00:00:00 --- log: started retro/13.01.19
08:16:37 <crc> tangentstorm: I cloned the noct repo; will try to take a look at it this afternoon
08:17:01 <crc> (I've been rather busy between work and dealing with sick kids at home all week)
08:30:09 * oPless is porting norvigs scheme to .net
08:30:35 <oPless> it now appears to work, unless you run r4rs test then it blows up
08:31:18 <oPless> not bad however for a few days work… there's a lot of annoying differences between the jvm and clr
08:46:50 --- join: ncv (~quassel@89.35.216.197) joined #retro
08:46:50 --- quit: ncv (Changing host)
08:46:50 --- join: ncv (~quassel@unaffiliated/neceve) joined #retro
08:55:06 --- quit: tangentsleep (Read error: Operation timed out)
09:03:00 --- join: tangentsleep (~michal@108-218-151-22.lightspeed.rcsntx.sbcglobal.net) joined #retro
09:26:43 --- nick: tangentsleep -> tangentstorm
09:28:10 * tangentstorm has been pretty sick all week himself. :/
09:51:36 --- quit: ncv (Read error: Operation timed out)
10:12:40 --- join: ncv (~quassel@89.35.216.197) joined #retro
10:12:40 --- quit: ncv (Changing host)
10:12:40 --- join: ncv (~quassel@unaffiliated/neceve) joined #retro
10:18:21 --- quit: ncv (Remote host closed the connection)
11:08:23 --- join: Necrs1mia84 (~Necrs1mia@78-47-111-140.cinfuserver.com) joined #retro
11:08:35 --- part: Necrs1mia84 left #retro
11:48:51 --- quit: karswell (Read error: Operation timed out)
13:12:36 --- join: Mat2 (~claude@91-65-144-133-dynip.superkabel.de) joined #retro
13:12:43 <Mat2> hello
13:12:56 <tangentstorm> heya!
13:25:52 <Mat2> gentstorm
13:26:00 <Mat2> hi tangentstorm
13:26:05 <tangentstorm> :)
13:26:26 <Mat2> what's up ?
13:26:37 <tangentstorm> working on an oberon07 compiler
13:26:47 <tangentstorm> just finished translaiting the PL/0 test suite
13:27:00 <tangentstorm> https://github.com/nickelsworth/noct
13:27:46 <tangentstorm> I need to add some tests for arrays, procedure variables, pointers....
13:28:07 <Mat2> oberon, nice. I have transformed ANSI C to it's Algol syntax roots
13:28:08 <tangentstorm> I think I may start by trying to target javascript or python.
13:28:32 <tangentstorm> I'm worried about garbage collection in retro. I've never tried to implement such a thing.
13:28:54 <tangentstorm> what do you mean? you wrote a compiler?
13:29:50 <Mat2> no, not really. GCC only needed some macros and some internal hints for supporting Algol syntax instead of the later C one
13:31:44 <tangentstorm> ah right :)
13:32:12 <Mat2> by the way, C is just Algol with some syntax short writings and symbolical obfuscation
13:33:54 <Mat2> int test (uInt8_t byte, void *p) {....; return ...} ->
13:34:55 <Mat2> function (integer) test (byte Byte, pointer p) begin .....; return ... end
13:35:17 <tangentstorm> :)
13:35:38 <tangentstorm> what about the whole "pass by name" thing in algol?
13:35:53 <Mat2> that's the same as in C
13:36:38 <tangentstorm> http://rosettacode.org/wiki/Man_or_boy_test#ALGOL_60_-_Knuth.27s_example
13:37:09 <Mat2> that's Algol 60 but C evolved over B from Algol W
13:37:21 <tangentstorm> http://rosettacode.org/wiki/Man_or_boy_test#C
13:37:26 <tangentstorm> oh i see
13:37:44 <tangentstorm> the call by name thing is all i really know about algol :)
13:39:16 <Mat2> Algol W was alike the system-programming variant of these language
13:39:34 <Mat2> Pascal evolved from Algol 60
13:40:06 <tangentstorm> yeah. I think the W is for Wirth :)
13:40:23 <Mat2> ah, no ?
13:41:56 <Mat2> but it was in fact the language-design of With
13:42:05 <Mat2> sorry Wirth
13:43:17 <tangentstorm> yeah. i was just guessing :)
13:44:27 <Mat2> https://en.wikipedia.org/wiki/ALGOL_W
13:44:53 <Mat2> please note the semicolon at end of the record defination
13:46:08 <Mat2> that's the reason why in C struct definations must be terminated by a semicolon (where otherwise the curly brackets are semantic symbols)
13:46:10 <tangentstorm> yeah. I think I'll stick with his later work :)
13:46:32 <Mat2> it was an experiment
13:46:52 <Mat2> the Algol syntax is much more readable in my opinion
13:47:13 <tangentstorm> than C? or than oberon/pascal?
