Norbert Preining wrote:

On Mon, 05 Feb 2007, Norbert Preining wrote:

...

Why why why? lmodern in Debian is at version 1.00 and Taco (or Hans) told me that the math fonts *are* included (and in fact I see them!) in the 1.00 version, too.

And when I take a look at the lm-hist.txt I *don*t see anything which could induce this, or am I blind (see below).

And then, please WHY does context not cry out loud something like character missing or so? I compile the document but the created documents actually misses characters. But in the log file no warning, no error, nothing. Is this the intention of ConTeXt, or is this a bug, or a feature?

missing chars are only reported by pdftex in the log file keep in mind that when tex processes a paragraph, it does not ask for each char to a macro package "am i present?" for tex, a char being defined in a tfm file is enough; only when a pdf is produced, pdftex will report missing bits and pieces (encoding vector, map files, font resources, chars in encoding but not in font, etc) such is tex Hans ----------------------------------------------------------------- Hans Hagen | PRAGMA ADE Ridderstraat 27 | 8061 GH Hasselt | The Netherlands tel: 038 477 53 69 | fax: 038 477 53 74 | www.pragma-ade.com | www.pragma-pod.nl -----------------------------------------------------------------

On Mon, Feb 05, 2007 at 06:07:47PM +0100, Norbert Preining wrote:

Good. Now that I know this I could track down the problem which was reported to something else but the lm fonts, namely to a buggy \pm command.

Same holds for \Omega. Actually, there is something wrong with the fonts; CMR fonts don't seem to be used. Here is an example: See caption for figures 4, 5, 6 etc. These should have a \Omega wherever there's a blank. Also move to the last page, the font rendering for math is far from OK. I am, in general, unhappy with the fonts: http://www.ee.iitm.ac.in/~ee03b091/ee03b091.pdf A similar document made with the old ConTeXt is: http://www.ee.iitm.ac.in/~ee03b091/ee03b091_nice.pdf The fonts render very well in Acrobat as well as have all symbols. This is what I was mentioning. Dr. Norbert Preining (Debian TeX maintainer) writes this: <quote> On Mon, 05 Feb 2007, Kumar Appaiah wrote:

OK, so it's the \pm stuff. But I am still not exactly happy with the way the fonts come in Acrobat. The characters look somewhat nicer even

Well, this is the switch to lmodern in all the fonts. Your (ee03b091_nice.pdf) document has these fonts: QGWSVP+CMR10 Type 1C yes yes no 10 0 GLMABO+CMCSC10 Type 1C yes yes no 8 0 KOSSLE+CMBX10 Type 1C yes yes no 12 0 TBOINI+CMR7 Type 1C yes yes no 29 0 IGJFUT+CMMI5 Type 1C yes yes no 27 0 YFUUGZ+CMSY7 Type 1C yes yes no 25 0 SZFXRB+CMMI7 Type 1C yes yes no 23 0 MXALVP+CMSY10 Type 1C yes yes no 21 0 KJSMVL+CMMI10 Type 1C yes yes no 19 0 DZLVPI+CMTT8 Type 1C yes yes no 53 0 KRZXUD+CMTT10 Type 1C yes yes no 65 0 while the new one ships: KOSSLE+LMRoman10-Bold Type 1C yes yes no 13 0 KHXVKD+LMRoman10-Regular Type 1C yes yes no 11 0 GLMABO+LMRoman10-CapsRegular Type 1C yes yes no 9 0 XJFWIW+CMR10 Type 1C yes yes no 25 0 VUKLKL+LMMathSymbols10-Italic Type 1C yes yes no 23 0 OPSIHW+LMMathItalic10-Italic Type 1C yes yes no 21 0 TBOINI+CMR7 Type 1C yes yes no 33 0 IGJFUT+LMMathItalic5-Italic Type 1C yes yes no 31 0 LLFODD+LMMathSymbols7-Italic Type 1C yes yes no 29 0 SZFXRB+LMMathItalic7-Italic Type 1C yes yes no 27 0 DZLVPI+LMTypewriter8-Regular Type 1C yes yes no 62 0 Complain to ConTeXt about this switch, or ask them to support the cmr fonts. This is probaly the only way to go. <unquote> So, what is the suggestion? Thanks. Kumar -- Kumar Appaiah, 462, Jamuna Hostel, Indian Institute of Technology Madras, Chennai - 600 036

On Tue, 6 Feb 2007, Kumar Appaiah wrote:

On Tue, Feb 06, 2007 at 07:09:38AM +0530, Kumar Appaiah wrote:

...

