00:00:00 --- log: started forth/21.06.28
01:06:51 --- quit: rain3 (Ping timeout: 120 seconds)
01:43:25 --- join: rain3 joined #forth
02:40:58 --- join: Glider_IRC__ joined #forth
02:43:49 --- quit: Glider_IRC_ (Ping timeout: 120 seconds)
03:34:10 --- quit: X-Scale (Ping timeout: 120 seconds)
05:55:26 --- join: Frek joined #forth
06:03:45 --- quit: Frek (Quit: )
09:20:53 <cbridge_> <remexre> siraben: ah, neat; I was watching the OPLSS security-typed-languages lectures the other day, sounds like this is the same thing
09:42:53 <proteus-guy> tabemann - regarding aligned word access being atomic on ARM - that's often not the case because of the odd caching model. You MUST use load exclusive followed by store exclusive to have any confidence your operation is atomic.
10:34:13 --- quit: rain3 (Ping timeout: 120 seconds)
11:58:53 --- join: X-Scale joined #forth
12:18:16 <cbridge_> <lispmacs[work]> I asked Dan a question about installing avrforth on atmega328P, and he responded by putting a new release on the site with support for 328*, 168p, and 88p:
12:18:20 <cbridge_> <lispmacs[work]> http://krue.net/avrforth/
12:18:35 <cbridge_> <lispmacs[work]> I forgot my programmer today, but was planning to try uploading that tomorrow
13:56:21 --- quit: cbridge_ (Quit: bye)
13:56:31 --- join: cbridge_ joined #forth
14:40:57 --- join: Glider_IRC_ joined #forth
14:43:51 --- quit: Glider_IRC__ (Ping timeout: 120 seconds)
14:59:18 <cbridge_> <dave0> maw
15:11:33 <cbridge_> <lispmacs[work]> wam
15:18:07 <cbridge_> <crc> maw dave0
15:24:31 <cbridge_> <eris[m]12> super maw io
15:26:41 <cbridge_> <dave0> maw crc 
15:27:38 <cbridge_> <eris[m]12> what is maw 
15:27:46 <cbridge_> <eris[m]12> like a condensed 'morning'? 
15:33:21 --- join: mark4 joined #forth
15:33:21 --- mode: ChanServ set +o :mark4
15:34:45 <mark4> http://dpaste.com/EJC2J9TKC  <-- can anyone here tell me wtf i screwed up in my list management code that makes the directory listing NOT sorted right?
15:34:53 <mark4> been banging my head against this wall for 4 days
15:50:16 <cbridge_> <dave0> mark4: that's not forth code!
15:50:56 <cbridge_> <eris[m]12> it could be, dave
15:51:02 <cbridge_> <eris[m]12> you just have to believe
15:52:01 <mark4> close enough for government work :)
15:54:23 <cbridge_> <dave0> mark4: could it be that add_entry cannot put d2 at the head of the list?
15:54:58 <cbridge_> <dave0> mark4: try adding entries in this order:    bbb ccc aaa
15:55:09 <cbridge_> <dave0> mark4: see if it comes out as  aaa bbb ccc
15:56:30 <mark4> add_entry seems to be right
15:56:39 <mark4> but if its not i cannot see HOW
15:57:01 <mark4> i have changed the sense on the sort too
15:57:06 <cbridge_> <jhswartz> mark4: readdir() is at the mercy of the underlying filesystem
15:57:13 <mark4> it changes the order but it does not result in a sorted list
15:57:23 <cbridge_> <jhswartz> if you want the entries ordered by name in any reliable sense, you have to do it yourself
15:57:53 <mark4> i am doing it myself
15:58:05 <mark4> but dirs and files while kept separate are not properly sorted
15:58:17 <mark4> i even wrote a custom version of strcmp to see if somehow there was a bug in that
15:58:24 <mark4> highly highly unlikely but i had to test
15:58:26 <cbridge_> <dave0> mark4: going by your comment on line 222, you always insert n2 /after/ n1, but what if n2 should be the new head of the list?
15:58:42 <mark4> thats taken care of
15:59:09 <mark4> node_insert() handles the case of an empty list
15:59:17 <mark4> so does the append() function
15:59:43 <cbridge_> <dave0> mark4: no no not an empty list... what if the list already contains   bbb ccc   and you try to add   aaa
16:00:22 <mark4> the strcmp should not allow out of order insertions
16:00:25 <mark4> thats the problem
16:00:33 <mark4> im getting out of order insertions or something
16:01:14 <cbridge_> <dave0> mark4: first time around the loop in add_entry, n1 points to bbb ... then you compare and aaa should come /before/ n1 ... but there's nowhere in the code to handle that case
16:01:55 <cbridge_> <dave0> mark4: i think that adding aaa to a list of bbb ccc, your current code would give bbb aaa ccc
16:02:00 <mark4> no. n1 points to the head of the list. if the new entry is greater than the current item n1 then we do n1=n1->next
16:02:11 <mark4> till the current scanned item is greater than the one were adding
16:02:25 <mark4> how?
