March 2008 - ntg-pdftex - NTG Mailing Lists

[pdftex-Feature Requests][662] avoid creating "fixed destination"
by pdftex-featurerequests＠sarovar.org 26 Mar '08

26 Mar '08

Feature Requests item #662, was opened at 2006-10-20 13:57 >Status: Closed Priority: 3 Submitted By: Ulrich Diez (ulrichdiez) Assigned to: The Thanh Han (hanthethanh) >Summary: avoid creating "fixed destination" >Resolution: Rejected Group: v1.50 Category: None Initial Comment: When referencing a non-existing destination, you get this warning: ! pdfTeX warning (dest): name{<something>} has been referenced but does not exist, replaced by a fixed one I always like the warning about non-existing destinations, but sometimes I'd prefer pdfTeX not to take action besides issuing a warning-message/I'd prefer pdfTeX not to create a "fixed" destination. E.g., when creating several separate pdf-files with pdfTeX (for later merging them together by means of external software like "pdftk" or "pdfcat") whereby the named destination is not known/available when compiling file A but is available/provided in file B so that in the result from merging these files, everything will be fine. Does it make sense/is it possible to implement some mechanism for forking if creation of fixed destination shall be avoided or not? (Either a new token/primitive similar to \globaldefs which can be used within TeX source or some command- line-option which can be specified at the command-line when running pdfTeX ?) Sincerely Ulrich ---------------------------------------------------------------------- Comment By: Ulrich Diez (ulrichdiez) Date: 2008-03-25 11:24 Message: This feature-request of mine is obsolete. By using \pdfliteral and \special you can easily create links to non-existing destinations which will not get replaced by "a fixed one" so that after pdf-concatenation links will work. I apologize for having send this feature-request. Sincerely Ulrich ---------------------------------------------------------------------- Comment By: Ulrich Diez (ulrichdiez) Date: 2008-03-25 11:23 Message: This feature-request of mine is obsolete. By using \pdfliteral and \special you can easily create links to non-existing destinations which will not get replaced by "a fixed one" so that after pdf-concatenation links will work. I apologize for having send this feature-request. Sincerely Ulrich ---------------------------------------------------------------------- You can respond by visiting: http://sarovar.org/tracker/?func=detail&atid=496&aid=662&group_id=106

1 0

[pdftex-Bugs][935] The primitive \pdfstrcmp is not documented in the manuals
by pdftex-bugs＠sarovar.org 25 Mar '08

25 Mar '08

Bugs item #935, was opened at 2008-03-25 10:40 Status: Open Priority: 3 Submitted By: Ulrich Diez (ulrichdiez) Assigned to: Nobody (None) Summary: The primitive \pdfstrcmp is not documented in the manuals Resolution: None Group: None Category: None Initial Comment: The primitive \pdfstrcmp is documented neither in pdftex-syntax.txt, nor in "The pdfTeX manual (a4/letter/screen). (http://sarovar.org/docman/index.php?group_id=106&selected_doc_group_id=307&…) At comp.text.tex I was advised to send a bug-report. Ulrich ---------------------------------------------------------------------- You can respond by visiting: http://sarovar.org/tracker/?func=detail&atid=493&aid=935&group_id=106

1 0

[pdftex-Feature Requests][662] avoid creating "fixed destination"
by pdftex-featurerequests＠sarovar.org 25 Mar '08

