Hacker News new | past | comments | ask | show | jobs | submit login
John Glenn popularized the word “glitch” (2019) (airspacemag.com)
101 points by Hooke 6 days ago | hide | past | favorite | 44 comments

"Glitch" is an interesting word. I've rarely heard it used to describe problems within the software industry, and almost never by actual software developers. Yet journalists seem to love it, and you often see headlines like "Software glitch costs bank $500m a day" or "Report says software glitch caused fatal accident", even when stronger-sounding words like "defect" or "failure" would seem more appropriate.

And politicians seem to really love it as well, perhaps because it is a funny-sounding, non-threatening word, implying a minor hiccup rather than a serious defect. So, whenever some new, high-profile software system crashes and burns upon launch, you get lots of folksy talk about glitches: there's just a few glitches, they're fixing the glitches, you've got to expect a few glitches, etc. Whenever I heard such talk, I like to imagine the difference between the attempts to downplay the problems, and the likely reality of stressed-out engineers scrambling to fix a system that was released in a half-finished state due to an unrealistic deadline.

It's very common in gaming circles (not game dev afaik, just ppl who play games). I found it weird wording too, at first; it seemed like a euphemism for 'bug' to me, but I think I now have a more nuanced understanding. I think a 'glitch' (to people outside the software industry) is something that makes software behave in strange or 'wrong', unintended ways, but not crash or give straight up gibberish results. Like, a game crashing is a 'bug' but a character being transported across the map is a 'glitch'.

Likewise, if software rounds bank transfers to the nearest thousand $ and loses the remainder, but only on Thursdays when there is an 'r' in the month, then that's a glitch; but when the software doesn't open that's a 'bug' (or rather, 'it doesn't work').

It's an inconsequential difference from a software engineering pov, but in my experience it's major to outsiders, to the point where they start looking for workarounds for 'glitches' or explain things away, but throw their hands up at the first sign of a 'doesn't work'. I don't understand the thinking, but at least trying to understand users in similar circumstances has made me realize that there are things where I just can't understand someone else's point of view, in which case I need to switch problem solving or design methods; because no matter how hard I try, my 'how is that not obvious' approach isn't going to work.

As a game developer I can confirm that not only is your description of what constitutes a glitch accurate, we use it that way in game development as well.

I'm a graphics programmer who previously used to work with UI and we often use the word glitch to refer to animation hiccups, graphics issues due to corrupted render buffers, etc.

In the Need For Speed games from Ghost, we had a visual effect of disturbances running across the screen which was shown when you were hit by an EMP or seriously damaged[0]. We sometimes referred to that visual effect internally as the "glitch shader".

edit: look at the UI elements: speedometer, mini map, etc.

[0] https://youtu.be/qE4fgixKLpE?t=387

It's funny how the weird behavior is a "glitch" until you narrow down exact circumstances when it occurs and then if the issue can be reproduced reliably, then it's a straight-up bug.

Yes, I think the implication is that there's some sort of weird, inconsistent behavior. (Which is consistent with something like voltage transients that are mentioned in the article.)

A glitch is unintended program behavior. The cause is undefined. It could be a programmer mistake, a content creator oversight (holes in 3D geometry, etc.), hardware failure, or a random cosmic ray causing one-shot disruption.

A bug is unintended program behavior generally caused by programmer mistake specifically.

Bug is a specific case of glitch.

Then there's also error, which is a glitch the computer itself is aware of. An exploit is a glitch that can be used to gain unfair advantage.

I've also seen similar usage of glitch versus bug, and think it's a useful distinction. But I'd say that actually it can be consequential from an engineering perspective, and can even run counter to the outsider intuition - that is, "doesn't work at all" is usually easy to diagnose and often not that hard to fix (e.g. "we just need to update the cert/restart the db/etc."), while "weird every-other-Thursday glitch" is harder to pin down and usually more intertwined with system logic ("why does our videogame physics system cause characters to float but only on levels 3 and 7?").

Basically, "complicated system doesn't work" is often a less intricate problem than "complicated system almost works." The perception may be opposite because it seems like more work to fix something that is "more broken", but that's the same logic as LOC as a good productivity metric. As with all things, there are exceptions, but I'd say the tendency is that glitches are harder to fix than bugs.

