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
orivej has quit [Ping timeout: 268 seconds]
karswell has joined #picolisp
orivej has joined #picolisp
karswell_ has joined #picolisp
karswell has quit [Remote host closed the connection]
_whitelogger has joined #picolisp
karswell_ is now known as karswell
patrixl has joined #picolisp
karswell has quit [Remote host closed the connection]
karswell has joined #picolisp
karswell has quit [Read error: Connection reset by peer]
<tankf33der> morning
<tankf33der> implementing working prototype for (upp?) to pass tests vs. python3
<Regenaxer> Good morning tankf33der
<Regenaxer> Wow! Have you read the books?
<tankf33der> yes and no. ive checked different languages and libs. i will report.
<tankf33der> btw. books are useless :(
<Regenaxer> Thought so ;)
rob_w has joined #picolisp
mtsd has joined #picolisp
<beneroth> hi tankf33der, Regenaxer :)
<beneroth> wow tankf33der, thank you so much!
<Regenaxer> Moin beneroth!
<beneroth> Regenaxer, I need your help. My db looks corrupted, probably the cause for the wrong ordering of the aux index output
<Regenaxer> oh
<beneroth> confirm, aux output should be ordered, right?
<Regenaxer> yes, eg with 'scan'
<Regenaxer> (sorry that I could not answer yesterday)
<beneroth> situation: I looked at the index, it is corrupted, danglings etc. collected all objects together. doing (rebuild List 'prop 'class). List is uniq. afterwards doing collect on the index -> double entries (same object!)
<beneroth> how can (rebuild) cause double inserts of the same objects? any ideas?
<Regenaxer> So the whole key (ie aux key plus object) exists several times?
<Regenaxer> This combination must be unique, yes
<beneroth> aye
<beneroth> hm
<beneroth> wait, maybe I just solved it now...
<Regenaxer> Really diff? Can you check with '=' on the keys?
<beneroth> solved it, never mind.. I think.. probably caused because I used (rebuild) with abstract parent class instead of the real indexed class... I guess, ending up two re-index calls which end up into the same index prop... no idea
<beneroth> I worked on a copy of the corrupt database, will try again on the real one
<Regenaxer> Great
<beneroth> but without double entries (same key, same value) in the index, the aux index seems to behave as expected xD
<Regenaxer> :)
<beneroth> thanks for your help, you get +1 in remote healing
<Regenaxer> hehe, I did nothing ;)
<beneroth> how is your time situation today? I might need some more support, might have more db problems. and played around a bit with form.l, high probability for questions there.
<Regenaxer> I have some duties, but in between you can ask any time :)
<beneroth> thank you :)
<Regenaxer> np :)
<beneroth> two indices rebuilt, dbCheck returns T :D :D
libertas_ has quit [Ping timeout: 268 seconds]
<beneroth> (same class... very prone to insertion errors, probably happened during migration)
libertas has quit [Ping timeout: 260 seconds]
libertas has joined #picolisp
libertas_ has joined #picolisp
libertas has quit [Ping timeout: 268 seconds]
libertas_ has quit [Ping timeout: 265 seconds]
<beneroth> Regenaxer, I <3 too.l
<Regenaxer> :)
libertas has joined #picolisp
libertas_ has joined #picolisp
xificurC has joined #picolisp
libertas has quit [Killed (moon.freenode.net (Nickname regained by services))]
libertas_ is now known as libertas
libertas_ has joined #picolisp
xificurC has quit [Ping timeout: 260 seconds]
xificurC has joined #picolisp
xificurC has quit [Remote host closed the connection]
mtsd has quit [Quit: Leaving]
rob_w has quit [Remote host closed the connection]
<tankf33der> 50% done
<Regenaxer> wuff
<Regenaxer> Wow!
<tankf33der> since i failed against python3 i think python3 has something special.
<Regenaxer> You did not use CaseFolding.txt ?
<tankf33der> nim programming and libutf dont use it
<tankf33der> i should digg in python34
<tankf33der> i should digg in python3
<Regenaxer> So this is a simple lookup table? Not code blocks with weird bit indexing?
<tankf33der> yeap
<Regenaxer> Hmm, but this needs a search for every character, right?
