What is the best free software for creating & editing PDFs nowadays

[Marion]

In a recent thread, the perennial topic came up, which needs updating:
*Software for creating and editing PDFs*
<https://www.novabbs.com/computers/article-flat.php?id=9874&group=alt.comp.os.windows-11#9874>

Funny story: Decades ago, in the Silicon Valley, I asked my company IT
department to consider PDFs and they wrote back vehemently that they
researched what PDF was (as it was brand new, but they knew about
PostScript) and they wrote a scathing denial email saying emphatically that
they do NOT want to "support yet another standard" (Microsoft Office being
their standard at that time). Heh heh heh... I wish I saved that email...

It has been a few years... I think collectively we need to update this
chart of the single best freeware for the stated PDF editing needs...

[?] Print book format PDF (FinePrint payware)
[x] Add or concatenate pages (pdftk, acrobat payware)
[x] Add signature (Adobe Reader Fill-and-sign sign-yourself tool)
[x] Archive sites (wkhtmltopdf, Acrobat payware,fastone scroll capture)
[x] Convert PDF to MSWord or any epub format & vice versa (Calibre)
[x] Create PDF new text (Irfanview or Paint.NET plugins + Ghostscript)
[x] Edit PDF existing text (Adobe Reader commenting, Acrobat payware)
[x] Extract images (PDF Exchange Viewer, PDF Shaper)
[x] Fast PDF reader: (Sumatra or Foxit)
[x] Globally search & replace PDF text (Libre Office)
[x] Merge PDFs (pdfsam, pdftk)
[x] OCR, PDF-Xchange, freeOCR (paperfile.net), GOCR (jocr.sourceforge.net)
[x] Online shrink PDF https://www.adobe.com/acrobat/online/compress-pdf.html
[x] PDF text to audio file (Balabolka)
[x] Print sans username in the properties (Libre Office Writer)
[x] Remove pages (pdfsam, pdftk)
[x] Remove restrictions (Ghostscript & Ghostview with ps2edit & pdfwrite or pdf2djvu)
[x] Renumber pages (Acrobat Reader)
[x] Reorder pages (mutool)
[x] Rotate pages (Acrobat Reader)
[x] Shrink PDFs (ImageMagick or Acrobat payware or rlvision shareware)
[x] Tile PDFs (i.e., to print large posters) (Posterazor)
[?] What other tasks do you do to edit or modify a PDF file?

What are your suggestions (so that everyone benefits from your knowledge)?

[Lawrence D'Oliveiro]

Post by Marion
... I think collectively we need to update this
chart of the single best freeware for the stated PDF editing needs...

I prefer Free software to freeware, myself.

<https://en.wikipedia.org/wiki/Freeware>
<https://en.wikipedia.org/wiki/Free_software>

[Marion]

Post by Lawrence D'Oliveiro

Post by Marion
... I think collectively we need to update this
chart of the single best freeware for the stated PDF editing needs...

I prefer Free software to freeware, myself.
<https://en.wikipedia.org/wiki/Freeware>
<https://en.wikipedia.org/wiki/Free_software>

Thanks for the clarification where I see, from your links, that "freeware"
is about cost, while "free software" is about user rights and freedoms.

While I appreciate the gentle word-use admonition, it's kind of like when I
ask people to use "lend" as a verb vs "loan" as a noun; or when I notice
people using "further" for "farther" in terms of distances; or when people
use "less" instead of "fewer" for things that can be counted; or when
people use "dirt" when they really mean "soil"; or when they call a "stone"
a "rock" when all of these things are actually not what people think.

But not many people know those distinctions, such as what it really means
for two people to be "Platonic", although it's not as bad as when people
say "I could care less" when what they mean is the exact opposite feeling.

Taking your kind advice in hand, I see The key difference between
"freeware" and "free software" lies in the concept of freedom, not just
cost, in terms of who retains copyright and controls distribution and
modification, specifically the freedom to run the program for any purpose,
such as to study how the program works and to maybe change it, and maybe
even redistribute modified versions.

I accept your suggestion to keep in mind that free software is often
available at no cost, not all software available at no cost is free
software.

With that taken care of to an appropriate level of clarification, what I
ask the team at large to help out for, is to flesh out this table.

What else is needed to be done with a PDF file & which programs do it?
[?] Print book format PDF (FinePrint payware)
[x] Add or concatenate pages (pdftk, acrobat payware)
[x] Add signature (Adobe Reader Fill-and-sign sign-yourself tool)
[x] Archive sites (wkhtmltopdf, Acrobat payware,fastone scroll capture)
[x] Convert PDF to MSWord or any epub format & vice versa (Calibre)
[x] Create PDF new text (Irfanview or Paint.NET plugins + Ghostscript)
[x] Edit PDF existing text (Adobe Reader commenting, Acrobat payware)
[x] Extract images (PDF Exchange Viewer, PDF Shaper)
[x] Fast PDF reader: (Sumatra or Foxit)
[x] Globally search & replace PDF text (Libre Office)
[x] Merge PDFs (pdfsam, pdftk)
[x] OCR, PDF-Xchange, freeOCR (paperfile.net), GOCR (jocr.sourceforge.net)
[x] Online shrink PDF https://www.adobe.com/acrobat/online/compress-pdf.html
[x] PDF text to audio file (Balabolka)
[x] Print sans username in the properties (Libre Office Writer)
[x] Remove pages (pdfsam, pdftk)
[x] Remove restrictions (Ghostscript & Ghostview with ps2edit & pdfwrite or pdf2djvu)
[x] Renumber pages (Acrobat Reader)
[x] Reorder pages (mutool)
[x] Rotate pages (Acrobat Reader)
[x] Shrink PDFs (ImageMagick or Acrobat payware or rlvision shareware)
[x] Tile PDFs (i.e., to print large posters) (Posterazor)
[?] What other common tasks do you do to edit or modify a PDF file?

[Marion]

Post by Marion
With that taken care of to an appropriate level of clarification, what I
ask the team at large to help out for, is to flesh out this table.

Password protecting a .pdf file.

Hi Rick,

I've written tutorials on how to REMOVE PDF restrictions, but I never
thought about *adding* PDF restrictions, such as password protection.

So I've added to the chart of things people want to do with a PDF:
[x] Offline encrypt PDF with a password (pdfencrypt)

<https://pdfencrypt.net/>
<https://pdfencrypt.net/files/setup.exe>
Name: setup.exe
Size: 5515925 bytes (5386 KiB)
SHA256: 7D5B37F986EBC374772CB749FAA4B4DA39D466D14A03BEE76B13800E59131992

What else is needed to do with a PDF that we haven't discussed in the past?

[Lawrence D'Oliveiro]

Post by Marion
[x] Add or concatenate pages
[x] Merge PDFs
[x] Print sans username in the properties
[x] Remove pages
[x] Reorder pages

Just note that the PikePDF toolkit is very handy for performing all these
tasks from a Python script. As an example use of it, I wrote this command-
line tool <https://gitlab.com/ldo/acrid>, which lets you examine and
change/add/delete the metadata associated with a PDF file -- both the old-
style format and the XMP format.

[Marion]

Post by Lawrence D'Oliveiro
Just note that the PikePDF toolkit is very handy for performing all these
tasks from a Python script. As an example use of it, I wrote this command-
line tool <https://gitlab.com/ldo/acrid>, which lets you examine and
change/add/delete the metadata associated with a PDF file -- both the old-
style format and the XMP format.

Thanks for the suggestion of PikePDF, which I was wholly unaware of, since
the list was taken from discussions on the windows newsgroups over time.

I'm surprised I missed a mention of PDF-related freeware, so I searched.

First, I searched the c.t.p & a.c.o.w-10 archives for mention of PikePDF:
<https://groups.google.com/g/comp.text.pdf>
<https://www.novabbs.com/computers/search.php?group=comp.text.pdf>
<https://www.novabbs.com/computers/search.php?group=alt.comp.os.windows-10>

The Google Groups c.t.p search returned zero hits for "PikePDF".
The Nova BBS c.t.p search returned one hit.
*How to remove a link in a PDF that is found in a thousand pages*
<https://www.novabbs.com/computers/article-flat.php?id=363&group=comp.text.pdf#363>
Likewise, the Nova BBS a.c.o.w-10 search returned that same hit.
<https://www.novabbs.com/computers/article-flat.php?id=79154&group=alt.comp.os.windows-10#79154>
The contents are (verbatim, in toto):
Write a program using a PDF-manipulation toolkit.
I have had good results writing Python code using pikepdf
<https://github.com/pikepdf/pikepdf>

At that web github location is the following description:
PikePDF: A Python library for reading and writing PDF, powered by QPDF
Documentation: <https://pikepdf.readthedocs.io/en/latest/>
"pikepdf is a library intended for developers who want to create,
manipulate, parse, repair, and abuse the PDF format.
It supports reading and write PDFs, including creating from scratch.
Thanks to QPDF, it supports linearizing PDFs and access to
encrypted PDFs. It is a low level library that requires knowledge
of PDF internals and some familiarity with the PDF specification.
It does not provide a user interface of its own."

First, I had to look up what "QPFD" was:
<https://sourceforge.net/projects/qpdf/>
"QPDF is a C++ library and set of programs that inspect & manipulate
the structure of PDF files. It can encrypt and linearize files,
expose the internals of a PDF file, and do many other operations
useful to end users and PDF developers."

Linearization, by the way, is optimizing (usually for the web).

Apparently QPDF is intended to perform content-preserving transformations
of PDF files by changing PDF structures without altering visual contents.

The description of PikePDF provides the following examples of what it does:
[x]Pikepdf would help you build apps that do things like:
A cartoon sketch of a pike
[x]Copy pages from one PDF into another
[x]Split and merge PDFs
[x]Extract content from a PDF such as images
[x]Replace content
such as replacing an image without altering the rest of the file
[x]Repair, reformat or linearize PDFs
[x]Change the size of pages and reposition content
[x]Optimize PDFs similar to Acrobat's features by downsampling images,
[x]deduplicating
[x]Calculate charges for a scanning project based on the materials scanned
[x]Alter a PDF to meet a target specification such as PDF/A or PDF/X
[x]Add or modify PDF metadata
[x]Add, remove, extract, and modify PDF attachments (i.e. embedded files)
[x]Create well-formed but invalid PDFs for testing purposes

Bingo! *Add or modify PDF metadata*

OK. I've confirmed what you've explained to us, which is the combination of
Python scripts, PikePDF and the underlying QPDF can remove PDF metadata.

I appreciate your suggested site <https://gitlab.com/ldo/acrid>, which will
help the programmers here who can take advantage of your kind offering.

Even though I go back to the sixties and seventies in programming (COBOL,
Fortran77 before there was a IV, IBM Assembly, Motorola 68701, etc.) I
swore off programming at some point, so I can't make use of these tools.

But others who know a lot more than I do about programming certainly can.
I would like to add it to the summary chart but it may be too eclectic.

A more readily available program to remove metadata might be LibreOffice.
Also PDFgear online/offline tools <https://www.pdfgear.com/>
Also PDF24 online tools <https://tools.pdf24.org/en/remove-pdf-metadata>
Also Sejda online tools <https://www.sejda.com/edit-pdf-metadata>

This is getting long so let's break off a tangent for metadata removal.
Suffice to say that removal of metadata is critically important, which
means it behooves us to find an easy way for everyone to be able to do it.

[Marion]

Post by Marion
A more readily available program to remove metadata might be LibreOffice.
Also PDFgear online/offline tools <https://www.pdfgear.com/>
Also PDF24 online tools <https://tools.pdf24.org/en/remove-pdf-metadata>
Also Sejda online tools <https://www.sejda.com/edit-pdf-metadata>
This is getting long so let's break off a tangent for metadata removal.
Suffice to say that removal of metadata is critically important, which
means it behooves us to find an easy way for everyone to be able to do it.

My goal, always, is to help everyone with every post, so this post will
delve into details since I'm all about solving the problem for everyone.

Let's break into a tangent for the removal of metadata, which is critically
important for privacy, where let's just state what can be in PDF metadata.