16:02:55 <mark4> oooh!
16:03:17 <mark4> if we nave bbb ccc ddd as the lsit and i try to add aaa there is nothing to handle that?
16:03:31 <mark4> i.e. the item being added is smaller than the item at the head of the list!
16:03:36 <cbridge_> <dave0> mark4: i think so
16:03:38 <mark4> that might be it!
16:03:42 <cbridge_> <dave0> mark4: yes :-)
16:03:51 <cbridge_> <dave0> i haven't tested your code, i just went by inspection
16:04:18 <cbridge_> <eris[m]12> working in languages without proper repls isnt fun :( 
16:06:31 <mark4> ok that was def part of it
16:07:26 <cbridge_> <dave0> eris[m]12: forth's repl is what drew me to forth :-)
16:08:08 <cbridge_> <dave0> i have only ever used edit-compile-run in c for so many years :-(
16:08:12 <mark4> so there must be more than one problem because fixing that only partially fixed the unsortedness
16:10:29 <cbridge_> <eris[m]12> a bytepusher?
16:10:55 <mark4> http://dpaste.com/5LMKVQFC7
16:11:05 <cbridge_> <eris[m]12> ah! 
16:13:42 <cbridge_> <jhswartz> mark4: from gdb, it looks like every entry falls through to line 223 where l->head != NULL && n1 == l->head - in other words every entry is being inserted after the first entry of the list, or becomes the first entry if the list is empty
16:13:56 <cbridge_> <eris[m]12> ah, this is one of the list of memory addresses 
16:14:20 <mark4> HOW? WHY?
16:14:31 <mark4> thats like saying the string compare is not working
16:14:42 <mark4> and if thats the case my rewrite of it failed identically
16:15:13 <cbridge_> <jhswartz> you can check this by setting a breakpoint on add_entry and stepping through
16:17:35 <mark4> but WHY is it doing that?
16:18:06 <mark4>         if(cmp <= 0)  that if is never true?
16:18:13 <mark4>   cmp = strcmp(d2->name, d1->name);
16:18:20 <mark4> then this call never returns a value <= zero
16:18:52 <cbridge_> <jhswartz> consider d1->name is always "."
16:19:07 <mark4> it cant be
16:19:11 <mark4> that can only happen once :)
16:19:20 <mark4> there can never be more than one . in a dir :)
16:20:18 <cbridge_> <jhswartz> ah wait, sorry, i screwed up that strcmp line when copying it
16:20:55 <mark4> :)
16:24:04 <cbridge_> <MrMobius> mark4, did you try printing out the whole list every time you insert something?
16:27:13 <mark4> i didnt think of that :)
16:35:28 <mark4> ok things are def getting added out of order :(
16:35:39 <mark4> http://dpaste.com/4NRS2YAUG
16:37:18 <mark4> it def looks like items are being appended onto the end of the list
16:37:53 <mark4> which tells me strcmp is returning wrong results but I absolutely do not believe that
16:49:44 <cbridge_> <jhswartz> mark4: https://dpaste.com/63DE2X72T
16:53:13 <mark4> that doesnt make sense
16:55:01 <cbridge_> <jhswartz> output looks correct here though
16:56:15 <mark4> yea but i dont understand it
16:57:12 <mark4> it IS correect now but i TOTALLY do not understand how you even came up with that or what it is doing
16:57:19 <mark4>   node_insert(n1, n2);
16:57:44 <mark4>     n2->payload = d2;
16:57:47 <mark4> and it invalidates that line
16:59:44 <mark4> how does that work?
17:00:02 <mark4> ive been fighting this for 4 days solid
17:00:12 <cbridge_> <jhswartz> if cmp < 0, then surely d2->name should come before d1->name?
17:00:37 <cbridge_> <jhswartz> you didn't have a node_insert_before, so that payload SWAP is an improvisation to get one
17:00:38 <mark4> yea i need to fix my list code to be insert before not insert after
17:00:42 <mark4> that was a mistake on my part
17:09:22 <mark4> ty :)
17:09:49 <cbridge_> <jhswartz> mark4: where can i send the invoice? :)
17:19:29 <mark4> to joe biden
17:20:12 <mark4> he lives on pensylvania avenye in washington dc :)
17:20:17 <mark4> pennsynvania?
17:20:59 <cbridge_> <crc> pennsylvania
17:21:35 <mark4> ya double n
17:21:48 <mark4> 1200 i thin but i could be wrong about hta
17:21:53 <mark4> 1700? i ferget
17:22:40 <cbridge_> <crc> 1600
17:25:04 <mark4> yea
17:25:14 <mark4> was too lazy to go look it up heh
20:53:56 <cbridge_> <dave0> maw
23:46:21 --- join: rain3 joined #forth
23:59:59 --- log: ended forth/21.06.28