Sometimes I wonder if it wouldn't be more efficient for people to just learn SQL instead of trying to build tools or layers on top of it that introduce more complexities and are harder to search for.
HN is inundated with posts announcing paper thin abstractions on top of existing technology or utilities that just move the goalpost of what you knowledge you need to be effective. It's a weird trend that seems almost entirely motivated by people wanting open source projects in their resume, or seek funding if its a startup.
> It's a weird trend that seems almost entirely motivated by people wanting open source projects in their resume
Developer here. I can't speak to what you see as the weird trend, but I can speak about sq's history:
- I created the first version of sq circa 2013 as my own personal dev tool to address some pain points (amongst other things: getting JSON out of various DBs from the command line in a consistent manner, and wanting an easy way to inspect DB schemas, i.e. "sq inspect")
- It was starting to be a minor PITA dealing with colleagues asking for the latest revision of the tool, so I open-sourced it. In 2016 or so I think?
- sq is FL/OSS and will remain so, no funding is sought, not even one of those "buy me a coffee" thingies
- I didn't create this HN post about sq, nor do I know the person who did. But thanks for sq's 15 mins of fame, kind stranger
Thank you for creating a composable, simple to grok Unix tool. sq is an example of what is so great about the Unix philosophy.
Personally, I can see a spot for sq in my developer toolbox. Instead of reaching for DBeaver’s bloated interface, or psql’s perplexing TUI, I like the idea of picking up an sq multitool to answer some simple questions quickly about my databases.
> It's a weird trend that seems almost entirely motivated by people wanting open source projects in their resume
That’s really harsh and misguided. If people didn’t do “paper thin abstractions” projects on their own time for the simple pleasure of doing it we wouldn’t have 90% of the successful projects we have today.
Let people have fun and don’t judge their motives when they’re making something Open Source. I can guarantee the person just thought “this would be cool to have” and implemented it.
What a cynical take. Most people working on side projects do so because they find them interesting or useful rather than to just score resume points.
They’re not moving anyone’s goalposts all usage is voluntary.
The fact that HN is mentioned is also confusing. If something had no value beyond paper thin abstraction I doubt we’d be seeing it on the front page with 500 votes.
Even if this project turned out to be less than hoped it seems counterproductive to complain about people creating and exploring. It’s a good thing to see and natural selection will do any sorting needed.
There are a several features here that go beyond 'learn SQL'. I don't see how you could jump to this conclusion if you read the docs.
Why does pgcli exist. What about dbeaver, datagrip, pgadmin, postico, LINQ, ORMs? It's almost as if people value different interfaces to databases.
Maybe it's not for you, but it's not hard to imagine that someone, beyond the author, might find it useful. Maybe it's just me but to dismiss those people as individuals that should just 'learn SQL' is a pretty rude thing to say.
I know sql pretty well and still find value in this kind of tool - creating schemas and inserting data is a clunky part of sql - the query language is where it really shines. So I can imagine using this to quickly insert some data or to get familiar with the schema and then dive in with normal sql queries.
Lots of people struggle to understand the declarative programming language paradigm. It is really pesky because since you're declaring what kinds of results you expect rather than dictating what must be done, you're forced to define your boundary conditions up-front. Much more fun and exciting to charge ahead without worrying about such things
Sometimes I wonder if it wouldn't be more efficient for people to just learn assembly instead of trying to build tools or layers on top of it that introduce more complexities and are harder to search for.
Actually, when it comes to SQL specifically, it almost demands using a language one higher up.
Any language that features the ability to embed/interpolate unescaped quotes is not secure enough to be used directly by professional developers. Just not worth it to track down all the potential injection attacks.
And if an abstraction lets you understand the structure of queries (eg for sharding them) and even writes the code for you and balances parentheses, flags other syntax errors at compile time etc. then it’s a net benefit.
And of course there is the Adapter Pattern to abstract away connection logic and differences between mysql, pg, etc.
I wrote the “Db” library 12 years ago in PHP and still use it.
I never released it on HN for reasons like this. But I might. Here it is, embedded in my larger library, but I might break it out:
It's funny because recently a "full stack" dev who's in reality 95% frontend was telling me he's not a fan of Tailwind and that he'd rather learn "proper" CSS.
And the irony is of course, he never wants to use a relational DB to avoid SQL, so No-SQL DB it is.
I've unfortunately met a number of "full stack" devs recently that are clearly 95% frontend as well, and have a very similar attitude as that person toward backend tech in general. We of course should not take such a small sample size and draw any sort of conclusions, but it's definitely a trend I'm looking out for now whenever someone claims to be "full stack." If the only backend stuff you've done is Firebase, or just a handful of serverless functions that don't do much, or if the "backend" you have in mind is just SSR-ing templates into HTML, then you really should qualify any claims about backend.