Some of the basic information hidden in PDF metadata can be:
Title: The name of the document.
Author: The person or entity who created the document.
Subject: A brief description of the document's content.
Keywords: Terms that describe the document's content, used for searching.
Creation Date: The date and time when the PDF was created.
Modification Date: The date and time when the PDF was last modified.
Creator: The application used to create the PDF.
Producer: The application used to convert the document to PDF.

While more advanced hidden information in PDF metadata might be:
XMP Metadata: More extensive and customizable metadata
Rights Management: Information related to copyright and usage permissions.
Security Settings: Details about encryption and access restrictions.
Accessibility Metadata: Assistive technologies to help interpret the PDF.
Embedded File Metadata: Embedded files (like images) have metadata too!

What irks me on metadata is that in my Adobe Acrobat version 6 payware, I
can't remove my PC username in the PDF metadata, which stinks for privacy.

However, what also is an issue are online metadata-removal tools, where,
well, having grown up during the Cold War where I had to duck and cover, I
am leery of anything online, especially an online privacy protection tool.

So if I discount the following online tools for removal of metadata
[x]PDFgear online tools <https://www.pdfgear.com/>
[x]PDF24 online tools <https://tools.pdf24.org/en/remove-pdf-metadata>
[x]Sejda online tools <https://www.sejda.com/edit-pdf-metadata>

That's leaves us, currently, with the following for metadata removal:
LibreOffice <https://www.libreoffice.org/download/download-libreoffice/>
PDFGear <https://www.pdfgear.com/pdfgear-for-windows/>

I don't need to delve into LibreOffice for most of the people here.
So let's take up how PDFGear does the removal of PDF metadata.
<https://downloadfiles.pdfgear.com/releases/windows/pdfgear_setup_v2.1.12.exe>
Name: pdfgear_setup_v2.1.12.exe
Size: 136412680 bytes (130 MiB)
SHA256: C8A19A4A06FB8D28812916FF1735CD4DC0F82BF16FBC5100BBEB71A44F32CCF9
Defaults to: C:\Program Files\PDFgear

Upon launching, PDFGear phones home via the default browser (e.g., TOR):
<https://www.pdfgear.com/congrats/?action=install>

Wow. I mean wow. This is kind of like Calibre, upon first inspection.
Just wow. It does a lot. Pretty much PDF to anything (e.g., PDF to Word).
[Of course, with varying levels of "doing something" but that's for later.]

Let's just figure out how to use PDFGear, offline, for metadata removal.
This is getting long in the tooth, so let's close this article with that.

[Marion]

Post by Marion
Let's just figure out how to use PDFGear, offline, for metadata removal.

Given everything I do is intended also for the benefit of everyone else,
first I need a sample PDF that has metadata that everyone can access.

Let's try this:
<https://www.hekatron.de/fileadmin/user_upload/testfolder/Sample.pdf>
Name: Sample.pdf
Size: 37545 bytes (36 KiB)
SHA256: 2534C7C146709BD2881BF2A791F44A3EFC31A7230EA0D400AA17E3E8FE5DE279

Good. In Adobe Acrobat 6 payware, there is metadata that we can test.
Acrobat6:File > Document Properties > Description >
Title: PDF Metadata Sample
Author: Nigel Maddocks
Subject: Test Document
Created: 8/21/2015 1:42:21 AM <== harder to remove than you'd think
Modified: 8/21/2015 1:45:31 AM
Application: Acrobat PDFMaker 15 for Word <== harder to remove also
etc.

Since not everyone has Acrobat payware, I'll describe how to look at the
metadata using PDFGear (so that everyone benefits from every action).

Looking at the UserGuide, this appears to be the procedure (simplified):
<https://www.pdfgear.com/windows-user-guide/introduction-pdfgear.htm>

a. C:\Program Files\PDFgear\PDFLauncher.exe
b. PDFgear:Open File > Sample.pdf
c. PDFgear:Help > Document Properties
d. You can manually remove some, but not all the document properties
[x]Title
[x]Author
[x]Subject
[x]Keywords
[_]Creator
[_]Producer
[_]Created
[_]Modified
etc.
e. Open in another tool to check if the metadata was removed.

Given this worked (for some degree of "working") to remove the most
egregious metadata, can I declare this a success for the team?
[x] Metadata removal (LibreOffice Writer, PDFGear offline)

[Lawrence D'Oliveiro]

Post by Marion
<https://www.hekatron.de/fileadmin/user_upload/testfolder/Sample.pdf>

ldo> acrid showinfo Sample.pdf
{
"/ModDate": "D:20150821094531+01'00'",
"/Subject": "Test Document",
"/Title": "PDF Metadata Sample",
"/Comments": "",
"/Author": "Nigel Maddocks",
"/CreationDate": "D:20150821094221+01'00'",
"/Keywords": "12345678",
"/Producer": "Adobe PDF Library 15.0",
"/Creator": "Acrobat PDFMaker 15 for Word",
"/Company": "",
"/SourceModified": "D:20150821084155",
"Metadata": «see below»
}

ldo> acrid getxmp Sample.pdf
<?xpacket begin="" id="W5M0MpCehiHzreSzNTczkc9d"?>
<x:xmpmeta xmlns:x="adobe:ns:meta/" x:xmptk="Adobe XMP Core 5.6-c015 81.157285, 2014/12/12-00:43:15 ">
<rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#">
<rdf:Description rdf:about=""
xmlns:xmp="http://ns.adobe.com/xap/1.0/"
xmlns:xmpMM="http://ns.adobe.com/xap/1.0/mm/"
xmlns:dc="http://purl.org/dc/elements/1.1/"
xmlns:pdf="http://ns.adobe.com/pdf/1.3/"
xmlns:pdfx="http://ns.adobe.com/pdfx/1.3/">
<xmp:ModifyDate>2015-08-21T09:45:31+01:00</xmp:ModifyDate>
<xmp:CreateDate>2015-08-21T09:42:21+01:00</xmp:CreateDate>
<xmp:MetadataDate>2015-08-21T09:45:31+01:00</xmp:MetadataDate>
<xmp:CreatorTool>Acrobat PDFMaker 15 for Word</xmp:CreatorTool>
<xmpMM:DocumentID>uuid:32f55d7c-7ef1-46ca-85c0-b5b91509ff82</xmpMM:DocumentID>
<xmpMM:InstanceID>uuid:4da38cf4-2b42-417c-b34e-f529c34ac6cc</xmpMM:InstanceID>
<xmpMM:subject>
<rdf:Seq>
<rdf:li>1</rdf:li>
</rdf:Seq>
</xmpMM:subject>
<dc:format>application/pdf</dc:format>
<dc:title>
<rdf:Alt>
<rdf:li xml:lang="x-default">PDF Metadata Sample</rdf:li>
</rdf:Alt>
</dc:title>
<dc:description>
<rdf:Alt>
<rdf:li xml:lang="x-default">Test Document</rdf:li>
</rdf:Alt>
</dc:description>
<dc:creator>
<rdf:Seq>
<rdf:li>Nigel Maddocks</rdf:li>
</rdf:Seq>
</dc:creator>
<dc:subject>
<rdf:Bag>
<rdf:li>12345678</rdf:li>
</rdf:Bag>
</dc:subject>
<pdf:Producer>Adobe PDF Library 15.0</pdf:Producer>
<pdf:Keywords>12345678</pdf:Keywords>
<pdfx:SourceModified>D:20150821084155</pdfx:SourceModified>
<pdfx:Company/>
<pdfx:Comments/>
</rdf:Description>
</rdf:RDF>
</x:xmpmeta>





















<?xpacket end="w"?>"

[Marion]

Post by Lawrence D'Oliveiro

Post by Marion
<https://www.hekatron.de/fileadmin/user_upload/testfolder/Sample.pdf>

ldo> acrid showinfo Sample.pdf

Definitely works nicely. I've added the suggestions that I think made sense
as a general use chart for Windows users, whose current version is below.

[?] Print book format PDF (FinePrint payware)
[x] Add or concatenate pages (pdftk, acrobat payware)
[x] Add signature (Adobe Reader Fill-and-sign sign-yourself tool)
[x] Archive sites (wkhtmltopdf, Acrobat payware,fastone scroll capture)
[x] Compress PDFs (ImageMagick, PDFgear, rlvision)
[x] Convert PDF to MSOffice (PDFgear, Calibre for MS Word only)
[x] Convert PDF to MSWord (Calibre, PDFgear)
[x] Convert PDF to epub format (Calibre)
[x] Convert PDF to PostScript (Calibre, Poppler)
[x] Converts PDFs to HTML (poppler)
[x] Converts PDFs to PNG, JPEG, etc (poppler) using Cairo graphics
[x] Converts PDFs to PPM/PGM/PBM image formats (poppler)
[x] Create PDF new text (Irfanview or Paint.NET plugins + Ghostscript)
[x] Edit PDF existing text (Adobe Reader commenting, Acrobat payware)
[x] Embeds files into a PDF as attachments (poppler)
[x] Extract images (PDF Exchange Viewer, PDF Shaper, PDFgear, poppler)
[x] Extract text (poppler)
[x] Extracts embedded files (attachments) from a PDF (poppler)
[x] Fastest PDF readers (Sumatra or Foxit)
[x] Globally search & replace PDF text (Libre Office)
[x] List fonts used in a PDF (poppler)
[x] Merge PDFs (pdfsam, pdftk, PDFgear, Poppler)
[x] Metadata display on command line (poppler)
[x] Metadata removal (LibreOffice Writer, PDFgear offline)
[x] OCR, PDF-Xchange, freeOCR (paperfile.net), GOCR (jocr.sourceforge.net)
[x] Offline encrypt PDF with a password (pdfencrypt)
[x] Online shrink PDF
https://www.adobe.com/acrobat/online/compress-pdf.html
[x] PDF text to audio file (Balabolka)
[x] Remove pages (pdfsam, pdftk)
[x] Remove restrictions (Ghostscript,Ghostview,ps2edit,pdfwrite,pdf2djvu)
[x] Renumber pages (Acrobat Reader)
[x] Reorder pages (mutool)
[x] Rotate pages (Acrobat Reader)
[x] Separates a PDF into individual pages (Poppler)
[x] Split PDFs (PDFgear, Poppler)
[x] Tile PDFs (i.e., to print large posters) (Posterazor)
[?] What other tasks do you do to edit or modify a PDF file?

I'm sure we're missing more important functionality than we have,
but so far I think this takes into effect the suggestions to date.

[Peter Flynn]

Post by Marion

Post by Lawrence D'Oliveiro

 <https://www.hekatron.de/fileadmin/user_upload/testfolder/Sample.pdf>

ldo> acrid showinfo Sample.pdf

Definitely works nicely. I've added the suggestions that I think made sense
as a general use chart for Windows users, whose current version is below.

[snip]

Thank you, this is a hugely useful list.

The one tool missing seems to be LaTeX, for creating PDFs, but perhaps
"create" in this context means "convert from some other typeset format"
rather than "typeset directly to PDF" (in industry terms, "originate")

Peter

[Lawrence D'Oliveiro]

Post by Peter Flynn
The one tool missing seems to be LaTeX, for creating PDFs, but perhaps
"create" in this context means "convert from some other typeset format"
rather than "typeset directly to PDF" (in industry terms, "originate")

Along those lines, is it also worth mentioning that the Cairo graphics
library includes the option for rendering drawing to PDF, among its range
of output surface types?

<https://www.cairographics.org/manual/cairo-PDF-Surfaces.html>

[Marion]

Post by Lawrence D'Oliveiro

Post by Peter Flynn
The one tool missing seems to be LaTeX, for creating PDFs, but perhaps
"create" in this context means "convert from some other typeset format"
rather than "typeset directly to PDF" (in industry terms, "originate")

Along those lines, is it also worth mentioning that the Cairo graphics
library includes the option for rendering drawing to PDF, among its range
of output surface types?
<https://www.cairographics.org/manual/cairo-PDF-Surfaces.html>

Regarding...
x] Converts PDFs to PNG, JPEG, etc (poppler) using Cairo graphics
Fist, I have to make two confessions to maintain my credibility.

