MetaPost mailing list

luigi scarso | 1 Dec 2010 11:46

$Picon$

Re: Problem with generating EPS from MetaFont source

On Wed, Dec 1, 2010 at 11:41 AM, Michael Ummels
<michael.ummels <at> rwth-aachen.de> wrote:
> Sorry, when stripping the test file, I removed some dimensions. If you want
> to compile test.mf using MetaFont, replace the line
> define_whole_pixels(equal_spread, arrow_horiz_len,
> arrow_vert_len, arrow_diag_len, arrow_spread, line_thickness,
> rule_thickness, side_bearing);
> by
> define_whole_pixels(equal_spread, arrow_diag_len, arrow_spread,
> line_thickness, rule_thickness, side_bearing);
> Then, if you run "mf '\mode:=localfont; input test'", you will see that it
> compiles just fine.
> Michael
It's ok now, thank you.
--

-- 
luigi

--
http://tug.org/metapost/

headers

Michael Ummels | 1 Dec 2010 11:41

$Picon$
$Picon$
$Favicon$
$Gravatar$

Re: Problem with generating EPS from MetaFont source

Sorry, when stripping the test file, I removed some dimensions. If you want to compile test.mf using MetaFont, replace the line

define_whole_pixels(equal_spread, arrow_horiz_len, arrow_vert_len, arrow_diag_len, arrow_spread, line_thickness, rule_thickness, side_bearing);

by

define_whole_pixels(equal_spread, arrow_diag_len, arrow_spread, line_thickness, rule_thickness, side_bearing);

Then, if you run "mf '\mode:=localfont; input test'", you will see that it compiles just fine.

Michael

On Wed, Dec 1, 2010 at 10:07, luigi scarso <luigi.scarso <at> gmail.com> wrote:

Not an answer, but a question (a bit off-topic..): how can you process
you test.mf with METAFONT ?
I've tried
# ./mf '&plain \mode:=localfont; mag=100.375; input test'

This is METAFONT, Version 2.718281 (Web2C 7.5.7)
(test.mf
>> 277.77776arrow_horiz_len#+0.5
! Not implemented: floor(unknown numeric).
<to be read again>
endgroup
<to be read again>
;
<for(arrow_horiz_len)> ...hround((SUFFIX0)#*hppp);
ENDFOR
l.13 ...e_thickness, rule_thickness, side_bearing)

--
luigi

--
http://tug.org/metapost/

headers

Yasir Malik | 2 Dec 2010 15:15

$Favicon$

Problems with passing parameters to macros

Hi,

I'm seeing strange behavior when I pass parameters to a macro.  Consider 
the following code fragment:

File macros.mp:
def someMacro(expr v, a, b, x, y, z) =
begingroup

show "inside func";
show v, a, b;
show "";

...

show "before exiting";
show v, a, b;
show "";

endgroup
enddef;

Another file:

input macros.mp

pair a, b, ap;

...

show "before calling";
show a, b, ap;
show "";
someMacro(a, b, ap, x, y, z);

show "after calling";
show a, b, ap;

Here's the output:

>> "before calling"
>> (-22.32306,38.6648)
>> (-128.55247,0)
>> (-37.20473,64.44066)
>> ""
>> "inside func"
>> (-22.32306,38.6648)
>> (-128.55247,0)
>> (-37.20473,64.44066)
>> ""
>> "before exiting"
>> (-22.32306,38.6648)
>> (-128.55247,0)
>> (-37.20473,64.44066)
>> ""
>> "after calling"
>> (-22.32306,38.6648)
>> (-128.55247,0)
>> (-128.55247,0) [33] )

As you can see, the value of ap is the same as b after coming out of the 
macro.  I'm not sure whether showing the entire code is relevent (there is 
a lot of code and all the parameters in the macro have the value of what I 
passed).  I can try to reproduce this issue with less code, but has anyone 
seen this problem before?

Thanks,
Yasir
--
http://tug.org/metapost/

headers

Yasir Malik | 2 Dec 2010 22:22

$Favicon$

Re: Problems with passing parameters to macros

I found the problem:

> File macros.mp:
> def someMacro(expr v, a, b, x, y, z) =
> begingroup
>
> show "inside func";
> show v, a, b;
> show "";
>
> ...
>
> show "before exiting";
> show v, a, b;
> show "";
>
> endgroup
> enddef;
>
>
>
> Another file:
>
> input macros.mp
>
>
> pair a, b, ap;
>
> ...
>
> show "before calling";
> show a, b, ap;
> show "";
> someMacro(a, b, ap, x, y, z);
>
> show "after calling";
> show a, b, ap;
>
In macro someMacro, I was declaring "pair ap" but not saving it first.  I 
was then assigning ap to another value that turned out to be the same as 
b.  ap in someMacro then conflicted with ap in the main file.  I guess 
this happens because Metapost expands the macro from whereever you called 
it from, even though the macro is defined in a different file.

I'm simplifying what happened, but this was difficult to trace because ap 
was declared in a macro that was many macros up in the stack of the 
top-level macro I was calling from the main file (macro calls within 
macro calls...).

