LCDproc development and user support list

Text archives Help

[Lcdproc] (h)Bars are weird

Chronological Thread 
  • From: bsdfan at (Markus Dolze)
  • Subject: [Lcdproc] (h)Bars are weird
  • Date: Sun, 07 Mar 2010 19:19:51 +0100


While working on the Crystalfontz driver for the 635, I had to take a
closer look at how (h)Bars work. I found that:

1. In the client language the length of a hBar is given as 'pixels',
for which the client has to take the cellwidth and calculate the
length of the bar.
2. In the driver API the length of the hBar is given as a promille
value relating to the total length of the hBar in characters (full
3. The server core translates from the client's 'pixel' value to the
driver's 'promille of cells' value. As it has no idea of how mayn
'cells' a hBar consists, it takes the X position (given by the
client) and the right border of the current frame (identical to
the screen in most cases).
4. Different drivers use different heights for a hBar, some use a
full cell height, some align the hBar with characters (which
usually are 1 row less than a full cell).
5. The "seamless hBar" compile time options switches between a custom
character and an icon charcater to display a full block. This
interferes with 4 in some cases.

What I would like to do:

* change 1 to be the same as 2, but this would require an
incompatible change to the protocol, therefore I will not do this
right now.
* make the hBar style (full cell / character height) a global
configuration option.
* at the same time remove 5 and always use a custom bar character to
display a full block. For displays which are known to be seamless,
the driver can automatically adjust.

Your ideas?


Archive powered by MHonArc 2.6.18.

Top of page