The first is: I cheat.

I always use the ancient payware Adobe Acrobat version 6 whenever I need to
convert a PDF into image formats - so I really don't know much about the
concept of converting a PDF to raster or vector graphics images.

The second confession is that I don't really know what other people want to
do when they "say" they want to convert a PDF into an image format.

But when I researched your poppler suggestion, I saw it did that well.

I had never heard of Cairo until I dug into your suggestion of using it.
Apparently poppler uses either the "Splash" or the "Cairo" graphics libs to
translate PDF instructions into graphical drawing commands.

Apparently Cairo drawing commands (i.e., pdftocairo) render the PDF onto
something called a "surface" which itself can be either a raster image
(PNG, JPEG, etc.) or a vector surface (DXF, SVG, etc.) for high quality.

Since I use the payware, I don't know what other solutions exist, so I dug
into the concept a bit to find that if you feed Inkscape a PDF, it attempts
to interpret the vector elements (lines, shapes, text) to make them
editable (which could be powerful for those changing PDF images).

In addition, it seems Inkscape can trace a bitmap to vectorize raster
images within a PDF into vector paths - which is useful for scaling.

On the other hand, I found in my searches today that ImageMagick is
apparently very good at converting PDFs into raster image formats
(like PNG, JPEG, PPM & TIFF).

Different from both Inkscape (vector) & ImageMagick (raster) is
Ghostscript, which can rasterize PDFs, but using a command line (unless you
combine it with GhostView) where Ghostscript splits & merges PDFs also.

Having confused myself with what I said above, I should ask what people are
trying to do with the image when they want to convert PDF to an image.

And do they want vector images out of the PDF. Or raster?
[x] Convert PDF to raster (Imagemagick,GhostScript,Poppler-pdftocairo)
[x] Convert PDF to vector (Inkscape, Poppler-pdftocairo)

Anything else (which is free to use)?

[Lawrence D'Oliveiro]

Post by Marion
Different from both Inkscape (vector) & ImageMagick (raster) is
Ghostscript, which can rasterize PDFs ...

Yes. It is essentially a full-function PostScript interpreter, it just has
lots of options for the output format. Rasterizing PDF was probably a
relatively minor function to add on top of that.

[Marion]

Thank you, this is a hugely useful list.'

Thanks. It needs updating but it covers most of what I've needed to do.
The hard part is keeping it to a single line per need, which necessitates
excessive shortening of the descriptions.

I think you're talking about this line item below:
[x] Create PDF new text (Irfanview or Paint.NET plugins + Ghostscript)
Where that's really about ADDING text to an existing PDF document.

Given the confusion inherent in the lousy way I wrote it, I'll change it to
[x] Add text to existing pdf (Irfanview or Paint.NET plugins + Ghostscript)

Moving forward on your point below with LaTeX, I agree with you that we
need a line item for creating PDFs from scratch using a markup language.

The one tool missing seems to be LaTeX, for creating PDFs, but perhaps
"create" in this context means "convert from some other typeset format"
rather than "typeset directly to PDF" (in industry terms, "originate")

Given LaTeX is the de facto standard for creating mathematical and
scientific documents, I agree with you that it belongs as a line item.
[x] Generate complex PDF using markup language (LaTeX via pdfTeX or LuaTeX)

When I researched what else that is no cost which generates PDFs, most were
programming libraries, such as ReportLab, PDFKit, jsPDF & PDFSharp.

Contrasting with those programming libraries (which require programming
code), LaTeX is a markup language and typesetting system. We write the
document's content and structure using LaTeX commands, and LaTeX handles
the visual formatting to PDF.

So I won't include the programming libraries in that new line, for now.
Does that clarify the two lines better for 'creating' & 'generating' PDF?

[x] Add text to existing pdf (Irfanview or Paint.NET plugins + Ghostscript)
[x] Generate complex PDF using markup language (LaTeX via pdfTeX or LuaTeX)

[Carlos E.R.]

Post by Marion
Given LaTeX is the de facto standard for creating mathematical and
scientific documents, I agree with you that it belongs as a line item.
[x] Generate complex PDF using markup language (LaTeX via pdfTeX or LuaTeX)

LyX is a visual editor (WYSIWYM, "what you see is what you mean"
approach), that can generate PDFs and other formats. It is related to
LaTeX but is not the same. Although it probably uses external libraries
to do the actual conversion.

Libre Office can also generate PDFs, including cryptographically signed
documents, probably using libraries. It can also edit PDFs.

[Peter Flynn]

[snip]

Post by Marion
Given LaTeX is the de facto standard for creating mathematical and
scientific documents, I agree with you that it belongs as a line item.

[...]> Contrasting with those programming libraries (which require
programming

Post by Marion
code), LaTeX is a markup language and typesetting system. We write the
document's content and structure using LaTeX commands, and LaTeX handles
the visual formatting to PDF.

There are two routes to PDF if you have XML documents (increasingly
common; and both Word and Libre Office are XML inside). Both use
Extensible Stylesheet Language (XSL) but in different ways

But these may be well outside the scope of your list as they are
two-stage processes.

• XSL-FO uses XSL to describe the transformation to Formatting Objects
(FO) and an FO processor converts that to PDF

• XSLT uses XSL to describe the transformation to any text format,
including LaTeX, which can then produce PDF.

XSL-FO is no longer being developed by the W3C; however both methods are
in common use in publishing.

Post by Marion
So I won't include the programming libraries in that new line, for now.
Does that clarify the two lines better for 'creating' & 'generating' PDF?
[x] Add text to existing pdf (Irfanview or Paint.NET plugins + Ghostscript)
[x] Generate complex PDF using markup language (LaTeX via pdfTeX or LuaTeX)

Yes, that looks fine, thanks.

Peter

[Lawrence D'Oliveiro]

Post by Peter Flynn
There are two routes to PDF if you have XML documents (increasingly
common; and both Word and Libre Office are XML inside). Both use
Extensible Stylesheet Language (XSL) but in different ways

Does anybody still use SGML? Remember, that gave birth to HTML.

[Peter Flynn]

Post by Lawrence D'Oliveiro

Post by Peter Flynn
There are two routes to PDF if you have XML documents (increasingly
common; and both Word and Libre Office are XML inside). Both use
Extensible Stylesheet Language (XSL) but in different ways

Does anybody still use SGML? Remember, that gave birth to HTML.

And TEI. And DocBook. And XML. And dozens of industrial vocabularies.

I believe a very small number of projects still use SGML for specialist
technical reasons, or possibly tied to obsolete software. Projects I was
associated with moved to XML the moment viable processing software
became available. But SGML still works, and so does the old software.

Peter

[Lawrence D'Oliveiro]

Post by Marion
Thanks for the suggestion of PikePDF, which I was wholly unaware of,
since the list was taken from discussions on the windows newsgroups over
time.

Yes, there is a difference in mentality between a gaggle of users
accustomed to isolated, monolithic applications versus one based on a
cooperating ecosystem of interlocking toolkits.

Here’s another PDF toolkit: Poppler. This is a more extensive one, that
covers both the creation and rendering of PDF files. For example, Inkscape
relies on Poppler when you ask it to import pages from a PDF file into
your illustration.

[Marion]

Here's another PDF toolkit: Poppler. This is a more extensive one, that
covers both the creation and rendering of PDF files. For example, Inkscape
relies on Poppler when you ask it to import pages from a PDF file into
your illustration.

You know your PDF tools for sure!

Apparently Poppler is a free and open-source software command-line tool
library used more on Linux than on Windows, who naming was inspired by
<https://en.wikipedia.org/wiki/The_Problem_with_Popplers>

But there are also Windows binaries, apparently.
<https://poppler.freedesktop.org/>
<https://github.com/oschwartz10612/poppler-windows>
<https://github.com/oschwartz10612/poppler-windows/releases/tag/v24.08.0-0>
Name: Release-24.08.0-0.zip
Size: 15090263 bytes (14 MiB)
SHA256: 58A6F9AE269756231D2F9AA6CBA39D75FEC6DEACAF3C4A50683383B5F3D5A527

The poppler-utils are used to
[x]Extract text from PDFs (pdftotext).
[x]Extract images from PDFs (pdfimages).
[x]Convert PDFs to other formats (pdftoppm, pdftocairo, pdftohtml).
[x]Get information about PDFs (pdfinfo).
[x]Merge and separate pdf files.

To test, I downloaded it and extracted it to Windows to see these:
pdfattach.exe: Embeds files into a PDF as attachments.
pdfdetach.exe: Extracts embedded files (attachments) from a PDF.
pdffonts.exe: Lists the fonts used in a PDF.
pdfimages.exe: Extracts images from a PDF.
pdfinfo.exe: Displays information about a PDF.
pdfseparate.exe: Separates a PDF into individual pages.
pdftocairo.exe: Converts PDFs to PNG, JPEG, etc. using Cairo graphics
pdftohtml.exe: Converts PDFs to HTML.
pdftoppm.exe: Converts PDFs to PPM/PGM/PBM image formats.
pdftops.exe: Converts PDFs to PostScript.
pdftotext.exe: Extracts text from a PDF.
pdfunite.exe: Merges multiple PDFs into one.
zstd.exe: Compresses or decompresses (requires Zstandard)

Let's just try "pdfinfo" on the sample PDF we previously downloaded:
<https://www.hekatron.de/fileadmin/user_upload/testfolder/Sample.pdf>

pdfinfo Sample.pdf
Title: PDF Metadata Sample
Subject: Test Document
Keywords: 12345678
Author: Nigel Maddocks
Creator: Acrobat PDFMaker 15 for Word
Producer: Adobe PDF Library 15.0
CreationDate: Fri Aug 21 02:42:21 2015 Mountain Daylight Time
ModDate: Fri Aug 21 02:45:31 2015 Mountain Daylight Time
Custom Metadata: yes
Metadata Stream: yes
Tagged: yes
UserProperties: no
Suspects: no
Form: none
JavaScript: no
Pages: 1
Encrypted: no
Page size: 595.32 x 841.92 pts (A4)
Page rot: 0
File size: 37545 bytes
Optimized: yes
PDF version: 1.5

The sample is too simple to extract images or detach files included with
the PDF, but I'm sure those other functions work, so it's a nice addition.

Seems like a nice tool. Too bad it doesn't remove metadata as it would be
nice to run a find for all pdf files and to strip out the metadata in them.

[Lawrence D'Oliveiro]

Perhaps another one I should mention is PDFMiner. This is a bit of a
specialist one, focused on extracting text items from a PDF page, and
using various heuristics to try to reassemble them into larger text
blocks.

[Marion]

Post by Lawrence D'Oliveiro
Perhaps another one I should mention is PDFMiner. This is a bit of a
specialist one, focused on extracting text items from a PDF page, and
using various heuristics to try to reassemble them into larger text
blocks.

Thank you for adding value to the spirit of this conversation where the PDF
experts and editing experts are involved, along with the Windows users.

Looking up what the PDFMiner Python tool can do for us, it's important to
note it's apparently designed for extracting information from PDF files.

I'm not quite sure how PDFMinor differs from any of the other text
extractors (e.g., PDF to TEXT) but it seems to gather layout data also.

While it can extract metadata, it seems to me it's mostly used to "mine"
large assemblages of PDF files for textual data of interest to the user.

