[Linrad] Re: Linrad MAP65 Communication II

[Leif Asbrink <sm5bsz.com; leif@xxxxxxxxxxxxxxxx>]

Hi Rein,
  
> I started this with the intention to use 2 computers and drifted
> eventually towards this solution with one computer..
OK. But two computers should actually be easier since there 
would not be any timing issues.

> Could not get the 2 computers to work together even after the
> introduction of the par_netsend_ip parameter.
Did you really write the address of the computer inside which 
you run MAP65 into the file?

> The introduction was via Joe's MAP65 I-Q package.
OK.
> 
> I am afraid the present problem is a timing problem in particular
> as I it working sort of in one of perhaps 15 trials.
> Shortage in CPU or memory.
Yes. Windows will use its swap file without giving any warning
and that may cause various problems. You can set all storage times
to minimum to minimize the memory used by Linrad.

> Here are a few more questions:
> 
> Given this 1 computer setup, should I select uni- or multicast?
Linrad does not give any choice. 

> You say Linrad is in multicast as default.  How could I have known
> this?
Start at SM5BSZ Home Page:
http://www.sm5bsz.com/index.htm
http://nitehawk.com/sm5bsz/index.htm
http://g7rau.demon.co.uk/sm5bsz/index.htm

On this page there are two links
Linrad <http://www.sm5bsz.com/linuxdsp/linrad.htm>
Linrad for newcomers <http://www.sm5bsz.com/linuxdsp/usage/newco/newcomer.htm>

The newcomer page has a link to <http://www.sm5bsz.com/linuxdsp/linrad.htm>
the same as the first link on the SM5BSZ Home Page.
The Linrad link goes to "Linrad Home Page" and here you should
search for network. The word network occurs two times on the page and
the second occurance is this:
LINRAD NETWORK <http://www.sm5bsz.com/linuxdsp/run/network.htm> 

That page starts like this:
General

Linrad uses the network to make raw or processed data available 
to other programs that may run on the same computer or on any 
other computer in a local network.

Linrad can multicast on IP addresses (groups) 239.255.0.0 to 
239.255.0.15 This means that several computers can multicast 
their data to the network simultaneously in different groups.

The Linrad base port has to be set between 50000 and 65000 
in steps of 10 and Linrad will send data to a port with an 
offset with respect to the base port that depends on the 
data format. 
 
> The only place I saw the " _cast" parameter selectable is in MAP65.
Yes. The Linrad network was not intended for MAP65 originally 
but for use in a multioperator contest station. I got 
enquieries about the use of two computers on the same antenna. 
Would it be possible to connect Delta44 cards in parallel or 
would one need two RX2500 units. (The answer is perhaps. 
There could be a significant increase in the noise floor but 
a really thick and short wire between the two Delta44 might 
help.) The network was originally implemented as a solution
to this problem.

> ( have been given by others no conclusive answers.,( also, it has come
> up in most of the exchanges)
Well, describing all options that would be possible and that
Linrad does not have would be quite a challenge. I do not know
enough about networking to have any idea about how to implement
a unicast solution because a pairwise computer communication 
is normally "error free" which means that the Linrad computer 
would hang if the map65 computer gets too busy. 

I think the terminology is confusing and presumably misleading.
What Linrad does is to send data of type SOCK_STREAM to the network
socket. The choice is based on this information:

SOCK_STREAM
Provides sequenced,  reliable,  two-way,  connection-based  byte
streams.  An out-of-band data transmission mechanism may be supâ
ported.

SOCK_DGRAM
Supports datagrams (connectionless,  unreliable  messages  of  a
fixed maximum length).

SOCK_SEQPACKET
Provides  a  sequenced,  reliable, two-way connection-based data
transmission path for datagrams of fixed maximum length; a  conâ
sumer is required to read an entire packet with each read system
call.

SOCK_RAW
Provides raw network protocol access.

SOCK_RDM
Provides a reliable  datagram  layer  that  does  not  guarantee
ordering.

SOCK_PACKET
Obsolete  and should not be used in new programs; see packet(7).

Read about multicast here:
http://en.wikipedia.org/wiki/Multicast

It seems to me that the transmission would be unicast if only one
computer joins the group (=the IP address to which Linrad transmits
the datagrams) while it becomes multicast if more computers join.
The transmitting Linrad does not get any information from MAP65
or or slaves as a response to the datagrams. Linrad does however
have a server to which other computers can send inquieries about
what calibration function to use etc. Linrad slaves use it for
unicast back to the master to say at what frequency they listen.  

> My interpretation was with a 2 computer system uni- 
> one communicates with 1 computer.
Yes. By only having two computers the communication could be called
unicast if one accepts the definition given by the illustrations
on wikipedia.

> multi_ with more than one "slave" computer.
Yes. But the difference is in the network. Not in Linrad.
When using SOCK_DGRAM one does not know if there is one listener
(=unicast) or more than one (=multicast) or wether there
is no listener at all. 

> For the same money, one could say uni- with a single
> computer, multi- with more than 1 computer in the system.
I do not know. It would be nice if someone who really knows
the terminology for networks would write a text about networks
with the relevant information for Linrad users that I could 
put at the top of the Linrad Network page.
 
> I would also think one would need more than one IP or computer
> name in order to communicate with one than one "slaved computer"
> 
> With just 2 computers waht should it really be?
I do not understand this question.

