Re: Problems with Pyramid-Display

On 30.10.2012 16:55, Manuel Dejonghe wrote:
> Hi,
>
> On 30.10.2012, at 07:28, Markus Dolze <[email protected]> wrote:
>
>> On 29.10.2012 19:02, Manuel Dejonghe wrote:
>>> Hi all,
>>> I am currently encountering some problems with some Pyramid Displays that we got. The displayed text seldom makes sense and reacts to keypresses only very very late.
>> Which version of LCDproc to you use? Versions prior to 0.5.5 had some
>> problems communicating with Pyramid displays.
> Yes, version 0.5.5
>
>> Which OS are you using? On Linux there are known problems perhaps
>> related to hi-resolution timers. Read
>> http://lists.omnipotent.net/pipermail/lcdproc/2009-September/013089.html.
> Thank you very much, that did the trick.
>
> As far as I understand, every user of "LCDd -d pyramid" has to expect this problem, since kernel 2.6.21. I suppose that http://lcdproc.sourceforge.net/docs/current-user.html#pyramid-howto as well as the LCDd.conf [pyramid] sections would be where a heads up warning could be searched for and read.
> Or is this a more general problem, rather than to be specific for the pyramid-display ?
>
> Is there possibly another fix for this problem ? Disabling the highres timers brings us trouble with the internal clock of virtual machines.
> Looking at the code of the pyramid driver (pylcd.c), I see things like "usleep(50);". That's microseconds. Could it be that the kernel WITH highres is actually able to do that, and the "old" code relies on the kernels behaviour beeing a little sloppy without highres ?
>
>>> Is there somewhere some documentation about the protocol spoken to the display ? I'd love to tap that connection for debugging, or also to simulate either side of the communication.
>> Yes, there is. I am currently checking if I am allowed to redistribute it.
> Thank you, I got an 8 pages "Pyramid Appliance Display USB-Version Telegrammbeschreibung" by now. The document does not have any information about copyright, license or distribution restrictions.
>
> kind regards,
> Manuel Dejonghe
>

Hi,

that is the correct one.

You may try what happens if you remove / comment out the small delay in
line 211 (that is in function real_send_tele). In most cases there is a
larger delay after calling that function.

If that does not help, there may be some tweaks to the TTY settings that
I am unaware of.

Regards,
Markus