The original PDFMiner has apparently been forked as pdfminer.six, which,
as far as I can tell from date stamps, is still actively being updated.
<https://github.com/euske/pdfminer>
<https://github.com/pdfminer/pdfminer.six>

Since it functions on windows, (within the python enviroment) and since it
does something useful (mine text in PDFs), I'll add it to the PDF chart as
[x] Extract text (poppler) or mine textual & metadata (pdfminersix)

Here's the current chart, where I simply ask for more things done to PDFs.
[?] Print book format PDF (FinePrint payware)
[x] Add or concatenate pages (pdftk, acrobat payware)
[x] Add signature (Adobe Reader Fill-and-sign sign-yourself tool)
[x] Archive sites (wkhtmltopdf, Acrobat payware,fastone scroll capture)
[x] Compress PDFs (ImageMagick, PDFgear, rlvision)
[x] Convert PDF to MSOffice (PDFgear, Calibre for MS Word only)
[x] Convert PDF to MSWord (Calibre, PDFgear)
[x] Convert PDF to epub format (Calibre)
[x] Convert PDF to PostScript (Calibre, Poppler)
[x] Converts PDFs to HTML (poppler)
[x] Converts PDFs to PNG, JPEG, etc (poppler) using Cairo graphics
[x] Converts PDFs to PPM/PGM/PBM image formats (poppler)
[x] Create PDF new text (Irfanview or Paint.NET plugins + Ghostscript)
[x] Edit PDF existing text (Adobe Reader commenting, Acrobat payware)
[x] Embeds files into a PDF as attachments (poppler)
[x] Extract images (PDF Exchange Viewer, PDF Shaper, PDFgear, poppler)
[x] Extract text (poppler) or mine textual & metadata (pdfminersix)
[x] Extracts embedded files (attachments) from a PDF (poppler)
[x] Fastest PDF readers (Sumatra or Foxit)
[x] Globally search & replace PDF text (Libre Office)
[x] List fonts used in a PDF (poppler)
[x] Merge PDFs (pdfsam, pdftk, PDFgear, Poppler)
[x] Metadata display on command line (poppler)
[x] Metadata removal (LibreOffice Writer, PDFgear offline)
[x] OCR, PDF-Xchange, freeOCR (paperfile.net), GOCR (jocr.sourceforge.net)
[x] Offline encrypt PDF with a password (pdfencrypt)
[x] Online shrink PDF <adobe.com/acrobat/online/compress-pdf.html>
[x] PDF text to audio file (Balabolka)
[x] Remove pages (pdfsam, pdftk)
[x] Remove restrictions (Ghostscript,Ghostview,ps2edit,pdfwrite,pdf2djvu)
[x] Renumber pages (Acrobat Reader)
[x] Reorder pages (mutool)
[x] Rotate pages (Acrobat Reader)
[x] Separates a PDF into individual pages (poppler)
[x] Split PDFs (PDFgear, Poppler)
[x] Tile PDFs (i.e., to print large posters) (Posterazor)
[?] What other tasks do you do to edit or modify a PDF file?

[G]

Post by Marion

Post by Lawrence D'Oliveiro
Perhaps another one I should mention is PDFMiner. This is a bit of a
specialist one, focused on extracting text items from a PDF page, and
using various heuristics to try to reassemble them into larger text
blocks.

Thank you for adding value to the spirit of this conversation where the PDF
experts and editing experts are involved, along with the Windows users.
Looking up what the PDFMiner Python tool can do for us, it's important to
note it's apparently designed for extracting information from PDF files.
I'm not quite sure how PDFMinor differs from any of the other text
extractors (e.g., PDF to TEXT) but it seems to gather layout data also.
While it can extract metadata, it seems to me it's mostly used to "mine"
large assemblages of PDF files for textual data of interest to the user.
The original PDFMiner has apparently been forked as pdfminer.six, which,
as far as I can tell from date stamps, is still actively being updated.
<https://github.com/euske/pdfminer>
<https://github.com/pdfminer/pdfminer.six>
Since it functions on windows, (within the python enviroment) and since it
does something useful (mine text in PDFs), I'll add it to the PDF chart as
[x] Extract text (poppler) or mine textual & metadata (pdfminersix)
Here's the current chart, where I simply ask for more things done to PDFs.
[?] Print book format PDF (FinePrint payware)
[x] Add or concatenate pages (pdftk, acrobat payware)
[x] Add signature (Adobe Reader Fill-and-sign sign-yourself tool)
[x] Archive sites (wkhtmltopdf, Acrobat payware,fastone scroll capture)
[x] Compress PDFs (ImageMagick, PDFgear, rlvision)
[x] Convert PDF to MSOffice (PDFgear, Calibre for MS Word only)
[x] Convert PDF to MSWord (Calibre, PDFgear)
[x] Convert PDF to epub format (Calibre)
[x] Convert PDF to PostScript (Calibre, Poppler)
[x] Converts PDFs to HTML (poppler)
[x] Converts PDFs to PNG, JPEG, etc (poppler) using Cairo graphics
[x] Converts PDFs to PPM/PGM/PBM image formats (poppler)
[x] Create PDF new text (Irfanview or Paint.NET plugins + Ghostscript)
[x] Edit PDF existing text (Adobe Reader commenting, Acrobat payware)
[x] Embeds files into a PDF as attachments (poppler)
[x] Extract images (PDF Exchange Viewer, PDF Shaper, PDFgear, poppler)
[x] Extract text (poppler) or mine textual & metadata (pdfminersix)
[x] Extracts embedded files (attachments) from a PDF (poppler)
[x] Fastest PDF readers (Sumatra or Foxit)
[x] Globally search & replace PDF text (Libre Office)
[x] List fonts used in a PDF (poppler)
[x] Merge PDFs (pdfsam, pdftk, PDFgear, Poppler)
[x] Metadata display on command line (poppler)
[x] Metadata removal (LibreOffice Writer, PDFgear offline)
[x] OCR, PDF-Xchange, freeOCR (paperfile.net), GOCR (jocr.sourceforge.net)
[x] Offline encrypt PDF with a password (pdfencrypt)
[x] Online shrink PDF <adobe.com/acrobat/online/compress-pdf.html>
[x] PDF text to audio file (Balabolka)
[x] Remove pages (pdfsam, pdftk)
[x] Remove restrictions (Ghostscript,Ghostview,ps2edit,pdfwrite,pdf2djvu)
[x] Renumber pages (Acrobat Reader)
[x] Reorder pages (mutool)
[x] Rotate pages (Acrobat Reader)
[x] Separates a PDF into individual pages (poppler)
[x] Split PDFs (PDFgear, Poppler)
[x] Tile PDFs (i.e., to print large posters) (Posterazor)
[?] What other tasks do you do to edit or modify a PDF file?

I suppose all are available for Windows, it would be useful to know which are
also for Linux or Mac.

G

[Marion]

Post by G
I suppose all are available for Windows, it would be useful to know which are
also for Linux or Mac.

Here's the Windows software, so far, that I've tested for PDF manipulation.
<Loading Image...>

However, your point about testing Linux/Mac software is well taken, given
the c.e and c.t.p newsgroups will have folks on the Linux & Mac newsgroups.

My history is that I cut my teeth on IBM assembly, cobol, fortran 77, etc.,
so I grew up on PDP11, DEC/VMS, SunOS, Solaris, etc., well before my first
real Linux (Redhat) & then Centos & Ubuntu, so I agree Linux is important.

And, in the Silicon Valley corporate atmosphere, long ago we all used the
Mac before the dual boot Windows/Redhat became our standard desktop PC.

Since I always download and test (almost) every suggested pgm, what's
needed for me to add Linux would be to dual boot to test these apps out.

For a long while I dual booted to Ubuntu in the Unity days and as an
indirect result, I personally abandoned the dual boot before Unity
(thankfully) died (although I had switched the desktop by then to KDE).

The point of that history being that while I *agree* Linux is important,
I'm not going to be who affirms what works and what doesn't, for Linux.

Likewise, while I have probably more Apple mobile devices than most people,
being a substitute teacher has taught me that the Mac isn't designed with
anywhere near, oh, shall we say "freedoms", as anything that I'm used to.

So, while both the Mac & Linux are widely available for PDF manipulation,
I'm not going to be the one to flesh out what works & what doesn't on 'em.

But someone else can take up the banner and run with it, as this is Usenet.
Here's a "dir /b" of my pspdf archive on Windows for software to consider:
acrobat
bullzip
calibre
cutepdf
fileoptimizer
fineprint
foxit
ghoststuff
msoffice_save_as_pdf
mupdf
ocr
pdf-xchange_viewer
pdf2office
pdf_text_to_audio
pdfcreator
pdfencrypt
pdfgear
pdfminersix
pdfsam
pdfshaper
pdftk
pdfxchange
pdfxv
poppler
posterazor
psutils
sumatra
wkhtmltox
wps_pdf2word
xpdf

[G]

Post by Marion
My history is that I cut my teeth on IBM assembly, cobol, fortran 77, etc.,
so I grew up on PDP11, DEC/VMS, SunOS, Solaris, etc., well before my first
real Linux (Redhat) & then Centos & Ubuntu, so I agree Linux is important.

I wrote my first program on punch card... for a IBM (maybe a 340?) and the lab
had half the punchers(?) from honeywell, which, of course had different
character set, Fun!.
And then the usual as you but avoided SunOs and SOlaris and went with Xenix
and HP-UX and of course CP/M and DOS. I installed my first Linux (Yggdrasil,
15 or so Floppy) on a Compaq Presario(I think) with a Massive 10MB HD!
Than I bought a RedHat box and kept with it, so Fedora Core and Fedora now
(KDE). I haven't used Windows for work for almost 30 years, so no ulcers.

[snip]

Thanks for your work here...

G

[Marion]

Post by G
I wrote my first program on punch card... for a IBM (maybe a 340?) and the lab
had half the punchers(?) from honeywell, which, of course had different
character set, Fun!.

Yeah. You reminded me of the computer rooms with the raised floors (for the
A/C) and the big magtapes (I still have one somewhere). When I wrote my
first program in school, it was in punched cards and Fortran 77 on an IBM,
oh, maybe an 1130?

I remember the punched tape machine sat there, unused, while we employed
the "more modern" boxes of punched cards. Do I remember correctly that a
box was about two thousand lines of Fortran code? Most of my code was about
a quarter to, at most, a half a box, so that's probably 500 to 1000 lines,
excluding the obligatory IBM JCL.

At some point (late seventies?) I was writing in PL/1 and about a decade
later in hex (when I wire wrapped my own Motorola 68701 microcontrollers).

At some point in time I swore off programming languages after concluding
that they all do teh same damn thing, only with different syntax. :)

I got sick of the syntax requirements the older I got. :)
Now the only thing I 'program' in is the Windows command line. :)

That's why when Lawrence mentioned Python, I shuddered. I still have
nightmares about having to look up "Error 52" or something like that in the
IBM 1130 documentation. Kids nowadays have no idea how that used to be! :)

[Lawrence D'Oliveiro]

Post by Marion
At some point in time I swore off programming languages after concluding
that they all do teh same damn thing, only with different syntax. :)

New ways of viewing the programming problem often lead to major
improvements in programmer productivity.

Remember Brooks’ Law (or one of them): once a code base reaches a certain
size, a skilled programmer that is familiar with it is only able to
contribute about 10 lines of suitably-debugged code per day. And that
applies across a wide range of language abstraction levels, from assembler
all the way up to what he called “metaprogramming” languages, or very-
high-level languages. Consider how little 10 lines of assembler can do,
versus 10 lines of shell script or Python or Lisp code.

So the only way to improve programmer productivity is to move to higher-
and higher-level languages.

[Tim Slattery]

Post by Marion

Post by G
I wrote my first program on punch card... for a IBM (maybe a 340?) and the lab
had half the punchers(?) from honeywell, which, of course had different
character set, Fun!.

Yeah. You reminded me of the computer rooms with the raised floors (for the
A/C) and the big magtapes (I still have one somewhere). When I wrote my
first program in school, it was in punched cards and Fortran 77 on an IBM,
oh, maybe an 1130?

For me it was an IBM 1620. This was the late '60s at Palo Alto High
School. The machine belonged to the school district. The school
district office was right next to the school, so we got to use their
machine.

Post by Marion
I remember the punched tape machine sat there, unused, while we employed
the "more modern" boxes of punched cards. Do I remember correctly that a
box was about two thousand lines of Fortran code? Most of my code was about
a quarter to, at most, a half a box, so that's probably 500 to 1000 lines,
excluding the obligatory IBM JCL.

A box of Hollerith (or IBM) cards held 2,000 cards. Each FORTRAN
statement would go on a separate card, so 2,000 FORTRAN statements is
right. And if you dropped your box and spilled your cards, good luck
getting them back in the correct order! BTW: the 1620 was pre-JCL.

