I recently contacted McGraw-Hill to see if they have a mechanism in place for printing out-of-print books that are still of interest. Specifically, I asked about "Power programming with Mathematica" by David Wagner, as I am personally interested in obtaining a copy, and suspect that others might also be interested.

Here is the literal response I received from McGraw-Hill:

Good morning Todd,

If there was a high enough demand for the book there is a possibility. If you can let me know how many you're looking for, and the name of the school or business you're with I can contact the editor and check to see if there's something that can be done.

McGraw-Hill Education

First, please don't get your hopes up, as nothing may come of this; however, I am committed to seeing this through if there is sufficient community support to get the publisher to make it available again, if only for a limited time.

If you would be interested in purchasing a copy of Wagner's text, please respond in the affirmative by making a comment to this question, such as "yes, I would like to purchase a copy." In this way, I can directly "show" our community's interest to the publisher.

At this time, I can't speak to cost, but it is clear that we need a "critical mass" to get the publisher's attention to make it worthwhile. Keep your fingers crossed and show your support!

  • 4
    $\begingroup$ Yes, I would very much like to purchase a copy and would also highly recommend this book to my university librarian. It's a pity that I cannot find a single copy of this excellent book in any of the University of California libraries and I have to borrow one via ILL from some tiny school 500 miles away from me (and I get to keep it for only 5 days!) $\endgroup$
    – rm -rf
    Commented Dec 17, 2012 at 18:44
  • 4
    $\begingroup$ Yes, I certainly would. Maybe inquire about making it available electronically since they could still sell it, but they would not be burdened with the costs associated with a physical paper print run. $\endgroup$
    – JohnD
    Commented Dec 17, 2012 at 18:44
  • 7
    $\begingroup$ This publisher needs to get with the program. They have the typeset copy of the book so for low demand or out of print books they should be able to print on demand using anyone of a number of POD services available. $\endgroup$ Commented Dec 17, 2012 at 20:25
  • 6
    $\begingroup$ I will buy two! $\endgroup$ Commented Dec 17, 2012 at 21:17
  • 7
    $\begingroup$ I have forwarded everyones' comments to McGraw-Hill and have attempted to make a strong argument for allowing the Mathematica user community to once again have access to this text. As soon as I know more, I'll forward it on. $\endgroup$
    – Todd Allen
    Commented Dec 18, 2012 at 18:52

5 Answers 5


To download a licensed copy of Power Programming with Mathematica by David B. Wagner, please click here:


Thank you to McGraw-Hill for granting me the license to scan and distribute this out-of-print text to the Mathematica community!

Thank you to Manfred Plagmann (aka matariki) for taking the time to carefully scan the entire text.

Thank you to Sophia Scheibe (aka halirutan's wife) for providing select scans of pages to Manfred to allow him to complete his work.

Thank you to Mr. Wagner for writing this text!

Happy computing everyone! Todd

  • 1
    $\begingroup$ First upvote! wheee $\endgroup$
    – s0rce
    Commented Apr 5, 2013 at 2:22
  • 10
    $\begingroup$ @Mr.Wizard Todd deserves a huge bounty for this, don't you think? ;) $\endgroup$ Commented Apr 5, 2013 at 9:47
  • 1
    $\begingroup$ If I can figure out how to do it easily, does your license allow for an outline of bookmarks to be included in the pdf? For instance, if I went through and generated the bookmarks for the chapters and subchapters, could you then distribute that? $\endgroup$
    – rcollyer
    Commented Apr 5, 2013 at 14:42
  • 7
    $\begingroup$ I can confirm that I have the notebooks and other files from the floppy. They open fine. Subject to us having permission to distribute the files, I am happy to host them on Verbeia.com. (I don't have a dropbox account, and anyway, the length of time I've hosted Ted Ersek's material shows that I'm a reliable host.) $\endgroup$
    – Verbeia
    Commented Apr 6, 2013 at 4:39
  • 2
    $\begingroup$ @Verbeia The license has no formal language answering "yes" or "no" to distributing the electronic files. If the files represent content printed in the text, then I think distributing them would be fine. I hesitate to contact McGraw-Hill again, as the first time was less than enjoyable. $\endgroup$
    – Todd Allen
    Commented Jul 1, 2013 at 1:06

Exciting news!

After nearly two months of agonizing communication, McGraw-Hill has granted me a license to scan one copy of "Power programming with Mathematica" for the purposes of distributing it (freely) throughout the Mathematica user community here on StackExchange.

First, thanks to everyone for showing support on this .... sorry it took so long.

Secondly, to make this a reality, I need to make contact with someone who is willing to sacrifice an original copy of the text (in good condition). It will be necessary to scan the entire book into a PDF file, and then I am required by the licence agreement to add author & publisher credit information to each page - which I know how to achieve using desktop publishing software that I have.

