSelection.py b/lib/python/Screens/EpgSelection.py
index 6db8c32..5987e89 100644
--- a/lib/python/Screens/EpgSelection.py
+++ b/lib/python/Screens/EpgSelection.py
@@ -36,7 +36,7 @@ class EPGSelection(Screen):
  self["key_red"] = Button("")
  self.closeRecursive = False
  self.saved_title = None
- self["Service"] = ServiceEvent()
+ self["Service"] = self["ServiceEvent"] = ServiceEvent()
  self["Event"] = Event()
  if isinstance(service, str) and eventid != None:
  self.type = EPG_TYPE_SIMILAR
@@ -161,7 +161,7 @@ class EPGSelection(Screen):
  l.moveToService(self.session.nav.getCurrentlyPlayingServiceOrGroup())
  elif self.type == EPG_TYPE_SINGLE:
  service = self.currentService
- self["Service"].newService(service.ref)
+ self["ServiceEvent"].newService(service.ref)
  if self.saved_title is None:
  self.saved_title = self.instance.getTitle()
  title = self.saved_title + ' - ' + service.getServiceName()
@@ -350,9 +350,9 @@ class EPGSelection(Screen):
  datestr = '%s %d.%d.'%(_("Today"), begTime[2], begTime[1])
  self["date"].setText(datestr)
  if cur[1] is None:
- self["Service"].newService(None)
+ self["ServiceEvent"].newService(None)
  else:
- self["Service"].newService(cur[1].ref)
+ self["ServiceEvent"].newService(cur[1].ref)

  if cur[1] is None or cur[1].getServiceName() == "":
  if self.key_green_choice != self.EMPTY:

[Dimitrij]

Checked.
Blue search button does not work.

[BuGless]

Checked.
Blue search button does not work.

Ok, thanks, back to the drawingboard.

[adri]

Seems like I found a problem.
I cannot seem to reproduce the GMEG up/down navigation problem, however, I *am* able to spot the Single EPG missing channel name.

It appears that the ServiceEvent in GMEG does not cause a problem. However the renamed Service -> ServiceEvent member in EPGSelection is causing the missing channel name.
So, if you could please reapply the ServiceEvent addition to GMEG which appears to be harmless, but apply the following patch to EPGSelection instead, and can verify that nothing breaks (I tested it here with different skins and the TMDB plugin, and it appears to work), I'd be much obliged.

BuGless,

After Littlesat reverted the commit, I updated to 25-01-2013 and the problem with the 'Single EPG' channel name is gone.
Good to see you can reproduce it yourself and a fix has been found.

Thanks,

Adri.

[mge]

i've got a strange behaviour. in the GraphMultiEPG with cursor right I can select every single event int the future. At the moment ,at zdf hd I can go til 03:30 am tomorrow, after that the Cursor right scrolls 2hours and not to the next event. Is this normal?

just closed the epg an reopened, now I can only scroll through the events till 19:45, after that it scrolls 2 hours at once.

???help???

ET4000, updated today

[Dimitrij]

install new image

[mge]

did you try to reproduce it?

[Dimitrij]

http://openpli.org/download/et4x00/

[mge]

ok, i flashed it and its still the same beahviour. So once again, did you try to reproduce the error?

the bug in short: scrolling with 'cursor right' through the events, at some point it scrolls in 2hours steps.