jan

I got something lispish in the back of my head, and already wrote a little writeup for it that sits somewhere on my hardrive.
It goes even much further than traditional lisps when it comes to DSLs and parsing to the point where the main language is little more than a rule set to link any kind of DSL together into a common framework, i.e. it is a DSL to combine DSLs and as such provides full control over the whole compiler/runtime stack from within the language itself. The type system is its own DSL, and the core language or compile target is a form of mill general assembly, also its own DSL, there is a DSL to define new DSLs etc.

But there are other things with higher priority to do. Mill is not in the language design business. My writeup may find its way onto the community wiki at some point though and when the mill architecture itself is a little more stable I may attempt an implementation.

I still think the first reiteration

is by far the best.

The one where the text is in the middle looks far too much like ancient greek columns.

I like this one too. And without the borders even better. Simple, and has the advantage of incorporating the name, and startups need name recognition anywhere they can get it. And the two rows of eight blocks can be subtly incorporated into a lot of designs and layouts even without the name and hopefully eventually be recognizeable.

Don’t know if I would like color in it. All the really good logos don’t depend on color. And I still remember when Deutsche Telekom tried to trademark the color magenta.