Re: [PATCH] kenwood drivers updates (into the bad)

Hi Alessandro, hello also to all others.

Sorry for the late answer but there was no time to test until today. I could
also only do some basic tests until now.

Let me report the state of the TS570 backend (which I can really test) and
give some comments to the kenwood backends at all.

The applied patches have done no good to the state of the rig and broke some
code. I got a lot of regressions. Lets see in more detail:

- get_rit is now broken. It reports the last engaged rit setting even if
  RIT is totally turned of. 

- At the moment there are new init and cleanup functions introduced. But in
  the ts570.c file they are only applied to one of both rigs in there.

- set_ant does not work properly. It switches the antenna but reports
  an error as it is waiting for an answer which will not be sent by the rig.

- Setting PTT does also fail. It switches TX and RX properly but the rig 
  responds always with a '?;' and the function reports an error in command
  execution.

The last two problems seems to affect all kenwood rigs.

I know these problems can be fixed and tested easily, but I fear that are
not the only ones.

You changed some other backend rigs besides your ts450. If you had not done

Configuration FRG 8800

------------------------------------------------------------------------------
This SF.net email is sponsored by:
SourcForge Community
SourceForge wants to tell your story.
http://p.sf.net/sfu/sf-spreadtheword

_______________________________________________
Hamlib-developer mailing list
Hamlib-developer <at> lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/hamlib-developer

Re: [PATCH] kenwood drivers updates (into the bad)

On Sun, 1 Feb 2009 20:16:24 +0100
Thomas Beierlein <tb <at> forth-ev.de> wrote:

> Let me report the state of the TS570 backend (which I can really test) and
> give some comments to the kenwood backends at all.
> 
> The applied patches have done no good to the state of the rig and broke some
> code. I got a lot of regressions. Lets see in more detail:

 Did you tested the CVS or the CVS plus the patches in the queue?

> - get_rit is now broken. It reports the last engaged rit setting even if
>   RIT is totally turned of. 

 will check.

> - At the moment there are new init and cleanup functions introduced. But in
>   the ts570.c file they are only applied to one of both rigs in there.

 The CVS has them on both rigs in ts570.c

> - set_ant does not work properly. It switches the antenna but reports
>   an error as it is waiting for an answer which will not be sent by the rig.

 good catch, will fix.

> - Setting PTT does also fail. It switches TX and RX properly but the rig 
>   responds always with a '?;' and the function reports an error in command
>   execution.
> 
> The last two problems seems to affect all kenwood rigs.

 fixed, thank you.

> I have no good idea how to fix the problems in short time to get ready for the
> next hamlib release (1.2.9). Maybe a rollback of all that and only changes to
> the backends, which you can really test, is the only quick solution to get a
> stable release.  Then all other fixes, cleanups and so on should be done
> stepwise afterwards without a hurry.

 If required, I could easily commit a rollback.

> One problem is that I am not able to do much work here in next 6..7 weeks as I
> will be away for some weeks. And it seems we both are the only ones who does
> work on kenwood rigs at the moment.

 Rob and Stephane are doing tests too

--

-- 

 Best regards,

 Alessandro Zummo,
  Tower Technologies - Torino, Italy

  http://www.towertech.it

------------------------------------------------------------------------------
This SF.net email is sponsored by:
SourcForge Community
SourceForge wants to tell your story.
http://p.sf.net/sfu/sf-spreadtheword

Re: [PATCH] kenwood drivers updates (into the bad)

On Sun, 1 Feb 2009 20:16:24 +0100
Thomas Beierlein <tb <at> forth-ev.de> wrote:

> - get_rit is now broken. It reports the last engaged rit setting even if
>   RIT is totally turned of. 

 regarding this point, Kenwood rigs have two xit/rit related controls. One
 is the frequency and the other is if it's engaged or not.

 the correct behaviour is to have both of them in hamlib. the old code
 was using a trick of returning 0 when not engaged. This is obviously
 wrong because the user should not be forced to engage it in order
 to set the frequency.

 This is, of course, a behavioural change, but the original implementation
 is not coherent with the rig's controls.

 I will add the the proper controls with ext_parms.

 other than that, I believe we have addressed all the open issues, so I will be committing
 the other patches in the queue, so that interested people can test it: it cannot do any
 further damage.

 If requested by the project managers, the recovery plan for a stable release is to
 branch the repository at a know working date and merge from the trunk any non-kenwood
 commit. I already gave a look at the procedure and involved files and should be able
 to easily complete it. I will then, obviously, keep the trunk up-to-date with any change
 to the stable branch until released.

