303 replies to this topic

[ims]

It is strange, but on some services is not in MEPG displayed picon (but there on Infobar and display is picon presented).
And when is zapped to service with this "bug", then running MEPG display always bouquet's services 1-6

eg:
in bouquet are services 1 - 20. For all are downloaded picons, but for service 15 is not in MPEG picon displayed.

When i watch service 15 and then i run MEPG, MEPG displaying services 1-6.
When i watch service 16 and i run MEPG, is correctly displayed 16 - 21

Edited by ims, 8 January 2012 - 11:29.

[awx]

It is strange, but on some services is not in MEPG displayed picon (but there on Infobar and display is picon presented).
And when is zapped to service with this "bug", then running MEPG display always bouquet's services 1-6

eg:
in bouquet are services 1 - 20. For all are downloaded picons, but for service 15 is not in MPEG picon displayed.

When i watch service 15 and then i run MEPG, MEPG displaying services 1-6.
When i watch service 16 and i run MEPG, is correctly displayed 16 - 21

I managed to reproduce something that has a similar behaviour to what you are seeing but I am not 100% sure that it is the same error you are seeing.
Here is an update with the potential fix

Changes
-------------------------

Workaround for service comparison
Add a yellow button as people may want to bind it to a function

Attached Files

•  MergedGraphMultiEpg.1.0.17.tar.gz   12.77KB   11 downloads

[theparasol]

@pliteam: Wanneer komt dit noeste programmeerwerk eens in de feed oid. Het voldoet voor mij nu al beter dan de code die er nu standaard inzit voor multiepg.

[littlesat]

It can be put in the feed... but what we do like/need is a diff agains our current git sources so we have a better impression about the exact changes - and also make it easier to commit...

[awx]

It can be put in the feed... but what we do like/need is a diff agains our current git sources so we have a better impression about the exact changes - and also make it easier to commit...

I have provide in a previous post a tar gzip with all the patches up to version 1.0.16

[ims]

still same...

I found, that happends on 2 services only. Tried change it's position - same, remove it's picon - same.
Tried remove one service from bouquet and add it again - begin works for this service.

I leave 2nd problematics for debug. Where is best position for debug, what happends ? You know better this code :-)

[awx]

still same...

I found, that happends on 2 services only. Tried change it's position - same, remove it's picon - same.
Tried remove one service from bouquet and add it again - begin works for this service.

I leave 2nd problematics for debug. Where is best position for debug, what happends ? You know better this code :-)

You want to look at the moveToService function and onCreate which calls moveToService passing in the currently Playing service reference.

[ims]

I saw on services in generateList()

for this service is INFO in E2: 1:0:1:7856: .... , but there in MEPG is for this service in generateList() displayed:

1:134:1:0:0:0:0:0:0:0:FROM BOUQUET "alternatives.viasat_history.tv" ORDER BY bouquet

problem in my bouquet

Edited by ims, 8 January 2012 - 18:12.

[theparasol]

Kdo nic nedělá, nic nezkazí!

Wie niets doet maakt ook geen fouten

[awx]

I saw on services in generateList()

for this service is INFO in E2: 1:0:1:7856: .... , but there in MEPG is for this service in generateList() displayed:

1:134:1:0:0:0:0:0:0:0:FROM BOUQUET "alternatives.viasat_history.tv" ORDER BY bouquet

problem in my bouquet

Ah ok, thanks for the update

[ims]

I think, when is run MEPG, is found this "alternatives" and then MEPG jump to 1st item in bouquet.
I dont know, how I get this, may be DreamboxEdit software ... but E2 knows work with alternatives well...

[awx]

I think, when is run MEPG, is found this "alternatives" and then MEPG jump to 1st item in bouquet.
I dont know, how I get this, may be DreamboxEdit software ... but E2 knows work with alternatives well...

What is the default behaviour when you look at the bouquet in the channel list?

[ims]

sorry, dont understand what you means...

[awx]

sorry, dont understand what you means...

Well you have an alternate bouquet included in your bouquet
Should all the services from the alternate bouquet be included in the current one or should something else happen?
For example you have to user bouquets
bouquet A
bouquet B

Bouquet A has

service 1
service 2

Bouquet B has

service 3
service 4
bouquet A

Should Bouquet B be displayed as
service 3
service 4
service 1
service 2

So when your watching tv on bouquet b and press down, what does the channel list look like?

[ims]

I have not alternative bouquet ... but I known, how i made it - long time ago, when i translate .po of E2 and tried items in menu.

You can try it: in service selector pres Menu - > add alternatives and select service.

