[drats_users] D-RATS 0.2.6b6

[David Hubbell]

Hi Dan,
Sitting here on a beautiful Autumn mor in NJ and wishing that there was 
a D-Star repeater that I could hit from my house. Is there a way that 
D-RATS could DIGI its messages in the Hinterland given a bunch of 
Dedicated simplex radios. I just can't seem to get the group here 
involved enough to buy a repeater. There are several of us who have 91ad's.
Dave
Can't thank you enough for all your work on D-RATS!
Dan Smith wrote:
> Hi all,
>
> I just posted a new beta with some changes that might be helpful for
> testing with gateways and reflectors.  The semantics of the changes are
> a little hard to describe, so I'll try to do it here and will be glad to
> answer questions and clarify as needed.  For those of you who have
> joined recently, the latest beta build can always be found here:
>
>   http://d-rats.danplanet.com/download/beta/
>
> Changes since the previous beta:
> - Added ability to set the color of "broken" text (i.e. the yellow
>   stuff)
> - Fixed up a few issues that should make startup a little quicker
> - Add options in settings dialog for block size and pipeline size[1]
> - Add COM port detection on Windows[2]
> - Make sure to limit users to 8-character callsigns
> - Add option for changing the map storage location
> - Add warmup[3] and forced delay[4] settings
>
> Notes:
> [1] When pipelined transfers are turned on, D-RATS sends multiple
> packets per "round".  If you have a weak link, more smaller packets are
> likely to yield better performance because a smaller amount of data
> needs to be resent in the event of a corruption.  Smaller packets give
> more overhead though, so fewer larger packets will give better
> performance on good links.  Right now, if you set these such that your
> radio transmits for too long at a time (about 120 seconds) other bits
> might time out, so just be aware of that for the moment.
>
> [2] Eric twisted my arm hard enough that I decided to put this in.  The
> method for doing this on Windows is dumb, takes about a second on a fast
> machine, and can't tell the difference between an in-use port and a
> non-existent port.  However, I don't use windows, so I won't feel the
> pain of this.  Please let me know if this behaves for you.
>
> [3] The warmup is a chunk of data sent before the real data to open up
> the power save on a handheld and potentially "prime" the gateway.  It
> has two settings, the size and the timeout.  The size is the number of
> bytes in the warmup block, and the timeout is how long we have to be
> silent in order to justify a warmup.  If we've transmitted recently, the
> power save on a handheld is probably still open, so setting this lower
> will help save bandwidth.  If you set the timeout to zero, it will
> always send the warmup.
>
> [4] The delay is a relatively un-intelligent delay I added to help
> diagnose some gateway issues.  It's a literal delay period in seconds
> before we start transmitting a series of blocks.  Setting it to a few
> seconds should prevent the ACKs from a file transfer from going out too
> quickly, and thus giving the gateway time to reset.
>
> For reflector use, I expect that setting D-RATS up for smallish blocks
> (~256), longer pipelines (maybe 8 or so), and a delay of maybe 4 seconds
> or so.  Performance will be dismal (compared to wide-open on simplex)
> but I think it might help a bit.
>
> VK3UR, please report back on the map storage location change.
> KD7CAO, please report back on the COM port detection piece.
> N5VRP, please report back on the broken text color selection bit.
>
> Thanks!
>
>