On Tue, Apr 15, 2008 at 02:50:15PM -0400, Simon Ruggier wrote:
> This isn't specific to this thread, but does the charging not work out
> of the box with berry_charge?

I haven't used the berry_charge module lately, but looking at the code
to the module in the latest 2.6.25 kernel sources, it appears that it should
work in most cases.

It is the corner cases that you don't have as much control with.  For example,
the kernel module will not let you set a Pearl device to 0001 instead of
0004, and this is sometimes useful for debugging, or getting at the
database in some devices.

I also remember something about the kernel being a bit too fast for some of
the older devices, and they got confused.  I know that even with bcharge,
if you run bcharge over and over on the same device in rapid succession,
it can cause the Blackberry to reset itself.  Fortunately, udev rules
prevent that for bcharge, while if it happens in the kernel, it's not
quite as easy to fix, in my experience.

Also, depending on your kernel version, the Blackberry's USB port would be
auto-suspended, and this would make for some funky behaviour with charging...
sometimes draining the battery instead of charging it, so I've heard.
Meanwhile, bcharge + proper udev rules will work with any kernel version
and try its best to do the right thing.

Short answer: if you're up to date with the kernel, you're probably fine,
but bcharge does more from userspace, and gives you more control.

- Chris


-------------------------------------------------------------------------
This SF.net email is sponsored by the 2008 JavaOne(SM) Conference 
Don't miss this year's exciting event. There's still time to save $100. 
Use priority code J8TL2D2. 
http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone
_______________________________________________
Barry-devel mailing list
Barry-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/barry-devel

Reply via email to