I remove it and all is ok now.

Edited by ims, 8 January 2012 - 19:17.

[mirakels]

I'll try merging the patches into the pli tree.
But I see the patches are not against the original pli codebase.
Also I see several patches that actually are part of a whole. E.g. patch 0002 and 0031 should be merged together.

Awx, can you reorder your patches and group then togeteher and base them on the pli code from start. I want to avoid putting in fixes upon fixes...

Also when you generate the patches also use sign-off so we know where the patches come from.

[awx]

I'll try merging the patches into the pli tree.
But I see the patches are not against the original pli codebase.
Also I see several patches that actually are part of a whole. E.g. patch 0002 and 0031 should be merged together.

Awx, can you reorder your patches and group then togeteher and base them on the pli code from start. I want to avoid putting in fixes upon fixes...

Also when you generate the patches also use sign-off so we know where the patches come from.

All patches are based off the PLi source.
I just generated them from a branch, rather than rebasing my changes on to master.
I will see what changes should be reordered or squashed, but if you have any other ones you think should be combined or reordered please let me know.

[mirakels]

if I look at the 0002 pathc it says:

- font = 0, flags = RT_HALIGN_LEFT | RT_VALIGN_CENTER,
- text = service_name,
- color = self.foreColorService,
- backcolor = self.backColorService,
- border_width = 1, border_color = self.backColor) ]
but the pli version contains:

font = 1, flags = RT_HALIGN_CENTER | RT_VALIGN_CENTER | RT_WRAP,
text = ev[1], color = foreColor, color_sel = foreColorSelected,
backcolor = backColor, backcolor_sel = backColorSelected, border_width = 1, border_color = borderColor))
so I conclude the patch is not against the base PLi code.
From the thread I see you started of with the dream version? and later on merged to the PLi version. Maybe some patches were not rebased properly?

About grouping/squashing patches. I guess you know the best what patches are related or fixes of previous patches. and waht belongs togethere to form an indendent patch as a whole. I don;t like to add patches like "a = None" and later on a patch that reverts it to" a =23001".

I don;t feel like going through 90 patches and figure out what belongs together... I hope you understand.

[awx]

if I look at the 0002 pathc it says:

- font = 0, flags = RT_HALIGN_LEFT | RT_VALIGN_CENTER,
- text = service_name,
- color = self.foreColorService,
- backcolor = self.backColorService,
- border_width = 1, border_color = self.backColor) ]
but the pli version contains:

font = 1, flags = RT_HALIGN_CENTER | RT_VALIGN_CENTER | RT_WRAP,
text = ev[1], color = foreColor, color_sel = foreColorSelected,
backcolor = backColor, backcolor_sel = backColorSelected, border_width = 1, border_color = borderColor))
so I conclude the patch is not against the base PLi code.
From the thread I see you started of with the dream version? and later on merged to the PLi version. Maybe some patches were not rebased properly?

About grouping/squashing patches. I guess you know the best what patches are related or fixes of previous patches. and waht belongs togethere to form an indendent patch as a whole. I don;t like to add patches like "a = None" and later on a patch that reverts it to" a =23001".

I don;t feel like going through 90 patches and figure out what belongs together... I hope you understand.

I will take a look and see, maybe the rebase didnt go well as you mentioned. Maybe the patch goes with a patch after that fixes the problem.
I will see which patches I can combine

[awx]

if I look at the 0002 pathc it says:

- font = 0, flags = RT_HALIGN_LEFT | RT_VALIGN_CENTER,
- text = service_name,
- color = self.foreColorService,
- backcolor = self.backColorService,
- border_width = 1, border_color = self.backColor) ]
but the pli version contains:

font = 1, flags = RT_HALIGN_CENTER | RT_VALIGN_CENTER | RT_WRAP,
text = ev[1], color = foreColor, color_sel = foreColorSelected,
backcolor = backColor, backcolor_sel = backColorSelected, border_width = 1, border_color = borderColor))
so I conclude the patch is not against the base PLi code.
From the thread I see you started of with the dream version? and later on merged to the PLi version. Maybe some patches were not rebased properly?

About grouping/squashing patches. I guess you know the best what patches are related or fixes of previous patches. and waht belongs togethere to form an indendent patch as a whole. I don;t like to add patches like "a = None" and later on a patch that reverts it to" a =23001".

I don;t feel like going through 90 patches and figure out what belongs together... I hope you understand.

rebase was fine you are looking at the wrong parts
The part you reference is from the event note the ev[1], the part that was being changed was from the service portion, which looks like it matches correctly to the PLi source
I will not take a look at combining the changes