Hacker News new | ask | show | jobs
by giraffe_lady 9 hours ago
I had to get into an old tcl program for work recently and had the same thought. I wouldn't necessarily pick it today but it was kind of nice in a way that's unfamiliar to me from modern development.
1 comments

The tcl syntax is fine. And modern tcl is fine.

But tcl 7.x and before was a pure string-based language. Everything was essentially a eval(). People would hit syntax errors on production code.

Fun, painful times.

The flip side: the interpreter is super simple and fun to write.

Tcl is still entirely stringly typed. That's never changed.

There are under-the-hood optimisations to make it less insanely slow but that only affects performance.

Tcl is a cool hack (the interpret is simple to write) but it's insane to actually use it. I wish the EDA industry would realise that.

Tcl 8 introduced dual-ported objects. Everything can exist as a typed object that is convertible back to a string in certain cases (some form of string operation typically). That plus the bytecode engine makes it work completely different than prior releases.
That's just a performance optimisation. It doesn't change the semantics at all. From the changelog:

> Tcl automatically manages these values so their type is transparent to the Tcl script writer.

https://www.tcl-lang.org/software/tcltk/whatsnew.tml

Saying it's "just strings" is disingenuous. You could make a js interpreter that uses strings internally too.
No. Tcl is all strings from a language point of view. To your Tcl code, everything is a string.

Internally in the main Tcl implementation, it isn't just strings. It's a bit cleverer so it can be faster. But that is completely invisible to the Tcl programmer.

JavaScript is not just strings. It has real types (even if it is happy to coerce them to strings at the drop of a hat).

To put it another way, JavaScript has `typeof` which returns the type of an object. Tcl doesn't have that because the answer would always be "string".