[Peter Flynn]

On 04/03/2025 15:23, Tim Slattery wrote:
[snip]

Post by Tim Slattery
A box of Hollerith (or IBM) cards held 2,000 cards. Each FORTRAN
statement would go on a separate card, so 2,000 FORTRAN statements
is right. And if you dropped your box and spilled your cards, good
luck getting them back in the correct order!

In my college, the computing centre had a card sorter, which was huge
and stood on cast-iron lion's feet which someone had painted gold :-)
But of course it only worked if your program statements or data lines
(cards) were numbered. You only drop a box of cards once.

(You may hear the voice of experience there :-)

Peter

[Don_from_AZ]

Post by G
[snip]

Post by Tim Slattery
A box of Hollerith (or IBM) cards held 2,000 cards. Each FORTRAN
statement would go on a separate card, so 2,000 FORTRAN statements
is right. And if you dropped your box and spilled your cards, good
luck getting them back in the correct order!

In my college, the computing centre had a card sorter, which was huge
and stood on cast-iron lion's feet which someone had painted gold :-)
But of course it only worked if your program statements or data lines
(cards) were numbered. You only drop a box of cards once.
(You may hear the voice of experience there :-)
Peter

I worked for Honeywell at a WWMMCCS site (World-Wide Military Command
and Control System) at the Washington Navy Yard as a tech support
guy. This was about 1973. I created new boot decks for the Honeywell 635
as needed, with new patch cards or new configurations. I was carrying a
tray of punched cards to the computer room, one hand on each end of the
card tray. I tried to hook the door handle with my little finger to pull
it open and lost my grip on the tray; cards all over the floor!
Embarrasing to say the least, and I didn't even try to put them back in
order, just punched out a new deck.

[Paul]

Post by Don_from_AZ

Post by G
[snip]

Post by Tim Slattery
A box of Hollerith (or IBM) cards held 2,000 cards. Each FORTRAN
statement would go on a separate card, so 2,000 FORTRAN statements
is right. And if you dropped your box and spilled your cards, good
luck getting them back in the correct order!

In my college, the computing centre had a card sorter, which was huge
and stood on cast-iron lion's feet which someone had painted gold :-)
But of course it only worked if your program statements or data lines
(cards) were numbered. You only drop a box of cards once.
(You may hear the voice of experience there :-)
Peter

I worked for Honeywell at a WWMMCCS site (World-Wide Military Command
and Control System) at the Washington Navy Yard as a tech support
guy. This was about 1973. I created new boot decks for the Honeywell 635
as needed, with new patch cards or new configurations. I was carrying a
tray of punched cards to the computer room, one hand on each end of the
card tray. I tried to hook the door handle with my little finger to pull
it open and lost my grip on the tray; cards all over the floor!
Embarrasing to say the least, and I didn't even try to put them back in
order, just punched out a new deck.

You could put numbers in column 72.

Loading Image...

You generally also need some spacing between card numbers, like use 10,20,30
then if you added a card it could be 25, then a card between 20 and 25, could
be card 23. You needed a means to support program edits.

If you needed a program label line number, that went on the left
of the card, while anti-spill card numbering went on the right of the card.
This is an example of an infinite loop, and if I spilled these two cards
on the floor, 25 comes before 30 and we're all good. The 100 on the left
is a program label.

100 GOTO 100 25
CONTINUE 30

For small student programs, nobody bothered with column 72. For larger
projects, it was considered a necessary evil.

The easiest way to get the cards numbered, would be to copy the deck to
the output punch and have column 72 content added automatically. I don't
think I ever used the output punch on a mainframe. Maybe my QDGS deck
had column 72, because that was a box-full (2000 cards for some purpose).
I don't recollect where that deck came from -- maybe someone had it
punched for me. The output came in a box (the Ice Queen probably put
the 2000 cards in a box, after they were punched). The ice Queen was
the emotionless computer operator, she wore a heavy sweater year round,
because it was like 50F inside the computer room. I'd been on a tour
of that computer room, and it's like working in the flash freezer
at the fish plant. It is COLD in there. That's why she was the
Ice Queen, as she had mastery over ice, and the ice could not get to her.
Nothing got to her. I don't think the expression on her face, changed
even once. She didn't even have a name! That's how emotionless she was.

Paul

[Daniel70]

Post by Paul

Post by Don_from_AZ

On 04/03/2025 15:23, Tim Slattery wrote: [snip]

Post by Tim Slattery
A box of Hollerith (or IBM) cards held 2,000 cards. Each
FORTRAN statement would go on a separate card, so 2,000 FORTRAN
statements is right. And if you dropped your box and spilled
your cards, good luck getting them back in the correct order!

In my college, the computing centre had a card sorter, which was
huge and stood on cast-iron lion's feet which someone had painted
gold :-) But of course it only worked if your program statements
or data lines (cards) were numbered. You only drop a box of cards
once.
(You may hear the voice of experience there :-)
Peter

I worked for Honeywell at a WWMMCCS site (World-Wide Military
Command and Control System) at the Washington Navy Yard as a tech
support guy. This was about 1973. I created new boot decks for the
Honeywell 635 as needed, with new patch cards or new
configurations. I was carrying a tray of punched cards to the
computer room, one hand on each end of the card tray. I tried to
hook the door handle with my little finger to pull it open and lost
my grip on the tray; cards all over the floor! Embarrasing to say
the least, and I didn't even try to put them back in order, just
punched out a new deck.

You could put numbers in column 72.
https://upload.wikimedia.org/wikipedia/commons/thumb/5/58/FortranCardPROJ039.agr.jpg/960px-FortranCardPROJ039.agr.jpg
You generally also need some spacing between card numbers, like use
10,20,30 then if you added a card it could be 25, then a card between
20 and 25, could be card 23. You needed a means to support program
edits.

That was the reason stated when I started BASIC Programming (1985'ish
.... O./K., so I was a late comer!!) "Number the Lines 10, 20, 30, etc,
so, if you need to add a bit extra, there were all those other line
numbers to use!!"

[Lawrence D'Oliveiro]

Post by Daniel70
That was the reason stated when I started BASIC Programming (1985'ish
.... O./K., so I was a late comer!!) "Number the Lines 10, 20, 30, etc,
so, if you need to add a bit extra, there were all those other line
numbers to use!!"

Some of us went by hundreds.

BASIC was designed as an integral part of an interactive timeshared system
for students and staff to use at Dartmouth. Line numbers served two
purposes: one as target labels for GOTOs, the other for ordering lines in
the editor.

Both uses are now obsolete.

By the way, did you have some kind of line-renumbering utility in your
BASIC system? That would also fix up GOTOs to correctly branch to the
renumbered lines.

[Daniel70]

Post by Lawrence D'Oliveiro

Post by Daniel70
That was the reason stated when I started BASIC Programming
(1985'ish .... O./K., so I was a late comer!!) "Number the Lines
10, 20, 30, etc, so, if you need to add a bit extra, there were all
those other line numbers to use!!"

Some of us went by hundreds.
BASIC was designed as an integral part of an interactive timeshared
system for students and staff to use at Dartmouth. Line numbers
served two purposes: one as target labels for GOTOs, the other for
ordering lines in the editor.

Hmm!! Probably O.T. but the 'Dartmouth' you mention .... was that
Dartmouth, Victoria, Australia or was that somewhere in U.K.?? Or
elsewhere??

Post by Lawrence D'Oliveiro
Both uses are now obsolete.
By the way, did you have some kind of line-renumbering utility in
your BASIC system? That would also fix up GOTOs to correctly branch
to the renumbered lines.

"line-renumbering utility"?? One's mind, maybe. ;-P

[Philip Herlihy]

In article <vqbhsv$2sjkp$***@dont-email.me>, ***@eternal-
september.org says...

Post by Daniel70
Hmm!! Probably O.T. but the 'Dartmouth' you mention .... was that
Dartmouth, Victoria, Australia or was that somewhere in U.K.?? Or
elsewhere??

Dartmouth College, Hanover, New Hampshire.

https://calltolead.dartmouth.edu/stories/celebrating-birth-basic-and-
beyond

[Frank Slootweg]

Post by Lawrence D'Oliveiro

Post by Daniel70
That was the reason stated when I started BASIC Programming (1985'ish
.... O./K., so I was a late comer!!) "Number the Lines 10, 20, 30, etc,
so, if you need to add a bit extra, there were all those other line
numbers to use!!"

Some of us went by hundreds.
BASIC was designed as an integral part of an interactive timeshared system
for students and staff to use at Dartmouth. Line numbers served two
purposes: one as target labels for GOTOs, the other for ordering lines in
the editor.
Both uses are now obsolete.

Was that by any chance one of the HP 2000 Series Timesharing Systems?

I supported those in the early 70s, mainly at the technical
'hogescholen' (one level below university) in The Netherlands.

Post by Lawrence D'Oliveiro
By the way, did you have some kind of line-renumbering utility in your
BASIC system? That would also fix up GOTOs to correctly branch to the
renumbered lines.

Wasn't 'renumber' (or 'ren'?) just a command? Or is my memory playing
tricks with me?

[Carlos E.R.]

Post by Frank Slootweg

Post by Lawrence D'Oliveiro
By the way, did you have some kind of line-renumbering utility in your
BASIC system? That would also fix up GOTOs to correctly branch to the
renumbered lines.

Wasn't 'renumber' (or 'ren'?) just a command? Or is my memory playing
tricks with me?

It could be an external program (in basic itself), that when you run it
loads some basic source (it is plain text) at renumbers it according to
some criteria. It has got to find all jumps and fill a table.

[Peter Flynn]

Post by Lawrence D'Oliveiro
By the way, did you have some kind of line-renumbering utility in your
BASIC system? That would also fix up GOTOs to correctly branch to the
renumbered lines.

AFAIR both HP BASIC and DEC-10 BASIC had line-renumbering (into 10s)
that updated GOTO statements. I believe the DEC-10 BASIC was a direct
descendant of Dartmouth BASIC, so presumably someone in DEC spotted the
need and added the feature.

Peter

[Lawrence D'Oliveiro]

Post by Peter Flynn

Post by Lawrence D'Oliveiro
By the way, did you have some kind of line-renumbering utility in your
BASIC system? That would also fix up GOTOs to correctly branch to the
renumbered lines.

AFAIR both HP BASIC and DEC-10 BASIC had line-renumbering (into 10s)
that updated GOTO statements. I believe the DEC-10 BASIC was a direct
descendant of Dartmouth BASIC, so presumably someone in DEC spotted the
need and added the feature.

I recall on RSTS/E, there was a separate utility program called RESEQ that
did this for BASIC-PLUS programs. It was a long time ago, but it could
have been that it did not operate on .BAS source code at all, but on the
byte-compiled executable .BAC files.

[Carlos E.R.]

Post by Paul

Post by Don_from_AZ

Post by G
[snip]

Post by Tim Slattery
A box of Hollerith (or IBM) cards held 2,000 cards. Each FORTRAN
statement would go on a separate card, so 2,000 FORTRAN statements
is right. And if you dropped your box and spilled your cards, good
luck getting them back in the correct order!

In my college, the computing centre had a card sorter, which was huge
and stood on cast-iron lion's feet which someone had painted gold :-)
But of course it only worked if your program statements or data lines
(cards) were numbered. You only drop a box of cards once.
(You may hear the voice of experience there :-)
Peter

I worked for Honeywell at a WWMMCCS site (World-Wide Military Command
and Control System) at the Washington Navy Yard as a tech support
guy. This was about 1973. I created new boot decks for the Honeywell 635
as needed, with new patch cards or new configurations. I was carrying a
tray of punched cards to the computer room, one hand on each end of the
card tray. I tried to hook the door handle with my little finger to pull
it open and lost my grip on the tray; cards all over the floor!
Embarrasing to say the least, and I didn't even try to put them back in
order, just punched out a new deck.

You could put numbers in column 72.
https://upload.wikimedia.org/wikipedia/commons/thumb/5/58/FortranCardPROJ039.agr.jpg/960px-FortranCardPROJ039.agr.jpg

I can not read punched cards.

The lab at uni used them just the year before me. I never had to use
cards or punched paper tape.