Same holds for \Omega. Actually, there is something wrong with the fonts; CMR fonts don't seem to be used. Here is an example:

See caption for figures 4, 5, 6 etc. These should have a \Omega wherever there's a blank. Also move to the last page, the font rendering for math is far from OK. I am, in general, unhappy with the fonts: http://www.ee.iitm.ac.in/~ee03b091/ee03b091.pdf

A similar document made with the old ConTeXt is: http://www.ee.iitm.ac.in/~ee03b091/ee03b091_nice.pdf The fonts render very well in Acrobat as well as have all symbols. This is what I was mentioning.

And the TeX source (to the second document actually) with EPS images, is at http://www.ee.iitm.ac.in/~ee03b091/compile.tar.bz2

Works fine here, Windows XP, latest ConTeXt standalone from pragma, and displays fine on Acrobat 7 C:\Windows\system32\cmd /c pdffonts ee03b091.pdf name type emb sub uni object ID ------------------------------------ ------------ --- --- --- --------- CXCDEM+LMRoman10-CapsRegular Type 1 yes yes yes 4 0 AHZACH+LMRoman10-Regular Type 1 yes yes yes 5 0 NEZHVQ+LMTypewriter10-Regular Type 1 yes yes yes 6 0 QGQEOY+LMRoman10-Bold Type 1 yes yes yes 7 0 DDNRFR+LMMathItalic10-Italic Type 1 yes yes yes 13 0 BWERNX+LMMathSymbols10-Italic Type 1 yes yes yes 14 0 AHZACH+LMRoman10-Regular Type 1 yes yes yes 15 0 TIIOKP+LMMathItalic7-Italic Type 1 yes yes yes 16 0 ZVJQLW+LMMathSymbols7-Italic Type 1 yes yes yes 17 0 DGERKI+LMMathItalic5-Italic Type 1 yes yes yes 18 0 LRNKJI+LMRoman7-Regular Type 1 yes yes yes 19 0 SPNZBI+LMTypewriter8-Regular Type 1 yes yes yes 32 0 I noticed that in my result the third column (uni) says yes while in yours it says no. I do not know if that is the cause of the problem Aditya

On Tue, Feb 06, 2007 at 09:05:22AM +0100, Taco Hoekwater wrote:

I see that file is created by dvips. To help with debugging, I am very interested source, dvi, tex log, dvips screen output, and ps2pdf screen output of this document.

I am also interested in the pdftex output and log.

I have tarred the whole directory and put it in the same file: http://www.ee.iitm.ac.in/~ee03b091/compile.tar.bz2 Of course, I couldn't make out much from the logs, but there do seem to be some odd things, like cont-sys.tex not being found etc., but that may be normal. Of course, more experienced persons: please throw some light. Thanks! Kumar -- Kumar Appaiah, 462, Jamuna Hostel, Indian Institute of Technology Madras, Chennai - 600 036

Kumar Appaiah wrote:

On Tue, Feb 06, 2007 at 02:07:02PM +0530, Kumar Appaiah wrote:

...

I have tarred the whole directory and put it in the same file:

http://www.ee.iitm.ac.in/~ee03b091/compile.tar.bz2

Of course, I couldn't make out much from the logs, but there do seem to be some odd things, like cont-sys.tex not being found etc., but that may be normal. Of course, more experienced persons: please throw some light.

And one more observation. dvipdfmx converts the math fonts correctly, and math symbols appear right. However, the font still doesn't look the same in Acrobat, like that which I get if I use the old ConTeXt.

how does it look when printed ... rendering is hinting dependent Hans ----------------------------------------------------------------- Hans Hagen | PRAGMA ADE Ridderstraat 27 | 8061 GH Hasselt | The Netherlands tel: 038 477 53 69 | fax: 038 477 53 74 | www.pragma-ade.com | www.pragma-pod.nl -----------------------------------------------------------------

Kumar Appaiah wrote:

On Tue, Feb 06, 2007 at 10:17:45AM +0100, Hans Hagen wrote:

...

how does it look when printed ... rendering is hinting dependent

Haven't printed it, yet, but I think it would be a fair assumption that printing would give a result as expected; i.e. it will print as well as any other document.

