开云体育

Bob,

Changing the mask has done the trick. The accesory decoder can now be
programmed with valid addresses between 1 and 511.

I've also made changes to the Pane Format of the CVPRoducts_AD4.xml
file and would like to submit them to you as Rev 2 but I'm having
trouble with a couple functions; I can't get the vslider or hslider
functions to work. In reference to the following line, can you see
what I'm doing wrong?

<display item="Output_A_On_Time" format="vslider" label="On Time
(1/10th Sec) "/>

Also, as the decoder supports only Direct and Ops Mode Programming,
I've made adjustments in the code that allow use of only those two
modes. Even though Paged and Register Mode are set to "no", I've
noticed that they are still selectable on the bottom of the
programming screen. Why is that? The code has been edited to read;

<programming direct="byteOnly" paged="no" register="no"
ops="yes"></programming>

I assume it to be correct but don't understand why the options are
still available?

Bob

--- In jmriusers@y..., Bob Jacobsen <Bob_Jacobsen@l...> wrote:

Oops.

Sorry, I misunderstood where the address is split between the two

CVs.

In the as-shipped AD4 file, you'll find this:

<variable label="Card Address" item="Short Address" CV="1"

mask="VVVVVVVV">

<splitVal highCV="9"/>
</variable>

That says that this value is to be split between CV1 and CV9, with
the low bits in CV1, and that _8_ bits are to be put in CV1.

That's

at least part of what's wrong.

Could you try changing the mask attribute to

mask="XXVVVVVV"

to put only 0-63 values in CV1?

Bob
--
--------------
Bob Jacobsen (Bob_Jacobsen@l..., 510-486-7355, fax 510-495-2957)
At CERN until August 10, replies may be slow.

Jon,

Now I think I'm getting confused!

Did a little checking and found a good explanation of "the numbers"
on Digitrax's Web; "Each DS54 has a stationary decoder address: 01-
511. Each DS54 has four independent function cells: A, B, C & D.
Each function cell has both output and input wires. Each cell has
its own switch address that controls the device attached to the
function cell. This means the DS54 has 4 switch addresses available
for each stationary decoder address."

If we ignore the fact that the AD4 doesn't have both input and
output wires and concentrate on the last sentence, we'll see
similarity between the DS54 and AD4. Specifically, reference to a
stationary decoder address and switch addresses. When programming an
AD4 accessory decoder, a "Decoder Address" is programmed into CV 1.
In cases where the address of an accessory decoder is greater that
63, CV9 is also used to store part of the decoder address.

Switch numbers, not decoder numbers are utilized to operate each of
the decoders' four function cells. Each decoder has four switch
numbers associated with its assigned address. The key here is not to
confuse a decoder address with switch address. I like to keep it
simple by always referring to the accessory decoders address when
programming is being discussed.

Through a little experimentation, I've found that Digitrax systems
will operate switches numbered greater than 999. However since the
throttle, which appears to be the limiting factor here, will only
address 999 switch addresses, a software solution is required. My
test was performed using RR&Co software on a PC connected to the
Loconet via a LocoBuffer interface, a DB150 and an AD4 programmed to
utilize switch addresses 2037, 2038, 2039 and 2040. Using this
hardware software combination, the above mentioned switch addresses
were operated without problem.

My conclusion is that as long as a means exists to establish
communication with the accessory decoders switch address, the AD4 is
fully functional with our Digitrax systems.

Bob

--- In jmriusers@y..., "Jon Miller" <atsf@i...> wrote:

Bob (black bart type<G>,
Notice also that DecoderPro has two boxes and they both read

the same

address which is not correct. The DCC address should be something

like

793-796 and the card number (address ) should be 199, as an

example. I not

sure what is supposed to read back. Also Digitrax will not

program the full

capacity of this card because it's limited to 999 stationary

decoder output

addresses and the AD4's will do 2044 output addresses.
So a Digitrax command station is only good for card numbers to

249 which

are output addresses 993 to 996.

And I'm getting totally confused about this unit<VBG>.

Also remember that the AD4 can have "ganged" addresses. Like

for

setting up a path through a yard. [The tomorrow file!]

Thanks! I _never_ would have found that.

Apparently, on a Windows machine you get popup menus when you "click
and release"; on a Mac, you "click and hold" to get a popup.

So do we need it on both events so that both the Mac and PC are covered?

Is this worth another installer right away, or can it wait a little
while? I'll probably get to rotating icons, etc, in a few days, can
put them out together then, along with whatever fixes the AD4 needs.

Not for me.

ALex

At 2:56 PM -0700 8/10/02, Robin Becker wrote:

Bob,

Another alternative to consider: just have a righthand and a lefthand
turnout, but allow the user to rotate them once they are on the panel. You
could do the same with blocks, etc. This would cut down on the number of
individual symbols in the list, and I don't think having to rotate the
symbol (from the elusive popup menu <G>) would be a big deal. Just a
thought.

