[prev in list] [next in list] [prev in thread] [next in thread] 

List:       openocd-development
Subject:    [Openocd-development] More reset_config docs
From:       zw () superlucidity ! net (Zach Welch)
Date:       2009-05-31 23:17:11
Message-ID: 1243811831.26434.684.camel () localhost
[Download RAW message or body]

On Sun, 2009-05-31 at 15:17 -0700, David Brownell wrote:
> On Sunday 31 May 2009, Zach Welch wrote:
> > 
> > > What I would like to see is that we push funky rules into board/target config
> > > files and keep the reset_config implementation as simple as possible.
> 
> One reason I dislike that model is the knowledge that
> configuration/installation errors are one of the top
> causes of pain with *any* software package.
> 
> Making configuration more complex ("funky" conditional
> rules certainly do that) guarantees more such errors.
> 
> 
> > > There are pieces missing today, such as the board/target config files
> > > ability to detect capabilities in the interface.
> > 
> > It seems simpler for the authors of the interfaces and boards files to
> > simply declare what the hardware can do, and OpenOCD should work out the
> > best capabilities to use.  That seems simpler for script authors than
> > trying make them detect things in the configuration files.  Or am I
> > missing something here?
> 
> I can't see that you're missing anything.

Well... Even assuming that this assessment is correct, I do not see the
best way to get from here to there.  I mostly see the big picture with
this issue, but the devil's in the details -- and there seem to be quite
a few of those to consider.

I hope others can step forward and explain how such suggestions would be
implemented properly, giving in-depth perspective of how we might build
up this kind of solution.  The Manual  would benefit from a patch to
give a full background on the subject.

Cheers,

Zach


[prev in list] [next in list] [prev in thread] [next in thread]