ChanServ changed the topic of #picolisp to: PicoLisp language | Channel Log: https://irclog.whitequark.org/picolisp/ | Check also http://www.picolisp.com for more information
_whitelogger has joined #picolisp
_whitelogger_ has joined #picolisp
_whitelogger has joined #picolisp
<tankf33der> so, i finished.
<tankf33der> UTF8 is ok.
<Regenaxer> Glad to hear that, thanks!
<Regenaxer> So I can release the end-of-year version soon
<Regenaxer> Currently testing on Debian/arm64 and Debian/ppc64
<Regenaxer> Debian/arm64 is done
<Regenaxer> Tested on qemu
<Regenaxer> Now installing new ppc64/qemu, takes time
<Regenaxer> tankf33der, anything else to test?
<tankf33der> sure, all except IRIX
<tankf33der> pil "@lib/test.l" +
<tankf33der> ?
<Regenaxer> yep
<Regenaxer> Whatever ideas you have ;)
<tankf33der> i will run huge.l on several machines
<Regenaxer> Great
<tankf33der> i specially created this.
<tankf33der> for future testings.
<Regenaxer> perfect
<Regenaxer> I'm also doing tests of OSM import on another machine
<Regenaxer> Takes a week or so in total to import all Germany
<Regenaxer> The XML is 52 GiB or so
<Regenaxer> I'm still trying to find the bottlenecks, so this will take longer and is not finished for this release
<tankf33der> openbsd pil64 - ok
<Regenaxer> ok
<tankf33der> i follow changelog and know what changed.
<tankf33der> only 4 new functions in 2018 and no core changes.
<tankf33der> safe to release.
<Regenaxer> good
<Regenaxer> I finish the ppc64 tests today, then release
<Regenaxer> The core changes were quite heavy though, internal stuff
<Regenaxer> coroutines, utf8, tsm
<Regenaxer> especially the coroutines are tough
<tankf33der> huge.l passed on linux and centos
<tankf33der> huge.l passed on linux and openbsd
<tankf33der> coroutines checked on huge.l after notes in changelog.
<Regenaxer> ok
<tankf33der> just add trivial coroutine
<tankf33der> code to test.l
<Regenaxer> good idea
<tankf33der> we discuss this every year, I: add more tests to test.l to be sure, You: test.l is not for testing, real life is best tester.
<tankf33der> :)
<Regenaxer> yes, as many things are difficult to test automatically
beneroth has joined #picolisp
<beneroth> hi guys
<tankf33der> o/
<beneroth> \o
<Regenaxer> Hi beneroth!
<beneroth> I've met alexshendi :)
<Regenaxer> Nice!
<beneroth> so I met 3 picolispers now (if we count vikid)
<beneroth> :)
<beneroth> see you around. have a nice day friends :)
beneroth has quit [Quit: Farewell]
orivej has joined #picolisp
razzy` has quit [Ping timeout: 245 seconds]
_whitelogger has joined #picolisp
<tankf33der> i found
<tankf33der> Regenaxer: what you could add to test.l about coroutines
<Regenaxer> yes?
<Regenaxer> Ah, I did not remember. There is even one test for coroutines in test/src/flow.l
<tankf33der> this one
<tankf33der> only 503 is too much for default 8k stack
<tankf33der> 64 is ok. tested.
<tankf33der> perfect example for test.l
<tankf33der> perfect test for test.l
<Regenaxer> Is this from Alioth?
<Regenaxer> threadring.l
<tankf33der> yea
<tankf33der> i have all your rings.
<Regenaxer> Good idea! Let me try
<Regenaxer> I don't remember well
<Regenaxer> I wonder why I did (let (N N ...
<Regenaxer> ah, of course
<Regenaxer> Need 'N' local in the coroutine ;)
<tankf33der> yea.
<Regenaxer> Hmm, but isn't this example too complicated for a unit test? What does it test
<Regenaxer> except for co performance?
<tankf33der> several lines tests co functionality.
