xcircuit-dev — XCircuit news and discussion

[Xcircuit-dev] An issue with text justification

[Svenn A. B. <sv...@bj...>]

Hi,

I have a happy hackin keyboard so I don't have the key pad. I have 
noticed when justifying text that if I change the top/middle/bottom 
property and the text is anything else but left justified, then it 
gets left justified.
-- 
Svenn

[Xcircuit-dev] name of page in page overview does not fit

[Svenn A. B. <sv...@bj...>]

HI,
I go to a page and save the page to a new name by changing the page 
and press apply. If I now hit Shift-p and go to the overview, the 
name of the page has not changed.


-- 
Svenn

[Xcircuit-dev] Problems with the bounding boxes in eps

[Svenn A. B. <sv...@bj...>]

Hi,

When writing one single page to eps, then the local pin labels get cut 
off if they are outside of a virtual box that contain all normal 
labels and lines.

-- 
Svenn

[Xcircuit-dev] RE: XCircuit netlist warnings

[R. T. E. <tim...@mu...>]

Dear John,

> I sometimes inadvertently duplicate my part numbers.  When this happens
> and I write a spice netlist from xcircuit, I see the appropriate warning
> messages.
>
> However if I choose to netlist again(without attempting to fix these
> errors), the warning messages go away.

The problem is that the warnings are generated when xcircuit creates
the netlist, not when writing it.  If nothing has changed in the
circuit, xcircuit will not re-create the netlist when writing netlists.

This could easily be corrected by setting an option to force the
netlist to be regenerated before all netlist writes.  I would also
find that useful, as I sometimes find myself testing a new netlist
feature by moving one wire back and forth to force the program to
regenerate the netlist.

					Regards,
					Tim

[Xcircuit-dev] More uses for XCircuit.

[Philip <sil...@us...>]

Okay, we all know how good XCircuit is for making schematic drawings.

And if we have browsed the website we also know it is good at music
typsetting/notation.

I use it _quite_ heavily for my non-dimentional (that is, not CAD)
drawing for audio/video system layout. Block diagram kind of things.

Well, the other day I found a new use

Some of the other family members were working on a cross-stitch project
and wanted to include the outline of a flower that didn't exist in the
pattern. Of course, they came to me for a clever solution! *grin*

First a bit of background. Cross-stitch is made using a needle and
thread on a grid-like material. The pattern is made of grid-line. The
tricky bit is that a cross-point on the pattern corresponds to a _hole_
on the material! Once I understood that the rest was easy.

I fired up XCircuit, set the grid quite coarse (1/3 inch, or so) and
proceded to connect gridpoints until I got a flower petal shape that the
artists approved. Copying, rotating, repositioning, coloring, etc the
petals until it looked like a flower was easy - and, I'll admit, a bit
fun.

I printed the final version, the project was finished, and now it's on
it's way to Grandma's!

Thank you Tim for such a _killer_ app!!

          Philip, "more sold-er than ever!"

   P.S. The cross-stitchers say "Thanks" too.

[Xcircuit-dev] duplicate part number warning message

[John D B. <sea...@ya...>]

Hi,
I sometimes inadvertently duplicate my part numbers.
When this happens and I write a spice netlist from 
xcircuit, I see the appropriate warning messages.

However if I choose to netlist again(without
attempting to fix these errors), the warning messages
go away.
If I fix one of the errors, the remaining messages
reappear after netlisting.

It would be great if this inconsistency could be
remedied.

Thanks






		
______________________________________________ 
Renovamos el Correo Yahoo!: ¡250 MB GRATIS! 
Nuevos servicios, más seguridad 
http://correo.yahoo.es

[Xcircuit-dev] compiling 3.3.11

[Uwe Z. <uwe...@sc...>]

Has anyone managed to compile the latest versions (3.3.10, 3.3.11)
with tcl under either Cygwin or SUSE v9.2?

If so: what am I doing wrong???

Under Cygwin I have compiled the tcl and tk libraries according to the
description on the homepage. Everything works fine, yet ./configure
still reverts to the non-tcl compile.

But even this does not go through. The next error message is an
missing library X11. Even though libX11 appears to be available in the
current search path.

I don't remember the error messages from the SUSE system at work -
I'll have to check on monday...

  Uwe.

