Thank you for your detailed explanation; I've tested on Raspberry Pi(slooow) though it's not that fast, yes, I found that java -jar clojure.jar thing is very fast compared to boot repl. Now I understand why it takes so long.
On Mon, Jun 20, 2016 at 12:43 AM Alex Miller <a...@puredanger.com> wrote: > This is a nice trick and its one I have done myself on Clojure CLIs > (particularly with "help" options) to defer loading and give a faster feel > to startup. It is great to have a repl that starts super fast for people > learning Clojure to get started, or to spin one up to try something. > > However, the time to start up a base Clojure repl (with no deps or app > code) is already subsecond - so you're saving no more than a second with > this approach. The problem is that (approximately) no one actually starts a > REPL this way. [You can find my ongoing notes on this subject at > http://dev.clojure.org/display/design/Improving+Clojure+Start+Time by the > way.] > > Most people start a REPL via Leiningen (or Boot), and the demo'ed > technique changes nothing about that experience. Leiningen still needs to > start the Clojure runtime, compile and load nrepl and all the tooling, > launch an nrepl server in a new JVM, connect to it, load all the middleware > code, and commonly even a vast amount of application code (based on common > patterns of project setup). If you look at the Luminus guestbook > application for example (not picking on this one, just one I happened to > look at), at dev time it loads env/dev/clj/user.clj prior to repl startup > which loads a vast number of dependency namespaces and application code, > which takes 10s of seconds, which dwarfs Clojure core load time. > > Because this a Clojure repl hack, it also does nothing to affect actually > running a Clojure application, which again must load the Clojure runtime > and your application code (which often involves compiling and loading a lot > of code) before it starts doing anything. > > So while this is a nice demo, I don't see that it actually changes > anything for the vast majority of actual "Clojure start time issues" people > experience on a daily basis. That said, this is an acknowledged pain point > that we still care about very much! I'm sure there will be more work on > this to come. > > On Saturday, June 18, 2016 at 9:25:37 PM UTC-5, Sungjin Chun wrote: >> >> Yes, I want this kind of instant speed (I hope this kind of speed is >> possible in >> emacs + side). >> >> https://www.youtube.com/watch?v=D3gpQth8uTI&feature=youtu.be >> >> Have you seen this? Why we simply cannot do this? >> > -- > You received this message because you are subscribed to the Google > Groups "Clojure" group. > To post to this group, send email to clojure@googlegroups.com > Note that posts from new members are moderated - please be patient with > your first post. > To unsubscribe from this group, send email to > clojure+unsubscr...@googlegroups.com > For more options, visit this group at > http://groups.google.com/group/clojure?hl=en > --- > You received this message because you are subscribed to a topic in the > Google Groups "Clojure" group. > To unsubscribe from this topic, visit > https://groups.google.com/d/topic/clojure/ejxIxPPYd84/unsubscribe. > To unsubscribe from this group and all its topics, send an email to > clojure+unsubscr...@googlegroups.com. > For more options, visit https://groups.google.com/d/optout. > -- You received this message because you are subscribed to the Google Groups "Clojure" group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en --- You received this message because you are subscribed to the Google Groups "Clojure" group. To unsubscribe from this group and stop receiving emails from it, send an email to clojure+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.