So, that means I shouldn't worry about the slightly bad rendering? That is all right. But I am yet to figure out why dvips/divpdf fail while dvipdfmx seems to get the math symbols right.

probably a map file problem ... does dvips have the option to be verbose (never use it here)? Hans ----------------------------------------------------------------- Hans Hagen | PRAGMA ADE Ridderstraat 27 | 8061 GH Hasselt | The Netherlands tel: 038 477 53 69 | fax: 038 477 53 74 | www.pragma-ade.com | www.pragma-pod.nl -----------------------------------------------------------------

Kumar Appaiah wrote:

On Tue, Feb 06, 2007 at 09:05:22AM +0100, Taco Hoekwater wrote:

...

I see that file is created by dvips. To help with debugging, I am very interested source, dvi, tex log, dvips screen output, and ps2pdf screen output of this document.

I am also interested in the pdftex output and log.

I have tarred the whole directory and put it in the same file:

http://www.ee.iitm.ac.in/~ee03b091/compile.tar.bz2

Of course, I couldn't make out much from the logs, but there do seem to be some odd things, like cont-sys.tex not being found etc., but

that's normal ... read cont-sys.rme for details (or copy that file to tex/context/usr/cont-sys.tex)

that may be normal. Of course, more experienced persons: please throw some light.

Thanks!

Kumar

-- ----------------------------------------------------------------- Hans Hagen | PRAGMA ADE Ridderstraat 27 | 8061 GH Hasselt | The Netherlands tel: 038 477 53 69 | fax: 038 477 53 74 | www.pragma-ade.com | www.pragma-pod.nl -----------------------------------------------------------------

Kumar Appaiah wrote:

On Mon, Feb 05, 2007 at 06:07:47PM +0100, Norbert Preining wrote:

...

Good. Now that I know this I could track down the problem which was reported to something else but the lm fonts, namely to a buggy \pm command.

Same holds for \Omega. Actually, there is something wrong with the fonts; CMR fonts don't seem to be used. Here is an example:

See caption for figures 4, 5, 6 etc. These should have a \Omega wherever there's a blank. Also move to the last page, the font rendering for math is far from OK. I am, in general, unhappy with the fonts: http://www.ee.iitm.ac.in/~ee03b091/ee03b091.pdf

can you take a look at the log file and see if it mentions missing fonts? are the lm math related map files installed? and the lm math related encoding files? (pdftex will mention missing files) Hans ----------------------------------------------------------------- Hans Hagen | PRAGMA ADE Ridderstraat 27 | 8061 GH Hasselt | The Netherlands tel: 038 477 53 69 | fax: 038 477 53 74 | www.pragma-ade.com | www.pragma-pod.nl -----------------------------------------------------------------

On Tue, Feb 06, 2007 at 02:53:41PM +0100, Norbert Preining wrote:

This is strange. I have these \Omegas occuring without any problem!

I also tried a minimal example: \starttext This should be Omega subscript s: $\Omega_s$

This should be 2 times pi cdot 11: $2\pi \cdot 11$ \stoptext and it showed up correct.

OK, so here's the diagnosis. The file you have quoted above is what I used. 1. texexec test.tex: Gives a PDF which renders all symbols. However, the font display in Acrobat isn't perfect (it's a bit light), but should be perfect when printed. 2. texexec --dvi test.tex;dvips test.dvi;ps2pdf test.ps: Same result as the previous. 3. texexec --dvi test.tex;dvips -Ppdf test.dvi;ps2pdf test.ps: Oops! \cdot is missing!!!! 4. texexec --dvi test.tex;dvipdf test.dvi: Oops! \cdot is missing again! If you still need test.log, I can upload it, but it doesn't have anything strange. Also, as Taco had suggested regarding the map file, I checked that original-public-lm.map has the entry for cmr10 etc. Finally, hoping that I'm being smart, I have bzip2ed the output of dvips -d 127 -Ppdf test.dvi. Please check the directories searched for the map file; the directory in which original-public-lm resides is /usr/share/texmf/fonts/map/pdftex/context and it doesn't appear on that list, but I'd like you to confirm. Get it at http://www.ee.iitm.ac.in/~ee03b091/dvips_output.bz2 Thanks. Kumar -- Kumar Appaiah, 462, Jamuna Hostel, Indian Institute of Technology Madras, Chennai - 600 036

On Wed, Feb 07, 2007 at 09:24:09AM +0100, Taco Hoekwater wrote:

Good thinking! But unfortunately, this won't work if you do it like this.