[Xcircuit-dev] RESOLVED an old startup problem ("missing colon in line 1:")

[Romano G. <ro...@de...>]

Tim, xcircuit users,

    I do not knw if you remember me, but I have found (testing xcircuit
3.3.10) from where derived the error saying

        loading history file ... 4 events added
        error in slave eval:
        missing colon on line 1
        Main console display active (Tcl8.4.5 / Tk8.4.5)
        (romano) 5 %

in the tcl console. The problem derived from a comment of the style

        /* something */

in the .Xdefaults file. It seems that tcl could not manage C-style comments
in .Xdefaults! I just removed them and it seems that all is working again.

I had another problem in 3.3.2 (no printing) which has desappeared, too. 

Great work!

Thank you

-- 
Romano Giannetti             -  Univ. Pontificia Comillas (Madrid, Spain)
Electronic Engineer - phone +34 915 422 800 ext 2416  fax +34 915 596 569

[Xcircuit-dev] Re: failing to compile 3.3.10 with tcl under cygwin

[The L. S. <lig...@rd...>]

I am also having this same problem.  For some reason XCircuit doesn=92t =
seem to like Cygwin.  It compiles fine without tcl however.

=20

Thanks,
The Lightning Stalker


--=20
Internal Virus Database is out-of-date.
Checked by AVG Anti-Virus.
Version: 7.0.300 / Virus Database: 266.5.0 - Release Date: 2/25/2005
=20

[Xcircuit-dev] failing to compile 3.3.10 with tcl under cygwin

[Uwe Z. <uw...@gm...>]

Hi Tim,

I am trying and constantly failing to compile 3.3.10 with TCL under
CygWin. I followed the instructions down to the letter, but always
./configure reverts to the non-TCL variant.

I wanted to try 3.3.10 because the pre-compiled Cygwin version often
fails with a segmentation fault when loading one ore more of my
schematics.

What am I doing wrong (except for not having a native Linux at hand)?

   Uwe.

[Xcircuit-dev] Re: the main menu doesn"t work

[Mark C. <mar...@gm...>]

It's probably the same bug listed here:
http://xcircuit.ece.jhu.edu/trouble.html

Quoted:
Menu Problems

   1. Problem: I can't get any menus!
      Solution: This seems to be a perennial bugaboo with the Xw
widget set. The menu buttons won't respond when Num Lock is on.
Fortunately, this can be cured by turning Num Lock off. One permanent
solution is to use the Tcl version of xcircuit (available in stable
form since xcircuit version 3.1 revision 3, February 10, 2003), which
does not use the Xw widget set.


-- 
Mark L. Chang

[Xcircuit-dev] the main menu doesn't work

[belahcene <bel...@ia...>]

Hi every body,
I ve just installed the  xcircuit , ( from sarge debian site)
but the menu on the top bar doesn't give any thing.
It open the main window with in right side the tools , but no posibility 
to save file, or to read ??
best regards
bela

[Xcircuit-dev] Re: SIM/SPICE flattened extraction bug

[Mark C. <mar...@gm...>]

That would be awesome. You, sir, are a great man. Save me from writing
a spice to sim converter for hierarchical spice!

Thanks again,
Mark


On Fri, 18 Feb 2005 12:37:28 -0500, R. Timothy Edwards
<tim...@mu...> wrote:
> Dear Mark,
> 
> > Tested in 3.3.6, 3.3.7, 3.3.9:
> >
> > Problem:
> > SIM netlist and flattened SPICE output does not preserve connections
> > or ports in the hierarchy.
> 
>    Okay, clearly I need to have a "test suite" as part of the make
> process so that I can immediately check after every code modification
> that this has or has not affected the schematic capture.
> 
>    Hopefully I can get around to fixing the error today or Monday.
> 
>                                         Regards,
>                                         Tim
> 
> 


-- 
Mark L. Chang

[Xcircuit-dev] Re: SIM/SPICE flattened extraction bug

[R. T. E. <tim...@mu...>]

Dear Mark,

> Tested in 3.3.6, 3.3.7, 3.3.9:
> 
> Problem:
> SIM netlist and flattened SPICE output does not preserve connections
> or ports in the hierarchy.

   Okay, clearly I need to have a "test suite" as part of the make
process so that I can immediately check after every code modification
that this has or has not affected the schematic capture.

   Hopefully I can get around to fixing the error today or Monday.

					Regards,
					Tim

[Xcircuit-dev] SIM/SPICE flattened extraction bug

[Mark C. <mar...@gm...>]

Tested in 3.3.6, 3.3.7, 3.3.9:

Problem:
SIM netlist and flattened SPICE output does not preserve connections
or ports in the hierarchy.

