116 replies to this topic

[littlesat]

Is it during the loading of e2 or during the downloading from the server/fallback box?
Sorry my epg.dat isn’t 12MB as I do not need epg from all possible channels I rarely tuned to! So I never saw .loading!?

Edited by littlesat, 3 September 2018 - 06:54.

[Rob van der Does]

It happens absolutely during reboot/restart; I've seen that quite a number of times (remember you have to have a quick look, as it's only briefly).

I'm not sure if the same happens due to the epg-save command that should be issued by the client box ti the server box, but it seems logical that it does.

[littlesat]

So when it says loading it is actually loaded by E2? The load should be performed after the download from the fallback/server box. If so I don't understand the issue IMS described.

 

I have just 2MB or so epg.dat... I never could see 'loading'.... (and as long the epg cache is in RAM I never recommend to put all those EPG in that you actually rarely use...)... 

Edited by littlesat, 3 September 2018 - 08:46.

[ims]

Fastscan Skylink 23.5 and zap over several services and you will to have big EPG too. Or I can send you 15MB epg. Put it to server and you can try it.

 

 

 

[littlesat]

I just checked on a box without epg fallback and it seems the .loading appears as soon the EPG is readed.... The latest patch should avoid that the just written EPG is read back during an E2 restart...

Edited by littlesat, 3 September 2018 - 09:28.

[littlesat]

https://github.com/O...vb/epgcache.cpp ->: Line 1300 e.v. In 1317 the file is renamed and in 1440 renamed back...

 

With my latest patch the load() should be skipped and performed actually later... I was doubting to move 

 

cleanLoop();

runLoop();

save();

 

at the end of load()... and should runLoop be included here? This is not clear to me yet.

 

Also wierd that save saves the epg.dat again and mayby that is the cullput here... as this still writes an 'empty' epg.dat at start of E2 when the fallback epg is enabled.... I think this may trigger an issue 'again'... as the file is written and another thread can also write in the same file.

 

When moving cleanloop...save to the end of load() then also at every load the data read from the epg is cleaned, runLooped() (is this needed?) and saved

Edited by littlesat, 3 September 2018 - 09:40.

[ims]

See output from my stupid script ( both remote, i am not home)

#!/bin/sh
n=1
echo "Start" > /media/usb/test.txt
rm /hdd/epg.dat
rm /hdd/epg.dat.loading
while [ $n -le 100000 ] 
do
    ls -e /hdd/epg.dat >> /media/usb/test.txt
    ls -e /hdd/epg.dat.loading >> /media/usb/test.txt
    n=$(( n+1 ))
    sleep .1
done

