Subject:
|
Re: lsynthcp community development project setup on sourceforge
|
Newsgroups:
|
lugnet.cad.dev
|
Date:
|
Tue, 21 Jan 2003 03:56:11 GMT
|
Viewed:
|
971 times
|
| |
|
|
In lugnet.cad.dev, Don Heyse writes:
> In lugnet.cad.dev, Kevin Clague writes:
> > In lugnet.cad.dev, Ross Crawford writes:
> > > In lugnet.cad.dev, Kevin Clague writes:
> > > > I think that positional parameters may
> > > > be a real hassle. How do you feel about named parameters
> > > > (FILENAME="foo.dat", LENGTH=20, UNITS=CM)?
> > >
> > > What about getopt() style parameters, eg:
> > >
> > > -f foo.dat -l 20 -u cm
> > > --filename foo.dat --length 20 --units cm
> >
> > Much better! I've been programming on Windows too much lately (isn't any
> > amount too much? ;^)
>
> Whoah, wait a minute! I like getopt() as much as anyone, but I
> thought we were talking about something to go *inside* the file on
> the SYNTH BEGIN line(s), not necessarily on the command line. Units
> may be universal for all synthesized parts in a file, but I think
> the length and name are best tied to a specific SYNTH part in the
> file. So how about this? Global defaults are set on the command
> line with getopt() style params. These can be overridden for each
> synthesized part by named parameters on the SYNTH BEGIN line of
> each synthesized part.
;^) Relax, we are talking about both on the command line that starts the
program, as well as on the LDraw meta-commands that start synthesis of a
specific part. Notice he said getopt() style parameters, not getopt() to
get the parameters. Getopt() style parameters is just like named parameters,
only the format is slightly different. I don't think there is much of a
precedence for either in the LDraw file format. For the sake of
consistancy, I vote we use getop() *style* arguments everwhere.
>
> I'd like to see something like --filename --outputfile --subfiles
> --onlysynthparts in addition to the length and units on the cmd line.
> I'm not sure how to work the default input and output. Perhaps
> make it stdin and stdout unless overridden by -f or -o. I guess
> for backwards compatibility I'd also allow the first cmdline param
> without a "-" to be the input file, and the 2nd without a "-" to
> be the output file. Informational messages should go to stderr.
It seems like the only required parameter is the source file since we're
going to allow synthesized output to be written to part files on a part by
part basis.
We could also have SYNTH GLOBAL meta-command where default values for units
and such could be placed, so that they would not have to be entered on the
command line every time.
>
> The --subfiles option would tell the program to copy the input file
> to the output file as is, and dump the synthesized parts to either
> numbered subfiles, or files named by the "NAME=" option.
I won't stop this from happening, but I don't understand the value of
this. Why have it if you have a --name <fname> option on the SYNTH BEGIN
meta-command line?
>
> The --onlysynthparts option would tell the program to skip everything
> but the new synthesized parts. These would either go to stdout, the
> named outputfile, or individual subfiles, depending on the other
> command line options.
>
> Does this make sense?
Why all the flexibility for output at the global level? Why not just leave
it at the local level? You must have a more far reaching vision than I do.
>
> Don
Kevin
|
|
Message is in Reply To:
19 Messages in This Thread:
  
 
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
