On 2 Jan 2006 00:48:53 GMT, Donnie Barnes
wrote:

It's all moot, IMHO. Neither compact or compress exist on nearly as many
systems as gzip.

"Smart" sysadmins may not allow you to install gzip on fairly
secure networks or machines. Nor compile it from the sourcecode.
I think you need multiple options.
--
Cliff

On Mon, 02 Jan 2006 01:41:30 +0000, (DoN.
Nichols) wrote:

According to Cliff :
On Sun, 01 Jan 2006 21:49:23 +0000, (DoN.
Nichols) wrote:

As for "compact" -- that may be truly antique, as I don't have
any examples of it.

Check HP-UX, probably among others.
"The command compact of UNIX implements the dynamic Huffman coding"

http://docs.hp.com/en/B9106-90007/compact.1.html

Hmm ... pretty old program, if the man page is still concerned
about filesystems which limit filenames to 14 characters (the old
filesystem which came with v7 unix, and early SysV variants as well.

I'm somewhat frustrated by the HP-UX (or the web page format)
eliminating the "Last change: 9 Sep 1999" entry at the bottom of each
page, which might make it easier to judge just how old compact(1)
happens to be. That one was from Solaris 10. Solaris 2.6 shows: "Last
change: 20 Dec 1996" Falling back to SunOs 4.1.4, I get: "Last change: 9
September 1987" and OpenBSD shows it as: "April 18, 1994". Of course,
those dates apply to the last change of the man page, not the program,
but at least some of the changes of the man page are to reflect changes
in the program.

Perhaps it worked without piles of changesG.

BTW, Compress & Gzip seem to get about the same compression
ratios, at least on the binary files I once tested such on. IIRC
Gzip *may* have been slightly faster in execution one way
(comressing) and slower the other (uncompressing).

I've seen gzip(1) usually giving a significant improvement over
compress(1) on most file types -- at least on SunOs and Solaris (even on
the Sun-3 (68020) machines).

IIRC I tested on HP-UX 9 but it might have been Sun. Memory fades
and too many platforms.
--
Cliff

On Sun, 01 Jan 2006 21:55:14 -0600, "Martin H. Eastburn"
wrote:

How about the Unix to Dos and Dos to Unix utilities!

Readily available from many sources if you don't already have them.
Unix2dos & dos2unix IIRC.
--
Cliff

On Sun, 01 Jan 2006 18:57:44 -0500, Cliff wrote:

Start out using IGES as your part file database.

Or consider STEP ....
--
Cliff

Wayne C. Gramlich wrote:

All:

Please note that the IGES 5.3 specification is not freely
available. If you want a copy of the IGES 5.3 specification,
be prepared to shell out $425 (US) for a downloadable PDF
and $495 (US) for a CD-ROM. Go to:

http://www.nist.gov/iges/

click on "Current Version" in the lefthand frame and
finally click on "US Product Data Assoiciation/IGES Catalog"
to get the price list.


Well of course we're talking about industrial standards here that
everybody understands not some make 'it-up-as-you-go-along' hackery.

DoN. Nichols wrote:

This makes me feel better. But there is still the problem of
what happens if someone is able to breach security on your web site.
I'm running my web site on an OpenBSD box (which runs the web server in
a chroot jail) and still don't trust it to hold anything truly
sensitive.

aye, nothing is 100% immune...

1/ that's why I decided to dost it on a "reseller" server with a pretty
reputable bunch who know their onions, I could have made it marginally
more secure myself but only by choosing between the expense of a
dedicated colo or the inconvenience of restricted bandwidth.

2/ that's why I'm looking for two admins with enough savvy to spot
potential problems as well as the ability to do periodic mysql dumps
etc

3/ that's why it's based on a fiarly good open source CMS, namely e107

4/ _if_ someone is able to breach security it is a lot of effort,
because everything worthwhile is open source anyway and freely
available, the people who use open source are likely to be able to make
things sticky for any wannabe cracker, and most of all it's a hell of a
lot of offort to get a bunch of email addresses from people who are
likely to be running their own domains anyway and quite able to create
one off throwaway email addresses.