25 Mar '08

Feature Requests item #662, was opened at 2006-10-20 11:57 Status: Open Priority: 3 Submitted By: Ulrich Diez (ulrichdiez) Assigned to: The Thanh Han (hanthethanh) Summary: avoid creating "fixed destination" Resolution: None Group: v1.50 Category: Initial Comment: When referencing a non-existing destination, you get this warning: ! pdfTeX warning (dest): name{<something>} has been referenced but does not exist, replaced by a fixed one I always like the warning about non-existing destinations, but sometimes I'd prefer pdfTeX not to take action besides issuing a warning-message/I'd prefer pdfTeX not to create a "fixed" destination. E.g., when creating several separate pdf-files with pdfTeX (for later merging them together by means of external software like "pdftk" or "pdfcat") whereby the named destination is not known/available when compiling file A but is available/provided in file B so that in the result from merging these files, everything will be fine. Does it make sense/is it possible to implement some mechanism for forking if creation of fixed destination shall be avoided or not? (Either a new token/primitive similar to \globaldefs which can be used within TeX source or some command- line-option which can be specified at the command-line when running pdfTeX ?) Sincerely Ulrich ---------------------------------------------------------------------- Comment By: Ulrich Diez (ulrichdiez) Date: 2008-03-25 10:24 Message: This feature-request of mine is obsolete. By using \pdfliteral and \special you can easily create links to non-existing destinations which will not get replaced by "a fixed one" so that after pdf-concatenation links will work. I apologize for having send this feature-request. Sincerely Ulrich ---------------------------------------------------------------------- Comment By: Ulrich Diez (ulrichdiez) Date: 2008-03-25 10:23 Message: This feature-request of mine is obsolete. By using \pdfliteral and \special you can easily create links to non-existing destinations which will not get replaced by "a fixed one" so that after pdf-concatenation links will work. I apologize for having send this feature-request. Sincerely Ulrich ---------------------------------------------------------------------- You can respond by visiting: http://sarovar.org/tracker/?func=detail&atid=496&aid=662&group_id=106

1 0

[pdftex-Feature Requests][662] avoid creating "fixed destination"
by pdftex-featurerequests＠sarovar.org 25 Mar '08

25 Mar '08

Feature Requests item #662, was opened at 2006-10-20 11:57 Status: Open Priority: 3 Submitted By: Ulrich Diez (ulrichdiez) Assigned to: The Thanh Han (hanthethanh) Summary: avoid creating "fixed destination" Resolution: None Group: v1.50 Category: Initial Comment: When referencing a non-existing destination, you get this warning: ! pdfTeX warning (dest): name{<something>} has been referenced but does not exist, replaced by a fixed one I always like the warning about non-existing destinations, but sometimes I'd prefer pdfTeX not to take action besides issuing a warning-message/I'd prefer pdfTeX not to create a "fixed" destination. E.g., when creating several separate pdf-files with pdfTeX (for later merging them together by means of external software like "pdftk" or "pdfcat") whereby the named destination is not known/available when compiling file A but is available/provided in file B so that in the result from merging these files, everything will be fine. Does it make sense/is it possible to implement some mechanism for forking if creation of fixed destination shall be avoided or not? (Either a new token/primitive similar to \globaldefs which can be used within TeX source or some command- line-option which can be specified at the command-line when running pdfTeX ?) Sincerely Ulrich ---------------------------------------------------------------------- Comment By: Ulrich Diez (ulrichdiez) Date: 2008-03-25 10:23 Message: This feature-request of mine is obsolete. By using \pdfliteral and \special you can easily create links to non-existing destinations which will not get replaced by "a fixed one" so that after pdf-concatenation links will work. I apologize for having send this feature-request. Sincerely Ulrich ---------------------------------------------------------------------- You can respond by visiting: http://sarovar.org/tracker/?func=detail&atid=496&aid=662&group_id=106

1 0

[pdftex-Feature Requests][881] Integrate pool file
by pdftex-featurerequests＠sarovar.org 23 Mar '08

23 Mar '08

Feature Requests item #881, was opened at 2007-11-13 10:48 >Status: Closed Priority: 3 Submitted By: Taco Hoekwater (taco) Assigned to: Taco Hoekwater (taco) Summary: Integrate pool file >Resolution: Accepted Group: v1.50 Category: web2c Initial Comment: So we don't forget (XeTeX is also pool-free nowadays). ---------------------------------------------------------------------- >Comment By: Martin Schröder (oneiros) Date: 2008-03-23 13:52 Message: Will be in 1.50 ---------------------------------------------------------------------- You can respond by visiting: http://sarovar.org/tracker/?func=detail&atid=496&aid=881&group_id=106

