Cannot set iso6937 for EPG

[nobody]

When trying to set the iso6937 charset Me-TV says "Failed to get active text value". (me-tv-1.3.2)

[Michael Lamothe (lamothe-deactivatedaccount-deactivatedaccount)]

Please try the latest.

[nobody (butrus-butrus)]

Unfortunatelly, 1.3.2 is the latest version in my distro (Fedora, even in the rawhide). Will try to make an updated package when I find the time (sorry, will not trash my system with not packaged programs)

[nobody (butrus-butrus)]

Ok, compiled an updated package (1.3.5), I can now set the charset to "iso9637" in preferences, but the EPG informations are still encoded wrong.

(BTW, why there is "auto" and "iso6937" choice, but no "iso-8859-1" like e.g. Kaffeine has?)

[nobody (butrus-butrus)]

Oh, It seems it works but it takes some time for the EPG to get updated. Couldn't the EPG data be cleared (or better: reencoded) directly after changing the settings?

[Michael Lamothe (lamothe-deactivatedaccount-deactivatedaccount)]

Me TV will not update EPG events that it already has stored, even if you change the text encoding setting, you need to delete those old events. It's actually way easier to delete the Me TV database and rescan your channels using the correct text encoding.

By default, the text encoding is sent in the DVB stream for iso-8859-?, but not in the case of iso6937. If it's iso-8859-1 then the DVB stream will tell me, you don't need to. You can put iso-8859-1 in the box if it makes you feel better, it's free text.

[Michael Lamothe (lamothe-deactivatedaccount-deactivatedaccount)]

This issue has already been fixed.

[nobody (butrus-butrus)]

> Me TV will not update EPG events that it already has stored, even if you change the text encoding setting,
> you need to delete those old events.

Sorry, but this is definitely a bug from the user-experience point of view. The EPG should either be cleared automatically after setting the encoding or the user must be warned that the change may take some time.

> By default, the text encoding is sent in the DVB stream for iso-8859-?, but not in the case of iso6937.

OK, wouldn't it be safer to choose automatically the iso-6937 encoding if nothing is sent in the stream? Or choose it according to the country selected?

[Michael Lamothe (lamothe-deactivatedaccount-deactivatedaccount)]

The "Failed to get active text value" bug is invalid because it has been fixed. It's bad practise to address 2 issues in the same bug report.

My statement was misleading, for iso-6937, something is sent in the stream, it's usually iso-8859-1. Choosing the encoding based on the country sounds like a neat idea, patches welcome.