You have to give the same commandline as context's texexec does. Dvips does not support mapfile \specials, so unless you feed it the needed files on the commandline, it will read the wrong map file.

The used commandline is included as a comment at the top of the generated postscript file, so the best approach is to do a run of texexec --dvi and then extract the commandline from the output (and insert the -d 127 , of course)

Done that. Now, it does seem that the map files are being looked at. However, doing this results in a new PS file, where the fonts for math look awful. The specimen is here: http://www.ee.iitm.ac.in/~ee03b091/test.ps And the dvips command extracted from the PS file is: dvips -d 127 -u +texnansi-public-lm -u +original-base -u \ +ec-public-lm -u +ec-base -u +lm-ec -u +lm-math -u +8r-base -u \ +t5-base -u +original-ams-base -u +original-ams-euler -u \ original-public-lm test And the dvips output is at the same place: http://www.ee.iitm.ac.in/~ee03b091/dvips_output.bz2 Hope I'm getting somewhere. Thanks for the support! Kumar -- Kumar Appaiah, 462, Jamuna Hostel, Indian Institute of Technology Madras, Chennai - 600 036

Taco Hoekwater wrote:

Hi,

Kumar Appaiah wrote:

...

Done that. Now, it does seem that the map files are being looked at. However, doing this results in a new PS file, where the fonts for math look awful. The specimen is here:

http://www.ee.iitm.ac.in/~ee03b091/test.ps

After looking at the files, I think I now know what is going wrong.

The map file line is

cmr12 CMR12

dvips takes this to mean: you should include both files, and then try to reencode the font CMR12 into the encoding specified in the enc file. When it starts writing the output document, it discovers there is no 'CMR12' font (the pfb defines the font LMRoman12-Regular), so it does not reencode anything, after all.

It should work out ok if you change the map file line to either:

cmr12 LMRoman12-Regular

or

cmr12 CMR12

Dvips' behaviour is probably correct or at least understandable, but the fact that it does not warn the user is probably a bug or at least worrying.

this also means that such a dvips file will *not* work with pdftex because that looks at the FontName when doing inserts, i.e. the CMR12 name is used for identifying the font and then lmr is used as replacement (else we end up with files where there can be lots of cmr fonts embedded); mostly related to graphics using cmr fonts Hans ----------------------------------------------------------------- Hans Hagen | PRAGMA ADE Ridderstraat 27 | 8061 GH Hasselt | The Netherlands tel: 038 477 53 69 | fax: 038 477 53 74 | www.pragma-ade.com | www.pragma-pod.nl -----------------------------------------------------------------

Kumar Appaiah wrote:

Well, here's another piece of information; the output of updmap:

(Whole output included).

<output> updmap: This is updmap, version 1122009795-debian updmap: using transcript file `/home/kumar/.texmf-var/web2c/updmap.log'

Does this have something to do with the problem?

Nope, not at all. Context does not use updmap-generated map files. Cheers, Taco

On Wed, Feb 07, 2007 at 08:25:22PM +0530, Kumar Appaiah wrote:

...

It should work out ok if you change the map file line to either:

cmr12 LMRoman12-Regular

This didn't work.

...

or

cmr12 CMR12

This worked. Now, things are quite all right, except that I am still unhappy with the rendering in Acrobat. But at least the things seem to be mostly solved with this one.