99.999% of the security comes from the rather obvious fact that there
is an inexhaustible supply of far more juicy targets out there in just
as easy reach.

On Mon, 02 Jan 2006 05:35:18 GMT, "Wayne C. Gramlich"
wrote:

Please note that the IGES 5.3 specification is not freely
available.

You might ask at your local library if they can get a copy ....
last I knew in the US there was the inter-library loan service.
An ISBN number might be handy, if it has one (I did not look).

On top of that try http://www.iges5x.org/ ,
http://www.nist.gov/iges/ and
http://www.iges5x.org/archives/version5x/ G.
--
Cliff

Cliff wrote:
On Mon, 02 Jan 2006 05:35:18 GMT, "Wayne C. Gramlich"
wrote:

Please note that the IGES 5.3 specification is not freely
available.

You might ask at your local library if they can get a copy ....
last I knew in the US there was the inter-library loan service.
An ISBN number might be handy, if it has one (I did not look).

On top of that try http://www.iges5x.org/ ,

[snip]

http://www.iges5x.org/archives/version5x/ G.

Definitely cheaper than shelling $400+ for the official
document.

With some digging around using Google I was able to find a
draft version of the 6.0 specification at:

http://www.devdept.com/code/ogif/version6.pdf

I have not found any copyright notice in the downloaded
the file above yet.

Thanks,

-Wayne

On Mon, 02 Jan, Cliff wrote:
"Smart" sysadmins may not allow you to install gzip on fairly
secure networks or machines. Nor compile it from the sourcecode.
I think you need multiple options.

This doesn't matter then for the purposes of what we're talking about
distributing since you wouldn't be able to install/build *that*, either.


--Donnie

--
Donnie Barnes http://www.donniebarnes.com 879. V.

The files I listed were in a directory on my disk.
Used to use them in file transport from/to work.
was an X11/Rn user from Sun 1 to Ultra 40 ? My Sun 10 was my real baby!
Dual CPU Three tape drives 4,8, 1/4 archive.

Martin
Martin Eastburn
@ home at Lions' Lair with our computer lionslair at consolidated dot net
NRA LOH & Endowment Member
NRA Second Amendment Task Force Charter Founder



Cliff wrote:
On Sun, 01 Jan 2006 21:55:14 -0600, "Martin H. Eastburn"
wrote:


How about the Unix to Dos and Dos to Unix utilities!


Readily available from many sources if you don't already have them.
Unix2dos & dos2unix IIRC.

----== Posted via Newsfeeds.Com - Unlimited-Unrestricted-Secure Usenet News==----
http://www.newsfeeds.com The #1 Newsgroup Service in the World! 120,000+ Newsgroups
----= East and West-Coast Server Farms - Total Privacy via Encryption =----

According to Martin H. Eastburn :
How about the Unix to Dos and Dos to Unix utilities!

UUENCode.exe - Wtar.exe Unix2Dos.exe Dos2unix.exe Hex40bin.exe

I have an archive of them.

Those are obviously the ones to run on DOS. The unix ones have
the same names (all lower case), without the ".exe" -- except for the
hex40bin one, which I think is for dealing with older Mac archives. And
you left out uudecode(1). (For those who don't know -- the "(1)" at the
end of the program names is a clue that they are in section one of the
unix man pages -- therefore they are commands. And -- it is not typed
when typing the command name to execute it. "(2)" is for system calls
in c, "(3)" is library functions in c. Various other meanings for the
other digits, which may vary from unix flavor to unix flavor.

All of the named programs come with most flavors of unix, except
the "hex40bin" one listed above.

Enjoy,
DoN.
--
Email: | Voice (all times): (703) 938-4564
(too) near Washington D.C. | http://www.d-and-d.com/dnichols/DoN.html
--- Black Holes are where God is dividing by zero ---
--
Email: | Voice (all times): (703) 938-4564
(too) near Washington D.C. | http://www.d-and-d.com/dnichols/DoN.html
--- Black Holes are where God is dividing by zero ---

Sticking with this thread.

