Re: [isabelle-dev] scala-2.11.0

2014-05-05 Thread Gottfried Barrow
On 14-05-05 15:37, Makarius wrote: That is why I pointed to $ISABELLE_HOME/src/Tools/jEdit/src/scala_console.scala before -- it shows how to turn the Scala compiler into a plain method String => String. This is the meaning of the word "interpreter". I think that occurred to me after the fact

Re: [isabelle-dev] scala-2.11.0

2014-05-05 Thread Makarius
On Mon, 5 May 2014, Gottfried Barrow wrote: I guess it's basically the same, in that the JVM is started up each time, which takes about 2 seconds, whether for Scala or JRuby. A short script that uses a C-based language, like Perl, is just the cost of the bash call, which is about 200ms. On a

Re: [isabelle-dev] scala-2.11.0

2014-05-05 Thread Gottfried Barrow
On 14-05-05 10:48, Makarius wrote: Here I guess you mean something like the isabelle_scala_script wrapper, which is indeed very slow on startup for two reasons: No, I meant something like ML{*Isabelle_System.bash "jruby my_script.rb"*}, where I was using scala instead of jruby. I guess it's

Re: [isabelle-dev] scala-2.11.0

2014-05-05 Thread Makarius
On Mon, 5 May 2014, Gottfried Barrow wrote: Cutting this short(er), part of my reject of Scala was because external calls to the JVM are slow, which is important for "REPL use", but I figured out it's not of ultimate importance, which led to me installing and using JRuby in my own distribution

Re: [isabelle-dev] scala-2.11.0

2014-05-05 Thread Gottfried Barrow
On 14-05-05 04:39, Makarius wrote: Here is a funny presentation on the Future of Programming from "1973" http://worrydream.com/#!/TheFutureOfProgramming Makarius, What's funny is that during the first minute of the video, I thought someone had inserted a laugh-track for what we would think is

Re: [isabelle-dev] scala-2.11.0

2014-05-05 Thread Makarius
Here is a funny presentation on the Future of Programming from "1973" http://worrydream.com/#!/TheFutureOfProgramming I could have put that on the thread "Remaining uses of Proof General", because it touches recurrent problems of getting stuck with accidental situations over decades, and the im

Re: [isabelle-dev] scala-2.11.0

2014-04-28 Thread Makarius
On Fri, 25 Apr 2014, Makarius wrote: trimming it down to the simplest possible approach. I have done this now in Isabelle/e1317a26f8c0 (and the row of changes before it). So most of the PIDE "operating system" functionality is now re-implemented by more basic means, in just 48 hours. It

Re: [isabelle-dev] scala-2.11.0

2014-04-25 Thread Makarius
On Wed, 23 Apr 2014, Makarius wrote: I am presently considering to discontinue actors altogether, and merely use the thread task pool, explicit threads, and typed channels between them. But I also need to see how Akka actors do it today, to get a few more ideas, before trimming it down to the

[isabelle-dev] scala-2.11.0

2014-04-23 Thread Makarius
The Scala guys have released scala-2.11.0, and we are using it since Isabelle/5d629da46f04. This requires the usual "isabelle components -a". Moreover, there is some chance that sticky JVM jam from old jars is getting in the way. In the case of odd Java/Scala startup failures (exceptions abo