IRC #oberon channel logs

[valexey_u]

Как все начиналось см. тут: http://oberspace.dyndns.org/index.php/topic,436.0.html

А пока логи никто не выкладывает, выложу я их здесь:

-Zorko- VERSION mIRC v6.02 Khaled Mardam-Bey
<tangentstorm> the ngaro instruction set is almost the same as the one from wirth's machine... except his uses registers, whereas ngaro uses the dual stack architecture
>dsar< CTCP VERSION
-dsar- VERSION HexChat 2.9.4 [x64] / Windows 7 [3.00GHz]
>harrison< CTCP VERSION
-harrison- VERSION Quassel IRC v0.6.3 (dist-de4dd62) (built on Oct 19 2011 21:33:07) -- http://www.quassel-irc.org
>impomatic< CTCP VERSION
-impomatic- VERSION Opera/9.80 (Windows NT 6.1; WOW64) Presto/2.12.388 Version/12.12
>lboasso< CTCP VERSION
-lboasso- VERSION ChatZilla 0.9.89 [Firefox 18.0/20130104151925]
>norayr< CTCP VERSION
* [norayr] is away (https://spyurk.am/u/norayr)
-norayr- VERSION Purple IRC
<dsar> :O
>StephenS< CTCP VERSION
-StephenS- VERSION XChat-WDK 1500 [x86] / Windows 7 [2,71GHz]
>tangentstorm< CTCP VERSION
-tangentstorm- VERSION WeeChat 0.3.2 (Jun 18 2010)
>valexey< CTCP VERSION
* Received a CTCP VERSION from valexey
-valexey- VERSION xchat 2.8.8 Linux 2.6.32-5-686 [i686/800.50MHz/SMP]
* norayr has quit (Quit: Leaving.)
<tangentstorm> hey there valexey
<tangentstorm> if i'm not mistaken, someone was just showing us a compiler you'd written...
<tangentstorm> https://github.com/valexey/Oberon-07-11-compiler
<valexey> It'is not my compiler. please read "read me"
<valexey> i just push it to github + some minor bugfix
<tangentstorm> Aha. Well, thanks for doing so.
<valexey> :-)
<tangentstorm> In any case, it's inspired me to get my own oberon compiler work up onto github. Working on that right now.
<valexey> Very well
* valexey waiting for link/url :-)
<tangentstorm> it'll be an hour or so... i'm writing a readme and i have to extract it from another codebase
<valexey> ok
<dsar> I'm also coding an oberon compiler (but only a frontend, I still have to learn how to write a code generator)
<dsar> but I'm going to split the parser in several modules, for every ast generator
<valexey> there is another one compiler under development. compiler to js-code. New market for Oberon :-)
<dsar> we could join in an unique project
<dsar> here there are at least 4 people making a compiler, ehehe
* Vapaamies (~Freeman@188.242.131.141) has joined #oberon
<valexey> :-)
<valexey> Oberon-07?
<dsar> yup
<valexey> what charset for string literals?
* Vapaamies (~Freeman@188.242.131.141) has left #oberon ("Black Mesa Research Center")
<dsar> Latin-1
<valexey>
<dsar> why?
<valexey> because in modern world de facto standard is unicode (utf8 for example)
<dsar> utf8 is a pain to support (the standard reference is about 600 pages :\)
<dsar> in my opinion it is better to leave this on the library side
<valexey> I think that for compiler implementation we need only information from wikipedia about utf8/unicode :-)
<dsar> in multilanguage programs, strings are read from a specific file
<valexey> ok. what size of char?
<dsar> 1 byte
<valexey> from Oberon-07/11 report size of char is unspecified.
<valexey> so, we get incompatibility between different Oberon-07 compilers.
<dsar> usually CHAR is 1 byte, UNICHAR 2 byte
<valexey> there is no unichar in Oberon-07 report :-)
<dsar> indeed
<valexey> and even 2 byte - it is only UCS-2. Full unicode is UTF-32 or UCS-4 (4 byte)
<dsar> by the way I'm an oberon purist I would had every extension (like exceptions through coroutines) via SYSTEM module
<dsar> UTF-8 via normal library
<dsar> I would add*
<valexey> sixe of char is unspecified, size of integer unspecified, so how to implement cross-compiler library for any size of char? ;-)
<dsar> there are a lot of unspecified and unclear parts in the report
<valexey> yep
<dsar> indeed there was a deep discussion about open array semantic in ocp.forum