In the second answer to this question above, a user named Matariki had stated that he had received permission to bulk scan a copy of the book. Does anyone know how to get a hold of him? If anyone else is interested in making this sacrifice for the user community, please contact me at my email address: [email protected]

Thanks, Todd

  • 1
    $\begingroup$ Saw your post mentioned in chat. I have a book in a reasonable condition. So far I have been processing single pages only. My intention is to turn the book into notebooks. I have the image processing done in Mathematica to identify and tag code segments, annotation, normal text etc. I am using Tesseract to do the OCR. Not quite ready for bulk processing yet. However, if we go for a standard scan then I'll go ahead and cut off the spine of the book and get it bulk scanned. I am rather busy at the moment but should be able to get this started next week. What resolution should I aim for? $\endgroup$
    – Matariki
    Commented Mar 16, 2013 at 4:16
  • 1
    $\begingroup$ For a bulk scan into PDF, I believe any resolution greater than or equal to 150 dpi should be sufficient. My licence only approves the scanning of the text into PDF, so if you are willing to do that for the community, I am sure everyone would be grateful. While I like your idea of turning the book into notebooks, please know that my license does not authorize that. Feel free to email me, Matariki. ([email protected]) Thanks. $\endgroup$
    – Todd Allen
    Commented Mar 16, 2013 at 15:01
  • $\begingroup$ I realise that the license isn't including a notebook form. This is just my private little project. I'll see that I can get the scanning on its way next week. I need to find a bookbinder to put the book back together again as this is part of the deal I made with the owner. I'll send regular updates on progress. $\endgroup$
    – Matariki
    Commented Mar 16, 2013 at 19:18
  • $\begingroup$ Todd, we all owe you a debt of gratitude. Thank you! @Matariki for scanned material, DjVu is generally a better choice of format than PDF. Particularly with the Adobe software, browsing a PDF consisting of the scanned pages of a book can be painfully slow and memory-consuming, whereas DjVu is very fast and convenient. Software such as SumatraPDF (Windows-only; based on MuPDF) can open both formats--although, to be fair, its performance with large PDFs is also nowhere near as bad as the Adobe Reader. $\endgroup$ Commented Mar 16, 2013 at 22:08
  • 1
    $\begingroup$ @Matariki sorry, I don't use a Mac so I don't have software recommendations for that platform. The DjVu specification is an open format and the reference implementation is DjVuLibre, from which a viewer (DjView4) has been implemented and is available for multiple platforms. Whether this is the best option for the Mac platform or if there are better tools available, I don't know. Perhaps for the sake of convenience it would be better to produce the PDF first and then convert to DjVu for those who want it. $\endgroup$ Commented Mar 17, 2013 at 6:27

I have packaged the electronic files that came on the original floppy and posted them in a ZIP archive on my web site, at: http://www.verbeia.com/mathematica/PowerProgMa.zip

File size is 119 kb.

Sorry for the delay. Enjoy!


Update (01-11-2013)

  1. I've contacted McGraw-Hill and was able to talk to one of the original editors involved in publishing this book. His response is that McGraw-Hill does not have the electronic (i.e. PDF) files necessary to offer this text again by print-on-demand. If the electronic files existed, they would be able to do it, but they can't because the book was published too long ago (1996).

  2. The editor did mention one possible solution to the problem: As a community, we could request permission from McGraw-Hill to photocopy the text (with a potential, as yet unknown fee attached) for our user community. Assuming McGraw-Hill gave permission, we would need someone in our user community that own's a copy of the text (in good condition), and is willing to take on the work of photocopying and distributing it. In essence, users would get a photocopied version suitable for placing in a 3-ring binder. I am willing to help coordinate the effort, but I don't own a copy of the text for photocopying purposes.

So, this brings up more questions:

A. Is there still interest among the users here to request permission from McGraw-Hill to photocopy the text? (I have the contact information to make the request on our behalf.)

B. Does anyone own the text who is willing to be involved with the photocopying and distribution of it?

Sorry I don't have more satisfying news at this moment. If you are still interested, please respond to Questions A & B in the comments section.

