LCDproc development and user support list

Text archives Help


[Lcdproc] Re: lcdproc-0.5.1 and serialized HD44780 on k-data wake-up 2.0 module


Chronological Thread 
  • From: matteo.pillon AT gmail.com (Matteo Pillon)
  • Subject: [Lcdproc] Re: lcdproc-0.5.1 and serialized HD44780 on k-data wake-up 2.0 module
  • Date: Wed Jan 24 15:55:01 2007

On Wed, Jan 24, 2007 at 12:26:41AM +0100, Bernd Büchs wrote:
>
> Thank you Matteo for your patch!
>
> I tried it as well. I use "lcdproc-0.5.1.tar.gz". I had some problems
> applying
> your patch to my version. I assume the ",v" at the end of the paths in the
> lines with "RCS file:" in the patch are a problem.

The patch is against CVS, you can apply it to today's nightly tarball
(my patch wasn't intended to replace Peter's changes, but to integrate
them, he did a good work ;-).

http://lcdproc.sourceforge.net/nightly/lcdproc-CVS-current.tar.bz2

> With the file "hd44780-serial.c" I got even a rejection (see attached
> file).
> The content that was in my version is as following:
>
> lcdproc-0.5.1.tar.gz [...
> unsigned int conf_bitrate;
> size_t bitrate;
>
> conf_bitrate=atoi( drvthis->config_get_string(drvthis->name, "Speed",
> 0, "0") );
> if (conf_bitrate==0)
> conf_bitrate=SERIAL_IF.default_bitrate;
> if (convert_bitrate(conf_bitrate, &bitrate)) {
> report(RPT_ERR, "HD44780: serial: invalid configured bitrate speed");
> return -1;
> ...]

In the CVS trunk, atoi call on the first line and the next two lines
are missing... It didn't even compile, here... My patch restores
what it's alreay fine in 0.5.1, this is why it didn't apply.

> That in mind, I think this migth be the reason for some of the problems
> that I
> discribed with the version from Peter!?

What you got is what I did ;-), the part that didn't apply doesn't
matter. I haven't spotted the backlight problem, yet.

> Anyhow here comes my test report of the current version:
>
> LCDd.conf with Backligth=yes switchs it on when starting LCDd and switchs
> it
> off when stoping LCDd - good!
> With Backlight=no I recofgnized that the Backligth is turned off when I
> stop
> LCDd! I wonder if this is a bug or a feature? (For me this behaviour it is
> fine!)

No, it's a bug...
But from the bug, a feature!!! :-D
We can implement it!

> I also experimented with lcdproc:
> lcdpro.conf: [...
> [Load]
> # Show screen
> Active=true
> # Min Load Avg at which the backlight will be turned off [default: 0.05]
> LowLoad=0.05
> # Max Load Avg at which the backlight will start blinking [default: 1.3]
> HighLoad=1.3 ...]
>
> This load screen turns of the backligth, even with LCDd.conf Backligth=no!
> The screen shows in the first line "hostname 0.13 0" and in
> the
> second line " 0". The first parts of the second line are
> filling from rigth to left with blocks/symboles.
> Unfortunately I have not experimented with the load screen before.
>
> All other screens activated in the standard lcdproc.conf are working
> correct.

I think I got the problem...
Logging what lcd recives shows escaped 0xC9 on load and the goodbye
screens...
Do you see "LCDprocand" on the last screen (without spaces)? Between
the two words there's 0xC0C9... (With a 20x4 should be "Thanksfor").

The attached patch discards every 0xC0C9 (not really useful,
but can help us spotting the bug, the backlight shouldn't
switch anymore with backlight=no, but the screen will easly get
scrambled ;-).

> I also tried Charmap=sed1278f_0b for curiosity and every thing still looked
> like with the default?! I have to admit that I never used char maps before
> and I have no clue how this could be tested in a propper way. If my
> understanding is correct this feature is somehow LCD dependent. I use the
> following LCD
> <http://www.reichelt.de/?SID=29wzKd6KwQAR0AAH7RiW092aed62a4d907ffedb39f64556e3e886;ACTION=3;LA=3;GROUP=A5212;GROUPID=3006;ARTICLE=53953;START=0;SORT=artnr;OFFSET=16>

You're right, it's lcd-dependent, every lcd controller has its own
charset, with compatibility to ISO-8859 only for letters, numbers
and punctation, so we need a translation table to show the correct
symbols on the lcd.
Your display should work with the ea_ks0073 charmap (you would notice
difference only with letters with umlauts, accents and particular
symbols, as ~| ).

Thanks for your detailed report ;-).

Bye.

--
* Matteo Pillon
http://hydra.ilc.pd.it/~pmatthew




Archive powered by MHonArc 2.6.18.

Top of page