Found an excellent source of CNC hardware, ball screws, linear
products, you name it, so done a quick article about them on the site
at http://www.open-source-cnc.com/content.php?article.6

This guy is Boeing certified, Open Source friendly, and only recently
set up himself after the company he used to work for outsourced
production from the Uk to the Far East.

There's a moral in there somewhere, if I could only work out what it
was.

cheers

"tanks" Don -

IIRC, I got those many moon ago when Schlumberger was funding some development
of these and modem utilities. 1980 - 85 ish. I did have that 3 ring binder - but it
likely got mildew and dumped. Rats. took a quick scan - and my little IT book is not
to be found. Such is life. At the time, IBM email was really different than many of
us on other nets - they changed externally - not sure internally due to the
software design. Kinda fun days in the early net - customers coming alive in more
ways than one!

Martin

Martin Eastburn
@ home at Lions' Lair with our computer lionslair at consolidated dot net
NRA LOH & Endowment Member
NRA Second Amendment Task Force Charter Founder



DoN. Nichols wrote:
According to Martin H. Eastburn :

How about the Unix to Dos and Dos to Unix utilities!

UUENCode.exe - Wtar.exe Unix2Dos.exe Dos2unix.exe Hex40bin.exe

I have an archive of them.


Those are obviously the ones to run on DOS. The unix ones have
the same names (all lower case), without the ".exe" -- except for the
hex40bin one, which I think is for dealing with older Mac archives. And
you left out uudecode(1). (For those who don't know -- the "(1)" at the
end of the program names is a clue that they are in section one of the
unix man pages -- therefore they are commands. And -- it is not typed
when typing the command name to execute it. "(2)" is for system calls
in c, "(3)" is library functions in c. Various other meanings for the
other digits, which may vary from unix flavor to unix flavor.

All of the named programs come with most flavors of unix, except
the "hex40bin" one listed above.

Enjoy,
DoN.

----== Posted via Newsfeeds.Com - Unlimited-Unrestricted-Secure Usenet News==----
http://www.newsfeeds.com The #1 Newsgroup Service in the World! 120,000+ Newsgroups
----= East and West-Coast Server Farms - Total Privacy via Encryption =----

If it's free, I'll take three….

The snow is back.

I am about to begin updating a program I wrote several years ago
that converted a v.11 AutoCAD ADI plotter file to G/M codes for a
EMCO F1 educational machining center, but
== before I do this I want to find out how much interest there
is. ==

The new version will convert a HPGL [not HPGL/2] file to ISO
standard cnc file format, and will be placed IN THE PUBLIC DOMAIN
UNDER THE GPL.

A google search indicates the existence of some commercial
programs that appear to do this, but these are expensive for
casual or 1 time use.

My main reason for using HPGL as source rather than dxf is
because this allows easy generation of characters from stroke
[Hershey] fonts and truetype [.ttf] fonts. DXF retains the text
as strings, while HPGL has the actual vectors making up the
characters. This will also let any program, not just cad or
cad/cam, that can generate a HPGL plotter file be used. In
windows, most all windows graphic programs will allow the use of
a HP plotter driver with output to file, which can be downloaded
for free from the HP web site. Some of the HP drivers are for
plotters that I have never seen (but would like to), such as a
roll feed unit that will plot 36 inches wide by 200 feet long.

I have looked at a large number of the HPGL files output by
windows, dos and other programs. None of these seem to use the
extend HPGL commands to draw arc, etc. Just the PU [pen up],
PD [pen down], IN [initialize to power up], and the end points of
the vectors. A few allow [but don't require] pen selection and
plotting speed specification. Curves are approximated by a large
number of short, straight line moves. The g code will thus have
only G0 [pen up] and G1[pen down] moves, with the vertical and
horizontal feed rates specified by the user.

HPGL files are in "plotter units" with 40 plotter units = 1 m/m
or 1016 plotter units = 1 inch. The first rewrite will be with
converted 0.001 inch output, with trailing zeros if required.