<valexey> url?
<dsar> if one is going to write an oberon compiler, should expand the report with more rules
* tcpip (ada418c1@gateway/web/freenode/ip.173.164.24.193) has joined #oberon
<tcpip> hi
<valexey> hi
<tcpip> not so many ppl is here
<tcpip> r u from russia?
<dsar> three days ago there were only two people (me and tangentstorm), ehehehe
<tcpip> hi dsar
<dsar> hola tcpip :-)
<valexey> tcpip, i'm from russia
>tcpip< CTCP VERSION
-tcpip- VERSION qwebirc v0.91, copyright (C) 2008-2011 Chris Porter and the qwebirc project -- Mozilla/5.0 (Macintosh; Intel Mac OS X 10_8_2) AppleWebKit/537.17 (KHTML, like Gecko) Chrome/24.0.1312.52 Safari/537.17
<tcpip> brb
<valexey> also there is russian jabber oberon conference:  oberon@conference.jabber.ru
<tangentstorm> why is oberon so popular in russia? is there some person/company that popularized it? or did it just kind of happen?
<dsar> I heard that modula-2 were used a lot in russia
<dsar> was
<dsar> for this reason, maybe, there are a lot of interested people in oberon
<valexey> some satellite firmware written in modula-2 (today)
<valexey> also there is XDS-compiler
<valexey> also BlackBox useful in some science/engineeric application
<valexey> messages per year on russian oberon forums: http://oberspace.dyndns.org/stats/stats_total.png
<valexey> http://oberspace.dyndns.org/stats/oberon_total.png
<dsar> :O
<valexey> more than 12000 messages last year :-)
<valexey> (not all messages was oberon-specific, but all written by russian oberon community )
<tcpip> looks like in Russia something start recovering finally
<valexey> yep :-)
<tcpip> glad to see it
<dsar> so russians are oberon lovers
<tcpip> if I remember M2 was the implementation language in some Russian space programs, e.g. Buran. Is it correct?
<valexey> I don't know about buran software, but yes, today some russian space satellite has firmware written on M2
<tcpip> also it was one of the language for Elbrus, monster parallel supercomputer system. Am I right?
<valexey> what Elbrus model? http://en.wikipedia.org/wiki/Elbrus_(computer)
<dsar> if you are sure about that, modify the page on the wikipedia :-)
<valexey> Yes, as I see, for Elbrus-1 and Elbrus-2 thare was Эль-76 language, Modula-2, Ada, Refal, Forth, Snobol and others.
<valexey> proof link (russian language): http://sevntu.com.ua/cgi-bin/irbis64r_72/cgiirbis_64.exe?Z21ID=&I21DBN=BOOK&P21DBN=BOOK&S21STN=1&S21REF=5&S21FMT=fullwebr&C21COM=S&S21CNR=20&S21P01=0&S21P02=0&S21LOG=1&S21P03=K=&S21STR=%D1%81%D0%B8%D1%81%D1%82%D0%B5%D0%BC%D0%B0%20%D0%9C%D0%BE%D0%B4%D1%83%D0%BB%D0%B0-2-%D0%AD%D0%BB%D1%8C%D0%B1%D1%80%D1%83%D1%81
<dsar> well, I trust you :-) ehehe
<tcpip> llong time ago in the collage where I was working Russian was the "must know" language for mathematicians
<tcpip> I was able to read papers and even new few spoken "special" words
<valexey> where it was?
<tcpip> in UNH
<valexey> University of New Hampshire?
<tcpip> yeap
<tcpip> then, later on, we had few professors from Russia and Ukraine
<tcpip> one maybe still working
<tcpip> he is not a programmer, just math
<valexey> I understand..
<tcpip> now nobody reads russian science magazines, if they still in place
<tcpip> all gone...
<valexey> Most publications in international magazines.
<tcpip> are you from Moscow?
<valexey> No, I'm from Nizhny Novgorod (it was known as Gorky).
<tcpip> no.
<tcpip> you had a great country in the past
<tcpip> I had a present from those folks with letters CCCP on it
<tcpip> If not your Sputnik I never had a doctor degree.
<valexey> :-)
<valexey> Everyone has ups and downs. Our country too.
<tcpip> The guy who really invent internet, not Al Gore, did it because your country
<tcpip> few years ago was his article in IEEE about it
<valexey> because nuclear weapon?