I think "glitch" is a word we apply to software that we rely on but can't fix, software that we cope with like we cope with the weather.

Hey I got an unrelated question for you. Can you email: g at colonist.io

For anyone old enough to have watched the original 1987 Verhoeven RoboCop, surely the word glitch is permanently associated for the hyper-violent scene where the enforcement droid ED-209 is demonstrated as a massive fail, only to be summed up by the project owner as a “glitch”. A scene that was outrageously violent in the theatrical release and even moreso in the director’s cut.

I know Verhoeven’s perspective was to use violent satire to speak about fascism, and the increasing militarization and technology-adoption of police forces — themes that are even more topical today.

I think your description of how glitch is used today, as a folksy (and ultimately dishonest) non-threatening way to describe a failure while stressed-out engineers scramble to prop-up a half-finished system, directly captures the ED-209 scene — where people seek to mislead about the true situation through the use of the word “glitch”.

While it might be uncommon among software devs, it has an established meaning for circuit designers. A glitch occurs when, say, two gates that are meant to switch at the same time actually have a small asynchronicity. If the output of a circuit is meant to be (A OR B), perhaps you'd expect it to stay high constantly while A drops to 0 and B rises to 1, but it briefly outputs 0 during the transition.

That pulse is a glitch and if your circuit isn't designed robustly it might go off and increment a counter somewhere or so something else it's not supposed to.

I would agree the 'glitch' it more appropriate for electronic system since they can be affected by an unexpected purturbance in physical reality. For instance if noise on the power supply causes some issue, I would consider that a glitch. Computers are more or less deterministic so when they crash it's because of some logic error, not a physical reality 'glitch'.

> Yet journalists seem to love it

Companies know this.

Journalists have been trained not to investigate the underlying (often, dysfunctioning organizational) causes of software meltdowns - and happily stop at "glitch!" - which seems to absolve everyone of responsibility.

Interesting point. 'Glitch' seems to have it's own agency.

Another good bit of etymology. Popularization of 'cool' / Lester Young.

To me "glitch" implies a one-off transient error, which ironically is the original meaning of the word "bug" (an actual insect trapped and blocking a switch or electrical connection).

Software people don't use glitch except for transient network issues or hardware reliability because software is deterministic. If it worked yesterday it will work the same way today, once you identify the actual bug-triggering conditions.

> "Yet journalists seem to love it"

Journalists have a predilection for using the same keywords and phrases in their headlines over and over again. Such as: slams, taps, full-throated, silence is deafening, soars, plunges, meltdown, and of course, glitch. I have no idea why they love these phrases so. Is it an established convention? A kind of shibboleth among professionals? Monkey-see, monkey-do?

On places HN people tend to focus on the high stakes kind of bugs that cause severe problems. But the living reality for most users is probably different. It is a constant low level of issues. More an annoyance than a threat. Glitch is a good word for this. For example I get kicked out of the work VPN. My cell providers app doesnt show my balance and just spins. Outlook search doesn't give any results. Small errors every day.

I guess it would never have occurred to me that the term wasn't commonly used in the software industry. I guess it's more associated with transient hardware issues but software certainly has no lack of weird hard to reproduce bugs too. (Or interactions between software and hardware of some sort.)

My 2¢: “bug” conjures up the actual physical creature to non-technical people (“what’s a software bug?”) whereas everybody immediately understands “glitch” is a software/hardware/computer issue. People tend to avoid overloading words so to speak, linguistically

Thats because the first time it was an actual bug https://www.nationalgeographic.org/thisday/sep9/worlds-first...

Alternative perspective: the glitch is the whole thing.


A glitch! The glitch, is related to the snark — but not, you see, the boojum.

Yes, it is often used to imply a software bug which may or may not exist, but without taking any responsibility (or rather, pushing the blame on developers).

Wow, that is interesting. If a colleague turned to me and described a bug/issue as a glitch I would be very confused.

If you hear somebody use it, it basically means they aren't a programmer.