My questions to the group.

(1) If its free and easy to use, would you like to have such a
program?

(2) Do you want the output to be a complete "stand alone"
program or simply a code fragment that you cut-n-paste into an
existing cnc program.

(3) Will a program that runs in a "DOS" box be acceptable, or do
you need/want one that runs in the windows environment?

(4) Does it make any difference what language the source code is
written in? I am leaning toward M/S qb4.5 as I have/use that
compiler. I will post both the source [.bas] and compiled [.exe]
files for download.

(5) How much group interest is there in a version that generates
metric output?

(6) How much group interest is there in a version that uses a
rotary indexer to generate text/graphics around a cylinder?

(7) Is there another output format that would be more
acceptable/useful to the group than ISO? If so which one, and
what are the specifications, i.e. number of characters in each
column, leading/trailing signs, decimal point inclusion, number
of places, etc. FWIW, the program output will be positive
absolute moves only.

(8) How would you use such a program?

Uncle George

On Tue, 10 Jan 2006 11:51:47 -0600, F. George McDuffee
wrote:

My questions to the group.

(1) If its free and easy to use, would you like to have such a
program?

Yes.

(2) Do you want the output to be a complete "stand alone"
program or simply a code fragment that you cut-n-paste into an
existing cnc program.

Doesn't really matter.

(3) Will a program that runs in a "DOS" box be acceptable, or do
you need/want one that runs in the windows environment?

Let's face the facts: Both DOS and Freddy Mercury are dead and it's
time to get over it. Windows are the way to go - just ask BD.
(besides, XP has an issue with DOS programs, IIRC)

(4) Does it make any difference what language the source code is
written in? I am leaning toward M/S qb4.5 as I have/use that
compiler. I will post both the source [.bas] and compiled [.exe]
files for download.

(5) How much group interest is there in a version that generates
metric output?

I *could* live with inch output and multiply the input with 25, but whynot
make it a user choice in the program?

(6) How much group interest is there in a version that uses a
rotary indexer to generate text/graphics around a cylinder?

No need here, as our mill has an option for rotary axis swapping.

(8) How would you use such a program?

Text engraving, frames, logos. The (DOS) program I'm using at the moment
has only one font, and is a PITA to work with.

From what I've seen of pen plotters and their seemingly random choice of
path, I'm not sure I would trust their code for pocketing or conturing.
--
Regards,
J. Nielsen

F. George McDuffee wrote:

My questions to the group.

(1) If its free and easy to use, would you like to have such a
program?

yes, in huge bold text


(2) Do you want the output to be a complete "stand alone"
program or simply a code fragment that you cut-n-paste into an
existing cnc program.

you said open source, so both really innit...


(3) Will a program that runs in a "DOS" box be acceptable, or do
you need/want one that runs in the windows environment?

platform independent like C would be best, failing that something that
could easily be recompiled into C, so lots of informative REM
statements


(4) Does it make any difference what language the source code is
written in? I am leaning toward M/S qb4.5 as I have/use that
compiler. I will post both the source [.bas] and compiled [.exe]
files for download.

see above

if the dll's are available if required linux users can always run it
under emulation.


(5) How much group interest is there in a version that generates
metric output?

honestly, it shoukd emuate g code and allow metric and imperial in the
same output


(6) How much group interest is there in a version that uses a
rotary indexer to generate text/graphics around a cylinder?

yes, radius compensation too for rotary table on mill stuff


(7) Is there another output format that would be more
acceptable/useful to the group than ISO? If so which one, and
what are the specifications, i.e. number of characters in each
column, leading/trailing signs, decimal point inclusion, number
of places, etc. FWIW, the program output will be positive
absolute moves only.

(8) How would you use such a program?

code library of gcode "objects" waiting to be made ;^)

On 10 Jan 2006 14:45:42 -0800, "Guy Fawkes"
wrote:


F. George McDuffee wrote:

My questions to the group.