13:47:27 <Mat2> at sure as C
13:48:45 <tangentstorm> I don't even hate curly braces all that much, but C hurts my eyes. I'm fine with D and Java and C# and even javascript.
13:49:12 <tangentstorm> it's the syntax for pointers I think
13:51:02 <Mat2> ther exist a problem fore the parser because the pointer prefix is also the multiply operator
13:51:17 <tangentstorm> yeah
13:52:05 <tangentstorm> it's also just ugly :) (at least to me)
13:52:43 <Mat2> you only need 10 preprocessor definations for the Algol W syntax
13:52:55 <tangentstorm> To implement oberon correctly in retro, I'll need to add a math coprocessor device.
13:53:20 <tangentstorm> :)
13:54:03 <Mat2> oh, AVM support integer, floating-point and vector types. I Will just grap your code and port it if finished :)
13:54:56 <tangentstorm> right now, the only part that works is oberon->oberon
13:55:29 <tangentstorm> but i already have all the code gen templates from PL/0, and retro supports about everything else already
13:55:53 <tangentstorm> well.. hmm.. inheritance too.
13:56:04 <tangentstorm> that one doesn't seem too hard
13:57:13 <Mat2> do you know a free oberon to c compiler ?
13:57:16 <tangentstorm> i want to use avm for speed but i also want this to run in a browser with javascript
13:57:19 <tangentstorm> yep
13:57:56 <tangentstorm> http://ooc.sourceforge.net/OOCref/OOCref_toc.html
13:58:08 <tangentstorm> it's oberon2 though
13:58:13 <tangentstorm> slightly different dialect
13:58:32 <tangentstorm> afk for a bit
13:58:35 --- nick: tangentstorm -> tangentaway
13:59:54 <Mat2> thanks, but it is not 64-bit clean
14:33:04 --- nick: tangentaway -> tangentstorm
14:33:14 <tangentstorm> oh
14:47:33 <Mat2> I'm just using vala in future
14:50:56 --- join: ncv (~quassel@89.35.216.197) joined #retro
14:50:56 --- quit: ncv (Changing host)
14:50:56 --- join: ncv (~quassel@unaffiliated/neceve) joined #retro
15:18:17 * oPless reads up
15:20:43 <oPless> avm?
15:44:47 <Mat2> my vm design
15:44:55 --- join: kumul (~Kumool@173.215.194.228) joined #retro
15:50:53 <oPless> ah
16:00:06 <oPless> ah ha.
16:00:35 <oPless> for some reason, the .net version of scheme I've ported over from java doesn't understand unquote
16:01:00 <oPless> alas unquote doesn't really appear anywhere in the source :)
16:02:34 <Mat2> probably limitations of the NET runtime are the reason for this
16:03:16 <oPless> nope, it a proper interpreter - I've probably missed a comparison somewhere
16:03:40 <Mat2> ok
16:03:42 <oPless> in java you can cast a string to an object, then use == to compare it to a string
16:03:46 <oPless> in .net you can't
16:04:31 <oPless> the equivalent is (someObject as string) == "some text"
16:04:55 <oPless> tis annoying.
16:07:20 <tangentstorm> Mat2: https://github.com/nickelsworth/OPCL another oberon compiler. it goes directly to native code though
16:08:14 <Mat2> ^thaknks, I will take a look
16:08:32 <tangentstorm> oPless: are you sure you can do that in java?
16:09:02 <tangentstorm> http://stackoverflow.com/questions/513832/how-do-i-compare-strings-in-java
16:10:38 <Mat2> for me, all of this (C# and Java) reads just like an unnecessary mess
16:11:42 <oPless> http://pastebin.com/bL4qviiE
16:11:52 <oPless> ^^ that's what I mean
16:12:35 <oPless> me too mat2
16:13:53 <oPless> and yeah, norvig's java code intern()'s strings
16:14:50 <Mat2> huh, I have now 4096 instruction pairs
16:17:43 <oPless> looks like a fairly minute thing to look for. 
16:17:58 * oPless goes to look for a translation tool
16:19:10 <Mat2> threading performance is now 4-6x of the Lua vm and bet gforth-fast by a factor of two
16:21:29 <Mat2> increasing the number of fused instructions seem to generate cache-misses to this is probably the optimum possible
16:22:56 <Mat2> without nativre-code generation
16:24:14 <tangentstorm> wow, Mat2. congrats :)
16:25:01 <Mat2> (that's on par with LuaJIT 1.2)
16:25:28 <Mat2> nice
16:30:28 <Mat2> get some sleep, ciao
16:30:33 --- quit: Mat2 (Quit: Verlassend)
18:05:17 --- quit: kumul (Quit: Leaving)
18:08:00 --- quit: ncv (Remote host closed the connection)
20:45:13 --- join: Fralinboy (~Fralinboy@ip70-188-28-6.rn.hr.cox.net) joined #retro
20:45:25 --- part: Fralinboy left #retro
21:49:04 --- nick: tangentstorm -> tangentaway
23:59:59 --- log: ended retro/13.01.19