:-D

The trouble was, there were not enough terminals for all the students. I
tried going somewhere where you could rent a computer by the hour.
Finally I decided I needed a computer of my own, a PC clone. I was
fortunate to have a family that could afford it.

[Paul]

Post by Carlos E.R.
I can not read punched cards.
The lab at uni used them just the year before me. I never had to use cards or punched paper tape.
:-D
The trouble was, there were not enough terminals for all the students. I tried going somewhere where you could rent a computer by the hour. Finally I decided I needed a computer of my own, a PC clone. I was fortunate to have a family that could afford it.

My hobby was self-financed as a student,
so everything had to be "cheep! cheep! cheep!" :-)

Real computers had to wait until I had a job.

Paul

[Paul]

Post by Carlos E.R.
I can not read punched cards.

This is a card from a Model 29. It has inked characters along the top,
so you could read the ASCII character equivalent of the 12 row Hollerith punch.

Loading Image...

Not all card schemes, were that friendly.

You could compare the inked characters on the card, to your 132 column line printer output.

Paul

[Carlos E.R.]

Post by Paul

Post by Carlos E.R.
I can not read punched cards.

This is a card from a Model 29. It has inked characters along the top,
so you could read the ASCII character equivalent of the 12 row Hollerith punch.
http://www.columbia.edu/cu/computinghistory/029-card.jpg

Yes, I have seen those (samples).

Post by Paul
Not all card schemes, were that friendly.
You could compare the inked characters on the card, to your 132 column line printer output.
Paul

[Frank Slootweg]

Carlos E.R. <***@es.invalid> wrote:
[...]

Post by Carlos E.R.
I can not read punched cards.
The lab at uni used them just the year before me. I never had to use
cards or punched paper tape.
:-D

I had colleagues who could - visually - read ASCII paper tapes. And
another one could visually read 9-track magtape, when the magnetzation
was made visible by some kind of fluid.

[...]

[Paul]

Post by Frank Slootweg
[...]

Post by Carlos E.R.
I can not read punched cards.
The lab at uni used them just the year before me. I never had to use
cards or punched paper tape.
:-D

I had colleagues who could - visually - read ASCII paper tapes. And
another one could visually read 9-track magtape, when the magnetzation
was made visible by some kind of fluid.
[...]

When you use paper tape regularly, you learn how to recognize
end of record marks. When a tape load reports a read error,
you can roll back a record and retry, which takes less time
than loading the paper tape all over again. Learning to do that,
is a "survival mechanism" :-)

Paul

[Marion]

Post by Paul
When you use paper tape regularly, you learn how to recognize
end of record marks. When a tape load reports a read error,
you can roll back a record and retry, which takes less time
than loading the paper tape all over again. Learning to do that,
is a "survival mechanism" :-)

Speaking of such survival mechanisms, when I was burning EPROMs for the
Motorola 68701 (probably mid 80's time frame) that I wire wrapped myself, I
would write down the Assembly Language instructions, at first, on my own.

Then, after a while, it was a "survival mechanism" to just use the hex
instead, as what's the difference between "load accumulator A" (LDAA) and
the hex (86) or for extended addressing, (B6) for the same command.

It's not that big of a stretch to remember 86 versus LDAA, and it helps a
lot when it came to burning the EEPROMs (which the 68701 had internally).

Those days are over and gone, never to return.
I won't need that EEPROM burner any more than a dwellmeter & timing light.

[Paul]

Post by Marion

Post by Paul
When you use paper tape regularly, you learn how to recognize
end of record marks. When a tape load reports a read error,
you can roll back a record and retry, which takes less time
than loading the paper tape all over again. Learning to do that,
is a "survival mechanism" :-)

Speaking of such survival mechanisms, when I was burning EPROMs for the
Motorola 68701 (probably mid 80's time frame) that I wire wrapped myself, I
would write down the Assembly Language instructions, at first, on my own.
Then, after a while, it was a "survival mechanism" to just use the hex
instead, as what's the difference between "load accumulator A" (LDAA) and
the hex (86) or for extended addressing, (B6) for the same command.
It's not that big of a stretch to remember 86 versus LDAA, and it helps a
lot when it came to burning the EEPROMs (which the 68701 had internally).
Those days are over and gone, never to return.
I won't need that EEPROM burner any more than a dwellmeter & timing light.

There are still EEPROMs. They're too convenient to throw away.

For example, you can store the program code for an FPGA
(Field Programmable Gate Array) inside an EEPROM.

Since they keep finding new uses for them, and they
keep going into different shaped packages, you'll never
really be rid of them.

Paul

[Frank Slootweg]

Post by Paul

Post by Frank Slootweg
[...]

Post by Carlos E.R.
I can not read punched cards.
The lab at uni used them just the year before me. I never had to use
cards or punched paper tape.
:-D

I had colleagues who could - visually - read ASCII paper tapes. And
another one could visually read 9-track magtape, when the magnetzation
was made visible by some kind of fluid.
[...]

When you use paper tape regularly, you learn how to recognize
end of record marks. When a tape load reports a read error,
you can roll back a record and retry, which takes less time
than loading the paper tape all over again. Learning to do that,
is a "survival mechanism" :-)

Indeed! When we were 'generating' (installing) an HP RTE (Real Time
Executive) system, we needed to load dozens of paper tapes, each of
which came in a about 10x10cm box, so rather long papertapes. When you
got a 'checksum error', which you were nearly guaranteed to get, you
didn't want to start all over again.

The system was installed *to* disk, but the (binaries) input came from
paper tape. Later we got disk drives with a fixed disk and a removable
diskpack, so we got 'Grandfather' diskpacks to generate a system,
instead of using papertape. Bliss!

[Carlos E.R.]

Post by Frank Slootweg
[...]

Post by Carlos E.R.
I can not read punched cards.
The lab at uni used them just the year before me. I never had to use
cards or punched paper tape.
:-D

I had colleagues who could - visually - read ASCII paper tapes. And
another one could visually read 9-track magtape, when the magnetzation
was made visible by some kind of fluid.

Wow.

Post by Frank Slootweg
[...]

[Anton Shepelev]

Post by Marion
It has been a few years... I think collectively we need to
update this chart of the single best freeware for the
stated PDF editing needs...
[?] Print book format PDF (FinePrint payware)

I believe LaTeX has packages for that. I have produced PDF
booklets from Postscrpt, with psbook and psnup. The
incoming PostScipt was mine, from either LaTeX or GNU Troff.

Post by Marion
[x] Convert PDF to MSWord or any epub format & vice versa (Calibre)
[x] Edit PDF existing text (Adobe Reader commenting, Acrobat payware)
[x] Globally search & replace PDF text (Libre Office)

PDF is meant to be a final format not meant for editing.
Keep it so.

Post by Marion
[x] Extract images (PDF Exchange Viewer, PDF Shaper)

SumatraPDF

Post by Marion
[x] Reorder pages (mutool)
[x] Rotate pages (Acrobat Reader)

pdftk of course.

Post by Marion
What are your suggestions (so that everyone benefits from
your knowledge)?

The obvious one -- typsetting software for producing PDFs
from text, e.g.: LaTeX, (GNU) Troff.

[Peter Flynn]

On 04/03/2025 16:13, Anton Shepelev wrote:
[snip]

Post by Anton Shepelev
PDF is meant to be a final format not meant for editing.
Keep it so.

Nevertheless, I have several times been able to make on-the-fly changes
and even introduce additional material like running headers or
paper-type change statements with

pdf2ps file.pdf | sed -e "<stuff>" | ps2pdf newfile.pdf

Peter

[Lawrence D'Oliveiro]

Post by Peter Flynn

Post by Anton Shepelev
PDF is meant to be a final format not meant for editing.
Keep it so.

Nevertheless, I have several times been able to make on-the-fly changes
and even introduce additional material like running headers or
paper-type change statements with
pdf2ps file.pdf | sed -e "<stuff>" | ps2pdf newfile.pdf

Remember that a PDF file is not required to contain anything resembling
readable text strings. If present, these are commonly stored separately
from the rendered glyphs you seen on-screen, in a form designed for
searching, together with an indication of which part of the page to
highlight to represent a selection of that text. But this is all optional.

[Marion]

Post by Anton Shepelev

Post by Marion
[?] Print book format PDF (FinePrint payware)

I believe LaTeX has packages for that. I have produced PDF
booklets from Postscrpt, with psbook and psnup. The
incoming PostScipt was mine, from either LaTeX or GNU Troff.

Thanks for that suggestion as, in the past, I printed booklets.

Printing a booklet requires arranging both sides of the pages in a specific
order so that when the 8.5x11-inch printed sheets are folded in half, the
pages appear in the correct sequence as if they were in a booklet.

Post by Anton Shepelev

Post by Marion
[x] Convert PDF to MSWord or any epub format & vice versa (Calibre)
[x] Edit PDF existing text (Adobe Reader commenting, Acrobat payware)
[x] Globally search & replace PDF text (Libre Office)

PDF is meant to be a final format not meant for editing.
Keep it so.

Understood. But sometimes you want to make minor edits when all you have is
the PDF and not the original document. This happens a lot, it turns out.

However, back to the printing of booklets, that's one thing I had trouble
finding free (as in no cost) software as printing a booklet from folded
8.5x11-inch paper is more complex than standard printing, especially when
dealing with double-sided printing and odd numbers of pages & title pages.

I'm aware of "pdfbook", but, alas, that requires Python (aurgh!, again!)
on Windows, but luckily, pdfbook should be easier to use on Linux & Mac.
<https://pdfbook.readthedocs.io/en/latest/Installation.html>

Unfortunately, the "examples" provided are, um, shall we say underwhelming?
<https://pdfbook.readthedocs.io/en/latest/Examples.html>

Digging a bit, I think something like this pdfbook command may work:
pdfbook input.pdf --paper letter --outfile output-booklet.pdf

Supposedly that pdfbook command will consider the number of pages in the
input.pdf to then automatically order the pages so that when folded, the
pages are in the correct order.

Can someone with Python installed test it out on a sample PDF for us?

Post by Anton Shepelev

Post by Marion
[x] Extract images (PDF Exchange Viewer, PDF Shaper)

SumatraPDF

Thanks for that suggestion. Googling it, apparently SumatraPDF can
*manually* copy an image which you can then paste into an image editor.

It turns out, I think, based on what I found anyway, that SumatraPDF uses
an underlying MuPDF library to extract images, so as a result of your
advice, I'll add muPDF to the line for extracting images.

While I was looking that up, I found that the free (no cost) PDF-XChange
Editor also can extract images from a PDF, so I'll add that too.

I think I'm going to have to give up on keeping it one line per item.
[x] Extract images (PDF Exchange Editor, PDF Shaper, PDFgear, poppler, muPDF)

Post by Anton Shepelev

Post by Marion
[x] Reorder pages (mutool)
[x] Rotate pages (Acrobat Reader)

pdftk of course.

Thanks for that suggestion. Checking rotate first, it seems that the
pdftoolkit rotation of 180 degrees is a great suggestion. Much appreciated.
pdftk input.pdf cat 1-endsouth output output.pdf

Looking that up, I found that mutool can also rotate, e.g., for 180 degrees
mutool convert -R 180 input.pdf output.pdf

I found out in that search that the GUIs for PDF-XChange Editor (free) and
PDF Arranger (free) can also rotate pages and save to a new PDF file.

Apparently Acrobat READER can only rotate the view, but it can't SAVE the
rotated results, so I'll remove Acrobat Reader from that rotation line.
[x] Rotate pages (pdftk, mutool, PDF-XChange Editor, PDF Arranger)

Now looking at the reordering of pages (which is really a duplicate of
inserting and deleting pages, isn't it?) the same programs can re-order
pages, but (as above) the latter two do it graphically, and pdftk is better
at it than mutool is, but all of them can reorder pages nonetheless.

For example, to flip the order of page 2 and 3 in a pdf using pdftk:
pdftk input.pdf cat 1 3 2 4-end output output.pdf
But it turned out to be difficult with mutool (possible but difficult).
So I removed muTool because it's just too complicated to reorder with it.
[x] Reorder pages (pdftk, PDF-XChange Editor, PDF Arranger)

Thanks for pointing out the omissions.

