xkapastel has quit [Quit: Connection closed for inactivity]
<tankf33der>
ok
<m_mans>
Hi all!
<Regenaxer>
Cheers m_mans!
rob_w has joined #picolisp
masak has joined #picolisp
mtsd has joined #picolisp
f8l has quit [Ping timeout: 265 seconds]
stultulo has joined #picolisp
stultulo is now known as f8l
_whitelogger has joined #picolisp
mtsd_ has joined #picolisp
lman has joined #picolisp
mtsd has quit [Ping timeout: 265 seconds]
<lman>
hi
<Regenaxer>
Hi lman!
<lman>
how are you doing?
<Regenaxer>
good
<lman>
how is picolisp going?
<Regenaxer>
Very good ;)
<lman>
no osx port yet?
<Regenaxer>
no
<lman>
why so
<Regenaxer>
wait for llvm version
<Regenaxer>
Follow the discussions
<lman>
ok
<lman>
do you have libraries to manage rest and json?
<Regenaxer>
lib/http.l lib/json.l
<lman>
cool thanks
<beneroth>
Hi Regenaxer, hi lman
<lman>
hi beneroth
<Regenaxer>
Hi beneroth!
<lman>
have you considered creating a javascript version of picolisp?
<Regenaxer>
I did not, but Jon Kleiser made a (partial) one several years ago
<lman>
I guess the main problem is how to be efficient, you may need to use arrays for cell emulation
beneroth has quit [Ping timeout: 260 seconds]
<Regenaxer>
right
<Regenaxer>
like the Java version
<DKordic>
JavaScript is a Weird Machine by design.
<DKordic>
It's User-proof Exploit.
beneroth has joined #picolisp
<lman>
unfortunately I'm very dependant on js
<lman>
js has conquered the world btw
<beneroth>
time to conquer it back
<Regenaxer>
indeed
mtsd_ has quit [Quit: Leaving]
orivej has quit [Ping timeout: 264 seconds]
mtsd has joined #picolisp
<beneroth>
Regenaxer, any pointers to common causes for an corrupted index? I have in one project a special "last-edit" index, which stores (+Aux +Date) (time). it regularly gets corrupted (the tree).
<beneroth>
no problem to rebuild, bug ugly
<Regenaxer>
hmm, happens seldom
<Regenaxer>
might be a bad db model
<Regenaxer>
sometimes in +Bag with several indexes
<Regenaxer>
but then it gives only 'dangling'
<beneroth>
yeah must be something like that..
<Regenaxer>
how much corrupted?
<beneroth>
dbCheck reports bad symbol
<Regenaxer>
oh
<beneroth>
always same index, always same pattern
<Regenaxer>
must be something
<Regenaxer>
can you pastebin a relevant piece?
<Regenaxer>
e/r
<beneroth>
I suspect my last-edit code (which is part of 'upd>) gets called multiple times for the same object (not really intended this way), and that breaks the tree
<Regenaxer>
or just paste here perhaps
<Regenaxer>
multiple times should be ok
<Regenaxer>
i.e. same value stored several times
<beneroth>
ok. cannot paste really, it's pretty involved, would need to assemble all relevant parts..
<beneroth>
maybe the issue is in method dispatching
<beneroth>
no idea
<beneroth>
will investigate in detail another time
<Regenaxer>
no idea either
<beneroth>
ugly, but no issue, I just correctly rebuild the index before doing things like dbgc ;-)
<beneroth>
thanks anyway :)
<Regenaxer>
should all not corrupt I think
<Regenaxer>
"dbCheck reports bad symbol"
<Regenaxer>
is it 'zap'ped?
<Regenaxer>
it is perhaps not the index itself
<Regenaxer>
dbCheck gives only "Bad Type" or "Bad Value"
<beneroth>
I just run it once again
<beneroth>
I just paraphrased before
<beneroth>
I think it was "Bad Value"
<Regenaxer>
ah, ok
<Regenaxer>
So the object is of wrong type
<Regenaxer>
oops
<Regenaxer>
no, 'has>' mismatches
<Regenaxer>
for "Bad Value"
<beneroth>
({C1446} 737805 42650 . {O21564}) -- Bad Value
<Regenaxer>
ok
<beneroth>
from log
<beneroth>
haven't looked into the symbols
<Regenaxer>
Could be a class mismatch
<Regenaxer>
'has>' for that attribute
<beneroth>
yeah smells like that, would fit me, you know I like to build class hierarchies ;-)
<Regenaxer>
yeah :)
<Regenaxer>
multi-path inheritances :)
<beneroth>
just tried to do the action which corrupted the tree before, but not corrupt this time
<beneroth>
aye
<beneroth>
not sure if I can reliably cause the bug
<Regenaxer>
Somehow an object of the wrong class gets stored
<beneroth>
well I also do set> operations, which even also update this lastedit relations
<beneroth>
so.. could well be
<beneroth>
I expect a stupid mistake in my code, not in pil ;-)
<Regenaxer>
:)
<Regenaxer>
{C1446} above does not have the attribute of the tree index
<Regenaxer>
So best to inspect it if you have it the next time
<Regenaxer>
(not just the log)
<beneroth>
aye
<beneroth>
I'm aware. Will do :)
<beneroth>
thanks for the ideas!
<Regenaxer>
I'm sure you will find quickly
<beneroth>
low-level debugging in pilDB is actually quite doable. much less hazzle than C pointer stuff. and not comparable to hazzle with typical database systems which do so much magic you have never a chance to look behind.
<Regenaxer>
glad to hear :)
<beneroth>
well the value of this property of pilDB cannot be overstated, I think :)
<beneroth>
lol
<beneroth>
my bug seems really to be a Heisenbug
<Regenaxer>
:)
<beneroth>
but that doesn't make sense at all
<beneroth>
it shouldn't be
<beneroth>
but yeah... doing the action which previously prompted the bug a second time, no bug
<Regenaxer>
oh
<beneroth>
and the action is basically "change many objects, than change them back again"
<beneroth>
(which prompts updates/puts to the lastedit relations)
<Regenaxer>
so the order in which it happens?
<beneroth>
that is the only variable, beside DB-internal structure of externals
<beneroth>
or so I believe currently
<Regenaxer>
Too bad. Once an error is reproducible in Pil, it is easy to debug
<beneroth>
T
<beneroth>
will ignore it for now, I have more urgent tasks, unfortunately
<Regenaxer>
ok
<beneroth>
it can be repaired and doesn't do damage, just looks ugly
<Regenaxer>
true
<beneroth>
thanks for the discussion Regenaxer :)
m_mans has quit [Quit: Leaving.]
<Regenaxer>
welcome :)
<beneroth>
ah noe
<beneroth>
now no "Bad Value" but danglings
<beneroth>
so maybe it IS reproducible :)
<Regenaxer>
still strange, different things ;)
<beneroth>
aye
mtsd has quit [Quit: Leaving]
xkapastel has joined #picolisp
rob_w has quit [Quit: Leaving]
Nistur has quit [Ping timeout: 256 seconds]
Nistur has joined #picolisp
orivej has joined #picolisp
xkapastel has quit [Quit: Connection closed for inactivity]