Test case:
Open logic8.ps and perform the SIM extraction from the highest level
of hierarchy (Page 1). You get this:

| sim circuit "logic" from XCircuit v3.30
p invert_1/In invert_1/Vdd invert_1/Out
n invert_1/In invert_1/GND invert_1/Out
p nand_1/In.1 nand_1/Vdd nand_1/Out
n nand_1/In.1 nand_1/ext13 nand_1/Out
n nand_1/In.2 nand_1/GND nand_1/ext13
p nand_1/In.2 nand_1/Vdd nand_1/Out
p nand_2/In.1 nand_2/Vdd nand_2/Out
n nand_2/In.1 nand_2/ext13 nand_2/Out
n nand_2/In.2 nand_2/GND nand_2/ext13
p nand_2/In.2 nand_2/Vdd nand_2/Out

Performing Flattened SPICE extraction you get this:
*SPICE (flattened) circuit "logic" from XCircuit v3.30

M1 invert_1/Out invert_1/In invert_1/Vdd Vdd pmos
M2 invert_1/Out invert_1/In invert_1/GND GND nmos
M3 nand_1/Out nand_1/In.1 nand_1/Vdd Vdd pmos
M4 nand_1/Out nand_1/In.1 nand_1/ext13 GND nmos
M5 nand_1/ext13 nand_1/In.2 nand_1/GND GND nmos
M6 nand_1/Out nand_1/In.2 nand_1/Vdd Vdd pmos
M7 nand_2/Out nand_2/In.1 nand_2/Vdd Vdd pmos
M8 nand_2/Out nand_2/In.1 nand_2/ext13 GND nmos
M9 nand_2/ext13 nand_2/In.2 nand_2/GND GND nmos
M10 nand_2/Out nand_2/In.2 nand_2/Vdd Vdd pmos

Notice nothing connected to "Pin.1" or "Pin.2", etc, which are the
highest-level pins.

Is this user error? Otherwise, seems to be a major-ish bug that will
probably be easily squished.

Thanks,
Mark

-- 
Mark L. Chang

[Xcircuit-dev] Re: XCircuit

[R. T. E. <ti...@op...>]

Dear Mark,

> Further, could you please add a short list to your webpages regarding 
> the system requirements needed before installing the x-circuit program?_

Like most open-source software, there are no specific requirements---
you need an operating system and a compiler.  Most systems, especially
Linux, come with these, but certain "minimal install" configurations
can be missing what is often called "development" packages, including
header files, link libraries, and possibly even the compiler itself.

> but everything goes wrong when i give the "make" command!
 ...
> BBoard.c:27:28: X11/IntrinsicP.h: No such file or directory
 ...

In your case, it appears that you are missing the parts of X11
(the graphics/windows system) needed to compile graphics programs.

Your choices are 1) go back to your Linux installer and add the "X11
development" packages (or "headers", "libraries", or whatever it
calls them), or 2) download a pre-compiled version of XCircuit (a
number of these are reasonably up-to-date).

Hopefully my comments will be of some help to you.

					Regards,
					Tim

[Xcircuit-dev] prerequisites for installing xcircuit !?

[Adriana & M. K. <m.k...@pl...>]

Hi there,

I 'm a newby to Linux and i'm really trying very, very hard to learn 
system adminstration, file mgt and installing new aplications on my 
linux machine on my own.

please help me with the installation of x-circuit!

Further, could you please add a short list to your webpages regarding 
the system requirements needed before installing the x-circuit program?_
_
This is what i did.  So far i have managed to succesfully:
(logged on as the root user)

    1 download the latest version of x-circuit (3.3.9);
    2 unzip it in a home/<user>/tmp directory;
    3 untar the file;
    4 succesfully did the ./configure command


but everything goes wrong when i give the "make" command!

Unfortunately i don't know about compiling and programming. Can you tell 
me what's going wrong? I have enclosed 2 text files showing the log for 
the configure and  make commands


Kind regards,

Mark Koetluk
The Netherlands

Re: [Xcircuit-dev] XCircuit on MacOSX-10.3

[Svenn A. B. <sv...@bj...>]

On Saturday 29 January 2005 23:08, Anthony Wilson wrote:
> Just wanted to let you know that I was able to successfully compile
> XCircuit 3.3.6.  I had to make a couple of changes to get it to
> compile.
>
> Setup
> --------
> Xcode 1.5
> gcc 3.3 (Apple build 1671 - latest update)
> Tcl 8.4.6 (installed from source from Fink)
> Mac OSX 10.3

What would be the effort to compile with AquaTcl? I don't like X11 on 
MacOSX. I once tried, but failed.
-- 
Svenn

Re: [Xcircuit-dev] Could use a little help. . .

