Hacker News Reader

Comment by asmithmd1 | Hacker News Reader

Being a mechanical engineer by training, I always assumed it had come from the requirement that all boilers must have a nameplate affixed with a bunch of not really useful information about how it was tested. The hot water heater in your house has a boiler nameplate, and nuclear power plants have a "nameplate" rated capacity:

https://www.nei.org/resources/statistics/us-nuclear-plant-ow...

View on HN

Regulation requiring these is newer than the term 'boilerplate' which was used to describe the boiler-making material.

https://www.etymonline.com/search?q=boilerplate

And the OED on the original 'boiler plate'

1. Usually as two words. A plate or sheet of metal of suitable thickness and strength to be used in constructing the shell of a boiler. Also as mass noun: metal in the form of such plates. Cf. boiler-iron n. at boiler n. Compounds 2.

1793 Star 13 Nov. (advt.) Table of the number and dimensions of Boiler Plates and weights thereof, suitable for Boilers, from eight to sixteen feet and a half diameter.

1860 D. K. Clark & Z. Colburn Recent Pract. Locomotive Engine i. i. 1/1 The earliest recorded trials of the strength of boiler-plate, are those of Mr. Fairbairn, made in 1838.

1874 J. H. Collins Princ. Metal Mining (1875) xiii. 74 The kibble is simply an iron bucket made of boiler plates, riveted together.

1915 J. Wedgwood With Machine-guns in Gallipoli i. 4 Our mechanics..lined her bridges with boiler plate and leaky sand-bags.

I don't think any of those show that the use of "boilerplate" to mean "required verbose writing" predates boiler nameplates though?

They don't but it explains why pre-set/cast type plates that were sent ready-made to newspapers were called 'boiler plate' which in turn explains the origin of 'boilerplate' a lot better than the nameplate thing.

A large slab of standardised content/dimensions ?

Yea, that's interesting, but I still bet this is likely a case of no one seriously looking into this term and knowing the full lexical space that well.

Boilerplate's early use was referring to boiler material, but it likely got semantically mixed together with the term for "standard form plate affixed to a boiler", as this term is actually more from the legal industry where it's literally a document with blanks you fill out, which more closely fits this definition.

I think people have looked into it and produced the most likely explanation. The name plate thing is a fun kind of 'folk etymology' theory but isn't supported by any evidence at all beside the hunches of people who've been around boilers. The other variant has a wealth of evidence supporting it.

Yea, I think this is actually the real answer, and OP is just wrong.

Here are old and modern examples where the boilerplate is a standardized plate that's filled out using punched letters/numbers.

https://lestaret.files.wordpress.com/2009/09/boiler-3a.jpg

https://www.steephillequipment.com/wp-content/uploads/2017/0...

(am also in the MechE space)

Assumptions are immaterial, however. Etymology of the term is as noted in the article. It is more likely that this use (as on water heaters and such) derives from the same typesetting usage.

It's always satisfying to be certain, but the meaning of words does shift over time, and sometimes words have multiple origins. Assuming that 70s or 80s programmers knew anything about printing presses is tenuous. Given the long assumption that boilerplate refers to plates on equipment and superior fit of "fill in the blank" documents, it seems more likely that this is the intended meaning. Perhaps the term was heard and misunderstood early on? But the proposed origin doesn't align with the community that's using it to refer to source code.

> Assuming that 70s or 80s programmers knew anything about printing presses is tenuous

True, but as you say the meaning of words shift. It is entirely plausible that they were exposed to the term in its bureaucratic meaning, and continued to use it without any connection to actual boilers - the same way we do today.

I remember asking my father in the 1980s what "boilerplate" meant, having encountered the term in a newspaper. He relayed it to me in the legal sense of the preliminary standard matter on a contract. He was a software developer starting in the 1960s so I have absolutely no doubt that the bureacratic usage was well known in those circles.

As to 70's and 80's programmers knowing about printing presses I'll point out that a significant early AI program (1968-1970) was named SHRDLU. One familiar with printing history might recognize those letters as part of the second column of moveable type characters on a Linotype machine (and other type-casting machines.) I didn't look up a reference, but I recall that the first row of bins for hand set hot type letters followed the same convention of letter frequency in English text (for english speaking countries that is.) etaoin shrdlu That string of characters became more well know due to its appearance in hot type set news papers of the era. The characters sometimes accidentally made it to press rather than being pulled as part of an erroneous line of text.

https://en.wikipedia.org/wiki/SHRDLU https://www.wikiwand.com/en/Etaoin_shrdlu

> Assuming that 70s or 80s programmers knew anything about printing presses is tenuous.