<valexey> My hometown is Arzamas-16 ( http://en.wikipedia.org/wiki/Sarov ) :-)
<tcpip> no. he said he decide to choose engineering cuz USSSR launch a sputnik
<tcpip> ... let me search for that article on Google
<valexey> ok
<tangentstorm> https://github.com/nickelsworth/noct README could probably use some work and I need to rebuild and upload the makefile, but it's a start.
<tcpip> http://www.computer.org/portal/web/internet/extras/Len-Kleinrock
<valexey> tangentstorm, wow! :-)
<valexey> tangentstorm, i must write about it to our forum :-)
<tcpip> I have a digital subscription for IEEE with full article
<valexey> tcpip, thanks
<tcpip> this link just pops up first
<tangentstorm>
<Zorko> valexey, from where you know about this channel, a little spion?
<Zorko> welcome
<valexey> from here : https://lists.inf.ethz.ch/pipermail/oberon/2013/006689.html
* impomatic (~digital_w@211.67.125.91.dyn.plus.net) has left #oberon
<Zorko> aha
<Zorko> mm.. and why akron1 not placed his compiler to github and ask you about it?
<valexey> because he don't know hot to do it
<valexey> *how
<Zorko> aaah. ok, I understand
<Zorko> interesting work
<Zorko> I would like to give him idea to add LLVM or C code generator
* tcpip has quit (Quit: Page closed)
<valexey> tangentstorm, wow, your using antlr! That's great :-) I like it :-)
<Zorko> and maybe native x86-32 may be easy modified to x86-64 code generator...
<tangentstorm> thanks yeah, antlr is a really nice tool
<Zorko> valexey, I'm Oleg N. Cher
<valexey> Zorko, I know
<Zorko> very satisfied with your forum
<Zorko> but it needs a many of time
<tangentstorm> i figured java made the most sense as a target since it already has inheritence, garbage collection, and modules (static classes)
<valexey> tangentstorm, are you use antlr workbench?
<Zorko> no time to Oberon coding
<valexey> Zorko, scheduling problem :-)
<Zorko> nono
<Zorko> it's a Dizzer problem - he do nothing
<valexey> Zorko, just ignore him
<tangentstorm> c and pascal are trickier because you have to implement garbage collection yourself, and of course pascal has the case sensitivity issue. there's a language called "D" though that might make sense as a target
<Zorko> it's not always easy - to re-calibrate mind between the tasks
<valexey> tangentstorm, and Go language
<Zorko> valexey, I can, of course
<tangentstorm> valexey: good point.
<Zorko> but why
<valexey> tangentstorm, and javascript :-)
<tangentstorm> go is actually a great choice. there's a version of the virtual machine written in go already, and it also runs natively on google app engine (which i do some work with)
<Zorko> dsar, have an idea - publish the channel logs on web
<tangentstorm> valexey: https://github.com/tangentstorm/ngaro-js
<Zorko> to be indexed by Google, to Oberoners find it by keywords
<Zorko> well guys, I go to sleep bye
<valexey> Zorko, try it
<valexey> and bye ;-)
<tangentstorm> seeya Zorko
<Zorko> try with Dizzer? and why he is there, he is not Oberoner
<Zorko> yes, thanks all
<valexey> try to publish this log.
<valexey> I think we need irc-bot for it
<Zorko> eh. I'm not a bot, my connect may be dropped
<Zorko> aha
<Zorko> well..
<tangentstorm> I have logs, with a few gaps... i've got some friends who do bot stuff
<tangentstorm> we'll get one hooked up
* Zorko is now known as Zorko_sleep
<valexey> good
* lboasso has quit (Quit: ChatZilla 0.9.89 [Firefox 18.0/20130104151925])
<valexey> tangentstorm, now there are several Oberon-07/11 compilers, I think that we need common test-set (set of oberon-sources) for all compilers.
<tangentstorm> makes sense to me
<tangentstorm> there's already one for oberon2 on sourceforge somewhere. maybe we could adapt that?
* tangentstorm is now known as tangentaway
<valexey> oo2c tests? I don't know. May be it is easy to make new from scratch. Oberon-07 have many small (and not small) diffs.