[Svenn A. B. <sv...@bj...>]

On Friday 04 February 2005 18:40, R. Timothy Edwards wrote:
> Hello everyone,
>
>    I just logged into the xcircuit project on sourceforge
> and realized that there were bug reports from a year ago
> that I hadn't looked at.  And there are others there that
> are open, which I have looked at but haven't done anything
> about, going back about two years. . .

With xcircuit it has so far been faster to send an email to the 
developer, which is better for a one man project. 

>
>    The last posted version of XCircuit on the SourceForge
> site is way out of date, and the CVS repository is even
> older.  Mostly I just try to point people toward the

I can't connect to cvs from work so I have appreciated the frequent 
tarbals. From home I can do anything but I'm mostly too tired of 
circuits....

> XCircuit website at JHU, which is the only one I really
> have the time to maintain.  And I can't really expect my
> former Ph.D. advisor there to keep that old server up and
> running forever (it has been remarkably long-lived as it
> is).

If it is working and you don't try to fix it, it will run forever. 
Does this mean that you are thinking about switching over to SF?

-- 
Svenn

[Xcircuit-dev] Could use a little help. . .

[R. T. E. <tim...@mu...>]

Hello everyone,

   I just logged into the xcircuit project on sourceforge
and realized that there were bug reports from a year ago
that I hadn't looked at.  And there are others there that
are open, which I have looked at but haven't done anything
about, going back about two years. . .

   The last posted version of XCircuit on the SourceForge
site is way out of date, and the CVS repository is even
older.  Mostly I just try to point people toward the
XCircuit website at JHU, which is the only one I really
have the time to maintain.  And I can't really expect my
former Ph.D. advisor there to keep that old server up and
running forever (it has been remarkably long-lived as it
is).

   Any volunteers for maintaining the XCircuit SourceForge
website?  Volunteers for development are welcome as well.

				Regards,
				Tim
   
+--------------------------------+-------------------------------------+
| Dr. R. Timothy Edwards (Tim)   | email: tim...@mu...     |
| MultiGiG, Inc.	 	 | web:   http://www.multigig.com      |
| 3600 Glen Canyon Road          | phone: (831) 440-0600 ext. 20       |
| Scotts Valley, CA 95066        | cell:  (240) 401-0616               |
+--------------------------------+-------------------------------------+

[Xcircuit-dev] XCircuit on MacOSX-10.3

[Anthony W. <an...@zo...>]

Tim,
Just wanted to let you know that I was able to successfully compile 
XCircuit 3.3.6.  I had to make a couple of changes to get it to 
compile.

Setup
--------
Xcode 1.5
gcc 3.3 (Apple build 1671 - latest update)
Tcl 8.4.6 (installed from source from Fink)
Mac OSX 10.3

Needed change
------------------------
Line 8519 in configure
old 	 LDDL_FLAGS="-dynamiclib -flat_namespace -undefined suppress 
-noprebind"
to
new	 LDDL_FLAGS="-dynamiclib -single_module -flat_namespace -undefined 
suppress -noprebind"

Secondly, in order to test the compile before install, I had to modify 
the lib/tcl/xcircuit.tcl script.  I had to modify the .so to .dylib.

Otherwise once you do a make install, everything was installed 
correctly.

Hope this helps.
-anthony.

[Xcircuit-dev] Re: XCircuit & ScriptEDA

[R. T. E. <tim...@mu...>]

Dear Anthony,

> As I have just finished browsing the Xcircuit site, I had a question 
> about your take on ScriptEDA.  Mainly it is a question on your choice 
> of using TCL over Python.  I see you had started with Python, but have 
> now elected to stick with TCL and drop the python support.

This was largely a matter of convenience.  Magic, IRSIM, and XCircuit
all had existing command-line interfaces whose syntax is similar to
Tcl (in the case of Magic, this is no coincidence, as Magic was written
by John Ousterhout, who then went on to create the Tcl language.

> I ask this because I am fond of python, but also because I would of 
> thought Python to be a better candidate than TCL for the ScriptEDA 
> philosophy.

Actually, the ScriptEDA "philosophy" was that the tools should be
interpreter-independent, or rather interpreter-interchangeable, and
clearly I've parted ways with this, though that was not my intent
from the beginning.

If there's one thing about Python that I prefer over Tcl, it is it's
ability to create nested dictionaries.  It's a much more sensible
approach than Tcl's array variables.

> 1.  Did you run into design problems or concerns in your choice of 
> using Python?

Mostly that I'm not familiar with the graphics part of Python,
although I learned to program in Python pretty well.  I wasn't
familiar with Tk graphics, either, but I had more examples on hand.
It was just too much work to support more than one interpreter.

