Phil Bagwells then started working in that direction also, see:
https://www.youtube.com/watch?v=K2NYwP90bNs
Some of Clojure data structure are pretty much from Okasaki, but Vector or HashMap.
Indeed there's novel stuff there.
Unless I'm misunderstanding, and book != thesis.
[EDIT] - Looking through the Amazon listing, it looks the book is certainly not just a republish of the thesis... Looks like I have even more reading to do
https://smile.amazon.com/Purely-Functional-Data-Structures-O...
also for reference some companies keep betting on CL today: https://github.com/azzamsa/awesome-lisp-companies (and I suppose that's just the emerging iceberg)
Wookie is one of those that does either-or -- the main competency is async io, and it offers multi-threaded perf with some plumbing (see the Threading guide[0]).
I don't mean to put down the libraries out there (lack/clack, hutchentoot, wookie, woo, ningle, etc), but for languages that aren't single-core-by-design/circumstance (i.e. node, python, ruby), I don't expect it to be this difficult to find a library that handles both multi-threaded operation and asyncio at the same time, abstracted cleanly so users don't have to think about it or add plumbing.
Performant web servers that take advantage of what your language is good at is table stakes these days, if things like this are not standardized/obvious choices at this point it tells me that I won't be able to find other things down the line. It's a circular problem though, someone has to build all the libraries that I'm expecting to find, and it could be me -- I just don't want to join the ranks of those committed to the CL ecosystem, right now at least.
> for reference some companies keep betting on CL today: https://github.com/azzamsa/awesome-lisp-companies (and I suppose that's just the emerging iceberg)
I didn't say that companies don't bet on CL, I just personally couldn't imagine recommending it to anyone as the language to do a new project/startup/whatever in. Lisp has value, but for me that value is much more slanted to the way of thinking than for practical usecases, even more than Haskell which is weird.
Popularity is not “innovative”.
However, if "Clojure" in that sentence is interpreted in a wider, more general and necessarily less concrete sense -- to represent the language, ecosystem, branding, etc -- good marketing/advertising, community management, and general operations-level can be rife with places to innovate.
An excellent language that never gets adopted has effectively zero value IMHO until it's corpse is ransacked by a language that does get adopted. People rediscover things that older languages have done all the while almost every week it seems -- creating a language and the requisite community to actually drive adoption can be considered innovative if the rest of the rest of the landscape is sufficiently similar. Clearly the Clojure project is doing something differently from the Common Lisp/Scheme/Racket crowd.
It is like Windows. When it appeared it was significantly inferior to Unix options, but it made a hundred or a thousand times more people to experience computers.
If a 1 to 10 percent of people is naturally innovative and creative, then the platform with most users ends innovating way more.
Lisp use at first was restricted only to Academics, it was so inefficient that you even required a very special(and expensive) purpose computer to run it. Those academics were not paying the machine with their paycheck so they didn't care.
More affordable solutions like c became popular and forced Lisp to improve(creating compilation pathways and so on), when it was obvious everyone was using the c toolchain(c++, objc, java).
Lisp was spreading in the 60s to everyone who had a mainframe or mini computer. In the 70s first applications were built and companies like BBN, Xerox, IBM started to use Lisp. In the 80s Common Lisp was standardized upon a DARPA initiative to have a single Lisp language for the applications they were prototyping or deploying in the military. The military got huge applications for battle management, battle simulation&training (like SIMNET), planning, design & construction developed by contractors.
> it was so inefficient that you even required a very special(and expensive) purpose computer to run it
Lisp developers required a personal workstation and thus they co-invented them in the 70s. Machines with 1 Megabyte or more RAM for a single user. Machines where one could develop and work with programs like Macsyma, an early computer algebra system widely used in maths and physics.
> creating compilation pathways and so on
efficient execution of Lisp on non-special-purpose hardware was available in the 70s for example with Maclisp and in the 80s with Common Lisp.
Characterized by the creation of new ideas or things
>> "It is like Windows ... but it made a hundred or a thousand times more people to experience computers."
Right, but that's completely irrelevant and not innovation.
Honestly, I think the word "innovation" needs to be tabooed from the discussion to get meaningful questions and answers.
> Popularity is not “innovative”.
But the way to reach popularity is innovative, at least for a lisp.
I don't pick Common Lisp/Scheme/Racket despite desperately wanting to because IMO Haskell has a better package ecosystem in my eyes, which is saying a lot. For example, try and find a HTTP server (by far the most common usecase for new projects IMO) that is both multi threaded and async IO capable (i.e. maximally using every core) -- it's surprisingly difficult, I found decent seeming libraries that did either-or, with varying levels of polish and paradigms involved.
I personally prefer CL to Clojure but I just can't imagine trying to convince an enterprise to use Common Lisp these days especially since I can't even really convince myself... Clojure would have the same problem but the ecosystem + buzz is the differentiator in my mind.
[0]: https://www.cs.cmu.edu/~rwh/theses/okasaki.pdf