Post by Anton Shepelev

Post by Marion
What are your suggestions (so that everyone benefits from
your knowledge)?

The obvious one -- typsetting software for producing PDFs
from text, e.g.: LaTeX, (GNU) Troff.

LaTeX seems to be what we have to fall back on when, for example, pdfbook
primarily focuses on the page reordering aspect of booklet creation
(although I'm confused since I saw mention that pdfbook is in the pdfjam
package, which can be installed within a TeX distribution so maybe it can
all be put together for everyone to easily output booket-style PDFs?).

As noted, LaTeX has sophisticated built-in features to analyze the content
of the PDF to more intelligently handle page breaks to avoid splitting
images or creating an awkward text flow.

Digging a bit into LaTeX (which I've never used myself), MiKTeX & TeX Live
seem to be free (no cost) Windows, Linux & Mac "modern" TeX distributions.
<https://miktex.org/howto/install-miktex>
<https://math.asu.edu/resources/computer-resources/texlive-windows>

Also TeXstudio or TeXworks appear to be free (no cost) LaTeX editors.
<https://www.texstudio.org/>
<https://www.tug.org/texworks/>

Since the expensive cost of free (no cost) software is in the trials and
tribulations to find the best ones that work, does anyone have experience
with any of the distributions above for creating the booklet style PDFs?

[Wolf Greenblatt]

Post by Marion
Can someone with Python installed test it out on a sample PDF for us?

I don't have Python but these are sample PDFs to run tests upon.
<https://sample-files.com/documents/pdf/>
<https://www.learningcontainer.com/sample-pdf-files-for-testing/>
<https://examplefile.com/document/pdf>
<https://getsamplefiles.com/sample-document-files/pdf>
<https://www.graydart.com/sample/documents/pdf>
<https://freetestdata.com/document-files/pdf/>
<https://onlinetestcase.com/pdf-file/>

[Anton Shepelev]

Post by Marion

Post by Anton Shepelev
I believe LaTeX has packages for that. I have produced
PDF booklets from Postscrpt, with psbook and psnup. The
incoming PostScipt was mine, from either LaTeX or GNU
Troff.

Thanks for that suggestion as, in the past, I printed
booklets.
Printing a booklet requires arranging both sides of the
pages in a specific order so that when the 8.5x11-inch
printed sheets are folded in half, the pages appear in the
correct sequence as if they were in a booklet.. [...]
However, back to the printing of booklets, that's one
thing I had trouble finding free (as in no cost) software
as printing a booklet from folded 8.5x11-inch paper is
more complex than standard printing, especially when
dealing with double-sided printing and odd numbers of
pages & title pages.

That's exactly what I did with psbook and psnup from
psutils. I produced an A4 booklet from an A5 document with
sequential pages. I printed my booklent by with a normal
single-sided printer, in two runs, without reordering the
sheaf in between. All the rearrangement was taken care of
during the generation of the PDF.

Before *roff and *tex, I used to print such booklets in
whatever software I had at hand, including MS Word '97 and
Adobe PageMaker. For Word, I had a simple Pascal program
that would generate two comma-spearated lists of page
numbers, ready to paste in into the Print window, for
printing the even and odd pages of the booklet.

The alrorithm is rather simple, IIRC. After you append
empty pages to make the total a multiple of four, the
following invariant holds true for each side of any quatro:

page_left + page_right = page_total + 1

For example, a twelve-page booklet will be printed on three
(12/4) sheets thusly:

even odd
12 1 verso 2 11
10 3 verso 4 9
8 5 verso 6 7

So, you first print the odd pages in increasing order, and
then odd ones in decreasing order, to end up with a set of
sheats ready to fold (IIRC). I still seem to have the ugly
ancient program in Pascal that I wrote in late school or
early University to perfrom that task:

https://paste.sr.ht/~shepton/4d8374ec6e2c543fa8caad43709596b1cae5cd94

It should compile in FreePascal compiler.

Post by Marion
As noted, LaTeX has sophisticated built-in features to
analyze the content of the PDF to more intelligently
handle page breaks to avoid splitting images or creating
an awkward text flow.

No, LaTeX and Troff are tools to author and typeset new
documents, rather than modify existing PDFs.

Post by Marion
Since the expensive cost of free (no cost) software is in
the trials and tribulations to find the best ones that
work,

Which is why I prefer to use time-honoured classics.

Post by Marion
does anyone have experience with any of the distributions
above for creating the booklet style PDFs?

I have used this one a long time ago:

https://ctan.org/pkg/booklet

And I have used psutils (with psbook and psnup) no so long
time ago:

https://github.com/rrthomas/psutils

Generally, I have found *roff much easier than LaTeX. I
have written several Groff macros myself, including those to
wrap text around images as shown in this newsletter:

https://corewar.co.uk/coreops/coreops02.txt

Both Groff and LaTeX have great and helpful communities.

[Zaidy036]

Post by Anton Shepelev

Post by Marion

Post by Anton Shepelev
I believe LaTeX has packages for that. I have produced
PDF booklets from Postscrpt, with psbook and psnup. The
incoming PostScipt was mine, from either LaTeX or GNU
Troff.

Thanks for that suggestion as, in the past, I printed
booklets.
Printing a booklet requires arranging both sides of the
pages in a specific order so that when the 8.5x11-inch
printed sheets are folded in half, the pages appear in the
correct sequence as if they were in a booklet.. [...]
However, back to the printing of booklets, that's one
thing I had trouble finding free (as in no cost) software
as printing a booklet from folded 8.5x11-inch paper is
more complex than standard printing, especially when
dealing with double-sided printing and odd numbers of
pages & title pages.

That's exactly what I did with psbook and psnup from
psutils. I produced an A4 booklet from an A5 document with
sequential pages. I printed my booklent by with a normal
single-sided printer, in two runs, without reordering the
sheaf in between. All the rearrangement was taken care of
during the generation of the PDF.
Before *roff and *tex, I used to print such booklets in
whatever software I had at hand, including MS Word '97 and
Adobe PageMaker. For Word, I had a simple Pascal program
that would generate two comma-spearated lists of page
numbers, ready to paste in into the Print window, for
printing the even and odd pages of the booklet.
The alrorithm is rather simple, IIRC. After you append
empty pages to make the total a multiple of four, the
page_left + page_right = page_total + 1
For example, a twelve-page booklet will be printed on three
even odd
12 1 verso 2 11
10 3 verso 4 9
8 5 verso 6 7
So, you first print the odd pages in increasing order, and
then odd ones in decreasing order, to end up with a set of
sheats ready to fold (IIRC). I still seem to have the ugly
ancient program in Pascal that I wrote in late school or
https://paste.sr.ht/~shepton/4d8374ec6e2c543fa8caad43709596b1cae5cd94
It should compile in FreePascal compiler.

Post by Marion
As noted, LaTeX has sophisticated built-in features to
analyze the content of the PDF to more intelligently
handle page breaks to avoid splitting images or creating
an awkward text flow.

No, LaTeX and Troff are tools to author and typeset new
documents, rather than modify existing PDFs.

Post by Marion
Since the expensive cost of free (no cost) software is in
the trials and tribulations to find the best ones that
work,

Which is why I prefer to use time-honoured classics.

Post by Marion
does anyone have experience with any of the distributions
above for creating the booklet style PDFs?

https://ctan.org/pkg/booklet
And I have used psutils (with psbook and psnup) no so long
https://github.com/rrthomas/psutils
Generally, I have found *roff much easier than LaTeX. I
have written several Groff macros myself, including those to
https://corewar.co.uk/coreops/coreops02.txt
Both Groff and LaTeX have great and helpful communities.

https://www.bookletcreator.com/how-it-works/

[Marion]

Post by Anton Shepelev

Post by Marion
[?] Print book format PDF (FinePrint payware)

I believe LaTeX has packages for that. I have produced PDF
booklets from Postscrpt, with psbook and psnup. The
incoming PostScipt was mine, from either LaTeX or GNU Troff.

Thanks for that suggestion as, in the past, I printed booklets.

Printing a booklet requires arranging both sides of the pages in a specific
order so that when the 8.5x11-inch printed sheets are folded in half, the
pages appear in the correct sequence as if they were in a booklet.

Post by Anton Shepelev

Post by Marion
[x] Convert PDF to MSWord or any epub format & vice versa (Calibre)
[x] Edit PDF existing text (Adobe Reader commenting, Acrobat payware)
[x] Globally search & replace PDF text (Libre Office)

PDF is meant to be a final format not meant for editing.
Keep it so.

Understood. But sometimes you want to make minor edits when all you have is
the PDF and not the original document. This happens a lot, it turns out.

However, back to the printing of booklets, that's one thing I had trouble
finding free (as in no cost) software as printing a booklet from folded
8.5x11-inch paper is more complex than standard printing, especially when
dealing with double-sided printing and odd numbers of pages & title pages.

I'm aware of "pdfbook", but, alas, that requires Python (aurgh!, again!)
on Windows, but luckily, pdfbook should be easier to use on Linux & Mac.
<https://pdfbook.readthedocs.io/en/latest/Installation.html>

Unfortunately, the "examples" provided are, um, shall we say underwhelming?
<https://pdfbook.readthedocs.io/en/latest/Examples.html>

Digging a bit, I think something like this pdfbook command may work:
pdfbook input.pdf --paper letter --outfile output-booklet.pdf

Supposedly that pdfbook command will consider the number of pages in the
input.pdf to then automatically order the pages so that when folded, the
pages are in the correct order.

Can someone with Python installed test it out on a sample PDF for us?

Post by Anton Shepelev

Post by Marion
[x] Extract images (PDF Exchange Viewer, PDF Shaper)

SumatraPDF

Thanks for that suggestion. Googling it, apparently SumatraPDF can
*manually* copy an image which you can then paste into an image editor.

It turns out, I think, based on what I found anyway, that SumatraPDF uses
an underlying MuPDF library to extract images, so as a result of your
advice, I'll add muPDF to the line for extracting images.

While I was looking that up, I found that the free (no cost) PDF-XChange
Editor also can extract images from a PDF, so I'll add that too.

I think I'm going to have to give up on keeping it one line per item.
[x] Extract images (PDF Exchange Editor, PDF Shaper, PDFgear, poppler, muPDF)

Post by Anton Shepelev

Post by Marion
[x] Reorder pages (mutool)
[x] Rotate pages (Acrobat Reader)

pdftk of course.

Thanks for that suggestion. Checking rotate first, it seems that the
pdftoolkit rotation of 180 degrees is a great suggestion. Much appreciated.
pdftk input.pdf cat 1-endsouth output output.pdf

Looking that up, I found that mutool can also rotate, e.g., for 180 degrees
mutool convert -R 180 input.pdf output.pdf

I found out in that search that the GUIs for PDF-XChange Editor (free) and
PDF Arranger (free) can also rotate pages and save to a new PDF file.

Apparently Acrobat READER can only rotate the view, but it can't SAVE the
rotated results, so I'll remove Acrobat Reader from that rotation line.
[x] Rotate pages (pdftk, mutool, PDF-XChange Editor, PDF Arranger)

Now looking at the reordering of pages (which is really a duplicate of
inserting and deleting pages, isn't it?) the same programs can re-order
pages, but (as above) the latter two do it graphically, and pdftk is better
at it than mutool is, but all of them can reorder pages nonetheless.

For example, to flip the order of page 2 and 3 in a pdf using pdftk:
pdftk input.pdf cat 1 3 2 4-end output output.pdf
But it turned out to be difficult with mutool (possible but difficult).
So I removed muTool because it's just too complicated to reorder with it.
[x] Reorder pages (pdftk, PDF-XChange Editor, PDF Arranger)

Thanks for pointing out the omissions.

Post by Anton Shepelev

Post by Marion
What are your suggestions (so that everyone benefits from
your knowledge)?

The obvious one -- typsetting software for producing PDFs
from text, e.g.: LaTeX, (GNU) Troff.