Later on you might want to allow copy operations on symbols already
displayed, and the copy could pick up the orientation as well which would
make it very easy. (assuming of course that the ID can be accessed from the
popup menu)

Those are both great ideas, thanks!

I think I can do those.

Bob
--
--------------
Bob Jacobsen (Bob_Jacobsen@..., 510-486-7355, fax 510-495-2957)
At CERN until August 10, replies may be slow.

At 12:20 PM +0000 8/9/02, original_black_bart wrote:

Bob,

With a clearer understanding of the need, I'd be willing to take on
the task of compiling the decoder information. I assume you'd want to
know the Decoders Manufacturer, Model Number and Version Number(s).
How do you wan the data collected - spreadsheet (Excel), database
(Access) or other means?

Thanks!

I'd like to match the manufacturer name & their model number (i.e. what people actually call the thing) with the values of CV7 and CV8. A specific model might have several values for these, and we should catch as many as we can.

Any way to store them is fine. I use a spreadsheet plus paper notebook. A while back I uploaded the contents to the "database" part of the Yahoo group



that's a convenient way to enter a line or two, or search for a particular model, but it's an inconvenient way to work with the whole thing.

Bob
--
--------------
Bob Jacobsen (Bob_Jacobsen@..., 510-486-7355, fax 510-495-2957)
At CERN until August 10, replies may be slow.

At 11:25 PM +1200 8/9/02, Alex Shepherd wrote:

Found it - the MouseEvent.isPopupTrigger() needs to be checked in the
Released event not the Pressed

I shifted showPopUp() to the Released event and it works as expected

Thanks! I _never_ would have found that.

Apparently, on a Windows machine you get popup menus when you "click and release"; on a Mac, you "click and hold" to get a popup.

For the future-Java-junkies out there, Java Swing handles this by providing a "isPopUpTrigger()" predicate that varies with the look&feel you're using. But you have to make sure to check it everywhere that it might come true. I was only checking it on some mouse events, not all, hence the problem.

Is this worth another installer right away, or can it wait a little while? I'll probably get to rotating icons, etc, in a few days, can put them out together then, along with whatever fixes the AD4 needs.

Bob

--
--------------
Bob Jacobsen (Bob_Jacobsen@..., 510-486-7355, fax 510-495-2957)
At CERN until August 10, replies may be slow.

At 3:26 AM +0000 8/10/02, sixth_rule wrote:

i downloaded Jbuilder
and got the zip file on the sourceforge pagei loaded up the code and
hit the run button and i get 5 erros
all involving this line

static org.apache.log4j.Category log =
org.apache.log4j.Category.getInstance(LocoNetMessage.class.getName());

I just tried it from scratch to see if I could reproduce your problem. The steps are:

download & unzip the jb1.0.3.zip file; that created a jb1.0.3 directory

cd jb1.0.3/java

jbuilder jmri.jpx
(This opened the JBuilder window with a tree of files, etc,
displayed down the left side)

click "rebuild" icon (little wheel with an arrow pointing
to 1's and 0's)

This gave warnings because of the filenames of certain files having been changed. I don't know whether this happened in the upload or in the download. The files it complained about are:

./test/jmri/jmrit/symbolicprog/CombinedLocoSelListP#52D2B.java
which needs to be renamed to:
./test/jmri/jmrit/symbolicprog/CombinedLocoSelListPaneTest.java

./src/jmri/configurexml/AbstractTurnoutManag#52982.java
which needs to be renamed to:
./src/jmri/configurexml/AbstractTurnoutManagerConfigXML.java

./src/jmri/jmrix/loconet/configurexml/LnSecurityElementMan#52B50.java
which needs to be renamed to:
./src/jmri/jmrix/loconet/configurexml/LnSecurityElementManagerXml.java

./test/jmri/jmrix/easydcc/EasyDccTrafficContro#52D5B.java
which needs to be renamed to:
./test/jmri/jmrix/easydcc/EasyDccTrafficControlScaffold.java

./test/jmri/jmrix/easydcc/EasyDccTrafficContro#52D5C.java
which needs to be renamed to:
./test/jmri/jmrix/easydcc/EasyDccTrafficControllerTest.java

I don't know what happened to these files, perhaps some filename-length-limit was hit somewhere.

After these are renamed, quit and restart JBuilder (some versions have a "refresh" option under the "Projects" menu that has the same effect). Clicking rebuild then gave me an expected warning about "Deprecated" features being used, and otherwise completed fine. I've uploaded a screenshot of this to the group photos area




From here, it could run DecoderPro, JBuilder, etc just fine.

