Front panel LEDs not work in Standby mode

[Benny]

Greetings,

In "Standby" mode on SPARK hardware front panel LEDs remains black regardless selected settings in VFD setup.
First suspect is neutrino-mp/source/driver/simple_display.cpp

Code: Alles auswählen

void CLCD::setled(void)
{
#if HAVE_SPARK_HARDWARE
	if(mode == MODE_MENU_UTF8 || mode == MODE_TVRADIO) {
		led_mode[SNeutrinoSettings::LED_MODE_STANDBY] = 0;
		led_mode[SNeutrinoSettings::LED_MODE_TV] = g_settings.led_mode[SNeutrinoSettings::LED_MODE_TV];
	} else if(mode == MODE_STANDBY) {
		led_mode[SNeutrinoSettings::LED_MODE_STANDBY] = g_settings.led_mode[SNeutrinoSettings::LED_MODE_STANDBY];
		led_mode[SNeutrinoSettings::LED_MODE_TV] = 0;
	}

Looks like when system goes to standby, "MODE_STANDBY" branch never reached. If i comment this out, LEDs work normally in standby mode, but not working anymore in TV mode.

martii, additional unrelated question: if i want to write cyrillic symbols on VFD, i only should write a custom conversions table for ShowText() in simple_display.cpp or there something else needs to be done?

[martii]

Hi Benny,

relocating the setledmode() call in neutrino.cpp should fix the standby LED issue (4022513465dafe18efd2904ecf03a752cec80095).

Regarding your other question: Yes, if suitable mappings for Cyrillic characters to the standard ASCII set supported by the AOTOM driver exist then extending the ShowText() function would be the way to go.

Cheers,

martii

[Benny]

relocating the setledmode() call in neutrino.cpp should fix the standby LED issue

Yes, confirm - fixed, at least on GM990. Thank you.

if suitable mappings for Cyrillic characters to the standard ASCII set supported by the AOTOM driver exist then extending the ShowText() function would be the way to go.

Ok, thanks. I'll try it.

[Tann]

Cyrillic characters not display 7162

gruß thomas

ersatzdisplay: http://www.imgbox.de/users/lboppi/thumb ... 1428_t.gif

[Benny]

Cyrillic characters not display 7162

Yes, i know - i'll plan check it on 7162 hardware.
As for me - it's pointless to draw complex characters on 7(8)-digits VFD, so i don't bother to try it for 7111.

[martii]

Hi Benny,

for 7162 displays modifying simple_display.cpp will not be sufficient. That would only work to map Cyrillic characters to ASCII. You'll have to extend the aotom driver to actually display Cyrillic characters and support UTF-8, too.

Cheers,

martii

[Benny]

You'll have to extend the aotom driver to actually display Cyrillic characters and support UTF-8, too.

I have some suspicions about it, thanks.
People report that in some Enigma2 firmwares with Cyrillic characters all ok on 7162, so at first i'll plan use aotom.ko (for appropriate kernel) from there.

[martii]

Hi Benny,

the attached archives includes experimental (compiles, but untested due to lack of suitable hardware) patches to add Cyrillic support. Care to try?

Cheers,

martii


[Edit: attachment removed, already commited and pushed]

[Benny]

Care to try?

Oh, thanks, of course i'll try =)
But patch failed for aotom_main.c:

patching file aotom_main.c
patching file aotom_main.c
Hunk #1 FAILED at 657.
1 out of 1 hunk FAILED -- saving rejects to file aotom_main.c.rej
patching file aotom_main.c
Hunk #1 succeeded at 30 with fuzz 2 (offset -10 lines).
Hunk #2 FAILED at 2215.
Hunk #3 FAILED at 2228.
2 out of 3 hunks FAILED -- saving rejects to file aotom_main.c.rej

I tried to remove /tdt completely and clone it again with "make preqs" - but still no go. Problem file attached below.

[martii]

Weird ... this should have worked, the patches were unmodified "git diff origin/master" output.

I've pushed the commits, feel free to git pull ...

Cheers,

martii

[Tann]

Code: Alles auswählen

People report that in some Enigma2 firmwares with Cyrillic characters all ok on 7162

7162 дисплей, он не может

[Benny]

7162 дисплей, он не может

See attach - if it work on Spark, then it can work elsewhere.

[Benny]

I've pushed the commits, feel free to git pull ...

Ok, thank you.

Considering

# The 7162 build is currently plain broken. I don't see why I should care. --martii
SPARK_ONLY=1

in environment.mk, aotom module build take more time than i thought =(

Anyway, nothing to report right now:
- i'm not 100% sure, than i fixed build pipeline properly (well, it builds... something =));
- still don't have 7162 hardware in reach.
I've posted aotom.ko and neutrino binaries at some Russian forum, and got feedback like "nothing changes", but again - until i don't see it on my own eyes or until i get feedback from trusted source - can't be sure about.

[martii]

The latest commit a couple of seconds ago should raise the aotom driver to a UTF-8 capable state.

Cheers,

martii

[Benny]

The latest commit a couple of seconds ago should raise the aotom driver to a UTF-8 capable state.

Yes, got a positive feedback - it works.
Great, and thanks a lot from russian-speaking users!

[Tann]

dann ist das ein neueres Modell mit anderen Display drinne

gibt auch modelle mit tuner dvb-t und welche mit dvb-t2
ebenso verschiedene Sattuner mittlerweile

[seife]

Meine 7162 ist zwar bis auf weiteres weg, aber ich sehe nicht, warum die das nicht anzeigen können sollte.

Diese 14 segment displays (in diesem Fall, wegen dem Punkt in der Mitte eigentlich 15 segment) sind ja so:

Code: Alles auswählen

 ---
|\|/|
 -*- 
|/|\|
 ---

Damit kann man das im Bild durchaus darstellen. Das ist zwar schlecht lesbar un d sieht scheisse aus, aber das tun auf diesem Display eigentlich alle Zeichen

In meinem alten Buildsystem in scripts/sparkvfd.pl ist übrigens der Algorithmus von bitmaske zu Anzeige beschrieben, allerdings in perl-code und nicht sonderlich lesbar

Code: Alles auswählen

~/test> scripts/sparkvfd.pl 0xd5 0x58
       
|     |
|     |
 --+   
|  |  |
|  |  |
 ----- 

~/test> scripts/sparkvfd.pl 0x59 0x1c
       
|    /|
|   / |
   +   
| /   |
|/    |

Das kann also problemlos angezeigt werden.

[martii]

dann ist das ein neueres Modell mit anderen Display drinne

Nein.