[valexey_u]

Маленькая, но важная добавочка:

<tangentstorm> valexey: the test suite i was thinking of is http://sourceforge.net/projects/hostess/
<valexey> ok. I should read it.

[valexey_u]

Логи. Продолжение:

<valexey> tangentstorm, your compiler really needs documentation how to build it (step by step) :-)
<tangentstorm> valexey: it needs the makefile!
<valexey> or ant-file :-)
<tangentstorm> i've never used ant
<tangentstorm> i'll upload the makefile as-is in a minute or two. the paths are all broken, but you'll at least be able to get the idea
<valexey> thanks
* tangentstorm has quit (Ping timeout: 256 seconds)
* tangentstorm (~michal@108-218-151-22.lightspeed.rcsntx.sbcglobal.net) has joined #oberon
<tangentstorm> valexey: okay. done. it's a complete mess and probably broken but it's up
* akastargazer (~stargazer@93-81-146-44.broadband.corbina.ru) has joined #oberon
<tangentstorm> ... and uploaded some test modules
* akastargazer (~stargazer@93-81-146-44.broadband.corbina.ru) has left #oberon
* akastargazer (~stargazer@93-81-146-44.broadband.corbina.ru) has joined #oberon
<akastargazer> hi all
<tangentstorm> hello
<tangentstorm> valexey: okay, just pushed a new version of the makefile. it runs a gUnit test suite for the Oberon07 parser... and passes all the tests, at least on my machine
<valexey> tangentstorm, ok. I'll test it
>akastargazer< CTCP VERSION
-akastargazer- VERSION Purple IRC
<valexey> tangentstorm, there is no imports in Hello module. Why?
<StephenS> I am Wirth
>StephenS< CTCP VERSION
-StephenS- VERSION XChat-WDK 1500 [x86] / Windows 7 [2,71GHz]
* Received a CTCP VERSION from StephenS
* Received a CTCP PING from StephenS
* Received a CTCP PONG from StephenS
* Received a CTCP TIME from StephenS
>TIME< CTCP STEPHENS
<akastargazer> Wirth?
<StephenS> Niklaus
<StephenS> Yes that is correct.
<valexey> from London?
* Ping reply from StephenS: 0.46 second(s)
<tangentstorm> valexey: looks like I was using hello.mod to test the pascal backend. Writeln is a predefined global in pascal.
<valexey> tangentstorm, ok. What rules for foreign IMPORT's? java, for example.
<tangentstorm> valexey: I'm not sure. I haven't looked at this code since April. I'm just dusting it off now.
<valexey> ok :-)
<tangentstorm> I suppose my interest was more in creating libraries that could be used in a lot of different places, not so much using other libraries from oberon.
<tangentstorm> so I probably don't have a good answer :/
<tangentstorm> It looks like used "!" as the type casting operator instead of Type(whatever) ... I consider that a bug now.
<valexey> but some basic IO, for example, is necessary.
<tangentstorm> The way the virtual machine i mentioned works is that there are ports that you can read and write to.
<tangentstorm> Then again, I don't think I'd discovered ngaro when I wrote this code...
<tangentstorm> I think at the time, my plan was to compile and run the actual oberon system from the project oberon book.
<valexey> hm.. ok :-)
<valexey> -----------------------------------------------------------------------
<valexey> executing testsuite for grammar:Oberon07 with 116 tests
<valexey> -----------------------------------------------------------------------
<valexey> 0 failures found:
<valexey> Tests run: 116, Failures: 0
<valexey> succ :-)
<tangentstorm> pretty sure i had support for DO..EXIT..LOOP at one point, so i could support the original oberon code. I'm not sure where it went.
<tangentstorm> yay!!!
<valexey> :-)
* akastargazer (~stargazer@93-81-146-44.broadband.corbina.ru) has left #oberon
<tangentstorm> Okay, added some notes to the makefile about running the pascal backend. Looks like there's still plenty of work to do.
<valexey> and I did't found how to generate java-code from O7 module
<tangentstorm> haven't tested that yet