-rwxr-xr-x    1 root     root       1122625 Mon Sep  3 10:27:32 2018 /hdd/epg.dat
-rwxr-xr-x    1 root     root       1122625 Mon Sep  3 10:27:32 2018 /hdd/epg.dat
-rwxr-xr-x    1 root     root       1122625 Mon Sep  3 10:27:32 2018 /hdd/epg.dat
-rwxr-xr-x    1 root     root       1122625 Mon Sep  3 10:27:32 2018 /hdd/epg.dat
-rwxr-xr-x    1 root     root       1122625 Mon Sep  3 10:27:32 2018 /hdd/epg.dat
-rwxr-xr-x    1 root     root       1122625 Mon Sep  3 10:27:32 2018 /hdd/epg.dat
-rwxr-xr-x    1 root     root             0 Mon Sep  3 10:27:56 2018 /hdd/epg.dat
-rwxr-xr-x    1 root     root             0 Mon Sep  3 10:27:56 2018 /hdd/epg.dat
-rwxr-xr-x    1 root     root             0 Mon Sep  3 10:27:56 2018 /hdd/epg.dat
-rwxr-xr-x    1 root     root             0 Mon Sep  3 10:27:56 2018 /hdd/epg.dat.loading
-rwxr-xr-x    1 root     root             0 Mon Sep  3 10:27:56 2018 /hdd/epg.dat.loading
-rwxr-xr-x    1 root     root             0 Mon Sep  3 10:27:56 2018 /hdd/epg.dat.loading
-rwxr-xr-x    1 root     root             0 Mon Sep  3 10:27:56 2018 /hdd/epg.dat.loading
-rwxr-xr-x    1 root     root             0 Mon Sep  3 10:27:56 2018 /hdd/epg.dat.loading
-rwxr-xr-x    1 root     root             0 Mon Sep  3 10:27:56 2018 /hdd/epg.dat.loading
-rwxr-xr-x    1 root     root             0 Mon Sep  3 10:27:56 2018 /hdd/epg.dat.loading
-rwxr-xr-x    1 root     root             0 Mon Sep  3 10:27:56 2018 /hdd/epg.dat.loading
-rwxr-xr-x    1 root     root             0 Mon Sep  3 10:27:56 2018 /hdd/epg.dat.loading
-rwxr-xr-x    1 root     root             0 Mon Sep  3 10:27:56 2018 /hdd/epg.dat.loading
-rwxr-xr-x    1 root     root             0 Mon Sep  3 10:27:56 2018 /hdd/epg.dat.loading
-rwxr-xr-x    1 root     root             0 Mon Sep  3 10:27:56 2018 /hdd/epg.dat.loading
-rwxr-xr-x    1 root     root             0 Mon Sep  3 10:27:56 2018 /hdd/epg.dat.loading
-rwxr-xr-x    1 root     root             0 Mon Sep  3 10:27:56 2018 /hdd/epg.dat.loading
-rwxr-xr-x    1 root     root             0 Mon Sep  3 10:27:56 2018 /hdd/epg.dat.loading
-rwxr-xr-x    1 root     root             0 Mon Sep  3 10:27:56 2018 /hdd/epg.dat.loading
-rwxr-xr-x    1 root     root             0 Mon Sep  3 10:27:56 2018 /hdd/epg.dat.loading
-rwxr-xr-x    1 root     root       2586480 Mon Sep  3 10:27:56 2018 /hdd/epg.dat.loading
-rwxr-xr-x    1 root     root       4410280 Mon Sep  3 10:27:56 2018 /hdd/epg.dat.loading
-rwxr-xr-x    1 root     root       6316980 Mon Sep  3 10:27:56 2018 /hdd/epg.dat.loading
-rwxr-xr-x    1 root     root       7145980 Mon Sep  3 10:27:56 2018 /hdd/epg.dat.loading
-rwxr-xr-x    1 root     root      10279600 Mon Sep  3 10:27:56 2018 /hdd/epg.dat.loading
-rwxr-xr-x    1 root     root      11416616 Mon Sep  3 10:27:56 2018 /hdd/epg.dat.loading
-rwxr-xr-x    1 root     root      11416616 Mon Sep  3 10:27:56 2018 /hdd/epg.dat.loading
-rwxr-xr-x    1 root     root      11416616 Mon Sep  3 10:27:56 2018 /hdd/epg.dat.loading

[littlesat]

I can also do not make some tests here right now... ... but at least I think we should also not do the save() when we initiate the thread.

[Rob van der Does]

My epg.dat is always around 10-12 MB. Just using some Dutch & UK services is all it takes to get that amount.

 

There is an other issue related to the import of EPG by a client box. It turns out that the box only looks for the epg.dat on HDD; if I have my epg on USB (not set as HDD replacement) the file is not found. IMHO the box should be looking at all normal places (internal, for in case no medium is attached at all, HDD, USB etc.).

[Rob van der Does]

I can also do not make some tests here right now... ... but at least I think we should also not do the save() when we initiate the thread.

IMHO the epg on the server should be saved when the client wants to import it. Otherwise epg.dat can be (very) old.