<tankf33der> binary search or prefix tree
<Regenaxer> Thats very slow
<tankf33der> right
<tankf33der> dlang uses prefix tree
<Regenaxer> The Kaffe thing does plain bit operations
<tankf33der> like musl
<tankf33der> algos unknown.
<Regenaxer> musl also uses bit indexing?
<tankf33der> yea
<Regenaxer> The algo must be documented somewhere
<Regenaxer> We need case stuff very fast, it is used heavily in DB searches
<Regenaxer> +IdxFold etc
<tankf33der> casemap.h and towctrans.c
<tankf33der> i think picolisp wil be equal to nim.
<Regenaxer> alpha.h has the tables
<Regenaxer> It is magic
<Regenaxer> So most implementers don't understand it
<Regenaxer> If doing a search (binary or tree), we could simply use CaseFolding.txt slightly parsed
<Regenaxer> But I want an efficient one, without any searching
<tankf33der> of course.
<tankf33der> it was prototype but failed.
<Regenaxer> Whe python tables are in which file?
<tankf33der> unknown yet.
<Regenaxer> Do you think the musl version is correct?
<Regenaxer> (we can never be sure probably ;)
<tankf33der> i would recommend subscribe to musl list and ask.
<Regenaxer> I don't even know what musl is ;)
<Regenaxer> There is #musl in Freenode
<tankf33der> musl is glib
<tankf33der> very stable and active, several distros based on it
<Regenaxer> ok
<Regenaxer> so not glib (gnu)
<Regenaxer> but libc
<tankf33der> musl do binarysearch
<Regenaxer> it is the one that broke select(), right?
<tankf33der> yeap
<Regenaxer> oh, search?
<Regenaxer> If musl is soooo efficient, their site should load faster ;)
<Regenaxer> Where do you find the search?
<tankf33der> in general i think current utf8 is ok in picolisp.
<Regenaxer> yeah, a while loop
<Regenaxer> hmm, in which places is the old kaffe version wrong?
<tankf33der> i wrote already here
<tankf33der> on right side from red line
<tankf33der> for example
<Regenaxer> Because it differs from python or others?
<tankf33der> from python
<tankf33der> python3
<Regenaxer> And you are sure python3 is correct? ;)
<tankf33der> i belive its correct
<tankf33der> but have something extended
<Regenaxer> We should also look how other Javas do it
<tankf33der> i will test agains ruby then
<Regenaxer> hmm
<Regenaxer> It is all a mess
<tankf33der> what javas you would prefer?
<tankf33der> not scala, right?
<Regenaxer> I would like to see the source, as in Kaffe
<tankf33der> i checked all versions and didnt found it
<tankf33der> you could write to author
<Regenaxer> Most Java is not free
<Regenaxer> No, we should write the Unicode people
<Regenaxer> There must be some explanation
<Regenaxer> If CaseFolding.txt is really the only one they have, we should find an algo
<Regenaxer> by analyzing it an finding a pattern
<tankf33der> casefolding is useless
<tankf33der> unicodedata file have everything
<Regenaxer> Why is it useless?
<Regenaxer> It shows all upper and lower chars and their mapping
<Regenaxer> Can be trivially filtered
<Regenaxer> But I have not fully checked
<tankf33der> i have got info from casefolding from unicodedata file and it still failed against python3.
<Regenaxer> ok
<Regenaxer> hmm, then it seems python3 is wrong
<Regenaxer> What if we check Kaffe against CaseFolding?
pointfree has quit [Ping timeout: 248 seconds]
michelp has quit [Ping timeout: 246 seconds]
michelp has joined #picolisp
pointfree has joined #picolisp
pointfree has quit [Excess Flood]
pointfree has joined #picolisp
pointfree has quit [Excess Flood]
michelp has quit [Read error: Connection reset by peer]
tankf33 has joined #picolisp
tankf33 has quit [Remote host closed the connection]
orivej has quit [Ping timeout: 260 seconds]
alexshendi has joined #picolisp
alexshendi has quit [Quit: Leaving]
alexshendi has joined #picolisp
beneroth has quit [Quit: Leaving]
alexshendi has quit [Remote host closed the connection]
alexshendi has joined #picolisp
tankf33der has quit [Quit: Connection closed for inactivity]
orivej has joined #picolisp
alexshendi has quit [Quit: Leaving]