<tangentstorm> actually, it should be the same as the rule for pascal ... but just change Pascal.stg to Java.stg
<tangentstorm> trying it now
<tangentstorm> hmm. yep. that's how it's meant to work but it doesn't seem to be working
* valexey going to sleep
<tangentstorm> me too, soon. have a good night, and thanks for trying this
* tangentstorm is now known as tangentsleep
* akastargazer (~stargazer@93-81-146-44.broadband.corbina.ru) has joined #oberon
* akastargazer (~stargazer@93-81-146-44.broadband.corbina.ru) has left #oberon
* akastargazer (~stargazer@93-81-146-44.broadband.corbina.ru) has joined #oberon
* akastargazer has quit (Ping timeout: 245 seconds)
<dsar> 'morning
* ChanServ has changed the topic to: Discussions related to the Oberon Operating System(s), Programming Language(s) and Philosophy.   ·   This channel is affiliated with #modula-2 for every discussion about the Oberon ancestor.
* akastargazer (~stargazer@93-81-146-44.broadband.corbina.ru) has joined #oberon
* akastargazer (~stargazer@93-81-146-44.broadband.corbina.ru) has left #oberon
* norayr (~noch@arnet.am) has joined #oberon
* norayr has quit (Client Quit)
* geniepro (~genieprox@84.54.123.54) has joined #oberon
* geniepro (~genieprox@84.54.123.54) has left #oberon
* geniepro (~genieprox@84.54.123.54) has joined #oberon
<geniepro> hi folks
<geniepro> version
<dsar> hola
* akastargazer (~stargazer@93-81-146-44.broadband.corbina.ru) has joined #oberon
* akastargazer (~stargazer@93-81-146-44.broadband.corbina.ru) has left #oberon
* Zorko_sleep has quit (Ping timeout: 246 seconds)
* geniepro (~genieprox@84.54.123.54) has left #oberon
* geniepro (~genieprox@84.54.123.54) has joined #oberon
<geniepro>  is there anybody out there?
* geniepro (~genieprox@84.54.123.54) has left #oberon
* Ericounet_portab (~ericp@mtr26-3-88-179-205-51.fbx.proxad.net) has joined #oberon
* Ericounet_portab (~ericp@mtr26-3-88-179-205-51.fbx.proxad.net) has left #oberon
* Ericounet_portab (~ericp@mtr26-3-88-179-205-51.fbx.proxad.net) has joined #oberon
* Zorko_sleep (~zorko@46.211.121.115) has joined #oberon
* Zorko_sleep is now known as Zorko
<Zorko> ah! I was sleep and I've a dream - Wirth is here
* norayr (~noch@arnet.am) has joined #oberon
<dsar> under an anonymous nickname? eheheh
* Ericounet_portab has quit (Quit: Ericounet_portab)
<Zorko> who can believe in his presence even under nick 'Wirth'
<Zorko> the invisible spirit of Wirth is here, guys
* Zorko is now known as Zorko_
<harrison> "veert by reference and worth by value"
<valexey> Project Oberon V4 on github (online sources): https://github.com/ilovb/ProjectOberonV4
* tangentsleep has quit (Read error: Operation timed out)
* Zorko_ has quit (Read error: Connection reset by peer)
* tangentsleep (~michal@108-218-151-22.lightspeed.rcsntx.sbcglobal.net) has joined #oberon
* akastargazer (~stargazer@93-81-146-44.broadband.corbina.ru) has joined #oberon
* Ericounet_portab (~ericp@mtr26-3-88-179-205-51.fbx.proxad.net) has joined #oberon
* akastargazer has quit (Client Quit)
* Ericounet_portab (~ericp@mtr26-3-88-179-205-51.fbx.proxad.net) has left #oberon
* Zorko (~zorko@46.211.123.6) has joined #oberon
* tangentsleep is now known as tangentstorm
<tangentstorm> valexey: nice
<tangentstorm> I have a small test suite from my PL/0 compiler : https://github.com/nickelsworth/PL0-Language-Tools/tree/master/tests
<tangentstorm> the runner is here: https://github.com/nickelsworth/PL0-Language-Tools/blob/master/test
<tangentstorm> it's a bash script that compiles and runs each program, then uses git diff to compare the actual output to the expected results
<tangentstorm> I think my goal for today will be to port these to oberon 07 and add in some extra stuff for arrays
<tangentstorm> and records and strings and everything else that oberon has but PL/0 doesn't