LaTeX seems to be what we have to fall back on when, for example, pdfbook
primarily focuses on the page reordering aspect of booklet creation
(although I'm confused since I saw mention that pdfbook is in the pdfjam
package, which can be installed within a TeX distribution so maybe it can
all be put together for everyone to easily output booket-style PDFs?).

As noted, LaTeX has sophisticated built-in features to analyze the content
of the PDF to more intelligently handle page breaks to avoid splitting
images or creating an awkward text flow.

Digging a bit into LaTeX (which I've never used myself), MiKTeX & TeX Live
seem to be free (no cost) Windows, Linux & Mac "modern" TeX distributions.
<https://miktex.org/howto/install-miktex>
<https://math.asu.edu/resources/computer-resources/texlive-windows>

Also TeXstudio or TeXworks appear to be free (no cost) LaTeX editors.
<https://www.texstudio.org/>
<https://www.tug.org/texworks/>

Since the expensive cost of free (no cost) software is in the trials and
tribulations to find the best ones that work, does anyone have experience
with any of the distributions above for creating booklet style PDFs?

[Peter Flynn]

On 04/03/2025 23:24, Marion wrote:
[...]

Post by Marion
Printing a booklet requires arranging both sides of the pages in a
specific order so that when the 8.5x11-inch printed sheets are
folded in half, the pages appear in the correct sequence as if they
were in a booklet.

This is called "imposition". Printing a book means arranging the pages
(usually) 16 (possibly 32) per side of a very large sheet (and therefore
16 the other side) making a "signature", laid out so that when folded
and folded and folded etc and trimmed, page 1 has page 2 on the back of
it; then repeat for the next 32 (or 64) pages, and repeat, etc until all
X00 pages are accounted for. Printed off a reel of paper (confusingly
called a web) nowadays, and slit to sheets before folding. Then stacked
together, the spines abraded and glued (or sewn with thread for fancy
books), then draw on the cover (printed separately on board), glue it,
and give it a final trim.

Post by Marion
However, back to the printing of booklets, that's one thing I had trouble
finding free (as in no cost) software as printing a booklet from folded
8.5x11-inch paper is more complex than standard printing, especially when
dealing with double-sided printing and odd numbers of pages & title pages.

And, domestically, having it all set up and running, and then the
cheapass paper-handling mechanism in the printer feeds two sheets
instead of one, and messes it all up.

Post by Marion
I'm aware of "pdfbook", but, alas, that requires Python (aurgh!, again!)
on Windows, but luckily, pdfbook should be easier to use on Linux & Mac.

I don't think anyone doing this seriously would consider Windows at all.
There is a massive collection of free text-manipulation tools known
collectively as "the Unix text tools" which work on Linux (including Mac
OSX) but which cause endless compilation trouble on Windows.

Post by Marion
Digging a bit into LaTeX (which I've never used myself), MiKTeX & TeX Live
seem to be free (no cost) Windows, Linux & Mac "modern" TeX distributions.

Correct. The canonical location is the TeX Users Group site (tug.org)

Post by Marion
Also TeXstudio or TeXworks appear to be free (no cost) LaTeX editors.

Both are excellent but there are lots of others, including (of course)
Emacs. https://latex.silmaril.ie/formattinginformation/editdis.html

Post by Marion
Since the expensive cost of free (no cost) software is in the trials and
tribulations to find the best ones that work, does anyone have experience
with any of the distributions above for creating booklet style PDFs?

My typesetting business has used LaTeX and the Unix text tools for the
last 30 years (mainly for books) without significant problems¹. There
are some technical aspects, such as ensuring that the L–R adjustment of
the text area on one page will correctly occupy the exact same space on
the back of the next page when printed and bound, but this really only
affects very large signatures where the thickness of the paper has to be
taken into account when folder 3–4 times; this is not really relevant
for booklets. Otherwise it's just a matter of using the right imposition
scheme and the right page-rearrangement software to implement it.

Peter

----------
¹ The real problems are in the copyediting and proofreading of the text,
and that's common to all systems. Maybe marginally easier in LaTeX
because the master source is plain text, but the real difficulties come
in dealing with semi-literate authors and technically ill-informed editors.

[Philip Herlihy]

Post by Peter Flynn
This is called "imposition". Printing a book means arranging the pages
(usually) 16 (possibly 32) per side of a very large sheet (and therefore
16 the other side) making a "signature", laid out so that when folded
and folded and folded etc and trimmed, page 1 has page 2 on the back of
it; then repeat for the next 32 (or 64) pages, and repeat, etc until all
X00 pages are accounted for. Printed off a reel of paper (confusingly
called a web) nowadays, and slit to sheets before folding. Then stacked
together, the spines abraded and glued (or sewn with thread for fancy
books), then draw on the cover (printed separately on board), glue it,
and give it a final trim.

I don't know what the international availability of this is, but this is
a BBC program detailing how hardback books are manufactured, with a good
discussion of "imposition".

https://www.bbc.co.uk/iplayer/episode/m0027f48/inside-the-factory-
series-9-5-hardback-books

(rejoin link with no whitespace)

[Carlos E.R.]

Post by Philip Herlihy

Post by Peter Flynn
This is called "imposition". Printing a book means arranging the pages
(usually) 16 (possibly 32) per side of a very large sheet (and therefore
16 the other side) making a "signature", laid out so that when folded
and folded and folded etc and trimmed, page 1 has page 2 on the back of
it; then repeat for the next 32 (or 64) pages, and repeat, etc until all
X00 pages are accounted for. Printed off a reel of paper (confusingly
called a web) nowadays, and slit to sheets before folding. Then stacked
together, the spines abraded and glued (or sewn with thread for fancy
books), then draw on the cover (printed separately on board), glue it,
and give it a final trim.

I don't know what the international availability of this is, but this is
a BBC program detailing how hardback books are manufactured, with a good
discussion of "imposition".
https://www.bbc.co.uk/iplayer/episode/m0027f48/inside-the-factory-
series-9-5-hardback-books
(rejoin link with no whitespace)

In Spain, I get:

Sorry, BBC iPlayer isn’t available in your region.

It looks like you’re outside of the UK. BBC iPlayer is only available in
the UK. If you are using a proxy or VPN, please turn off any of these
services and try again.
For help, please go to iPlayer help and FAQs

[Peter Flynn]

Post by Philip Herlihy
I don't know what the international availability of this is, but this is
a BBC program detailing how hardback books are manufactured, with a good
discussion of "imposition".

I watched that the other night. Very well done, although it omitted the
slicing, folding, and stacking in signatures. I used to live near there.

Post by Philip Herlihy
https://www.bbc.co.uk/iplayer/episode/m0027f48/inside-the-factory-series-9-5-hardback-books

That's an iPlayer link, AFAIK not available outside the UK.
The normal BBC programme description is here:
https://www.bbc.co.uk/programmes/m0027f48

Peter

[Marion]

Post by Anton Shepelev

Post by Marion
[?] Print book format PDF (FinePrint payware)

I believe LaTeX has packages for that. I have produced PDF
booklets from Postscrpt, with psbook and psnup. The
incoming PostScipt was mine, from either LaTeX or GNU Troff.

Thanks for that suggestion as, in the past, I printed booklets.

Printing a booklet requires arranging both sides of the pages in a specific
order so that when the 8.5x11-inch printed sheets are folded in half, the
pages appear in the correct sequence as if they were in a booklet.

Post by Anton Shepelev

Post by Marion
[x] Convert PDF to MSWord or any epub format & vice versa (Calibre)
[x] Edit PDF existing text (Adobe Reader commenting, Acrobat payware)
[x] Globally search & replace PDF text (Libre Office)

PDF is meant to be a final format not meant for editing.
Keep it so.

Understood. But sometimes you want to make minor edits when all you have is
the PDF and not the original document. This happens a lot, it turns out.

However, back to the printing of booklets, that's one thing I had trouble
finding free (as in no cost) software as printing a booklet from folded
8.5x11-inch paper is more complex than standard printing, especially when
dealing with double-sided printing and odd numbers of pages & title pages.

I'm aware of "pdfbook", but, alas, that requires Python (aurgh!, again!)
on Windows, but luckily, pdfbook should be easier to use on Linux & Mac.
<https://pdfbook.readthedocs.io/en/latest/Installation.html>

Unfortunately, the "examples" provided are, um, shall we say underwhelming?
<https://pdfbook.readthedocs.io/en/latest/Examples.html>

Digging a bit, I think something like this pdfbook command may work:
pdfbook input.pdf --paper letter --outfile output-booklet.pdf

Supposedly that pdfbook command will consider the number of pages in the
input.pdf to then automatically order the pages so that when folded, the
pages are in the correct order.

Can someone with Python installed test it out on a sample PDF for us?

Post by Anton Shepelev

Post by Marion
[x] Extract images (PDF Exchange Viewer, PDF Shaper)

SumatraPDF

Thanks for that suggestion. Googling it, apparently SumatraPDF can
*manually* copy an image which you can then paste into an image editor.

It turns out, I think, based on what I found anyway, that SumatraPDF uses
an underlying MuPDF library to extract images, so as a result of your
advice, I'll add muPDF to the line for extracting images.

While I was looking that up, I found that the free (no cost) PDF-XChange
Editor also can extract images from a PDF, so I'll add that too.

I think I'm going to have to give up on keeping it one line per item.
[x] Extract images (PDF Exchange Editor, PDF Shaper, PDFgear, poppler, muPDF)

Post by Anton Shepelev

Post by Marion
[x] Reorder pages (mutool)
[x] Rotate pages (Acrobat Reader)

pdftk of course.

Thanks for that suggestion. Checking rotate first, it seems that the
pdftoolkit rotation of 180 degrees is a great suggestion. Much appreciated.
pdftk input.pdf cat 1-endsouth output output.pdf

Looking that up, I found that mutool can also rotate, e.g., for 180 degrees
mutool convert -R 180 input.pdf output.pdf

I found out in that search that the GUIs for PDF-XChange Editor (free) and
PDF Arranger (free) can also rotate pages and save to a new PDF file.

Apparently Acrobat READER can only rotate the view, but it can't SAVE the
rotated results, so I'll remove Acrobat Reader from that rotation line.
[x] Rotate pages (pdftk, mutool, PDF-XChange Editor, PDF Arranger)

Now looking at the reordering of pages (which is really a duplicate of
inserting and deleting pages, isn't it?) the same programs can re-order
pages, but (as above) the latter two do it graphically, and pdftk is better
at it than mutool is, but all of them can reorder pages nonetheless.

For example, to flip the order of page 2 and 3 in a pdf using pdftk:
pdftk input.pdf cat 1 3 2 4-end output output.pdf
But it turned out to be difficult with mutool (possible but difficult).
So I removed muTool because it's just too complicated to reorder with it.
[x] Reorder pages (pdftk, PDF-XChange Editor, PDF Arranger)

Thanks for pointing out the omissions.

Post by Anton Shepelev

Post by Marion
What are your suggestions (so that everyone benefits from
your knowledge)?

The obvious one -- typsetting software for producing PDFs
from text, e.g.: LaTeX, (GNU) Troff.

LaTeX seems to be what we have to fall back on when, for example, pdfbook
primarily focuses on the page reordering aspect of booklet creation
(although I'm confused since I saw mention that pdfbook is in the pdfjam
package, which can be installed within a TeX distribution so maybe it can
all be put together for everyone to easily output booket-style PDFs?).

As noted, LaTeX has sophisticated built-in features to analyze the content
of the PDF to more intelligently handle page breaks to avoid splitting
images or creating an awkward text flow.

Digging a bit into LaTeX (which I've never used myself), MiKTeX & TeX Live
seem to be free (no cost) Windows, Linux & Mac "modern" TeX distributions.
<https://miktex.org/howto/install-miktex>
<https://math.asu.edu/resources/computer-resources/texlive-windows>

Also TeXstudio or TeXworks appear to be free (no cost) LaTeX editors.
<https://www.texstudio.org/>
<https://www.tug.org/texworks/>

Since the expensive cost of free (no cost) software is in the trials and
tribulations to find the best ones that work, does anyone have experience
with any of the distributions above for creating a booklet style PDF?