I remember listening to this when it was recorded and I still remember details from it years later – probably my favorite Changelog podcast ever. It's obviously very foolish to make software choices based on how much you like its creator as a person. But Richard Hipp comes off as a guy so likable that goshdarnit, I hope his little project succeeds. Jokes aside, though, I don't think it's entirely coincidence that SQLite, something that is so good and reliable, was made by someone who seems so conscientious and thoughtful.
Though I have to admit, I was and am still disappointed to learn that the official way to pronounce SQLite is "S-Q-L-ite", i.e. "like a mineral".
Conscientiousness and empathy are rare commodities in software development. It's difficult for somebody to get out of their own head and consider a problem fresh, and it's not always personally rewarding to do so. It's also expensive, if you want to back it up with user testing.
You see it, even, in ostensibly public-consumption open-source projects, where the goal is clearly not merely to scratch one's own itch but to market the project for uptake for this reason or that. I feel like SQLite is a great example of what happens when a team is able to foster the kind of empathy to really understand how the user wants it to be.
I think there's a case for disputing the official pronunciation because it's based off an "official" pronunciation of SQL. I've heard it ess-queue-elle and sequel, but personally side with sequel. It was originally called “Structured English Query Language”, abbreviated and pronounced "SEQUEL"; later shortened to "SQL" but same pronunciation. So I figure sequel-ite is a perfectly legitimate pronunciation.
That's how I do it. I understand the etymological reasons for "S-Q-L-ite", but SQLite is just too ubiquitous in my everyday work for me to speak 4 syllables when referring to it.
PostgreSQL is the fun one out because they officially say it is Postgres-Q-L.
The "it's always S-Q-L and never `sequel`" is IBM's fault and an early trademark infringement issue in computing. (IBM was told it couldn't call it "SEQUEL" by an aircraft company.)
For anyone curious about the trademark infringement, I hunted down the excerpt from the book that Wikipedia uses as a source (a book called Databases Demystified) and this is what it says in the book:
>> The forerunner of SQL, which was called SEQUEL (for Structured English Query Language), first emerged in the specifications for System R, IBM’s experimental relational database, in the late 1970s. However, two other products, with various names for their query language, beat IBM to the marketplace with the first commercial relational database products: Relational Software’s Oracle and Relational Technology’s Ingres. IBM released SQL/DS in 1982, with the query language name shortened to “SQL” after IBM discovered that “SEQUEL” was a trademark of the Hawker-Siddeley Aircraft Company. When IBM released its next generation RDBMS, called DB2, the SQL acronym remained. To this day, you will hear the name pronounced as an acronym (S-Q-L) and as a word (see-quel), and both are considered correct pronunciations.
I was hoping for a bit more interesting or detailed story.
I've heard that, but assumed it came from splitting up the name as "Postgre"+"SQL"; if you leave off the SQL part you get a word ending in E, and "postgreh" can't be right, can it? :)
Which is why I've also heard developers that either assumed the 'g' was silent or a transposition problem and you get "poster SQL", "postreg SQL", or worst of all "posgret SQL".
Somewhere, I believe in an HN comment, I saw a Postgres developer say that one of the biggest regrets of the project naming was capitalizing that "shared S".
The norm is to pronounce it by how likely you think you might get yelled at for trademark infringement by a holding company of the assets of an old aircraft company: https://en.wikipedia.org/wiki/Hawker_Siddeley
The name confusion is all IBM's fault.
> Chamberlin and Boyce's first attempt of a relational database language was Square, but it was difficult to use due to subscript notation. After moving to the San Jose Research Laboratory in 1973, they began work on SEQUEL. The acronym SEQUEL was later changed to SQL because "SEQUEL" was a trademark of the UK-based Hawker Siddeley Dynamics Engineering Limited company.
I looked it up, you're right, it is officially my-ess-que-ell. This leads to another question, how do you pronounce "NoSQL"? I couldn't find a convincing answer on the web. I used to say "sequel" for everything SQL related but clearly I was wrong.
One of the many reasons I love sqlite and redis are because of the attitude of the developers. Sqlite has just an uncanny amount of testing, I wonder if it's the most formally tested piece of user software (outside of formally proven software and the space/medical industry I suppose).
And Salvatore (redis) has a way of simplifying issues that gives me a great deal of confidence in the decisions he makes and the code he writes. I really like the code and reasoning behind his 'linenoise' library in particular
> Jokes aside, though, I don't think it's entirely coincidence that SQLite, something that is so good and reliable, was made by someone who seems so conscientious and thoughtful.
The same could probably be said about Redis and Salvatore Sanfilippo.
Though I have to admit, I was and am still disappointed to learn that the official way to pronounce SQLite is "S-Q-L-ite", i.e. "like a mineral".