In summary, I didn't recreate your problem. It would be helpful to have the exact text of the messages. They sound like what you had happening before, when the libraries were not connected. Could you check whether there's a "JMRILibraries.library" in the directory where you start JBuilder? That directory should look like:

[karl:~/Desktop/jb1.0.3/java] jake% ls-F
CVS/ jmri.jpx src/
JMRILibraries.library* jmri.jpx.local test/
classes/ jmri.jpx~*

(src, test, CVS are directories; classes might not be there if you haven't built yet; you also might not have a jmri.jpx~ file)

If that looks OK, I think we'll need to get some help from other people on the list who use Windows machines.

Bob
--
--------------
Bob Jacobsen (Bob_Jacobsen@..., 510-486-7355, fax 510-495-2957)
At CERN until August 10, replies may be slow.

Oops.

Sorry, I misunderstood where the address is split between the two CVs.

In the as-shipped AD4 file, you'll find this:

<variable label="Card Address" item="Short Address" CV="1" mask="VVVVVVVV">
<splitVal highCV="9"/>
</variable>

That says that this value is to be split between CV1 and CV9, with the low bits in CV1, and that _8_ bits are to be put in CV1. That's at least part of what's wrong.

Could you try changing the mask attribute to

mask="XXVVVVVV"

to put only 0-63 values in CV1?

Bob
--
--------------
Bob Jacobsen (Bob_Jacobsen@..., 510-486-7355, fax 510-495-2957)
At CERN until August 10, replies may be slow.

Bob (black bart type<G>,
Notice also that DecoderPro has two boxes and they both read the same
address which is not correct. The DCC address should be something like
793-796 and the card number (address ) should be 199, as an example. I not
sure what is supposed to read back. Also Digitrax will not program the full
capacity of this card because it's limited to 999 stationary decoder output
addresses and the AD4's will do 2044 output addresses.
So a Digitrax command station is only good for card numbers to 249 which
are output addresses 993 to 996.

And I'm getting totally confused about this unit<VBG>.

Also remember that the AD4 can have "ganged" addresses. Like for
setting up a path through a yard. [The tomorrow file!]

Bob,

Please don't see this as being trouble. For me, it's pure education
and enjoyment!

My Digitrax DB150 won't allow read back, thus the need to use the PR-
1.

With an intended address of 65, the CV pane shows CV7= and CV9=.
Looks like a bug that I think I validated by using the Monotoring
tool to see what data is being sent across the loconet to CV9.
Regardless of the address it looks like CV9 is always being
programmed with a zero. If I program an address of 65, the result is
an address of 1. CV7=1, CV9=0.

Bob

Ok, I'm starting to figure this out, I think. AD4's only program by
"card number" so a card number of say 26 has decoder addresses of 101, 102,
103, and 104. So using the addresses of 1 to 63 in CV521 (CV1) shows me a
card address in DecoderPro, not the output addresses. Remember now that
Card address #64 is skipped, which are output addresses of 253, 254, 255,
and 256.
Now when I put a number in CV521 (CV9) I start getting numbers that
resemble output addresses. Remember I said I put a 10 and a 1 in their
respective CV's and got the number 266. That's starting to sound like an
output address not a card number.

Bob,
I thought everything worked yesterday but today is a different day. I
had no luck with the PR1DOS and MS100. Using DecoderPro it didn't want to
read either but I could change the address. Using the single bit programmer
would change the address everytime but to strange values.
I bought a booklet from Keith when I order the AD4 so I could read
it.CV513 (CV1*) is the and CV521 (CV9*) is the but it says this. "Long"
ard numbers are accessory decoder card numbers above 63. When higher than
63 must be split into two values--sic--- strange split because CV513 only
has 6 bits (NMRA specs). 3 more bits are stored in CV521. Thus 9 bits
gives 511 unique decoder addresses.

I tested and put a 10 in CV1 and a 1 in CV9 and this resulted in an
address of 266, as if I had a 10 and a 4 programmed. Keith uses a special
macro in his command station to do it.

*Digitrax Chief CV's

Jon Miller
AT&SF
For me time has stopped in 1941
Digitrax DCC owner, Chief system
NMRA Life member #2623
Member SFRH&MS

At 9:31 PM +0000 8/10/02, original_black_bart wrote:

To Jon and other testing CVP's AD4 DecoderPro file,

I managed to get a little time this afternoon to modify the screen
format for the AD4. I've run into a little bug and wondered if
someone would be kind enough to program their AD4 to an address
greater than 63, then read the data back and confirm the write.

Based on what I've seen while testing this out this afternoon, any
address greater than 63 is not written properly to the decoder. I
don't think CV9 is being set correctly but I can't tell for sure as
I'm working with a PR-1 to read back my data. Your confirmation of
such would be appreciated.