> SM0ERR is running linrad on a corporate system.
> 
> How does he handle the IP issue without a par_netsend_ip
> and or par_netrec_ip parameter?
Oooh! He does use these files. What he wrote was his response
when I introduced these files.

> This question generated in my mind because you say
> the introduction of a blank par_net*_ip is not such a
> good idea.
I can not follow your argument here.

Mart has created a par_netsend_ip on the computer that is connected
to the radio hardware. In that file he has written an IP address
that is legal within his corporate network. Perhaps the IP
address of the computer he has in his office, perhaps some other
address.

He has then a file par_netrec_ip with the same address in each
computer that he wants to use for the Linrad slave. He might
have one in his office and another in the lab and perhaps even one
at home. I actually do not know - but perhaps he will send
a mail to the list and give some info if he reads this:-)

> I hope the answer is not: "he is running Linux and tunes his files
> and the program directly to make at  work.?"
No No. Linrad is the same in Linux and Windows.
Before I added the par_xxx_ip files Linrad users had to change
the IP address in the source code in network.c and recompile
in order to use MAP65. That was the same under Linux and Windows. 
Just two commands "configure" and then "make":-)

> It is hard for me to imagine that he is able and to get away with
> this in a corporate environment with out REALLY understanding
> networking. Or, more likely that he is the corporate top IT person
> himself!
It is not complicated at all. I think you have been confused
by far too much information....

> I still like to hear from somebody whether in a 2 computer system
> I should have "shared hard drives" for his?
There is absolutely no reason that I can find out. You can
have it if you want to have access to files that are stored on the
other computer. I do not know to what extent that might cause
loss of data for MAP65. 

> More to the point,  WHY or why not.
For Linrad and MAP65 or Linrad slaves there is absolutely 
no reason.

> Getting back to your idea of a problem in the USB start/stop
> data flow. You are right I am sure.
> 
> 
> In using the parameters from Joe's MAP65 I-Q fresh loaded
> for SSB mode only.
> 
> I start Linrad, S and U, are done by Joe;s parameters
> 
> I press D, Linrad is working ( stand alone )
OK.

> The next step is real confusing to me, given that the program
> started with "netsend on" and it is suggested to do the
> saving "W"
???????

The W means that you have changed one of the computer related
parameters. (One of those in par_userint) That will happen if
you use R, S, T, U, V or N in the main menu. Linrad assumes
something is changed as soon as you enter one of these functions
and shows the W message even if you actually did not change 
any parameter value.

> All I can do is, toggle T to remove the netsend on message,
> toggle T again and do a W and then click D
No. You have choices:

1) You can ignore the warning which means that the changes you 
   made will be valid for the current Linrad session but next 
   time you start Linrad it will use the unchanged parameters 
   from your par_userint file.

2) Just press W, then D.
 
> Thinking make sure it is in "T" and  then save it.
> 
> One point of the confusion is that linrad comes in the
> netsend on mode?
Why is that? 

Step 1: Exexcute the selfextracting file that Joe supplies.
Step 2: Click the Linrad icon that now appears on the desktop.
Step 3: Press D.

That is all if you run MAP65 in the same computer.


> Anyway. Next is I start MAP65 I-Q and hope to not
> to see "no RX data" and see a real dB number in the
> noise indication window.
> 
> It leaves the selection of the cast mode in MAP65
> as well as the setting of the diagnostics also in
> the MAP65. program.
Presumably you should just start the program.

> I hope somebody can tell me: you are all wrong it
> should be this way.
> 
> Another issue is that the above is working most of
> the time but not 100% solid,  I getting "no RX data"
> at times.
You mean "No input" in Linrad?

> I then start over. I have not bern able to discover
> a way to correct the data stream issue with one or
> 2 key strokes.
When USB hangs, the only cure is to exit from Linrad.
It should never happen though. Do you know what other
programs are running? If you have visual effects
enabled (Default in Windows) you will easily overload 
the CPU in case your videocard does not have appropriate 
hardware acceleration.

> Next question
> 
> Once MAP65 goes into transmit ( generating
> tones via the sound card ( selected in MAP65 )
> and also selected earlier for Linrad ( by Joe's
> parameter file set.)
> 
> Should at that point the USB data stream form the
> SDR I-Q box into linrad stop or not?
The USB data stream should not stop. Linrad does not
know that MAP65 is listening. 

> The incoming data are useless at that point anyway.
Hmmm, you can use it to see if you are transmitting
at the desired frequency and you can use Linrad to monitor
your frequency drift. You could also connect your receiver
to a directional coupler during transmit and use the
Linrad S-meter to check your WSVR. 

> Should at the end of the send period in MAP65 the
> USB data stream resume?
NO. USB should run all the time.

> Suppose I did away with the USB SDR I-Q all togheter,
> 
> And change back to the delta44 what should happen
> then in this respect ( linrad keeps working during the
> send period? I have removed the USB factor now?
That should solve the problem. You could also connect
a second computer for MAP65.

> Going to put MAP65 I/Q on 3 other XP bench tops
> and see what happens in transmit.
You mean Linrad + MAP65-IQ ?

> In conlusion, I get the impression of being an idiot having
> to ask all these questions in the presence of 150 other
> users!
It does help:-)
Your questions have lead to one improvement in the Linrad setup
dialogue already. Hopefully there will be someone who will 
help by writing a suitable introduction to networking for
Linrad users to be placed on the Linrad network page....

73

Leif




--~--~---------~--~----~------------~-------~--~----~
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