  • 1
    $\begingroup$ Why can't a scanned PDF be used for the print-on-demand? $\endgroup$
    – P. Fonseca
    Commented Jan 11, 2013 at 19:42
  • $\begingroup$ McGraw Hill does not have a paper copy of the text left. I did not inquire if we (the user community) could scan it and give that to McGraw Hill for print on demand. I can ask, but let's see what others think first. $\endgroup$
    – Todd Allen
    Commented Jan 11, 2013 at 20:18
  • 2
    $\begingroup$ I think we should get more info on the possibilities (or degree of freedom) with a scanned copy. Ideal would be that we're allowed to send it out to users of this community. Perhaps we could have a one time payment as a community? If a paper copy is required we could perhaps distribute the copy to several people located in the different countries to keep shipping costs low etc. I'm willing to help this community. $\endgroup$
    – Lou
    Commented Jan 11, 2013 at 20:23
  • 8
    $\begingroup$ I don't own a book, but here's an idea: We can split up the text into chunks of say 5-10 pages and get people to type it verbatim into a Mathematica notebook. We already spend time plenty of time typing answers for SE; at least this would be for a much bigger cause. If there's interest, we can probably organize this in a dedicated chat room and distribute the task to users who have access to the book. Once everyone is finished (1-2 months), others who don't have access to the book (me) can help with proofreading, verifying examples, typesetting, prettifying it to make it more of a book, etc. $\endgroup$
    – rm -rf
    Commented Jan 12, 2013 at 0:31
  • 4
    $\begingroup$ (continued...) This way, we have it in an electronic format, with interactive examples/code, searchable/editable notebook for tinkering, etc. Maybe we can even hand that over to McGraw-Hill who can then sell that for a fee, but allows those who put in effort to creating it to have a copy for free. I would be willing to help co-ordinate this effort and help with putting the whole thing together. If there's support for this idea, you can perhaps try contacting McGraw-Hill to see if it'll fly. $\endgroup$
    – rm -rf
    Commented Jan 12, 2013 at 0:33

I have scanned today an randomly chosen page (157) from the book to get a first feeling of how to tackle the problem. I ran it through the OCR software that comes with the VueScan scanner application. I had to use a rather high scan resolution of 600 dpi to minimize recognition errors. Has anyone experience with some better OCR software than those freebees that come with scanners? The recognized text file is below:

Pattern Building Blocks 1 57
That got rid ofthe constant diff[3 x"2 - 2 x + 1, x]
term in the previous result. 2
-2 diff[x, X] + 3 diff[x , x]
The derivative of X’ is n2("1. Note that the use of “n_ . ” for the exponent means that
the first parameter also matches the expression xx
diff[x_"n_., x_] /; FreeQ[n, x] := n x"(n - 1)
Sofarsogood... diff[3 x"2 — 2 x + 1, x]
-2 + 6 x
but diff doesn't work on diff[(x + 1)"2, X]
expressions of this form. 2
diff[(l + x) , X]
We need to implement the chain rule. Below, fx is a mnemonic for “function of x.”
diff[fx_"n_, x_] /; FreeQ[n, x] && !FreeQ[fx, x] :=
n * fx"(n - 1) * diff[fx. x]
diff[(x + 1)"2. x]
2 (1 + x)
diff[(x"2 + 2 x + 1)"3, x]
3 <2 + 2 X) (1 + 2 X + X2)2
Chain rule seems to eliminate the need for the power rule, so we might try to UnSet
the latter:
Note that I0 UnSet 3 ru|e, diff [x_"n_, , x_] /; FreeQ [n , x] =,
the left-hand side must be
typed verbatim, including 7 diff
conditions. G loba 1‘ di f f
diff[(c_)*(f_), x_] /; FreeQ[c, x] := c*diff[f, x]
diff[(f_) + (g_), x_] := diff[f, x] + diff[g, X]
diff[c_, x_] /; FreeQ[c, x] := 0
diff[(fx_)"(n_). x_] /:
FreeQ[n, x] && !FreeQ[fx, x] :=
n*fx"(n — l)*diff[fx, x]
Unfortunately, our rules now leave out one important case: diff [x , x] .
diff[(x"2 + 2 x + 1)"3, x]
3 (1 + 2 X + X2)2 (2 diff[x, X] + 2 X diff[x, X])
This is easily fixed by the following simple rule:

There are still a few misses. In particular the ^ is interpreted as ". Exponents of output cells end up somewhere else etc. The annotations Wagner uses on the left margin are also, as expected, interspersed with the following line. However, after deleting all output cells it was a matter of a few minutes to clean up the page and recreate the output cells. Certainly possible to do and the result will be much more useful than a photocopy.

enter image description here

  • $\begingroup$ Occasionally I use Adobe Acrobat for OCR, which is not so bad, but I have never tested it on complex stuff. $\endgroup$
    – Yves Klett
    Commented Jan 16, 2013 at 7:13
  • $\begingroup$ @YvesKlett Thanks, I'll see if Acrobat works better. $\endgroup$
    – Matariki
    Commented Jan 16, 2013 at 17:18
  • $\begingroup$ Good news! Please see this answer from Todd. He's trying to get in touch with you $\endgroup$
    – rm -rf
    Commented Mar 16, 2013 at 2:09
  • $\begingroup$ @rm-rf Thanks, just saw it in chat. $\endgroup$
    – Matariki
    Commented Mar 16, 2013 at 4:19

Not the answer you're looking for? Browse other questions tagged or ask your own question.