McCarthy - Dijkstra short polemics in 1976.







It doesn't really fit in my previous post -  What Edsger W. Dijkstra blogged about Lisp  because Dijkstra doesn't comment Lisp in this polemics. However, McCarthy does. The original is distributed by Dijkstra, republished in EWD Archive and I believe that this copy is "fair use."





Original photo from Infolab Stanford Museum



To Professor John McCarthy

Dear Colleague,

thank you for your amazing letter of August 18, 1976. But you really don't need to be sorry for me because the output of your xerographic printer hurts my eye. (For after all, all of us need our daily dose of irritation: I satisfy my needs in this respect by exposure to audible wallpaper, xerographic computer output and similar would-be services to the consumer.)

I am puzzled by your ban on my use of the verb "to degrade": I thought I had used it correctly, in the sense of "to impair the quality of". But let me propose a deal: you allow me to dislike your poor xerographic printer output (and to say so), and I promise not to mind if you call me a snob.

As far as the mechanics of manuscript production are concerned, I can warmly recommend to all prospective authors to train themselves to get their texts nearly right the first time: it is a fast, efficient, and cheap way of working that almost always gives great esthetic and intellectual satisfaction.

And finally: if you have the text of your recent letter still in your computer files, you could do me three favours: inserting the missing "W." in the first line of the address, and correcting the spelling errors in the next two lines of the address.

As EWD581 - 1 your letter will get the same distribution as EWD574, my letter to Zohar.

Greetings and best wishes!
Yours ever,
Edsger W. Dijkstra
Dear Professor Dijkstra:

Zohar Manna showed me your letter of 26 July. I am sorry you "vehemently abhor" computer produced manuscripts in different type fonts. I wish our xerographic printer had higher resolution and didn't make smudges when it hasn't recently been adjusted; we do the best we can. You are entitled to your tastes, but when you claim that people with different tastes are using computers to "degrade our lives", you are contributing to an atmosphere of snobbery that has done much to degrade discussion of programming style. Consider this also a protest against the language used in your campaign against gotos. (Sorry our boldface isn't bolder).

As the inventor of one of the first languages (LISP) that allowed programming without gotos, I nevertheless consider that they have their uses. However, my main objection is to the character of the campaign you launched against them, which seemed to be based on an appeal to snobbery and seemed to encourage more snobbery.

As you see, justified right margins are optional in computer produced documents. Many people share your preference for unjustified lines, but it is wrong to make a moral issue of it. As a matter of taste, I find the backwards words fi and od unpleasing and would prefer parentheses or begins and ends for resolving ambiguities.

I am glad you like the content of Manna's and Waldinger's report. Manna would be an excellent expositor even if his reports had to be incised in clay tablets and baked, but the polished style and the prompt appearance may owe something to the opportunity for frequent revision afforded by computer produced documents. Certainly, this was one of our main reasons for pioneering them at this laboratory. Moreover, we cannot afford the secretarial time required to make several versions of a typescript.

If you felt like distributing this protest to your EWD list, I would be grateful.

Sincerely,
John McCarthy
transcribed by Carl Ludwigson
revised Sun, 9 Sep 2007

Edsger W. Dijkstra, A somewhat open letter to Professor John McCarthy, 1976, EWD551.






---

2 comments:

  1. It's very fascinating to me, as a fan of typography & programming, to see two geniuses sperging out over fonts. I think typography is underrated in importance by most people.

    ReplyDelete
  2. I tend to agree with McCarthy in just about everything he says in the letter.

    I'd really like to know what Dijkstra seems to think is wrong with right justification, thought. Maybe he has a point, but I'm inclined not to think so.

    On the post you mentioned at the beginning of this article, he clearly does not understand Lisp nor does he devote much effort in doing so, although he goes on to despise the language based on a single poor written language manual. Had he tried to see beyond the thin layer of Lisp's "atrocious syntax", as he calls it, I believe he would've been able to appreciate the giant leap forward that was Lisp in comparison to what was in use to that day (and actually much of what we still have today).

    ReplyDelete