> 2.  Do you think it would be difficult to maintain the Python portion 
> of Xcircuit?

No, but to be "done correctly", the Python version should be
redesigned like the Tcl version, making the program an extension
of the interpreter, instead of embedding the interpreter.  But
that's a LOT of work.  Getting the Xcircuit w/embedded Python to
compile and making sure it still works with the latest Python version
should not be all that difficult.  I may have broken things here or
there with a misplaced #ifdef, but fundamentally there is nothing that
will prevent it from working.

> 3.  Do you think it would be difficult to add Python support to Magic?

Difficult, no;  a lot of work, yes.  Magic is better modularized than
XCircuit, so there are not very many Tcl calls from inside the source
code.  Most of these calls are to set the result for the interpreter.
A good thing to do would be to rewrite all the functions in a more
general way, like "InterpSetResult()", and then define this function
for Python or for Tcl at compile time (or SCHEME or. . .)

If you cd to the magic top level directory and do:

	grep Tcl_ */*.c

you'll see just how much work is involved.  It shouldn't be difficult,
though, in the sense that most of the basic problems of compiling the
program as an interpreter extension and creating the GUI as an
interpreter script have been solved (for all the programs).

					Regards,
					Tim

[Xcircuit-dev] ScriptEDA and design tools

[Anthony W. <an...@zo...>]

Tim,
As I have just finished browsing the Xcircuit site, I had a question 
about your take on ScriptEDA.  Mainly it is a question on your choice 
of using TCL over Python.  I see you had started with Python, but have 
now elected to stick with TCL and drop the python support.

I ask this because I am fond of python, but also because I would of 
thought Python to be a better candidate than TCL for the ScriptEDA 
philosophy.  I was also considering porting ngspice to something 
similar to tclspice.  But at the same time, I like the idea of the 
complete tool set being in a "single" environment.

So I would like to ask a couple of questions:
1.  Did you run into design problems or concerns in your choice of 
using Python?
2.  Do you think it would be difficult to maintain the Python portion 
of Xcircuit?
3.  Do you think it would be difficult to add Python support to Magic?

As I ask these questions, I am torn between my goals of wanting a 
unified design environment and using Python.  I am leaning heavily 
toward the unified design environment because I want to leverage off 
the work that is currently taken place to make sure that the tools 
compile and work for MacOSX.

Thanks for your time.
-anthony wilson
an...@zo...

[Xcircuit-dev] Re: xcircuit compile failure

[R. T. E. <ti...@op...>]

Dear Dick,

> Today I tried building 'xcircuit-3.3.4.tgz' 
> using my usual procedure and got the following failure during the 'make 
> tcl' step:

Indeed, line 1378 of tclxcircuit.c should be moved below the type
declarations, say to line 1382.

The interesting thing about this is that it does not cause a compile
failure on my machine, suggesting that gcc-3.X-something allows one
to get away with more C++ syntax like type declarations in the middle
of procedural blocks.  I may have to change my compile flags to make
sure that I'm not breaking the compilation on older versions of gcc.

					Regards,
					Tim

[Xcircuit-dev] problem with latest revision ...

[Richard F. {x. C. <rg...@sg...>]

Hi!

I've been using xcircuit off and on now for about a year and really 
appreciate it! I use it for schematic capture (spice) and general 
drawing, check your website often and build the latest version as soon 
as I find a new release. Today I tried building 'xcircuit-3.3.4.tgz' 
using my usual procedure and got the following failure during the 'make 
tcl' step:
  ...
  tclxcircuit.c: In function `xctcl_netlist':
  tclxcircuit.c:1381: parse error before `int'
  tclxcircuit.c:1385: `instofname' undeclared (first use in this function)
  tclxcircuit.c:1385: (Each undeclared identifier is reported only once
  tclxcircuit.c:1385: for each function it appears in.)
  tclxcircuit.c:1386: `x' undeclared (first use in this function)
  tclxcircuit.c:1386: `y' undeclared (first use in this function)
  make[1]: *** [tclxcircuit.o] Error 1
  make[1]: Leaving directory `/home/moose2/rgf/xcircuit-3.3.4'
  make: *** [tcl] Error 2
  ...

OS/HW is Redhat linux/Dell Optiplex

Dick

-- 
 Richard Finstad                E-mail: rg...@sg...
 SGI                            Voice : (715)726-7483
 1168 Industrial Blvd           Fax   : (715)726-7662
 Chippewa Falls, WI 54729-0078  Room  : CF2/269B
--------------------------------------------------------------------------