* Ericounet_portab (~ericp@mtr26-3-88-179-205-51.fbx.proxad.net) has joined #oberon
* Ericounet_portab (~ericp@mtr26-3-88-179-205-51.fbx.proxad.net) has left #oberon
<dsar> OK
<dsar> Wirth replied to my email about open array semantic
* norayr has quit (Quit: Leaving.)
* Necrs1mia84 (~Necrs1mia@78-47-111-140.cinfuserver.com) has joined #oberon
<dsar> http://www.ocp.inf.ethz.ch/forum/index.php/topic,744.msg2694.html#msg2694
* Necrs1mia84 (~Necrs1mia@78-47-111-140.cinfuserver.com) has left #oberon
<dsar> tangentstorm
<tangentstorm> sorry... was writing test cases
<tangentstorm> hm
<tangentstorm> okay
<tangentstorm> dsar: i don't understand the problem well enough. why is it difficult for open arrays to be assignable?
<dsar> they are easily susceptible to runtime errors
<dsar> it is better to protect them
<tangentstorm> maybe i've been spoiled by free pascal
<tangentstorm> https://github.com/nickelsworth/noct/commit/cf140fc39dc375ef414c07b507b0be5d61895513  <- first crack at an oberon-07 test suite
<tangentstorm> it's missing a test for CASE and one for the built-in functions
<dsar> a := b; with a as open array, the assignment to a in the procedure is limited by the length of b, it is impossible to check at compile time (because the length of a is not fixed) and it could cause runtime error that are not easy to find
<tangentstorm> also doesn't do anything for REAL, SET, LONGINT, RECORD, POINTER...
<tangentstorm> hmm
<dsar> nice
<tangentstorm> dsar: I picture array assignment as a procedure call, with a runtime bounds-checking component, and reallocation and whatnot. Probably that's way more magic than should be happening in an oberon system
<dsar> surely
<dsar> by the way oberon has fixed arrays (no dynamic arrays), they can be checked at compile time
<dsar> only open arrays are checked in runtime
<tangentstorm> not sure i believe fixed arrays can be checked at compile time
<dsar> give me an example
<tangentstorm> PROCEDURE whatabout( x : INTEGER ) : INTEGER; BEGIN RETURN my_fixed_array[ x ] END whataboutthis;
<tangentstorm> argh. underscores aren't allowed, but you get the idea
<dsar> a smart compiler would ask for a guard in every array access
<tangentstorm> yeah
<tangentstorm> that's why i see all this stuff happening at run-time
<tangentstorm> in fpc, all arrays and strings are on the heap, and they all have a length field
<dsar> (by the way I was talking about array assignment, I know that indexing has runtime support)
<dsar> a := b; both fixed arrays
<dsar> a compiler can check for the validity of assignment by length
<tangentstorm> hmm
<tangentstorm> I can definitely see the benefit of an error when the sizes are fixed.
<tangentstorm> because you might want to read a file with fixed sized records or something.
<tangentstorm> i guess free pascal makes a distinction between dynamic arrays and open arrays.
<tangentstorm> open arrays being arguments to a function (and you can pass in fixed size or dynamic array variables)... dynamic arrays are basically pointers to arrays
<dsar> yup
<tangentstorm> i think the two concepts could be merged, but maybe i'm not thinking clearly.
<tangentstorm> anyway. anybody want to help with these test cases?
<tangentstorm> I'm kind of thinking the easiest target will be D, Go, Java, or Javascript
<tangentstorm> pascal and retro are what i really want, but i'll need to implement garbage collection first and i'm not comfortable with that yet
* Ericounet_portab (~ericp@mtr26-3-88-179-205-51.fbx.proxad.net) has joined #oberon
* Ericounet_portab (~ericp@mtr26-3-88-179-205-51.fbx.proxad.net) has left #oberon
* tangentstorm is now known as tangentaway
* tangentaway is now known as tangentstorm
<tangentstorm> I've thought about it, and I've decided to focus on targeting haxe.
<dsar> haxe?
<tangentstorm> http://haxe.org/