I don't think that's the case at all. Lots of work was done computerising newsrooms in that period.

I find more likely that the programming "boilerplate" comes from the legal "boilerplate".

The replies seem confused.

No one ever called it “boilerplate code” for any reason having to do with boilers or typesetting or water heaters.

It’s a language idiom. It entered common use through reasonably well documented avenues, and that is via typesetting.

We use it because that’s just how language works. All these “it makes more sense to me if it referred to such and such” misunderstand language.

Merriam Webster[1] online disagrees with you and agrees with OP. That said, etymology is often contentious.

https://www.merriam-webster.com/dictionary/boilerplate

This looks much more in line with programming boilerplate with literal fill-in-the blanks.

I would say the article captures early etymology but this could be why we call it boilerplate.

What I'd like to learn is the earliest computer related usage of the term. Wikipedia says[0]:

> The term arose from the newspaper business. Columns and other pieces that were distributed by print syndicates were sent to subscribing newspapers in the form of prepared printing plates. Because of their resemblance to the metal plates used in the making of boilers, they became known as "boiler plates", and their resulting text—"boilerplate text". As the stories that were distributed by boiler plates were usually "fillers" rather than "serious" news, the term became synonymous with unoriginal, repeated text.[2][3]

> A related term is bookkeeping code, referring to code that is not part of the business logic but is interleaved with it in order to keep data structures updated or handle secondary aspects of the program.

[0] https://en.wikipedia.org/wiki/Boilerplate_code#Origin

At a certain point, the slang term "boilerplate" branched away from the derogatory sense in reference to newspapers to the more neutral sense of formulaic. The first "neutral" sense reference in the OED is from 1949:

Navy Contract Law (U.S. Bureau Naval Personnel) ix. 212/2 -- "This type of clause has proved so valuable that it is presently standard ‘boilerplate’ not only in shipbuilding contracts but also in almost every kind of contract."

The first reference in the OED for it being used w/r/t computer stuff is 1990:

L. Wall & R. L. Schwartz Programming Perl vii. 379 -- "Like mus itself, man2mus is not 100% effective, but can save you a lot of time producing the initial boilerplate."

The actual man2mus program appears to be not accessible to the easily searchable public internet anymore, but there's a few references to it - "usub/man2mus A manual page to .mus translator". It's Perl, apparently.

Neat.

The reference to "Programming Perl" was no doubt helped along by Perl developer Jesse Sheidlower who worked at the OED in the early 2000s. From his Wikipedia page: "Although not a computer programmer by training, Sheidlower introduced Perl to the North American offices of Oxford University Press and developed tools for data manipulation when no programmers were available. He is also one of the core developers of Catalyst, a popular Perl web development framework."

> all boilers must have a nameplate affixed with a bunch of not really useful information

Lest we forget why those nameplates and the standards they reflect matter: https://en.wikipedia.org/wiki/Sultana_(steamboat)

> Sultana was a commercial side-wheel steamboat which exploded and sank on the Mississippi River on April 27, 1865, killing 1,169 people in what remains the worst maritime disaster in United States history.

Additionally "boilerplate" was used in legal circles before the invention of the linotype ("boiler plate") machine, so the linked article is almost certainly wrong.

https://news.ycombinator.com/item?id=33617106

Not a mechanical engineer, but similar assumptions. Chances are the term wouldn't have become a fixture in software jargon without that folk etymology. I'm genuinely surprised that it's quite literally about copy/paste, and not about formalities at all!

I have only ever known boiler plate code as that kind of code that is needed to stitch things together, initialize libraries and so on. So not exactly copy/paste, but rather "formulaic" code that has not yet been deemed worthy of an abstraction

I believe the term came about in the early days of steam power when designers had a habit of “overcompensating” because boilers had a history of catastrophic and deadly failures. This lead to excessive and unnecessary overbuilding of the boiler walls which ironically made it MORE dangerous when it exploded.

I had always assumed that it referred to the practice of making mock-ups or stand-ins out of cheap steel in the aerospace industry.

https://afspacemuseum.org/artifacts/apollo-boilerplate-capsu...

That was always my assumption as well.

Reply on news.ycombinator.com