--

-- 

 Best regards,

 Alessandro Zummo,
  Tower Technologies - Torino, Italy

  http://www.towertech.it

------------------------------------------------------------------------------
This SF.net email is sponsored by:
SourcForge Community
SourceForge wants to tell your story.
http://p.sf.net/sfu/sf-spreadtheword

[tb <at> forth-ev.de: Re: [PATCH] kenwood drivers updates (into the bad)]

On Mon, Feb 02, 2009 at 09:24:01PM +0100, Alessandro Zummo wrote:
> On Sun, 1 Feb 2009 20:16:24 +0100
> Thomas Beierlein <tb <at> forth-ev.de> wrote:
> 
> > - get_rit is now broken. It reports the last engaged rit setting even if
> >   RIT is totally turned of. 
> 
>  regarding this point, Kenwood rigs have two xit/rit related controls. One
>  is the frequency and the other is if it's engaged or not.

I do know about that. Thats why I implemented the former version which
respects both of these settings.
>  
>  the correct behaviour is to have both of them in hamlib. the old code
>  was using a trick of returning 0 when not engaged. This is obviously
>  wrong because the user should not be forced to engage it in order
>  to set the frequency.

Sorry, but I think I have to disagree. From the users point of view if he ask
for the actual setting of RIT (or XIT) he wants to know which value is really
in effect. And then get_rit should report a value of 0 without regard if it is
so because RIT is switched off or because RIT is on and set to zero. That is
the same for him.

Same goes for set_rit: If he choose a RIT value of 1 kc RIT has to be switched 
on and set to 1 kc. If he request a value of 0 it is locigal correct to switch
it off.

73, de Tom DL1JBE.

--

-- 
"Do what is needful!"
Ursula LeGuin: Earthsea
--

------------------------------------------------------------------------------
Create and Deploy Rich Internet Apps outside the browser with Adobe(R)AIR(TM)
software. With Adobe AIR, Ajax developers can use existing skills and code to
build responsive, highly engaging applications that combine the power of local
resources and data with the reach of the web. Download the Adobe AIR SDK and
Ajax docs to start building applications today-http://p.sf.net/sfu/adobe-com

Re: [PATCH] kenwood drivers updates (into the bad)

On Tue, 3 Feb 2009 15:45:21 +0100
Thomas Beierlein <tb <at> forth-ev.de> wrote:

> Sorry, but I think I have to disagree. From the users point of view if he ask
> for the actual setting of RIT (or XIT) he wants to know which value is really
> in effect. And then get_rit should report a value of 0 without regard if it is
> so because RIT is switched off or because RIT is on and set to zero. That is
> the same for him.
> 
> Same goes for set_rit: If he choose a RIT value of 1 kc RIT has to be switched 
> on and set to 1 kc. If he request a value of 0 it is locigal correct to switch
> it off.

 That means I have to re-set the xit/rit every time I want to activate it,
 while I might want it to remember the last value, which is what it's done
 if your manually operate the rig.

 That's way you have two controls on the rig instead of only one.

--

-- 

 Best regards,

 Alessandro Zummo,
  Tower Technologies - Torino, Italy

  http://www.towertech.it

------------------------------------------------------------------------------
Create and Deploy Rich Internet Apps outside the browser with Adobe(R)AIR(TM)
software. With Adobe AIR, Ajax developers can use existing skills and code to
build responsive, highly engaging applications that combine the power of local
resources and data with the reach of the web. Download the Adobe AIR SDK and
Ajax docs to start building applications today-http://p.sf.net/sfu/adobe-com

Re: [PATCH] kenwood drivers updates (into the bad)

On Tue, 3 Feb 2009 15:55:36 +0100
Alessandro Zummo <azummo-lists <at> towertech.it> wrote:

>  That's why you have two controls on the rig instead of only one.

 And, btw, that's why Kenwood implemented two commands,
 one to set the level and one to turn it on or off.

