[Linrad] Re: Linrad MAP65 Communication II

[Joe Taylor <Princeton.EDU; joe@xxxxxxxxxxxxxxxx>]

Hi Rein, Leif, and all,

I have read the recent traffic about Rein's attempts to make the 
Linrad+MAP65-IQ combination work properly together.  It has been a busy 
week for me, so I've not had time to respond before today.

First, let me offer a few preliminaries.

Designing and implementing these software packages has required major 
efforts for Leif and me -- both Linrad and MAP65 represent many years of 
work.  Neither of us is a computer expert, nor a professional 
programmer.  We bring genuine enthusiasm for Amateur Radio and some 
relevant technical knowledge from our earlier professional lives to the 
problems at hand, but for the most part we must figure out the necessary 
computer details by experimentation, trial and error, and lots of 
reading.  We have both learned enough about "sockets" programming to 
implement network communication effectively in Linrad and MAP65, but 
neither one of us would claim to be anything close to an authority in 
this area.  (Obviously I speak mainly for myself here, but I believe 
these comments apply to Leif equally well.)

Speaking for myself: I've gone to great effort to make WSJT easy to set 
up and use.  For most people, program installation is a few mouse 
clicks, and configuration is simple as well.  However, such a level of 
simplicity is not possible for the Linrad+MAP65 combination, mostly 
because of the wide range of radio hardware and computer systems on 
which they can be used.  For example: many different kinds of baseband 
converter; sound cards and/or RF A-to-D converters; real or complex 
sampling; single- or dual-polarization systems; Windows, Linux, FreeBSD, 
Mac OS-X, ... operating systems.  Even if we had "documentation 
departments" to help write instruction manuals (which of course we 
don't), covering all these possibilities would be a huge task. 
Therefore, in many cases individual users must invest significant effort 
to learn what's needed to make Linrad+MAP65 work with their own 
hardware, computer(s), and operating system(s).

Of course it's desirable to have clear instructions available, to help 
new users to get these powerful but complicated programs to work.  But 
the range of computers and OS versions is vast, so detailed instructions 
applicable to a particular system are nearly impossible.

In general, if you want to use a computer for these radio-communication 
purposes you should configure it in something approaching a "no-frills" 
way.  Turn off most visial and audio effects; use settings that optimize 
performance, rather than appearance; optimize scheduling of Windows 
background services; and so on.  In general I do not advise using 
firewalls, virus scanners, and the like.  (If these work for you, and/or 
you know what you are doing, fine; but these programs can cause many 
hassles -- and there are so many possibilities that you are unlikely to 
get any useful help from the likes of Leif or me.)

A few simple cases are relatively easy to deal with.  As far as I know, 
anyone with an SDR-IQ or SDR-14 can get Linrad+MAP65-IQ running under 
Windows with little more effort than required to install and run WSJT. 
A few clicks will install both Linrad and MAP65-IQ, with most everything 
configured and "ready-to-go".

Systems using two computers and/or different RF hardware are somewhat 
more complicated, but nevertheless many people are using Linrad+MAP65 on 
them regularly.  If you can ping each computer from the other one, and 
therefore know their IP addresses, the two-computer network setup is not 
difficult.

Now, on to some particular points raised by others.

>> You did not define what you understand uni 

> ... I never used the word uni. As seen from within Linrad the
> transmission of data is multicasting. Linrad does not know
> whether there is anyone listening at all or whether many
> computers are receiving the data. 

Communication between Linrad and MAP65(-IQ) is one-way, Linrad to MAP65. 
  Linrad sends UDP packets, and does not care what (if anything) happens 
to them.  As I understand it, IP addresses in the range 224.0.0.0 
through 239.255.255.255 are treated as "multicast" addresses by routers 
and the like; packets sent to other addresses (such as that of a 
particular target machine, or the "loopback address", 127.0.0.1, for a 
single-machine system) are not forwarded to addresses where they were 
not intended to be routed.

As far as Linrad is concerned, the only difference between these two 
cases is the requested IP address.  In MAP65, a slightly different set 
of parameters is sent to the system routine "setsockopt" in order to 
receive multicast (as opposed to target-machine-directed) packets.

