I'm a non-native english speaker and, in my case, it doesn't bother me at all, but as I was reading it made me realise that when I see names like those, I just read them in my native language. I've always read imgur like "eem-gur", NGINX as "ngheenx", and SPDY was the only one I quickly spotted the "speedy" lookalike. For the rest I usually just say the letters (I always say S-Q-L and not sequel). I think that not having a native english developer community around me meant that nobody really cared about how you pronounce things.

SQL has the opposite problem in that it's legitimately an acronym (Structured Query Language) but is often colloquially pronounced as a word (sequel). However "S-Q-L" (spoken as letters) is technically correct.

I don't know if the "Sequel" pronunciation is a recent thing (relatively speaking to the age of the language itself) but it's only been the latter half of my career that I've heard people favouring that over "S-Q-L". In fact "Sequel" really threw me at first.

I remember failing a telephone interview once because I didn't know what `chmod 777` did. However I've always known it as "chi-mod" and the interviewer phrased the question as "What is C-H-M-O-D 7-7-7" so I started thinking it was an ISO / ANSI type standard I hadn't heard of and didn't make the connection it was a UNIX command. I remember being rather pissed off about that considering, at that time, I'd had > 10 years of UNIX systems administration so was intimately familiar with `chmod`.

I've always mentally pronounced chmod as changemod (and changepass, changeroot, changegroup), I guess it was sort of intuitive. Going back to the comment about dyslexia and naming, the unix core utils are probably a good example of "a lot" of abbreviated names that might be difficult to read for people with dyslexia. I wonder if there is a project to automatically create long named aliases for those (list, move, copy, remove...)

Pronunciation of shortened technical jargon is an interesting thing. I've noticed certain common variations but outside of very localized "communities" (like the employees of a particular company) it's not clear to me that there is a particular regional or community pattern for most of them.

* I've always assumed the conventional way to pronounce `chmod` was something like "ch-mod" (i.e., "chŭ-mod", with a short u). Is that less common than I think?

* `char` (like the C-style character variable type) is an interesting example. I've always read it as - and usually heard it as - "car". But there are groups where "char" (like the word meaning "burn") is more common. Strangely no one seems to say "care", which would be the corresponding part of the root word "character", right?

* "JSON" is another curious one. I've always just read that as "Jason" (like the name) but others seem to emphatically drop the missing "A", saying it more like "j-SON".

* For that matter, I've always read `fsck` like "f-seek", but apparently the creators intended it as "fisk".

Is there some kind of ethnographic study on this topic? Surely there are identifiable communities of these, but I've never been able to discern the broader pattern. I'm guessing it was once relatively regional, but it's probably more broadly distributed (and maybe domain/community-based) in the post-YouTube/podcast era.

I always read clang as "C-lang", then heard a LLVM developer said "cling". I feel like if the pronunciation is not obvious, the project should at least specify it somewhere offically.

Also, NGINX is pronounced "engine X".

> Also, NGINX is pronounced "engine X".

It absolutely is, but my first interpretation of that word was more like "n-jinks".

I mean, I understand that it's meant as "n-gin-x", but glommed together that still seems like an unnatural way to pronounce that word to me.

For what it's worth I am a native English speaker (and I assume nginx creator Igor Sysoev isn't), but maybe that's just me.