--

-- 

 Best regards,

 Alessandro Zummo,
  Tower Technologies - Torino, Italy

  http://www.towertech.it

------------------------------------------------------------------------------
Create and Deploy Rich Internet Apps outside the browser with Adobe(R)AIR(TM)
software. With Adobe AIR, Ajax developers can use existing skills and code to
build responsive, highly engaging applications that combine the power of local
resources and data with the reach of the web. Download the Adobe AIR SDK and
Ajax docs to start building applications today-http://p.sf.net/sfu/adobe-com

Re: [PATCH] kenwood drivers updates (into the bad)

------------------------------------------------------------------------------
Create and Deploy Rich Internet Apps outside the browser with Adobe(R)AIR(TM)
software. With Adobe AIR, Ajax developers can use existing skills and code to
build responsive, highly engaging applications that combine the power of local
resources and data with the reach of the web. Download the Adobe AIR SDK and
Ajax docs to start building applications today-http://p.sf.net/sfu/adobe-com

_______________________________________________
Hamlib-developer mailing list
Hamlib-developer <at> lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/hamlib-developer

Re: [PATCH] kenwood drivers updates (into the bad)

On Tue, 3 Feb 2009 11:47:47 -0500
Martin Ewing <martin.s.ewing <at> gmail.com> wrote:

> The fact that Kenwood provides two commands does not mean that Hamlib needs
> to do so. I note that the frontend code says the following for rig_set_rit:
> "Sets the current RIT offset. A value of 0 for rit disables RIT."   It seems
> to mean that Hamlib rolls the RIT on/off  together with the freq. setting,
> and this is what backends "ought" to do.  (You can ask what is the
> difference between RIT=0 & RIT "on" and RIT=0 & RIT "off".  On the TenTec
> Orion, the difference is that the RIT LED indicator is on or off!)

 The same is on the kenwood. you have XIT and RIT indicators on the display
 plus the frequency.

> This is a discussion we have had for some time:  Is it our purpose to
> provide access to all the features of each rig, or should we provide a
> standardized "virtual rig" API so that application writers do not have to
> tailor their code to support many different rigs?

 I believe the api should be as standardized as possible but also as wide
 as possible.

 With two commands you allow the frontend to present an option
 in any comfortable way, by forcing only one you leave no options at all.

 
--

-- 

 Best regards,

 Alessandro Zummo,
  Tower Technologies - Torino, Italy

  http://www.towertech.it

------------------------------------------------------------------------------
Create and Deploy Rich Internet Apps outside the browser with Adobe(R)AIR(TM)
software. With Adobe AIR, Ajax developers can use existing skills and code to
build responsive, highly engaging applications that combine the power of local
resources and data with the reach of the web. Download the Adobe AIR SDK and
Ajax docs to start building applications today-http://p.sf.net/sfu/adobe-com

Re: Hamlib PTT for TS-570

On Sat, 31 Jan 2009 08:54:53 -0800
Rob Frohne <frohro <at> wallawalla.edu> wrote:

> One thing that probably doesn't bear on the present question, but in 
> case it does, I'm going to mention it. It is that I need to use RTS/CTS 
> for my  hamlib control cable.  I have them wired together at present, 
> which is probably why I had to set the timeout on Kenwood rigs 

 This might or might not be the problem, but it will lead to all sort
 of glitches, so I'd suggest to fix it.

> excessively long in hamlib, and that should probably be reversed, but 
> someone (probably me) needs to check and see if using RTS/CTS correctly 
> solves the problem of the messups when you are tuning for example with 
> VFO while rig commands are happening.  (The TS-850 at least won't send 
> or receive rig control commands while the VFO is spinning, and also 
> likely some other front panel interface things are happening).

 According to the docs, the rig is supposed to slow down the serial operations
 while the dial is operated. If kenwood implemented it correctly, 
 rtc/cts is used to tell the host to wait.

 you might want to check the cvs for a couple of kenwood and ts850 related fixes
 (ptt, ctcss)

 Newer rigs have separate PTT commands to operate the ACC2 PTT, so it's credible
 that on the 450, 690 and 850 it must be done via ACC2 PTT.

 I might be wrong.. if you find that you can do it with any windows software,
 we can just snoop it so see how it's done.

 I was thinking about programming a microcontroller to listen for commands
 on the same serial line as the main micro and operate a few missing functions,
 like TS450's NB, PROC, AGC, NOTCH and ACC2's PTT. 

--

-- 

 Best regards,

 Alessandro Zummo,
  Tower Technologies - Torino, Italy

  http://www.towertech.it

------------------------------------------------------------------------------
Create and Deploy Rich Internet Apps outside the browser with Adobe(R)AIR(TM)
software. With Adobe AIR, Ajax developers can use existing skills and code to
build responsive, highly engaging applications that combine the power of local
resources and data with the reach of the web. Download the Adobe AIR SDK and
Ajax docs to start building applications today-http://p.sf.net/sfu/adobe-com