Re: [ANN] Clojure 1.7.0-alpha3 now available

Encrypted fs have a file path limit around 140 bytes vs 255 w/o encryption. I hit this problem a couple of months ago with a library and AOT. The library had pretty long fn signatures, some generating class file names around 110 bytes long. The author tweaked these a bit to shorten the class file

Re: [ANN] Clojure 1.7.0-alpha3 now available

As Alex Miller mentioned earlier in the discussion (it can be confusing with multiple subjects being discussed), the change that caused this behavior was due to ticket http://dev.clojure.org/jira/browse/CLJ-1330 which has been reopened, and they plan to do something to improve the situation in the

Re: [ANN] Clojure 1.7.0-alpha3 now available

I hit this error when moving to a new box that had an encrypted FS. Might be related to your case as well. Good luck. On Wednesday, October 29, 2014 1:34:28 AM UTC+11, Stefan Kamphausen wrote: > > Hi, > > have there been any changes how fns with a name and recursion are > compiled? One of my pr

Re: [ANN] Clojure 1.7.0-alpha3 now available

Picking up the variadic arguments discussion, it seems that in a simple definition like (fn [& args] (apply f arg1 args)) One could conceivably put some sort of preprocessing smarts into the fn macro that notices that a) this is a variadic arglist b) the variadic arg (args) is a symbol, not so

Re: [ANN] Clojure 1.7.0-alpha3 now available

It's a big change and will require a fair amount of screening and analysis, none of which has been done yet. On Wednesday, October 29, 2014 7:25:20 AM UTC-5, Max Penet wrote: > > Just curious, why http://dev.clojure.org/jira/browse/CLJ-1517 didn't make > it. It seemed complete at first glance. >

Re: [ANN] Clojure 1.7.0-alpha3 now available

Just curious, why http://dev.clojure.org/jira/browse/CLJ-1517 didn't make it. It seemed complete at first glance. On Tuesday, October 28, 2014 9:53:05 PM UTC+1, Alex Miller wrote: > > We will definitely be rolling this particular aspect of CLJ-1330 back in > the next release. > > Alex > > On Tu

Re: [ANN] Clojure 1.7.0-alpha3 now available

We will definitely be rolling this particular aspect of CLJ-1330 back in the next release. Alex On Tuesday, October 28, 2014 10:44:51 AM UTC-5, Sam Ritchie wrote: > > Yeah, I'm getting this too with a file named "wparankings.clj". I haven't > investigated which function is causing this, but thi

Re: [ANN] Clojure 1.7.0-alpha3 now available

Reid McKenzie writes: > For the special case of a "reduction" over varargs as in #'clojure.core/+ > or #'clojure.core// where you're just generating Duff's I think the macro > should be pretty straightforward but it won't be general at all. I tend to agree with the lack of generality -- from a l

Re: [ANN] Clojure 1.7.0-alpha3 now available

Phillip, Alan was talking about this on IRC yesterday with me an Alex... I think in general you can't "just" macro this, because hand unrolling is generating the nth duff's device by partially applying the "base" case over N elements from the "head" of varargs. This is pretty straightforward for c

Re: [ANN] Clojure 1.7.0-alpha3 now available

Ah, if only some one had implemented one. I guess there is no reason NOT to implement this? It's still only a single class, and single object per function? The class is going to be longer, because method bodies like... public Object invoke(Object arg1, Object arg2) { return throwArity(2

Re: [ANN] Clojure 1.7.0-alpha3 now available

Yeah, I'm getting this too with a file named "wparankings.clj". I haven't investigated which function is causing this, but this is definitely troubling. Stefan Kamphausen October 28, 2014 at 8:34 AM Hi, have there been any changes how fns with a name and recursion a

Re: [ANN] Clojure 1.7.0-alpha3 now available

On Sunday, October 26, 2014 11:43:23 PM UTC-4, Alex Miller wrote: > > ### 1.4 update - like update-in for first level > > `update` is a new function that is like update-in specifically for > first-level keys: > > (update m k f args...) > > Example use: > > user> (update {:a 1} :a inc) > {:

Re: [ANN] Clojure 1.7.0-alpha3 now available

Hi, have there been any changes how fns with a name and recursion are compiled? One of my projects has a function which does not compile with 1.7.0-alpha3 anymore, but did fine with 1.6.0. I tried to create a minimal example at https://github.com/ska2342/nested-fn-breaks-clojure-17 (I know t

Re: [ANN] Clojure 1.7.0-alpha3 now available

http://dev.clojure.org/jira/browse/CLJ-731 ? On Tue, Oct 28, 2014 at 2:39 AM, Phillip Lord wrote: > Reid McKenzie writes: > > This suggests that |apply| is immensely expensive in general, > > and that such arity unrolling even for trivial functions would be a good > > thing. Albeit hard to buil

Re: [ANN] Clojure 1.7.0-alpha3 now available

Reid McKenzie writes: > This suggests that |apply| is immensely expensive in general, > and that such arity unrolling even for trivial functions would be a good > thing. Albeit hard to build. Wonder whether it is macroable. Something like (def new-function (with-arities [20] [args]

Re: [ANN] Clojure 1.7.0-alpha3 now available

Oops. Linked an old benchmark version. https://www.refheap.com/92391 Reid On 10/27/2014 03:13 PM, Reid McKenzie wrote: For grins I slapped together this benchmark , which tries to compare calls with callsite static arity to application both for fully unrolled a

Re: [ANN] Clojure 1.7.0-alpha3 now available

For grins I slapped together this benchmark , which tries to compare calls with callsite static arity to application both for fully unrolled and partially unrolled functions. The results I see indicate that in both cases the cost of |apply| is almost equal for all

Re: [ANN] Clojure 1.7.0-alpha3 now available

Both apply and variadic functions have some overhead. This helps address both. On Mon, Oct 27, 2014 at 11:50 AM, Alex Miller wrote: > On Monday, October 27, 2014 9:36:02 AM UTC-5, Phillip Lord wrote: >> >> >> And the performance boost is just to avoid the list creation in the >> variadic argumen

Re: [ANN] Clojure 1.7.0-alpha3 now available

On Monday, October 27, 2014 9:36:02 AM UTC-5, Phillip Lord wrote: > > > And the performance boost is just to avoid the list creation in the > variadic argument? > Yes. There's some perf timings in the ticket. -- You received this message because you are subscribed to the Google Groups "Clojure

Re: [ANN] Clojure 1.7.0-alpha3 now available

> an arity of 20, but >> haven't yet, just because it seems a step too far. >> >> Phil >> >> >> >> From: clo...@googlegroups.com [clo...@googlegroups.com >> ] on behalf of Alex Miller [al...@puredanger.com >> ] >> Sent: 27 October 2014 03:

Re: [ANN] Clojure 1.7.0-alpha3 now available

lf of Alex Miller [al...@puredanger.com > ] > Sent: 27 October 2014 03:42 > To: clo...@googlegroups.com ; cloju...@googlegroups.com > > Subject: [ANN] Clojure 1.7.0-alpha3 now available > > Clojure 1.7.0-alpha3 is now available. > > Try it via > - Download:

RE: [ANN] Clojure 1.7.0-alpha3 now available

] Sent: 27 October 2014 03:42 To: clojure@googlegroups.com; clojure-...@googlegroups.com Subject: [ANN] Clojure 1.7.0-alpha3 now available Clojure 1.7.0-alpha3 is now available. Try it via - Download: http://central.maven.org/maven2/org/clojure/clojure/1.7.0-alpha3/ - Download securely: htt

[ANN] Clojure 1.7.0-alpha3 now available

Clojure 1.7.0-alpha3 is now available. Try it via - Download: http://central.maven.org/maven2/org/clojure/clojure/1.7.0-alpha3/ - Download securely: https://repo1.maven.org/maven2/org/clojure/clojure/1.7.0-alpha3/ - Leiningen: [org.clojure/clojure "1.7.0-alpha3"] For users of Clojure 1.7.0-alpha2