At 22:36 20/04/2004, you wrote:

still gives different results here for pdftex 1.11b and 1.20a: 1.20a gives only 0.5 in left/upper borders. Meanwhile I believe, that 1.20a is correct, and 1.11b and before were wrong, unnoticed. One has to set the \pdf... dimensions new behind \magnification.

not that any people use magnification, it was needed in the bitmap times to get a bigger output but still in good quality; with todays outlines and fit to paper options of printer drivers ... no need for such thing Hans

On Wed, 21 Apr 2004, Hans Hagen wrote:

not that any people use magnification, it was needed in the bitmap times to get a bigger output but still in good quality; with todays outlines and fit to paper options of printer drivers ... no need for such thing

It's one robust way to make big posters with pdftex. Just do the poster layout on a smaller design page (e. g. 250mm * 350mm, maybe multi columns, grid) where the 10/11/12pt fonts (even math) work right natively without any font size tuning. Then at the document begin: 1. Calculate all page dimensions non-true, 2. set \mag, and 3. reset page dimensions to true dimens --- and you get a big poster, with fonts correctly scaled. One-file/one-step process, no PDF embedding step for ximage magnification required. Also get a small handout version fitting to A4, if \mag < 1000. Even with clickable links, which is not yet possible with pdfximage (pdfpages package)! Alternative would be poster macro packages, where all fonts are scaled in the macro package relative to the final poster size, IMHO not flexible for new fonts, much more work than just let pdftex recalculate a few page related dimensions. Regards, Hartmut

Seems \mag gives problems with seminar.cls, see comp.text.tex below. Seminar.cls uses \mag and \pdfpageheight, \pdfpagewidth, and \pdfhorigin and \pdfvorigin are not mentioned, which might result in an additional shift... Regards, Hartmut

From: "Peter Van Impe" wrong mediabox when using ifmslide/powersem/pdftex Date: Thu Apr 22 15:46:50 CEST 2004 Newsgroups: comp.text.tex Subject: Re: wrong mediabox when using ifmslide/powersem/pdftex

I'm using powersem with ifmslide on a MikTeX system to make my presentations. Everything has worked well until I performed an update just recently (yesterday acutally). Now I get pdfs with the wrong mediabox, it's as if my presentation has been zoomed in. Does anyone have an idea what would have changed to make the calculation of the page size different from the way it was before ? Neither powersem or ifmslide have changed during the last update. I'm wondering which package is responsible for this behaviour.

On 2004-04-23 15:07:33 +0200, Martin Schröder wrote:

On 2004-04-23 02:05:43 +0200, Hartmut Henkel wrote:

...

Seems \mag gives problems with seminar.cls, see comp.text.tex below. Seminar.cls uses \mag and \pdfpageheight, \pdfpagewidth, and \pdfhorigin and \pdfvorigin are not mentioned, which might result in an additional shift...

Is MikTeX distributing 1.20a?

Yes, rc1. :-{ Best regards Martin -- Martin Schröder, ms@artcom-gmbh.de ArtCom GmbH, Lise-Meitner-Str 5, 28359 Bremen, Germany Voice +49 421 20419-44 / Fax +49 421 20419-10 http://www.artcom-gmbh.de

On 2004-04-21 17:15:09 +0200, Martin Schröder wrote:

It seems reasonable to assume the same for pdf as for dvi: The origin is not moved by \magnification, i.e. the behaviour when producing pdf and dvi is the same. This would mean that rc2 is buggy.

More food for thought: The TeXbook on p.60 says: ----------------------- \ddanger Magnification is actually governed by \TeX's ^|\mag| primitive, which is an integer parameter that should be positive and at~most~32768. The value of\/ |\mag| is examined in three cases: (1)~just before the first page is shipped to the |dvi| file; (2)~when computing a |true| dimension; (3)~when the |dvi| file is being closed. Alternatively, some implementations of \TeX\ produce non-|dvi| output; they examine |\mag| in case~(2) and also when shipping out each page. Since each document has only one magnification, the value of\/ |\mag| must not change after it has first been examined. ----------------------- Best regards Martin -- Martin Schröder, ms@artcom-gmbh.de ArtCom GmbH, Lise-Meitner-Str 5, 28359 Bremen, Germany Voice +49 421 20419-44 / Fax +49 421 20419-10 http://www.artcom-gmbh.de

On Wed, 21 Apr 2004, Martin Schröder wrote:

The discussion on true (on p59f) implies that after a \mag one has to set \pdf*origin new (as plain does with \hsize and \vsize).

Yes. Or set them to right non-true dimen (truedimen / \mag) before \mag. All \pdfhorigin \pdfvorigin \pdfpagewidth \pdfpageheight should be handled consistent to TeX's other dimensions like \hoffset \voffset... Seems 1.20a does this ok. Also, no idea what \pdfhorigin and \pdfvorigin are really needed for. Regards, Hartmut