Thanks,
Yasir
--
http://tug.org/metapost/

headers

Dan Luecking | 4 Dec 2010 00:02

$Picon$
$Favicon$

Re: Problem with generating EPS from MetaFont source

[Forgot to copy the list]

Michael Ummels <michael.ummels <at> rwth-aachen.de> wrote:

 >
 > I am trying to use MetaPost to generate EPS files for each
 > character in a
 > MetaFont as described in Appendix B of the MetaPost manual.
 > However, some
 > strokes do not come out correctly or are missing at all in the
 > resulting EPS
 > files. I would be glad if somebody could tell me what is going wrong.
 >
 > To replicate my results, run "mpost '&mfplain \mode:=localfont;
 > mag=100.375;input test'" on the file test.mf, which is given below.
 > In the resulting EPS
 > files test.2 and test.3, strokes are missing/incorrect, whereas
 > test.0 and
 > test.1 are fine.
 >
 > Interestingly, the problem does not occur when using pencircle
 > instead of
 > pensquare.

Metapost has longstanding bugs in the handling of polygonal pens.
It tries to calculate the path of the outermost corner, which can switch
from one corner to another along a curve. Unfortunately, this is a rather
unstable mathematical problem and metapost frequently gets it wrong.
Sometimes an infinitessimal rotation of the pen causes it to come out
right.

You would be advised to calculate these paths yourself and  fill the
resulting outline.

I believe Taco may know more about these issues.

Taco: I have wondered a lot about metapost's problems with polygonal
pens. If someone could point out the code that handles it (and perhaps
a description of the algorithm), I would like to take a whack at figuring
out at least where and/or why it goes wrong.

Dan

Daniel H. Luecking
Department of Mathematical Sciences
Fayetteville, Arkansas
http://www-cs-faculty.stanford.edu/~knuth/iaq.html 

--
http://tug.org/metapost/

headers

luigi scarso | 4 Dec 2010 02:12

$Picon$

Re: Problem with generating EPS from MetaFont source

On Sat, Dec 4, 2010 at 12:02 AM, Dan Luecking <luecking <at> uark.edu> wrote:
>
> [Forgot to copy the list]
>
> Michael Ummels <michael.ummels <at> rwth-aachen.de> wrote:
>
>>
>> I am trying to use MetaPost to generate EPS files for each
>> character in a
>> MetaFont as described in Appendix B of the MetaPost manual.
>> However, some
>> strokes do not come out correctly or are missing at all in the
>> resulting EPS
>> files. I would be glad if somebody could tell me what is going wrong.
>>
>> To replicate my results, run "mpost '&mfplain \mode:=localfont;
>> mag=100.375;input test'" on the file test.mf, which is given below.
>> In the resulting EPS
>> files test.2 and test.3, strokes are missing/incorrect, whereas
>> test.0 and
>> test.1 are fine.
>>
>> Interestingly, the problem does not occur when using pencircle
>> instead of
>> pensquare.
>
> Metapost has longstanding bugs in the handling of polygonal pens.
> It tries to calculate the path of the outermost corner, which can switch
> from one corner to another along a curve. Unfortunately, this is a rather
> unstable mathematical problem and metapost frequently gets it wrong.
> Sometimes an infinitessimal rotation of the pen causes it to come out
> right.
>
> You would be advised to calculate these paths yourself and  fill the
> resulting outline.
>
> I believe Taco may know more about these issues.
>
> Taco: I have wondered a lot about metapost's problems with polygonal
> pens. If someone could point out the code that handles it (and perhaps
> a description of the algorithm), I would like to take a whack at figuring
> out at least where and/or why it goes wrong.
The example doesn't compile with
MetaPost, version 1.503 (kpathsea version 6.0.0)
Log says
reloading the plain base, version 0.99: preliminaries,
 basic constants and mathematical macros,
 macros for converting units,
 macros and tables for various modes of operation,
 macros for drawing and filling,
 macros for proof labels and rules,
 macros for character and font administration,
and a few last-minute items.)
! Attempt to close the bottom level file!.
l.584 numeric mode,mag;
                            % but leave mode and mag undefined

It's ok with
This is MetaPost, version 1.211 (TeX Live 2010) (kpathsea version 6.0.0)

--

-- 
luigi

--
http://tug.org/metapost/

headers

Taco Hoekwater | 4 Dec 2010 09:48

$Gravatar$

Re: Problem with generating EPS from MetaFont source

On 12/04/2010 02:12 AM, luigi scarso wrote:
> The example doesn't compile with
> MetaPost, version 1.503 (kpathsea version 6.0.0)
> Log says

> ! Attempt to close the bottom level file!.

This is easily fixable: when you use mpost 1.503, you also need to use
the mfplain.mp that comes from mpost 1.503's distribution (or add 'dump'
at the end of the file you have on disk already).

For me, it runs past this point (but then it segfaults at finish,
which is a different, totally unrelated bug).