1 0

[pdftex-Bugs][933] synchronization crashes
by pdftex-bugs＠sarovar.org 23 Mar '08

23 Mar '08

Bugs item #933, was opened at 2008-03-19 02:41 >Status: Closed Priority: 3 Submitted By: Robert (schlcht) Assigned to: Nobody (None) Summary: synchronization crashes >Resolution: Works For Me >Group: v1.50.0-alpha >Category: synctex Initial Comment: pdftex 1.50 crashes on either of the following here (cygwin): % pdflatex \synctex=1 \documentclass{minimal} \stop % pdftex \synctex=1 \font\x=cmr12 x \bye ---------------------------------------------------------------------- >Comment By: Martin Schröder (oneiros) Date: 2008-03-23 13:49 Message: I can't reproduce this. Maybe it's a Cygwin issue? ---------------------------------------------------------------------- You can respond by visiting: http://sarovar.org/tracker/?func=detail&atid=493&aid=933&group_id=106

1 0

Fwd: [Sarovar.org] Downtime on 15 Mar 2008
by Martin Schröder 12 Mar '08

12 Mar '08

---------- Forwarded message ---------- From: Sarovar. org Admin <noreply(a)sarovar.org> Date: 12.03.2008 16:01 Subject: [Sarovar.org] Downtime on 15 Mar 2008 To: Martin Schröder <martin(a)oneiros.de> Hello, Sarovar.org will be down for maintenance on 15th March 2008 from 10 AM till 10 PM IST. We will be preforming a software as well as hardware upgrade during this time. While we do expect the migration to go smoothly and all data and settings to be migrated to the new server, please let us know if there are any problems after migration is complete. Thanks for using Sarovar.org and sorry for the inconvenience. Regards, Sarovar Team

1 0

Looking at a PDF/DVI while it is being produced
by David Kastrup 08 Mar '08

08 Mar '08

Hi, sometimes on a long run one would like to be able to take a look at the current state of a PDF being produced. I don't know whether there is something like a progressive PDF format: likely not. But if I could send a signal to my PDFTeX/LuaTeX process, and this would cause it (at least if it would not currently be in a \shipout) to flush the current buffers, create a copy of the current PDF/DVI file under a snapshot name, properly finish/close it (writing reference table/DVI trailers) and then resume regular work, this would be quite dandy. I often have the situation of starting a dead slow 1400 page compilation where I really need the full output, but can probably see after looking at a dozen pages whether or not the whole compilation will be rubbish, needs some minor change and restart. Being able to use a signal for "give me a viewable snapshot PDF/DVI for the current state of the document" would be rather invaluable. It would be conceivable to offer this as an option to the TeX interactive prompt, by typing >mysnap.pdf Then one could just interrupt PDFTeX using ^C as usual, and type this command for getting a snapshot. -- David Kastrup, Kriemhildstr. 15, 44793 Bochum

4 4

Re: [NTG-pdftex] [pdftex] PDF Outlines in different versions of MiKTeX
by Heiko Oberdiek 07 Mar '08

07 Mar '08

On Fri, Mar 07, 2008 at 07:56:18AM -0600, Brad Burkman wrote: > I run pdfTeX 3.141592-1.20a (MiKTeX 2.4) on one computer, and pdfTeX > 3.141529-1.40.4 (MiKTeX 2.6) on another. When I compile the following > text: > > \pdfcatalog{/PageMode/UseOutlines} > > \pdfdest name{(-19,-17),(5,25),(17,19)} fit > > \pdfoutline goto > name{(-19,-17),(5,25),(17,19)}{(-19,-17),(5,25),(17,19)} > on the earlier version, everything works just fine. With the later > version, however, I get an outline pane but no bookmarks. Just a blank > pane. Has there been a syntax change? Perhaps different versions of > Acrobat? Windows XP v/s Vista? Error message of xpdf: Error (54): Dictionary key must be a name object And the PDF shows: 1 0 obj << /S /GoTo /D (-19,-17),(5,25),(17,19) >> endobj The surrounding parentheses are missing. pdftex.web, procedure pdf_print_str(s: str_number): ... if (str_pool[i] = '(') and (str_pool[j] = ')') then begin pdf_print(s); return; end; ... * First the test is insufficient, as the example shows. The test for a hex string is better. * Also it looks like a design error. I doubt, a user would suspect that \pdfdest name {abc} fit and \pdfdest name {(abc)} fit and \pdfdest name {<616263>} fit are using the same destination name. * Another BUG: \pdfdest name{a} fit \pdfdest name{(x)} fit \end The result is a wrong destination name tree: /Names [(x) 5 0 R (a) 4 0 R] /Limits [(x) (a)] `x' is sorted as destination name "(x)", but written as destination name "x". I think, pdfTeX shouldn't assume special meanings of the destination names. The byte string from the token list of keyword "name" should be the destination name without special meanings. And if pdfTeX writes the name, it takes care of the protection of special characters. Otherwise the user would have to use \pdfdest name{(\pdfescapestring{...})}... \pdfdest name{<\pdfescapehex{...}>}... to get control over the exact name. This isn't a good design, because implementation issues leaks into the user interface. Yours sincerely Heiko <oberdiek(a)uni-freiburg.de>