There may well be a bug there, as that's the new code that I wrote without a way to test directly.

There are a couple of things that might help figure out what it's doing wrong. The "Simple Programmer" from the tools menu will let you read back the value from CV9 to see what it was sent to. Using the "Comprehensive" programmer format will give you a CV pane where you can see the intended contents (although you might not want to trust it, as the bug could have effected that too).

And adding the line:

log4j.category.jmri.jmrit.symbolicprog=DEBUG

to the default.lcf file will generate trace information from the programming operations. (You might want to send that to a file; there are instructions at the top of the file for how to do that)

Sorry for the trouble.

Bob
--
--------------
Bob Jacobsen (Bob_Jacobsen@..., 510-486-7355, fax 510-495-2957)
At CERN until August 10, replies may be slow.

Bob,

Another alternative to consider: just have a righthand and a lefthand
turnout, but allow the user to rotate them once they are on the panel. You
could do the same with blocks, etc. This would cut down on the number of
individual symbols in the list, and I don't think having to rotate the
symbol (from the elusive popup menu <G>) would be a big deal. Just a
thought.

Later on you might want to allow copy operations on symbols already
displayed, and the copy could pick up the orientation as well which would
make it very easy. (assuming of course that the ID can be accessed from the
popup menu)

Robin

To Jon and other testing CVP's AD4 DecoderPro file,

I managed to get a little time this afternoon to modify the screen
format for the AD4. I've run into a little bug and wondered if
someone would be kind enough to program their AD4 to an address
greater than 63, then read the data back and confirm the write.

Based on what I've seen while testing this out this afternoon, any
address greater than 63 is not written properly to the decoder. I
don't think CV9 is being set correctly but I can't tell for sure as
I'm working with a PR-1 to read back my data. Your confirmation of
such would be appreciated.

Bob

i downloaded Jbuilder
and got the zip file on the sourceforge pagei loaded up the code and
hit the run button and i get 5 erros
all involving this line

static org.apache.log4j.Category log =
org.apache.log4j.Category.getInstance(LocoNetMessage.class.getName());

Chad Spickett

Mine is revision 62. I see the data. The 4, 5, and 6 are probably
major revisions and the second digit are minor revisions. Just a guess
until I hear from Keith. Will put a note on the Yahoo site and see what
comes for an answer.
I know it says to be compatible with the Chief it needs to be 5.0 or
greater.

Jon Miller
AT&SF
For me time has stopped in 1941
Digitrax DCC owner, Chief system
NMRA Life member #2623
Member SFRH&MS

John,

Information found in the AD4 Accessory Decoder Manual [See page 5,
Accessory Decoder CV's, CV 519 (7)]indicates software versions 4, 5
and 6 have shipped since the manual was printed in 1999. What
version number do you read via CV7?

Bob

--- In jmriusers@y..., "Jon Miller" <atsf@i...> wrote:

Once I got my head screwed on straight I tested my AD4's.

Little things

like "Direct Mode only" and removing the jumper help<G>. As mine

are for

slow motion machines I don't have the pulse part on my boards.

Programming

seems to be fine. I didn't test the flash rate part.
There is a manufactures ID and version number for AD4's CV's

519 (7) and

520(8). CV8 works just fine and identifies my unit as CVP

products but CV7

doesn't pick up my revision of 62. I guess we would have to check

with CVP

to determine the revision numbers associated with the AD4's, which

I will do

sometime today. Bob (Black Bart) as you are playing with the AD4

file I

will get the information to you as some as I can!

Jon Miller
AT&SF
For me time has stopped in 1941
Digitrax DCC owner, Chief system
NMRA Life member #2623
Member SFRH&MS

Once I got my head screwed on straight I tested my AD4's. Little things
like "Direct Mode only" and removing the jumper help<G>. As mine are for
slow motion machines I don't have the pulse part on my boards. Programming
seems to be fine. I didn't test the flash rate part.
There is a manufactures ID and version number for AD4's CV's 519 (7) and
520(8). CV8 works just fine and identifies my unit as CVP products but CV7
doesn't pick up my revision of 62. I guess we would have to check with CVP
to determine the revision numbers associated with the AD4's, which I will do
sometime today. Bob (Black Bart) as you are playing with the AD4 file I
will get the information to you as some as I can!

Jon Miller
AT&SF
For me time has stopped in 1941
Digitrax DCC owner, Chief system
NMRA Life member #2623
Member SFRH&MS

Bob,

With a clearer understanding of the need, I'd be willing to take on
the task of compiling the decoder information. I assume you'd want to
know the Decoders Manufacturer, Model Number and Version Number(s).
How do you wan the data collected - spreadsheet (Excel), database
(Access) or other means?

Bob
afticarr@...