As a German, I just realized that my intuitive understanding of the (seemingly English) word "glitch" seems to be far better than that of native English speakers, because the Yiddisch word "glitshn" comes from German "glitschen", which comes from "gleiten" - to slide, slip. So to me, it always meant exactly that: the program slipped and fell.

French has a related word with the same meaning, "glisser". Both the German and the French term appear to come from the same original word. The French word "glacier" and the English word "glide" are also derived from it.

As is the German "Gletscher" (glacier).

I used to see a Metrobus on the Washington, DC, streets, with an advertisement from the German government and listing a lot of German words in common American use. My impression is that most of them made their way into use via Yiddish.

Something like a quarter of English words are Germanic in origin, but mostly not modern German. Many of the identifiably German words in English are more recent inventions (eg kindergarten) or yeah, Yiddish.

"And" is Germanic, but we don't think about that.

I'd love to import the word 'gitch' into German (like the word 'cringe' has been incorporated into youth slang), but it doesn't work because there already exists 'glitsch(ig)' - a close but yet very different word.

I especially like the phrase "someone is glitching out". I think this originated in scifi (think a bot), but you can use it to describe mental problems or just wierd behavior. It's impossible to use in German: "Jemand glitsht" - wait, why are they slippery?

I wonder if he picked it up from the German scientists working for NASA

I doubt those particular scientists were big on Yiddish :/

It's possible! Adolf Galland (Luftwaffe fighter ace) used to greet pilots with "Hals- und Beinbruch", the equivalent of "break a leg", which descends from the Hebrew -> Yiddish blessing "hatslokhe u brokhe".

This write-up seems to skip the part of the story where John Glenn popularized the word glitch.

Paragraph 4.

No; that describes how Glenn learned the term. Nothing in the article describes how he popularized it.

Well, TFA notes that John Glenn introduced what was previously NASA jargon to a mainstream audience in a book about the Mercury Seven.

The book's publication date was during a very exciting year of the "Space Race" — first American in orbit (John Glenn), first orbital solar observatory, first spacecraft to impact the far side of the Moon, first active communications satellite, etc. — so was presumably very popular.

I'm not seeing any evidence the book was particularly widely read; it's barely googlable and shows up on no historic bestsellers list I can find. Apparently I'm not the only skeptic:

> One might surmise from this that Glenn and his fellow astronauts took a highly technical term and "adopted" it into a more general term for any malfunction. But that overlooks the history of the term predating the space program — a history only now becoming clear.


I wouldn’t call myself a tech expert. I’m more of what is sometimes called a prosumer. But I know enough that friends and family often rely on me for light tech support.

As any experienced tech would, the first thing I do in most cases is just turn off the offending device, feature, or program, and then turn in back on again.

As I doubt I have to tell hn, that solves a great number of issues, but you rarely know why. The person I’m helping will ask what exactly was wrong in the first place, and generally all I can say is: “Dunno, something must have glitched.”

Google Books Ngram Viewer [1] shows a usage blip (glitch?) in 1962, a exponential growth starting in 1964.

[1] https://books.google.com/ngrams/graph?content=glitch&year_st...

Considering the large number of German engineers in the space program, via Werner von Braun and his team, are we sure that the use of "glitch" in the space program wasn't of German rather than specifically Yiddish origin? The original meaning would be the same.

I like to imagine Neil Armstrong saying "stop trying to make glitch happen, John"

I wish that word would die. It is used far too often to imply that software made a mistake (presumably had a bug) when in fact the user misunderstood something or pushed the wrong button.

Its overuse leads people to mistrust systems more than they probably should.

It also allows people to excuse actual faults as if they are unavoidable realities of life, ultimately leading to increasingly relaxed standards and more future flaws.

Similar examples are "thing", which you use to refer to something you don't know the name of or can't be bothered to name specifically (because you are mentally lazy or you haven't tried to learn and remember), and the new popular "fake news" -- the phrase you can use when you don't believe what was said, or you don't want to believe it even though you know it may be true, or perhaps you are unwilling to accept the information and want to suggest to your listener that they should also ignore the information.

Glitsh, slippery slope indeed.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact