2016-02-10 13:06 GMT+01:00 Gary Verhaegen <gary.verhae...@gmail.com>:
> Please excuse the very naive question, but if the main problem seems to be > compilation of loaded namespaces, what about aot? Does it help at all? If > not, why? As far as I understand the problem, it does help a little bit, but much less than expected. The main slowdown is not in generating the byte code, but in (statically) constructing all the bits and pieces, that constitute a namespace: var objects, fn objects, metadata. This is done on every start, because JVM lacks any kind of data literals (AFAIK, even arrays are constructed piecemeal on the JVM bytecode stack) > As far as I know, the JVM already does lazy loading with Java classes > essentially in the way that Mike described. It does, but when require'ing a namespace, every class it comprises is loaded, because the vars + their contents must be reconstructed. This is where Mike's proposal comes in. BTW Mike: I think your proposal might just work, if we do that kind of lazy loading only for vars containing plain fn forms. It would have to be done very carefully indeed, since even a metadata expression can have side-effects. Makes me wonder if that kind of elaborate effect-tracking, couldn't be used more generally, to do closed-world tree-shaking on clojure and clojurescript in addition to lazyfying side-effect free code-paths, killing multiple birds with one stone. Clojurescript already has GClosure, I hear you say? Right, then why does the lower bound for meaningful CLJS programs seem to be at around 600K (minified, uncompressed)? -- 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.