(1) If its free and easy to use, would you like to have such a
program?

yes, in huge bold text


(2) Do you want the output to be a complete "stand alone"
program or simply a code fragment that you cut-n-paste into an
existing cnc program.

you said open source, so both really innit...


(3) Will a program that runs in a "DOS" box be acceptable, or do
you need/want one that runs in the windows environment?

platform independent like C would be best, failing that something that
could easily be recompiled into C, so lots of informative REM
statements


(4) Does it make any difference what language the source code is
written in? I am leaning toward M/S qb4.5 as I have/use that
compiler. I will post both the source [.bas] and compiled [.exe]
files for download.

see above

if the dll's are available if required linux users can always run it
under emulation.


(5) How much group interest is there in a version that generates
metric output?

honestly, it shoukd emuate g code and allow metric and imperial in the
same output


(6) How much group interest is there in a version that uses a
rotary indexer to generate text/graphics around a cylinder?

yes, radius compensation too for rotary table on mill stuff


(7) Is there another output format that would be more
acceptable/useful to the group than ISO? If so which one, and
what are the specifications, i.e. number of characters in each
column, leading/trailing signs, decimal point inclusion, number
of places, etc. FWIW, the program output will be positive
absolute moves only.

(8) How would you use such a program?

code library of gcode "objects" waiting to be made ;^)
=============
ref item 2 --
cut-n-paste option would have only the actual g0/g1 moves.
header of program to turn on spindle CW, set absolute, etc, and
footer turning everything off would be written by the user, who
then uses an editor to insert the generated code into the
complete program. From some emails I have received I think i
will do it this way as you can do a zero offset to position the
text/artwork just where you want it.

As several people have pointed out it is trivial to allow
selection of either inch or m/m output, so will include that.

I will get the "flat" version up and running and then try for the
cylinder model.

Uncle George

On 10 Jan 2006 14:45:42 -0800, "Guy Fawkes"
wrote:


F. George McDuffee wrote:

My questions to the group.

(1) If its free and easy to use, would you like to have such a
program?

yes, in huge bold text


(2) Do you want the output to be a complete "stand alone"
program or simply a code fragment that you cut-n-paste into an
existing cnc program.

you said open source, so both really innit...


(3) Will a program that runs in a "DOS" box be acceptable, or do
you need/want one that runs in the windows environment?

platform independent like C would be best, failing that something that
could easily be recompiled into C, so lots of informative REM
statements


(4) Does it make any difference what language the source code is
written in? I am leaning toward M/S qb4.5 as I have/use that
compiler. I will post both the source [.bas] and compiled [.exe]
files for download.

see above

if the dll's are available if required linux users can always run it
under emulation.


(5) How much group interest is there in a version that generates
metric output?

honestly, it shoukd emuate g code and allow metric and imperial in the
same output


(6) How much group interest is there in a version that uses a
rotary indexer to generate text/graphics around a cylinder?

yes, radius compensation too for rotary table on mill stuff


(7) Is there another output format that would be more
acceptable/useful to the group than ISO? If so which one, and
what are the specifications, i.e. number of characters in each
column, leading/trailing signs, decimal point inclusion, number
of places, etc. FWIW, the program output will be positive
absolute moves only.

(8) How would you use such a program?

code library of gcode "objects" waiting to be made ;^)
===============
Item 3/4
I will provide both source and compiled [.exe] files that should
run on any late windows box. I am most comfortable working with
the older basic dialects in DOS [and I still like cp/m], however
as someone pointed out its time to move on as XP has problems
with some DOS "box" programs, so I an updating to visual basic
for windows.

Uncle George

On 31 Dec 2005 14:10:30 -0800, "Guy Fawkes"
wrote:
Yes I checked the charters and I don't believe this is in breach.
Basically following recent discussions on alt.machines.cnc and
rec.crafts.metalworking it became apparent to me that there is a lot of
interest in the Open Source aspect of CNC in general, but no real
central repository of information.