I'll add tracker items for each of these three problems; the one
with the polygonal pen is the hard one to fix. Dan, I'll get back
to you on the algorithm later.

Best wishes,
Taco
--
http://tug.org/metapost/

headers

luigi scarso | 4 Dec 2010 13:28

$Picon$

Re: Problem with generating EPS from MetaFont source

On Sat, Dec 4, 2010 at 9:48 AM, Taco Hoekwater <taco <at> elvenkind.com> wrote:
> On 12/04/2010 02:12 AM, luigi scarso wrote:
>>
>> The example doesn't compile with
>> MetaPost, version 1.503 (kpathsea version 6.0.0)
>> Log says
>
>> ! Attempt to close the bottom level file!.
>
> This is easily fixable: when you use mpost 1.503, you also need to use
> the mfplain.mp that comes from mpost 1.503's distribution (or add 'dump'
> at the end of the file you have on disk already).
yes, sorry for the noise. Still lots to learn.

> For me, it runs past this point (but then it segfaults at finish,
> which is a different, totally unrelated bug).
true
--

-- 
luigi
--
http://tug.org/metapost/

headers

Boguslaw Jackowski | 4 Dec 2010 15:17

$Picon$
$Favicon$

Re: Problem with generating EPS from MetaFont source


Hi,

DL> Metapost has longstanding bugs in the handling of polygonal pens. It
DL> tries to calculate the path of the outermost corner, which can switch
DL> from one corner to another along a curve. Unfortunately, this is a 
DL> rather unstable mathematical problem and metapost frequently gets it 
DL> wrong. Sometimes an infinitessimal rotation of the pen causes it to 
DL> come out right.

... and vice versa. :)

It's true that the problem of calculating the envelope of the polygonal
pen is mathematically unstable (it is not an exception in the realm of
discrete geometry), but this does not explain the bugs -- the 
representation of the path (pen's envelope) may change unstably but 
not its visual appearance. Or I am mistaken?

Cheers -- Jacko

Ps. Some theoretical details about polygonal ("Metafont-oriented") pens 
can be found in: John D. Hobby, Rasterizing Curves of Constant Width, JACM 
36(2), 1989.

--

-- 
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
  Bogus\l{}aw Jackowski: B_Jackowski <at> GUST.ORG.PL
----------------------------------------------------------------
  Hofstadter's Law: It always takes longer than you expect, even
                    when you take into account Hofstadter's Law.
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

--
http://tug.org/metapost/

headers

Stefan Witzel | 3 Dec 2010 18:01

$Picon$
$Picon$

increase numerical tolerance

Hi,

I'm drawing a picture where I want to cut ellipses inscribed into a circle before and after they meet the
circle. Mostly this works but some ellipses are not cut presumably because metapost thinks that they do
not intersect. Is there any way to increase the numerical tolerance for two paths to intersect? Since I
know for sure that the paths intersect, I could then write an operator that increases the tolerance until
the paths finally intersect. Thanks in advance!

Best regards,
Stefan Witzel
--

-- 
GRATIS! Movie-FLAT mit über 300 Videos. 
Jetzt freischalten unter http://portal.gmx.net/de/go/maxdome
--
http://tug.org/metapost/

6	May 2013
17	April 2013
33	March 2013
13	February 2013
9	January 2013
2	December 2012
7	November 2012
32	October 2012
21	September 2012
109	August 2012
13	July 2012
34	June 2012
7	May 2012
5	April 2012
14	March 2012
18	February 2012
14	January 2012
9	December 2011
35	November 2011
34	October 2011
9	September 2011
4	August 2011
46	July 2011
64	June 2011
15	May 2011
16	April 2011
14	March 2011
17	February 2011
22	January 2011
21	December 2010
42	November 2010
19	October 2010
80	September 2010
17	August 2010
1	July 2010
29	June 2010
34	May 2010
7	April 2010
6	March 2010
16	February 2010
42	January 2010
49	December 2009
28	November 2009
21	October 2009
5	September 2009
9	August 2009
21	July 2009
34	June 2009
56	May 2009
21	April 2009
46	March 2009
25	February 2009
28	January 2009
12	December 2008
16	November 2008
16	October 2008
10	September 2008
40	August 2008
41	July 2008
9	June 2008
25	May 2008
60	April 2008
37	March 2008
43	February 2008
22	January 2008
12	December 2007
95	November 2007
15	October 2007
48	September 2007
19	August 2007
12	July 2007
28	June 2007
27	May 2007
25	April 2007
32	March 2007
45	February 2007
18	January 2007
54	December 2006
51	November 2006
16	October 2006
30	September 2006
10	August 2006
3	July 2006
1	June 2004
5	May 2004
1	February 2004
8	January 2004
1	December 2003
7	November 2003
10	October 2003
28	September 2003

Mon	Tue	Wed	Thu	Fri	Sat	Sun

		1	2	3	4	5
6	7	8	9	10	11	12
13	14	15	16	17	18	19
20	21	22	23	24	25	26
27	28	29	30	31