[Ads-l] software "patch"

Ben Zimmer bgzimmer at GMAIL.COM
Fri Dec 20 04:33:54 UTC 2024


Garson may have missed the OED3 entry for "patch" as a noun, which predates
the verb. The earliest cite is from the ACM's First Glossary of Programming
Terminology (1954): "_Patch_, a section of coding inserted into a routine
(usually by explicitly transferring control from the routine to the patch
and back again) to correct a mistake or alter the routine."

Nothing to suggest it's based on actual patches on punch cards. The
glossary is online here:
https://archive.computerhistory.org/resources/text/Knuth_Don_X4100/PDF_index/k-8-pdf/k-8-u2741-2-ACM-Glossary.pdf

Interestingly, the ACM's Committee on Nomenclature was chaired by Grace
Murray Hopper, and she writes an introduction to the glossary. Hopper is of
course associated with the famous computing etymythology for "bug" -- see
e.g. my 2013 WSJ column archived here: https://archive.is/YTkMn

The "bug" story is actually mentioned in the YouTube video starting around
2m45s -- one of the participants says "I know that the first bug was an
actual bug," but Tom Scott says that's "a little apocryphal," adding, "I
think there was a bug, but I think the term was already in use."

I would guess the "patch" story is similar, just as the physical bug that
Grace Hopper found isn't the origin of the term "bug."

--bgz

On Thu, Dec 19, 2024 at 11:32 PM ADSGarson O'Toole <
adsgarsonotoole at gmail.com> wrote:

> The Oxford English Dictionary lists the pertinent sense with an
> initial citation in 1962.
>
> [Begin excerpt from OED]
> patch verb
> 8.b. transitive. Computing. To correct or improve (a program or
> routine) by inserting a patch.
>
> 1962 Patch,..(2) to insert corrected coding.
> Automatic Data Processing Glossary (U.S. Bureau of Budget) 38/2
>
> 1984 The version distributed with Kaypro automatically patches the
> Kaypro operating system with its own cursor control codes.
> Australian Microcomputer Magazine January 31/3
> [End excerpt from OED]
>
> It is unclear (to me) what mechanism was used to "insert corrected coding".
>
> I used punch cards to program a computer in 1976. The programmer would
> use a typewriter-like device to create each punch card. Many cards
> were combined to form a deck which was fed into a card reader. The
> program data was sent to the mainframe computer which executed several
> programs in a batch.
>
> If I found an error then I would use the typewriter-like device to
> create a new card. I would not attempt to put tape on the card because
> that might cause the card reader to malfunction. It was easier to
> create a new card and insert it into the deck. Of course, other people
> may have had a different experience.
>
> Punch cards were old-fashioned in 1976. I had already experienced more
> advanced technology. When I was a youngster back in 1972 I used an
> IBM- Selectric-like device (with a special type-ball) as a computer
> terminal. Programs were composed in a language called APL. Punch cards
> were not needed. To change a program you typed into the terminal and
> specified the modification. There was no physical patching.
>
> Garson
>
> On Thu, Dec 19, 2024 at 10:30 PM mr_peter_morris at outlook.com
> <mr_peter_morris at outlook.com> wrote:
> >
> > A recent Youtube piece on the origin of "patch" in software terms.
> >
> > https://www.youtube.com/watch?v=Nfm1HoTiYAY
> >
> > Allegedly,  it comes from putting a literal patch over an incorrect
> > punched card.
> >
> > This sounds like folk etymology to me.  I mean, the word "patch" is
> > widely used for various things that repair a minor defect. You can have
> > a patch to repair a small tear or hole in clothing, you can patch a leaky
> > roof, you can patch a flat bike tyre, you can put a patch over a missing
> eye,
> > and so on.  A software patch just seems like a natural extension of that.
> >
> > Also, it seems to me like it would be quicker just to make a brand new
> > punched  card, than to repair an old one. And I don't know what the extra
> > thickness would do to the mechanism. Would it still feed through, or
> would
> > it jam?
> >
> >
> > Can the word sleuths of ADS  help? What is the true origin of the term.
> What
> > is the first known use of patch in a  software context? Did the practice
> of sticking
> > patches on punched cards happen at all?
> >
> >
> >
> > ------------------------------------------------------------
> > The American Dialect Society - http://www.americandialect.org
>
> ------------------------------------------------------------
> The American Dialect Society - http://www.americandialect.org
>

------------------------------------------------------------
The American Dialect Society - http://www.americandialect.org


More information about the Ads-l mailing list