1 0

[ pdftex-Bugs-741 ] wrong colors for included PDF-images (page groups)
by noreply＠sarovar.org 04 Mar '08

04 Mar '08

Bugs item #741, was opened at 2007-03-01 16:50 You can respond by visiting: http://sarovar.org/tracker/?func=detail&atid=493&aid=741&group_id=106 Category: PDF inclusion Group: v1.30.0 Status: Closed Resolution: Fixed Priority: 5 Submitted By: Uwe Stöhr (stoehr) Assigned to: Martin Schröder (oneiros) Summary: wrong colors for included PDF-images (page groups) Initial Comment: Attached is a PDF 1.4 image. When you include this to a document the output generated by pdfetex 1.30.6 the images looks different: The colors are much darker, see attached PDF result. ---------------------------------------------------------------------- Comment By: Erich Hoover (compholio) Date: 2008-03-04 20:43 Message: Logged In: YES user_id=4885 If you are using a distribution that has not caught up with the new release then I have found a work-around that may help. You can "fix" a PDF you would like to embed by using the PDF Writer backend for cups to convert your PDF to CMYK. Simply open the PDF with acroread and then use the "PDF" printer, this will place the CMYK version of the file in the "~/PDF/" folder. In a somewhat related issue, I've found that embedding some PNG files causes the same conversion problem. However, I do not see why some PNG files would do this and others would not. If someone can point me in the right direction for solving this then I'd appreciate it. ---------------------------------------------------------------------- Comment By: Martin Schröder (oneiros) Date: 2007-11-29 07:49 Message: Logged In: YES user_id=421 A fix for this will be included in the next release (1.40.6/1.50.0). ---------------------------------------------------------------------- Comment By: Martin Schröder (oneiros) Date: 2007-10-03 12:51 Message: Logged In: YES user_id=421 Fix is in the works. ---------------------------------------------------------------------- Comment By: Martin Schröder (oneiros) Date: 2007-09-30 23:08 Message: Logged In: YES user_id=421 Adding /Group to the /Page is not an option for us (and probably wrong): What should we do when we have more than one included pdf on a page and they have different /CS entries? And why does transparency in an /XObject need an entry in the /Page? The entry in the /XObject should suffice. I wouldn't be surprised if AR is faulty here. ---------------------------------------------------------------------- Comment By: Adrian Johnson (ajohnson) Date: 2007-09-30 18:04 Message: Logged In: YES user_id=4699 After adding /Group << /CS /DeviceRGB /Type /Group /S /Transparency >> to the /Page object in newfile2-1.pdf it displays correctly in Acrobat. ---------------------------------------------------------------------- Comment By: Martin Schröder (oneiros) Date: 2007-09-30 17:04 Message: Logged In: YES user_id=421 The attached file newfile2-1.pdf has a copied page group, which as I now know, should be correct. Still, AR8@Linux shows different colours for it and TestSkizze.pdf . ---------------------------------------------------------------------- Comment By: Uwe Stöhr (stoehr) Date: 2007-09-30 16:51 Message: Logged In: YES user_id=3168 > Why is this WONTFIX? Sorry, this bug was once closed when this additional info was not available. -> I reopened this bug. ---------------------------------------------------------------------- Comment By: Uwe Stöhr (stoehr) Date: 2007-09-30 16:46 Message: Logged In: YES user_id=3168 Why is this WONTFIX? I mean: > The file TestSkizze.pdf specifies the color blending space > with "/CS /DeviceRGB" in the transparency group. In > the PDF produced by pdftex this has been stripped out. So the bug is that pdftex strips out the color space. If you really think it's WONTFIX, then tell me what is the correct way of specifying the color space so that I can forward this to the Inkscape