Experimental Physics and
Industrial Control System

"Ernest L. Williams Jr." <[email protected]> · Wed, 01 May 2002 12:16:32 -0400

Hi,

I just wanted to follow up and mention the need for the VDCT source
distribution to have a build/Make system.
I would really prefer that it be wrapped with EPICS Make system e.g as jca,
jdct, and jprobe are.
We will use VDCT as an EPICS Extension here at the SNS.

In the meantime, I thought I would try using "ant" the java based build
tool.
Well,  ant did not work since the source distribution did not have a
build.xml file.

Any thoughts out there on the VDCT source distribution?  Of course, we are
still using the binary distribution to get started.

Thanks,
*******************************************
Ernest L. Williams Jr.
SNS Control Systems Group, ORNL
701 Scarboro Rd., MS 6473
Oak Ridge, TN 37830
Phone 865-241-9071
e-mail: [email protected]
Fax 865-574-4799
******************************************
----- Original Message -----
From: "Jim Thomas" <[email protected]>
To: <[email protected]>
Cc: <[email protected]>; <[email protected]>; <[email protected]>
Sent: Tuesday, April 30, 2002 9:27 PM
Subject: Re: VDCT - Visual DCT

> >>>>> "John" == John Maclean <[email protected]> writes:
>
>  John> A new build (1228) of VDCT is available from
>
>  John> http://kgb.ijs.si/VisualDCT/
>
> In the interest of saving others some time, plus a couple comments, ...
>
> (environment: W2000SP1, JSE2 1.4.0, VDCT 1232)
>
> It would help if the VDCT page provided a bit more guidance as to the
> minimum things to download for trial (plain binary w/o JAXP and user
> manual, I think?).
>
> The user manual says to point at kgb.vdct.VisualDCT .  The current class
> name seems to be com.cosylab.vdct.VisualDCT ?  Then the usage line should
> be:
>
>      java com.cosylab.vdct.VisualDCT [<DBD>] [<DB>]
>
> , right?  (strings is your friend :-)  Could it be made a bit more
obvious?
>
> In the Edit menu, "Move/rename" might be more obvious if it said
> "mv/rename".  I tried to get it to move an object and wondered "how do it
> know" :-)  (But then I have the same gripe about symed Draw Circle/Arc.)
>
> After creating a few objects (2 ao's and an ai) the create dialog box
seems
> broken.  If I try to create another object of a different type, after I
> select the new type, I can no longer get the cursor to appear in the Name
> field.  Fortunately, it has switched types so clicking Cancel and trying
> again does work, but ...  Anyone else have this problem?
>
> It seems to get a bit confused with where the connecting "wires" should
> go.  Make an ai.  Make an ao to the left of it.  Connect ao.OUT to ai.VAL
> .  Make another ai to the left of the ao, lining up the tops of the two
> symbols, and putting them fairly close together (I've got 2 grid dots
> showing).  Then hook ao.DOL to ai2.VAL .  The "wire" I get comes across
> from ai2 almost all the way to ao , then down and seemingly to the left
> through the NPP NMS, as if ai2 really wanted to be further to the left.
>
> When I printed the result, the leftmost vertical line was dropped, both on
> page preview and on paper.  I think this is because the line was off the
> window edge.  The preview mode was "fit to paper".  Should this perhaps
> read "fit view to paper"?  Should there also be a "fit drawing to paper"?
> Why is the default preview scale 25%?  For speed?
>
> Overall, in spite of the above comments, it looks really nice.  But.....
it
> really really needs text blocks for annotations.  The user manual says
> that's on the list.  I'm hesitant to start using it without being able to
> comment things visibly on the drawing.  There doesn't even seem to be any
> way to label a whole page.
>
> What is the intention for printing hardcopies?  I'd like something that
> came out at a predictable scale/size.  CapFast does that by borders and
fit
> drawing to paper.  Since there are no borders here there's nothing
> equivalent.  I just tried mode 1:1, and it previewed my drawing on 4
> sheets.  Those 4 did not seem to have any relation to the drawing area
> shown on the navigator inset.  IMHO things need to be predictable to work
> well - either by being "obvious" or by providing some kind of training to
> clue me in to what to expect :-)
>
> Next I tried scale and setting 200%.  That doubled the drawing scale as
> expected.  I switched back to "fit to paper".  The 200% field
specification
> disappeared, but that scale is still being used^H^H^H, no it was used once
> on "fit to paper", but I can't reproduce it.
>
> In the Save As dialog, most Windows programs add the "type"/"extension"
> automatically (even if one doesn't want it :-) .  Here one has to type
> foo.db to get it named .db .
>
> Thanks for listening,
> Jim
>
> =================================== 8< -----------------------------------
>
> #! Generated by VisualDCT for Java v2.1
>
> record(ao,g1:ao2) {
>   field(OUT,"g1:ai1.VAL")
>   field(DOL,"g1:ai2.VAL")
> }
>
> # main analog output
> record(ao,g1:ao1) {
>   field(SCAN,"10 second")
>   field(FLNK,"g1:ai1")
>   field(OMSL,"supervisory")
> }
>
> # my analog input
> record(ai,g1:ai1) {
>   field(DESC,"analogue guy")
>   field(FLNK,"g2:ai1")
> }
>
>
> record(ai,g1:ai2) {
> }
>
> #! Further lines contain layout data used by VisualDCT
>
> #! Group(g1,2300,2460,0,"")
> #! Record(g1:ao2,2280,2540,0,1,"g1:ao2")
> #! Field("g1:ao2.OUT",0,1,"g1:ao2.OUT")
> #! Link("g1:ao2.OUT","g1:ai1.VAL")
> #! Field("g1:ao2.DOL",0,0,"g1:ao2.DOL")
> #! Link("g1:ao2.DOL","g1:ai2.VAL")
> #! Record(g1:ao1,2280,2360,0,1,"g1:ao1")
> #! Field("g1:ao1.FLNK",0,1,"g1:ao1.FLNK")
> #! Link("g1:ao1.FLNK",g1:ai1)
> #! Record(g1:ai1,2640,2420,0,0,"g1:ai1")
> #! Field("g1:ai1.VAL",0,0,"g1:ai1.VAL")
> #! Field("g1:ai1.FLNK",0,1,"g1:ai1.FLNK")
> #! Record(g1:ai2,2060,2540,0,1,"g1:ai2")
> #! Field("g1:ai2.VAL",0,1,"g1:ai2.VAL")
>

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System