The sq and jq tools are both neat command line gimmicks but in my workflow their usefulness is very short. I can't imagine using sq on a query involving a handful of tables and some inner/outer joins. How would I know its outputting the correct SQL? If you mess up joins you end up with bad output.
Here's my theory: some developers see simple languages that are easy to learn and want to build something more complex to output that language. Maybe its a sub-conscious thing.
HTML is another one, if explained simply, HTML and CSS basic use cases should be easy-ish to pick up by nearly anyone. The fact that we have so many over-engineered frameworks and libraries out there that generate these is evidence of over complicating simple things. Maybe its called resume driven development? Maybe people see genuine useful frameworks that get wide adoption and are wannabes?
If I have to go look at log files to see if the intermediary tool is doing the right thing, I might as well just write the SQL myself, right? jq has more use cases dealing with JSON that could come in handy. sq seems to handle trivial use cases but for complicated data models its going to be a hindrance.
I think it'd be a moot point if SQL wasn't painful and awkward to work with in the first place. But database purists control it and won't let go, so we will have to live with everyone else inventing layers to make their lives easier.
For me this feels like the complaints about error handling in Go. People who work with it all the time, don't even think about it past the first week. If you are starting out it might bother people because they are not used to it.
Personally I really like working with SQL and find it quite elegant, I always encourage people to use it as it really is a job-superpower if you can just dig up issues directly in the DB quickly. It has a long history so for every question you have there's many answers or avenues to ask them.
I agree with the GP’s point that people should probably just learn SQL, but there are things about SQL that are objectively bad, especially from the perspective of a software developer.
It’s not a very composable or consistent language, so I think it makes total sense that we see so many abstractions over it.
The issue with the abstractions over SQL is that while they fix some problems they always introduce a bunch of new problems so in the end SQL is still preferable. I have yet to see an example where that is not the case.
The fundamental issue is that they have to generate SQL at the end of the day, so there’s a hard limit on how much you can really change. I don’t know why every database treats SQL as the only API, even Postgres. Even the extension systems, which have the opportunity to hook directly into DB internals (and has no standardization to bother meeting) end up with SQL as the API to do actual db operations.
> Even the extension systems, which have the opportunity to hook directly into DB internals (and has no standardization to bother meeting) end up with SQL as the API to do actual db operations.
FWIW, nothing forces an extension to do so. I'm pretty sure there are several that do DML using lower level primitives.
For me query builders are the quintessential example. Not ORMs, just thin layers that allow you build up a query in pieces.
If you have cases where you might need to conditionally join, or have any reasonably complex filtering based on input, building up a plain SQL statement using string interpolation at the call site gets very messy very quickly.
I have used many query builders and experienced both the upsides and the painful downsides. The by far best one I have used has been JOOQ (Lukas Eder is a genius who really understands SQL well) but even that often causes more pain than it helps. Compared to the issues caused by them I in most cases prefer string interpolation of SQL.
Take the simple example of joining via foreign key to another table in a parent child relationship. SQL makes you redundantly specify the other table even though it is fully defined by the foreign key relationship. Purists will say "what if you actually want to join on some other column" or "what if there are multiple foreign key relationships" and a plethora of other "what ifs" and all these ignore the reality that 99.99% of the time it is completely unambiguous and they are just happy that millions of developers all around the world are uselessly typing redundant terms into queries. This is what leads people to eventually say "screw it, I'm scratching that itch" and we have "yet another abstraction layer".
I like Go a lot and can work with it's error handling paradigm, but I still often wonder if the same semantics could be accomplished with less verbosity in a way that makes the underlying algorithm more clear when reading the code.
Verbosity is in the eye of the beholder; To me, it's the verbosity of error handling that makes the algorithm clear. Of course I recognize with others that opinion probably changes depending on whether the person reading a given bit of code views error handling as part of the algorithm.
> People who work with it all the time, don't even think about it past the first week.
I think SQL is often quite awkward, at least when you look at people exploring alternatives, like https://prql-lang.org/
Not just “the standard” variety, but also all of the vendor specific varieties that you’ll use in practice (MySQL/MariaDB, PostgreSQL, SQLite, Oracle, SQL Server etc.) and how the features offered by each differ, how the support for custom types differs, how the procedural languages differ, the actual experience of using them, seemingly half baked error messages when compared to most imperative languages, varying support for getting the query optimiser to do what you want (e.g. hints), query plans that aren’t pleasant to stare at, often no support for automated analysis of how things are running and suggestions for indices (e.g. Oracle had that sort of feature, it is sometimes helpful, but the automatically generated indices are or at least last I checked were treated as their own special thing and you couldn’t easily delete them). Even things like varying support for working with geospatial data, JSON or time series data.
