Replacement similarities

In my last post I stated that hits on posts of my blog went up by 20% compared to last year. The reason I checked is clearly vanity. The reason I checked on that day is an article about the increasing popularity of Rust. As a spiritual exercise I turned the different reasons given by programmers who ditch the language they use into a check list.

[X] genuinely new features
[X] rule of least surprice (mentioned often)
[ ] no footguns
[x] language designers listen to community
[x] complete tooling (mentioned often)
[X] central repo for libraries/modules
[ ] ideomatic code is fast
[X] FFI interface
[x] many platforms support
[X] memory sefety/type safety/data race safety
[ ] fast/low memory footprint (mentioned often)
[X] backward compatible to itself
[ ] proper error handling
[ ] does not force OOP

The items are from the article and apply to Rust. The ticks are where I see Raku excels to. A capital X is a strong match for us. The domain of the two languages are quite different. I strongly doubt anybody would like to write code for a microcontroller in Raku. Yet what programmers care about forms a sizeable overlap.

Another aspect of the Rust article is the general tenor of the participants of that questionary. All users come from a different language and where driven by pain to reach safer grounds. That means all of them are replacing $language with Rust and like it that way. The success of Rust lies as much in getting things right as in other languages getting things wrong. That was pretty much the goal for Perl 6 in the year 2000. We can tick that box too.

The core team of Rust sees a very strong open community but little resources for beginners. The same is still true for us. Maybe because writing beginners guides is really really hard. We have the same weakness. There are plenty of books for Raku already. I never really liked that idea. I can understand why a contributor with an academic background would want to write a book. You aren’t really an academic if you don’t. I can also see why academics don’t really have a problem with paywalls. And yes, books are paywalls. I used to climb that wall too. There is Stroustrup’s The C++ Programming Language in a bookshelf right behind me. Yet I doubt that a paywall is going to get us more users. A beginners guide must be online and to be reached by one click from raku.org.

There are also differences. Rust was meant to replace C/C++ as a tool to make security critical software at Mozilla. When Raku was release we didn’t really want it to replace anything. Not because there are plenty of dynamic languages out there that need replacing real quick in a hurry. Anybody here who loves PHP? Fancy another outtake of Node.js because a module gets pulled? No, we didn’t want to replace things because we want to be nice. The whole point of marketing is to be nice to the folk you want to reach and be nasty to your competitors. And it’s not like those Rust folk are actually good at marketing. When I read rust I think of old oil barrels, decaying machines and another unwelcome reason to wash my hands. Names seam not to matter all that much for programmers.

At the beginning of 2018 I gave a very brief motivational speech. I still stand by what I wrote back then. Yet I’m not happy with the success we had. Maybe it’s time to stop learning from the mistakes of others and have a look of what swiftly growing language communities are doing right.