<tankf33der> 11:59 <Regenaxer> The core changes were quite heavy though, internal stuff
<tankf33der> 12:01 <Regenaxer> coroutines, utf8, tsm
<tankf33der> 12:02 <Regenaxer> especially the coroutines are tough
<Regenaxer> well, only 'co' and 'yield'
<tankf33der> to test before and after release.
<tankf33der> and protect co run only on pil64
<tankf33der> `(== 64 64)
<tankf33der> :)
<Regenaxer> yes, but those changes involved internal issues
<Regenaxer> garbage collection concerning namespaces in coroutines
<Regenaxer> Very difficult to test in unit tests. Heisenbugs
<Regenaxer> The other was co in combination with catch
<Regenaxer> frame inconsistencies
<Regenaxer> I had test scripts for those
<Regenaxer> but not meaningful in general tests
<Regenaxer> The thread ring is more a stress test than one for correctness
<tankf33der> ### member ###
<tankf33der> (test (3 4 5 6) (member 3 (1 2 3 4 5 6)))
<tankf33der> (test NIL (member 9 (1 2 3 4 5 6)))
<tankf33der> (test '((d e f) (g h i))
<tankf33der> (member '(d e f) '((a b c) (d e f) (g h i))) )
<tankf33der> the same as this one.
<tankf33der> just testing if all ok.
<tankf33der> then add co and thread ring.
<Regenaxer> yes, but a test for co and yield is already there
<tankf33der> ok.
<Regenaxer> So threadring with a smaller number?
<Regenaxer> eg 602 to get 100?
<Regenaxer> Why does it start 503 instances? I don't remember the task
<tankf33der> just big number
<tankf33der> i offer 64
<tankf33der> to be able to run on smaller machines too.
<Regenaxer> or even less, 20?
<tankf33der> of course.
<tankf33der> 24 :)
<Regenaxer> ok
<Regenaxer> or 12, one dozen
<tankf33der> sure
<tankf33der> and what number i will pass for ring ?
<tankf33der> 1024 ?
<Regenaxer> 20
<Regenaxer> to get 9
<Regenaxer> No need to let it run so long
<tankf33der> ok
<Regenaxer> So how about this: http://ix.io/1wUe
<Regenaxer> The coroutines need to be stopped too
<Regenaxer> not to disturb the following tests
<tankf33der> link above for co-yield
<tankf33der> 16:06 <Regenaxer> yes, but a test for co and yield is already there
<tankf33der> i thought you will add thread ring.
<Regenaxer> yes
<Regenaxer> So we need perhaps: http://ix.io/1wUh
<Regenaxer> Or better first make a list of tags?
<Regenaxer> (ring1 ... ring12)
<Regenaxer> hmm
<Regenaxer> perhaps better not
<Regenaxer> This is a too complicated test. Not clear what it does
<Regenaxer> I think unit tests must be simpler, so that is easily understandable what they do so one can verify the result
<Regenaxer> This one is too involved
<tankf33der> ok
<tankf33der> afk.
<Regenaxer> thanks anyway!
orivej has quit [Ping timeout: 268 seconds]
orivej has joined #picolisp
<Regenaxer> ppc64 tests succeeded
<Regenaxer> PicoLisp 18.12 released
<Regenaxer> PilBox also
<Regenaxer> and I asked the maintainers of Debian and Termux to update too
<tankf33der> ok
orivej has quit [Ping timeout: 245 seconds]
orivej has joined #picolisp
ubLIX has joined #picolisp
orivej has quit [Ping timeout: 272 seconds]
ubLIX has quit [Quit: *cackles*]
<tankf33der> it should, i cant imagine how it could fail
<tankf33der> pil is so stable and production ready.
<Regenaxer> T
orivej has joined #picolisp
_whitelogger has joined #picolisp
orivej has quit [Ping timeout: 250 seconds]
orivej has joined #picolisp
orivej has quit [Ping timeout: 250 seconds]
shpx has joined #picolisp
rob_w has joined #picolisp
orivej has joined #picolisp
shpx has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
shpx has joined #picolisp
shpx has quit [Client Quit]
ubLIX has joined #picolisp