developers. ---------------------------------------------------------------------- Comment By: Adrian Johnson (ajohnson) Date: 2007-09-30 07:34 Message: Logged In: YES user_id=4699 The file TestSkizze.pdf specifies the color blending space with "/CS /DeviceRGB" in the transparency group. In the PDF produced by pdftex this has been stripped out. When the color space is not specified Acrobat defaults to CMYK. This results in RGB colors being converted to CMYK, blended in CMYK color space, then converted back to RGB. This will produce different results compared to blending the colors in RGB space. This problem was also reported to the cairo list: http://lists.cairographics.org/archives/cairo/2007-September/011471.html ---------------------------------------------------------------------- Comment By: Uwe Stöhr (stoehr) Date: 2007-03-03 11:41 Message: Logged In: YES user_id=3168 > To my eye the Harlequin results are the same for both > file, Yes and the result is correct. > while Jaws has to different results. In my eyes the Jaws results are identical but wrong. So it seems that pdftex is not the problem here, but it would be nice if pdftex could handle page groups in one of the next releases: >"PDF inclusion: Page Group detected which pdfTeX can't > handle. Ignoring it." So I'm not sure to close this bug or mark it as enhancement. Thanks for your help. ---------------------------------------------------------------------- Comment By: Martin Schröder (oneiros) Date: 2007-03-03 07:48 Message: Logged In: YES user_id=421 I've attached the results of two professional RIPs: Global Graphics Jaws & Harlequin. To my eye the Harlequin results are the same for both file, while Jaws has to different results. :-{ ---------------------------------------------------------------------- Comment By: Martin Schröder (oneiros) Date: 2007-03-02 05:16 Message: Logged In: YES user_id=421 Your document is using Page Groups (see PDF Reference 7.3.6). Including pdfs that use this is a grey area; see the notes in NEWS and pdftoepdf.cc. It works most of the times and with most RIPs. I'll look into it again, but I don't have much hope. Btw: pdftex warns you: "PDF inclusion: Page Group detected which pdfTeX can't handle. Ignoring it." ---------------------------------------------------------------------- Comment By: Uwe Stöhr (stoehr) Date: 2007-03-02 05:05 Message: Logged In: YES user_id=3168 > problem in AR7 - but only there. But how could this be a bug in Acrobat when Adobe set the standards? Why is this correctly shown before the pdftex-run. The display is the same here in Acrobat 5 - 8 so I think they would have fixed it when it's a bug since Acrobat 5. Hmm but OK, when I open the result with GSview besides the GS bug, the color tones are correctly be displayed. So close this bug report again, when you are sure that pdftex follows here exactly the PDF-specification. ---------------------------------------------------------------------- Comment By: Uwe Stöhr (stoehr) Date: 2007-03-02 04:57 Message: Logged In: YES user_id=3168 > It's a bug in AR; No! Acrobat is correct: I created this image with Inkscape 0.45 and exported it the as PDF. I designed e.g. the green as light green. And Acrobat 7 and 8 shows this exactly as designed while the pdftex result has wrong color tones, see the attached screenshots. (Ghostscript is btw. not able to render the PDF-image correctly - I reported this to the GS people who confirmed it.) ---------------------------------------------------------------------- Comment By: Martin Schröder (oneiros) Date: 2007-03-01 17:05 Message: Logged In: YES user_id=421 You don't say it, but I see the same problem in AR7 - but only there. It's a bug in AR; try kpdf or ghostscript instead. ---------------------------------------------------------------------- You can respond by visiting: http://sarovar.org/tracker/?func=detail&atid=493&aid=741&group_id=106

1 0