On Thu, Oct 02, 2014 at 08:18:58AM -0700, Dirk Hohndel wrote:
So it seems that the old B1 command is still working, and without any DTR/RTS changes. You told me none of that worked for you, so I wonder what is going on here.
That's strange. I kept trying B1 and couldn't get it to work on mine. I will try again with my latest build to see if this was some weird cross dependency on other changes that I made... If B1 does work I can completely get rid of the cache, so that would be nice.
Oh FUN. I think I now get it. Using B1 appears to be much more timing sensitive than B8 on the A300CS that I have access to.
Specifically, in my setup my first attempt to use B1 almost always fails a couple of times (but the algorithm then simply retries and eventually succeeds). I almost never see failures with B8.
The way I tested things initially I saw the failures and didn't retry but assumed (looking at the trace of what the Aeris software does, which ALWAYS uses B8) that I should try B8... and that worked. And then when I tried B1 again at a later point, it once again failed and I figured they had removed this. Turns out that I'm wrong.
This is excellent news. I can rip out a big chunk of the new code that I wrote :-)
/D