<tangentstorm> note the "multiplatform" part
<dsar> oh nice
<dsar> license?
<tangentstorm> not sure
<tangentstorm> http://haxe.org/doc/license
<tangentstorm> i don't care about the GPL here though, because i'm just using it as a black box until i have time to write my own code gen templates for those other languages
<dsar> burp
<dsar> (sorry)
<tangentstorm> I'm trying to compile oxford oberon 2 compiler but it keeps segfaulting in the build
<tangentstorm> maybe because i'm on a 64 bit platform...
<dsar> compile it with -m32
<tangentstorm> how?
<tangentstorm> in the configure script?
<dsar> yup
<dsar> as an argument
<tangentstorm> but how?
<tangentstorm> it won't take -m32
<tangentstorm> it's got to be something like CFLAGS='-m32'  right?
<dsar> yup
<tangentstorm> not working :/
<dsar> your gcc is able to generate 32bit code?
<tangentstorm> no idea
<tangentstorm> http://pastebin.com/8KvX9FNX
<tangentstorm> might just be easier to use my 32 bit machine. i was going to try compiling my test suite under oberon-2
<dsar> yup
<tangentstorm> huh. I didn't know oberon-2 had the WHILE/ELSIF construct.
<tangentstorm> once i fixed the syntax errors, all of my tests run fine under obc
<dsar> oberon-2 has WHILE/ELSIF?
<tangentstorm> oxford oberon2 does anyway
<dsar> oh nice
<dsar> where did you see it?
<dsar> just tried the syntax?
<tangentstorm> the "what's new in oberon-07" PDF presents the diffs from classic Oberon... i thought it was new in oberon-07
<tangentstorm> it's part of oberon07. i just wasn't expecting it to be in oberon2
<tangentstorm> http://www.inf.ethz.ch/personal/wirth/Articles/Oberon.html has an article on the differences between oberon and oberon-07, and i had misremembered it as being the differences since oberon 2
<tangentstorm> so far all my oberon07 example code runs just fine under obc
<dsar> I know that article
<dsar> oh yeah, it works!
<dsar> very nice
* tangentstorm is now known as tangentaway
* akastargazer (~stargazer@93-81-146-44.broadband.corbina.ru) has joined #oberon
* Zorko has quit (Ping timeout: 246 seconds)
* Zorko (~zorko@46.211.123.6) has joined #oberon
<Zorko> guys, Oberon-2 has WHILE/ELSIF
<Zorko> may be dsar is a purist and so big fan of O-07 because he didn't use O-2 ?
<Zorko> btw, has Oberon-07 read-only export ( a- : INTEGER ) ?
<dsar> in Oberon-07 exported variables are readonly and only of scalar type
<valexey> In Oberon-07 ALL expoer is read only
<valexey> *export
<dsar> not public record field they are writtable
<Zorko> heh. not very clear
<Zorko> need to make wrappers
<valexey> all global variable is read only. write permission granted only for members of the same module
<Zorko> PROCEDURE SetVar* (val: INTEGER); BEGIN locvar := var END SetVar;
<valexey> yep. and it is cool
<Zorko> why?
<valexey> good practice
<Zorko> I don't use global vars a lot
<Zorko> but it is better if the export specified explicitly
<valexey> because module can control this variables
<valexey> full control
<dsar> imho, writtable global variables are usefull as input parameter of the module, in some cases (a large buffer) it is good
<Zorko> no one bothers to have the same full control in Oberon-2 / CP
<Zorko> this problem present only in Oberon-1
<valexey> Zorko, ok. How to log all variable changes in Oberon-2?
<Zorko> do not export the variables to write
<valexey> It is Oberon-07 way :-)
<Zorko> valexey, to have full export of vars is more traditional way
<Zorko> and each of these approaches has its advantages and disadvantages
<Zorko> but O-2 allows both ways.
<valexey> Oberon-07 too. Just export global record variable.
<valexey> * pointer to record
<valexey> Oberon-07 structures intermodule scheduling.
<Zorko> I for some reason to think, Oberon-2 is not so popular, and O-07 will be even less, with its highly directional style.