Thanks for all the help, but one final question before I give up! IF I use pdflatex, this is the header of a PS file I get upon doing dvips: %!PS-Adobe-2.0 %%Creator: dvips(k) 5.95b Copyright 2005 Radical Eye Software %%Title: myfile.dvi %%Pages: 1 %%PageOrder: Ascend %%BoundingBox: 0 0 595 842 %%DocumentFonts: CMR17 CMR12 CMBX12 CMR10 CMMI12 CMEX10 CMSY8 CMMI8 CMR8 %%+ MSBM10 CMR7 CMBX10 However, for ConTeXt, the output I get is: %!PS-Adobe-2.0 %%Creator: dvips(k) 5.95b Copyright 2005 Radical Eye Software %%Title: test.dvi %%Pages: 1 %%PageOrder: Ascend %%BoundingBox: 0 0 595 842 %%DocumentFonts: LMRoman12-Regular CMR12 LMMathItalic9-Italic %%+ LMMathItalic12-Italic LMMathSymbols10-Italic Now, is there any way to force the fonts to map on to the old CMR names? That is because I feel it is because of the CMR/lmodern issue that pdflatex generated PDFs appear properly, while ConTeXt ones don't. What should the map file look like? Also, please find the Debian default map file here: http://www.ee.iitm.ac.in/~ee03b091/original-public-lm.map.gz Thanks for bearing with me. I want to try my best before giving up, failing which I will be forced to return to teTeX (with the old ConTeXt) and miss out on some new ConTeXt niceties, such as \startalign etc. Thanks. Kumar -- Kumar Appaiah, 462, Jamuna Hostel, Indian Institute of Technology Madras, Chennai - 600 036

On 2/8/07, Kumar Appaiah wrote:

On Wed, Feb 07, 2007 at 08:25:22PM +0530, Kumar Appaiah wrote:

...

...

It should work out ok if you change the map file line to either:

cmr12 LMRoman12-Regular

This didn't work.

...

or

cmr12 CMR12

This worked. Now, things are quite all right, except that I am still unhappy with the rendering in Acrobat. But at least the things seem to be mostly solved with this one.

Thanks for all the help, but one final question before I give up!

IF I use pdflatex, this is the header of a PS file I get upon doing dvips:

%%DocumentFonts: CMR17 CMR12 CMBX12 CMR10 CMMI12 CMEX10 CMSY8 CMMI8 CMR8 %%+ MSBM10 CMR7 CMBX10

However, for ConTeXt, the output I get is:

%%DocumentFonts: LMRoman12-Regular CMR12 LMMathItalic9-Italic %%+ LMMathItalic12-Italic LMMathSymbols10-Italic

This is (should be) correct. With LaTeX you can get the same with \usepackage{lmodern}

Now, is there any way to force the fonts to map on to the old CMR names?

There is (I guess that it might be enough to include the old map files for cmr and to remove all \definefontsynonym [cmr12][lmr10]-like definitions from type-*), but you don't want to do that! LatinModern is considered standard today. It has far more glyphs than cmr. OK, you cannot use them all with a single encoding anyway, but you don't have to bother about anything either. Support for LatinModern math has been added recently, so it might be that there are some bugs, but if there are bugs, it means that they have to be resolved before others have the same problems. With direct conversion to PDF (so, without the --dvi switch), your problems seem to disappear.

That is because I feel it is because of the CMR/lmodern issue that pdflatex generated PDFs appear properly, while ConTeXt ones don't. What should the map file look like?

What do you get if you use \usepackage{lmodern} with LaTeX?

Thanks for bearing with me. I want to try my best before giving up, failing which I will be forced to return to teTeX (with the old ConTeXt) and miss out on some new ConTeXt niceties, such as \startalign etc.

The bug has to be resolved, but why not converting your EPS graphics into PDF and using pdfTeX (in pdf mode) instead? There should be an "epstopdf" script present in the TeX distribution among other "binaries". Mojca

On Thu, Feb 08, 2007 at 08:41:31AM +0100, Mojca Miklavec wrote:

This is (should be) correct. With LaTeX you can get the same with \usepackage{lmodern}

Well, for some strange reason, it doesn't use the LM fonts.

Support for LatinModern math has been added recently, so it might be that there are some bugs, but if there are bugs, it means that they have to be resolved before others have the same problems.

With direct conversion to PDF (so, without the --dvi switch), your problems seem to disappear.

Correct.

The bug has to be resolved, but why not converting your EPS graphics into PDF and using pdfTeX (in pdf mode) instead? There should be an "epstopdf" script present in the TeX distribution among other "binaries".

Did that, and found that things are all right. I'll stop this thread now, since my problems seem to have been resolved; all but the ``slightly poorer'' display which I seem to have on Acrobat, which other's don't. But, as you say, using PDF graphics from epstopdf made things all right. Thanks a lot to all those who patiently replied! Kumar -- Kumar Appaiah, 462, Jamuna Hostel, Indian Institute of Technology Madras, Chennai - 600 036