[Pr2]

Please keep in mind that some people would like to use this feature with capabilities purely local to the client box (when the overwrite local bouquet and lamedb will we tackle down), so on the client box epgimport will be installed and configured to fill up EPG for channels that are not available on the server box. So please don't assume that the EPG from the server is the only one that we want to have on the client box!

[littlesat]

There is an other issue related to the import of EPG by a client box. It turns out that the box only looks for the epg.dat on HDD; if I have my epg on USB (not set as HDD replacement) the file is not found.

 

This is indeed a 'pain' and I'm aware of it...

 

The code is now made that it will only 'grab' the epg.dat from flash or from /hdd...  I still think the owif need a change so we can give it a command to download a epg,dat from the server/fallback tuner and the owif arranges that the correct file is derived from the server/fallback tuner. When doing this on a client box we need to download and parese the /etc/enigma2/settings file from the server/fallback tuner box. 

 

In addition the UI of openPLi does not allow to relocate the epg.dat to an USB stick or so at this moment (you can only do it manually) So for now the code is right. 

 

I'm afraid the change in OWIF will never be done....  so we only can download the settings file and parse/grab the epg.location....

 

or do you thing to 'scan' all media for an epg.dat is also fine...as work-a-round? Then I prefer that

Edited by littlesat, 3 September 2018 - 10:59.

[littlesat]

IMHO the epg on the server should be saved when the client wants to import it. Otherwise epg.dat can be (very) old.

It is saved via a command over the OWIF. And after it is saved it is copied to the client box also over OWIF.

[ims]

Btw - there in EpgRefresh plugin can be changed path for epg.dat => can be placed to usb too...

[littlesat]

Brrrr. are they really using os.system in EPGRefresh...

https://github.com/O...gcache_filename

[Rob van der Does]

Please keep in mind that some people would like to use this feature with capabilities purely local to the client box (when the overwrite local bouquet and lamedb will we tackle down), so on the client box epgimport will be installed and configured to fill up EPG for channels that are not available on the server box. So please don't assume that the EPG from the server is the only one that we want to have on the client box!

Do I understand correctly that the import (of services/bouquets/EPG) is done on top of the existing data? So the import can be used irrespective of the use of local tuners (e.g. when the server box has DVB-S and the client has DVB-T; the client can use both the local tuner + EPG and the remote tuner(s) + EPG)?

If so, that would be a great feature!

 

 

There is an other issue related to the import of EPG by a client box. It turns out that the box only looks for the epg.dat on HDD; if I have my epg on USB (not set as HDD replacement) the file is not found.

 

This is indeed a 'pain' and I'm aware of it...

 

The code is now made that it will only 'grab' the epg.dat from flash or from /hdd...  I still think the owif need a change so we can give it a command to download a epg,dat from the server/fallback tuner and the owif arranges that the correct file is derived from the server/fallback tuner. When doing this on a client box we need to download and parese the /etc/enigma2/settings file from the server/fallback tuner box. 

 

In addition the UI of openPLi does not allow to relocate the epg.dat to an USB stick or so at this moment (you can only do it manually) So for now the code is right. 

 

I'm afraid the change in OWIF will never be done....  so we only can download the settings file and parse/grab the epg.location....

 

or do you thing to 'scan' all media for an epg.dat is also fine...as work-a-round? Then I prefer that

 

I think scanning all media should be fine.

 

That is: in most cases. Because it can happen that an old, no longer used epg-dat is available on one of the devices. Such a situation can easily arise when later on a medium has been added.

[littlesat]

It seems parsing the server/falbacktuner's settings file was a peace of cake...  Can be done in a 'oneliner'...

Edited by littlesat, 3 September 2018 - 13:06.

[Rob van der Does]

I take it you're referring to https://github.com/O...dd300072bbc9992

[littlesat]

Yep

[Rob van der Does]

FMI (as I don't have a dec-image): which issues have now been fixed?