Yes there are specific websites such as LinuxCNC org, but that won't
help you with finding ballscrews locally for your own CNC project, or
help you with a specific G code problem, or suchlike.

So, I have today registered http://www.open-source-cnc.com/ and it is
now up and live in a very basic form, work very much in progress, but
there is enough there to register as a user and most importantly
participate. Yes folks, this is not a site about me, but about CNC so
it needs users but most importantly in the early days it needs
moderators and admins.
====================
snip
Open source cnc software - HPGL to g code programs

My major interest in this was to allow easy use of fonts to
generate cnc programs.

As the sign says "so soon old - so late smart."

As part of writing the converter I did intensive web research.
Google did not directly point to this, but through a series of
links I discovered that such an application already exists and
has long been used in the Unix/Linux community, and has been
ported to Windows. [After I blew the butter-and-egg money on a
new compiler that will generate console apps that will work in XP
-- did get a cool t-shirt though. see:
http://www.powerbasic.com/]

You need to download and install hp2xx. There are several sites
and mirrors. Among many see
http://www.gnu.org/software/hp2xx/hp2xx.html

The version I used is hp2xx-3.4.4.2.exe. This is an
automatically unpacking install file. It will install the base
hp2xx.exe program, most likely in c:\program Files\GnuWin32\bin

My small contribution to the group is to make you aware of this
program and provide you with an easy way to run it with the
required options. Basically I wrote a dos bat file called
makecnc.bat. This needs to go all on one line, and reads as
follows.: (note that this is from the unix/linux world and thus
the toggles are case sensitive)

c:\progra~1\GnuWin32\bin\hp2xx -m nc -n -N -z -2.00 -Z 2.00 -l
makecnc.log -f cnccode.out %1

You can put this batch file in what every your working directory
is [directories are].
The lead-in accesses hp2xx in its "home" directory

-m nc tells the program you want cnc code out

-n tells the program not to fill polygons

-N tells the program to ignore the PS or pen size command

-z -2.0 tells the program to plunge 2.0 m/m below the x-y plane
on a pen down PD. You can set this to whatever you want as long
as the pen up xy is greater than the pen down xy

-Z 2.0 tells the program to raise 2.0 m/m over the x-y plane on a
pen up PU. You can set this to whatever you want as long as it
is greater than the pen down xy.

-l makecnc.log tells the program to write progress
information/errors to makecnc.log in the directory where you
called it from. You can change this to whatever you want but use
the 8.3 format.

-f cnccode.out tells the program to write the output to
cnccode.out in the directory where you called it from. You can
change this to whatever you want but use the 8.3 format.

The %1 is what every you type after makecnc. For example if I
input "makecnc plaque.plt" at the command line, the %1 becomes
"plaque.plt". This allows easy conversion.

Be sure and download [and read] the documentation.

You will want to experiment with the PS [pen size] and the -n/-N
fill toggles.

Note that all moves are absolute [G90] metric [G71]
Use the G71 [not included] to set to metric at the start of the
block and G70 to return to inch at the end of the block.
You will have to set the feed rate to the slower of the plunge or
traverse, and will need to have a header to start your machine
and a footer to shut everything down.

I have tried this with the HP7550 plus, the HP Draftmaster SX,
and the HPGL/2 plotter drivers in Windows2000 with good results
from several graphics and cad programs including CMS Intellicad
5.1 PE+. As some of these plotters can go to E size and/or can
use 36-inch wide roll paper, they can generate some files for big
parts [anyone with a cnc spar mill out there?] Note that HP2xx
is mainly for HPGL and not HPGL/2, but it accommodates a large
subset of HPGL/2 that included everything I tried.

The downside is that while this does indeed let you use
stroke/Hershey and TrueType (ttf) fonts, the code does not use
anything but G0 and G1 moves. All arcs are approximated by an
enormous number of short line segments, so the output files are
huge. However, as most people have the ability to cut-n-paste
this should not present a problem. Most likely not the thing to
use for high volume production, but may be just what you need for
a one-off special for the boss.

Let us know how this works and what you used it for.

Uncle George