Maybe I was wrong to use the term "unicast" for the latter situation, I 
don't really know.  The packets are still UDP packets -- as opposed to 
TCP packets, which require acknowledgement of receipt, etc.

>> Believe me Leif, I have been told by others that I have been e-mailing with,
>> both settings to use. For the same situation.
> Now you refer to settings in MAP65 and I do not have any
> idea in what way they differ. I hope Joe is reading this and
> can explain to us:-)

MAP65 receives on port 50004, MAP65-IQ on port 50024.  (The "base ports" 
are 50000 and 50020.)  I used different addresses simply because I once 
wanted to run both programs simultaneously, on the same computer.

>> Leif, I really want to be positive and creative here. Really.
> I appreciate that - but I can not explain why Joe adopted the terminology
> of uni- and multi-casting.

Maybe this usage is wrong.  (See above.)  Please, anyone, suggest 
something different if you know the correct terminology better than I do.

>> The use of Linrad with MAP65 is at this point a much more important 
>> application in amateur cicles I think. There is ton of amateurs on
>> 2 M EME with JT65 these days, virtually around the clock 7 * 24 * 30/(31)
>> They all could use Linrad/MAP and the way people spend effort
>> ( and money ) these days many try to obtain the very best within there means.
>> The nature of EME in a 100 Khz wide band makes finding each other 
>> a difficult problem and time consuming problem , it ties in with the 
>> hated use of the EME loggers etc etc.
>> MAP65 allows to operate without help from loggers, sked lists, the phone.
>> you name it.

> Yes. I have no idea why there is a problem in the first place. I find
> it hard to believe it is within Linrad or its setup procedures.
> 
> It could be the operating system: Visual effects, firewalls, other
> softwares that destroy the latency, whatever. 
> 
> It could be the network itself. Routers, switches, whatever.
> 
> It could also be MAP65. 

I agree with all of the above.  However, I have not found any reasonable 
single-computer or two-computer setup, Windows or Linux, that could not 
run Linrad and MAP65 successfully.  Yes, sometimes it takes a bit of 
messing around to get things configured right.

>> The 512 MB machine runs fine with MAp65, there  are no interruptions of the MAP65
>> data stream reception when the WSJT65B send cycle starts. Linrad keeps on sending
>> data it seems and as said it is working. Except for a minor point.
>> It I open a pull down menu in MAP65 the data steam reception of MAP65 stops
>> as in the previous laptop version when MAP goes in "send"

> Visual Effects may cause such problems.

Yes.  If pulling a menu down causes an interruption of UDP packets, then 
your system is spending *FAR* too much time at doing screen updates at 
high priority.  I don't know how you may have got it configured in such 
a way; but it seems that you are asking for trouble when trying to use 
such a computer for tasks with real-time requirements.

With modern computers Linrad runs easily with only modest CPU demands. 
MAP65 does its heavy computing while decoding, at the end of an Rx 
minute; but it does so at reduced priority, so that real-time CPU needs 
are still easily met.

>> The 2d benchtop is just working! During send Linrad keeps working and 
>> sending audio to the sound card. it keeps doing this even when MAP65
>> goes into send. WSJT tones and receiver audio are both present.
>> Sound card gets obvious signal from 2 sources. Ansd it is not causing
>> any problems.

It is NOT a good idea to have Linrad sending its audio output to the 
sound card used for MAP65 Tx audio.  Your Tx audio will then have sounds 
and squawks from Linrad superimposed on the Tx tones.

> OK. I do not know much about all the softwares that typically
> run in the background under XP. Having both Linrad and MAP65 on
> a single computer is time critical and unforgiving because of
> the way USB behaves. The problem does not exist under Linux....

In my experience, running Linrad and MAP65 on the same computer works 
fine.

>>    That's why I keep on asking about Shared

I don't see that shared disk drives have anything to do with Linrad or 
MAP65 or the transfer of data from Linrad to MAP65.

...

I hope these comments will be of some use!

	-- 73, Joe, K1JT

--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups "Linrad" group.
To post to this group, send email to linrad@xxxxxxxxxxxxxxxx
To unsubscribe from this group, send email to linrad+unsubscribe@xxxxxxxxxxxxxxxx
For more options, visit this group at http://groups.google.com/group/linrad?hl=en
-~----------~----~----~----~------~----~------~--~---


LINRADDARNIL