Not just that, but also the tooling (or the lack of it) - good luck debugging the execution of stored procedures in your DB or placing breakpoints in there, good luck hooking up observability/tracing solutions as easily as you would for your back end, good luck debugging why your database link calls take 100x longer when executed through JDBC in your app but not in the developer tooling.
Not that ORMs or NoSQL make everything much better, you’d just trade one set of annoyances for another, especially when you end up trying to generate dynamic SQL with the likes of myBatis XML mappers. Don’t get me started on people over fetching or ending up with N+1 problems with most ORMs, or executing bunches of queries against the DB just to retrieve some data that you might as well get in one go with a well crafted DB view. Not that you can’t make your DB schema a mess as well with a liberal application of EAV or OTLT (or just attempting to make polymorphic links).
I think SQL as a language makes sense, but PRQL often feels more sensible. I feel like the current RDBMSes out there are very powerful, but that there are also aspects about them (and the differences between the ones you’ll use) that absolutely suck. It feels like where back end languages are getting progressively better DX, databases instead gave us… CTEs? Some syntactic sugar for JSON in PostgreSQL? Feels like they move painfully slow.
For what it's worth, that's why I welcome every new tool or attempt at improving things, even if they won't replace the usual enterprise stacks of having some RDBMS, a JDBC/ODB/whatever driver and probably an ORM on the back end.
I truly have never understood this mentality. To me, SQL is very convenient and quick to work with. It's extraordinarily painless.
I've done data processing in Perl and Python before. THAT is painful. What I think people don't realize is that doing procedural "querying" in a typical PL is, like, 10x the amount of code as an equivalent SQL query. I don't see how the alternative is much better.
SQL just doesn't work like a typical PL, but that's a good thing.
While I agree with your general idea my lists are quite different.
I would move Zig (innovates a lot in language and compiler design) and JMAP (IMAP is horrible and needs to be replaced) up and CMake and Bazel down (I count them the same as Meson). I would say the jury is still out on Google's tweak and on seL4.
SQL interop is where the pain is at. Using standard tooling of most database systems, best you can get is CSV with all the pains this shithole of a data transfer format brings.
That doesn't match my experience at all. Every database I'm aware of has multiple options for data export and import for batch and command processing in the standard tooling. CSV is almost never "the best you can get". What is going to be best depends on your use case, and indeed at a bare minimum you can almost always change the field and line seperators and get something that works just fine on the commandline and avoids most of the stuff people find hard about CSV.
Additionally, if people like jq, they don't need a special tool like this. They can just get the database to output json and use json/jq - another tool isn't needed. In postgres you can do something like this
select array_to_json(array_agg(row_to_json (r))) from (
... put your sql query here...
) r;
...and postgres will output a json array for you where each item in the array is a json map of a single row of the output.
I'm sure other databases have similar functionality.
Out of the random shit I had to deal with CSV file wrangling in the last two years:
- no defined encoding, so it may be anything from US-ASCII over ISO-8859-x to UTF-8 and it's always a guesswork what it actually is
- no definition of the separator sign, usually it can be a literal comma, a semicolon, or tabs
- escaping of the separator sign in column values is always fun
- escaping of newlines is even MORE fun
- line endings are not specified
Every piece of software there is has their own expectations on what specific kind of format it expects, and to make it worse, using "Microsoft Excel" as a target/source isn't foolproof either because it behaves differently between the Mac and the Windows version!
JSON in contrast is clearly defined in all of these issues and has battle-tested implementations in every programming language there is.
Most of these issues while real don't actually arise in this case, because we're not trying to ETL some random file, we are the ones talking to the database so we get to choose exactly how the data gets formatted on extract.
For example, here's your list fully handled in postgres:
1. SET CLIENT_ENCODING TO 'value'; (eg 'UTF8')
2. COPY ... with FORMAT CSV DELIMITER 'delimiter_character' QUOTE 'quote_character'
Now the output format is fully specified and everything just works fine (including for input into excel)
The values in each record are separated by the DELIMITER character. If the value contains the delimiter character, the QUOTE character, the NULL string, a carriage return, or line feed character, then the whole value is prefixed and suffixed by the QUOTE character, and any occurrence within the value of a QUOTE character or the ESCAPE character is preceded by the escape character.
Well not every application can read any file and either way "sq" isn't going going to make any difference. Like I said they are real issues but don't arise in this case.