<Zorko> we need to develop a good habit of coding, gentlemen.
<Zorko> O-07 will help us with this.
<valexey> Oberon needs killer app :-)
<valexey> application /= software in this case.
* akastargazer (~stargazer@93-81-146-44.broadband.corbina.ru) has left #oberon
<Zorko> applications are fat and bad designed soft for Windows?
<Zorko> hehe don't use Linux - it's not better
<valexey> application /= software
<dsar> linux sucks :PP
<dsar> ehehe
<valexey> but windows sucks more :-)
<Zorko> only AOS is rulezzz 4eva
<valexey> but AOS site is slo-o-ow.
<Zorko>
<Zorko> valexey, haven't you a wish to adapt Oberon-07/11 to Linux? to make ELF linker
<Zorko> as I think, it is must not be too hard
<Zorko> as a base you can give ELF linker from BlackBox
<valexey> On April-May, may be.
<Zorko> valexey, do you know Alexander Shiryaev?
<Zorko> he's man who may be interested in the same
<valexey> Yes, I do.
<Zorko> eh no one like to teach ELF format and write linker
<Zorko> teach=learn
<valexey> just use gold :-)
* Zorko has quit (Ping timeout: 244 seconds)
* norayr (~noch@arnet.am) has joined #oberon
<dsar> hola norayr
<dsar> :-)
<norayr> dsar: hey
<norayr> and what did he say?
<dsar> pvt
* Zorko (~zorko@46.211.123.6) has joined #oberon
* akastargazer1 (~stargazer@93-81-146-62.broadband.corbina.ru) has joined #oberon
* akastargazer1 (~stargazer@93-81-146-62.broadband.corbina.ru) has left #oberon
* akastargazer (~stargazer@93-81-146-62.broadband.corbina.ru) has joined #oberon
* akastargazer1 (~stargazer@93-81-146-62.broadband.corbina.ru) has joined #oberon
* akastargazer has quit (Ping timeout: 240 seconds)
* dsar has quit (Quit: Leaving)
* tangentaway is now known as tangentstorm
<tangentstorm> hey all
<valexey> tangentstorm, hi
<tangentstorm> heya
<tangentstorm> I added a very primitive hand-written test suite based on the one i used for PL/0
<tangentstorm> haven't gotten to records, pointers, and arrays yet though
<tangentstorm> I've got some kind of error in the OberonEmitter grammar.
<tangentstorm> kind of wondering if it's worth upgrading to antlr4 after all
<valexey> ok. I'll try it.
<valexey> thanks
<tangentstorm> valexey: try what?
<tangentstorm> i mean which part ?
<valexey> tests
<valexey> I like to use this tests for other compiler :-)
<tangentstorm> ah cool
* Ericounet_portab (~ericp@mtr26-3-88-179-205-51.fbx.proxad.net) has joined #oberon
* impomatic (~digital_w@211.67.125.91.dyn.plus.net) has joined #oberon
* Ericounet_portab (~ericp@mtr26-3-88-179-205-51.fbx.proxad.net) has left #oberon
<tangentstorm> I added a little script to dump the AST for a module as a lisp-style s-expression:  https://github.com/nickelsworth/noct/blob/master/parse-oberon07
<tangentstorm>
<tangentstorm> T00Output now compiles and runs under haxe.
<tangentstorm> https://gist.github.com/4581941
<tangentstorm>
* norayr has quit (Quit: Leaving.)
* egp_ (~egp_@31.47.170.115) has joined #oberon
* sinuhe (~sinuhe@kaptah.oberon07.com) has joined #oberon
* sinuhe has quit (Client Quit)
* dsar (~dsar@host217-170-dynamic.36-79-r.retail.telecomitalia.it) has joined #oberon
<dsar> 10 users.. wow
<tangentstorm> hey, we're just getting started
<tangentstorm> so... haxe does not have pass-by-reference semantics for primitive types, nor does it have a CHAR type. so i guess i'm going to have to make wrapper objects for everything
<tangentstorm> it does have macros though.. hmm..
<dsar> :O
* akastargazer1 (~stargazer@93-81-146-62.broadband.corbina.ru) has left #oberon

[Geniepro]

Слишком много мусора типа "has joined", "has left" -- трудно читать полезный текст. Ниасилил стену, короче ))

[valexey_u]

Слишком много мусора типа "has joined", "has left" -- трудно читать полезный текст. Ниасилил стену, короче ))

Да, без подсветки "синтаксиса" читать тяжко.

[vlad]

Еще логи от Олега  (не смотрел, почитаю на досуге)