2009 Jul 20 00:02:15  <inimino> hey
2009 Jul 20 00:02:40  <inimino> aamar, mariorz, around?
2009 Jul 20 00:03:00  <mariorz> hey
2009 Jul 20 00:03:02  <aamar> hey
2009 Jul 20 00:03:04  <inimino> hey
2009 Jul 20 00:03:50  <mariorz> everyone good?
2009 Jul 20 00:03:59  <mariorz> did not make to the end
2009 Jul 20 00:04:03  <inimino> I didn't get too far these last two weeks
2009 Jul 20 00:04:06  <mariorz> not even close, working on 2.24
2009 Jul 20 00:04:09  <inimino> brb phone call
2009 Jul 20 00:04:18  <mariorz> er 3.24
2009 Jul 20 00:05:38  <aamar> 3.30
2009 Jul 20 00:05:52  <aamar> busy week for me as well
2009 Jul 20 00:07:09  <mariorz> ive been playing around with clojure and that ate up some of my sicp allocated time :P
2009 Jul 20 00:08:05  <aamar> been meaning to try out clojure
2009 Jul 20 00:08:09  <aamar> maybe in an upcoming week
2009 Jul 20 00:08:10  <aamar> :)
2009 Jul 20 00:09:36  <mariorz> http://paste.lisp.org/display/83831
2009 Jul 20 00:09:41  <mariorz> its cool
2009 Jul 20 00:09:58  <mariorz> but im having trouble wrapping my head around the whole java archtecture
2009 Jul 20 00:10:31  <mariorz> everything feels over engineered i dont know
2009 Jul 20 00:11:30  <mariorz> that builds graph of wikipedia articles
2009 Jul 20 00:11:52  <mariorz> but there is some kind of memory leak
2009 Jul 20 00:12:26  <mariorz> the sax parser seems to want to keep the whole doc in memory
2009 Jul 20 00:15:08  <aamar> it's not just the node_ids and the nodelinks, but the files themselves?
2009 Jul 20 00:16:18  <mariorz> i think so, the node_ids alone take about 500mb which i foudn believable (the whole file is 2.5 GB) but then on that seond test run for the node links, if completelly fills the heap
2009 Jul 20 00:16:32  <mariorz> a 2GB heap
2009 Jul 20 00:16:49  <mariorz> and well the links are only being stored as keys
2009 Jul 20 00:21:48  -*- inimino finally off the phone
2009 Jul 20 00:22:01  -*- inimino reads
2009 Jul 20 00:22:06  <aamar> time to start?
2009 Jul 20 00:22:18  <inimino> yeah, I got to 3.25 only
2009 Jul 20 00:22:46  <mariorz> cool i was worried
2009 Jul 20 00:22:53  <mariorz> its like no child left behind
2009 Jul 20 00:22:58  <mariorz> thanks guys!
2009 Jul 20 00:23:11  <inimino> hehe
2009 Jul 20 00:23:15  <aamar> haha
2009 Jul 20 00:23:41  <aamar> ok, should we start with 3.20 ?
2009 Jul 20 00:23:42  <inimino> alright, sure, let's start
2009 Jul 20 00:24:34  <mariorz> heh
2009 Jul 20 00:24:49  <inimino> http://inimino.org/~inimino/projects/2009/SICP/chap_3/3.20.svg
2009 Jul 20 00:25:02  <mariorz> oh dman
2009 Jul 20 00:25:17  <mariorz> i didnt do the diagrams
2009 Jul 20 00:28:08  <aamar> http://github.com/aalearn/aalearn-sicp/blob/6ce0edfd7bf669eaedd50d2249a4433457ee0453/ch3-outline.scm#L449-461
2009 Jul 20 00:28:12  <mariorz> but yours looks right
2009 Jul 20 00:28:48  <inimino> ok
2009 Jul 20 00:29:11  -*- inimino looks at aamar's
2009 Jul 20 00:29:33  <aamar> mine's different -- trying to remember what exactly I meant
2009 Jul 20 00:30:51  <inimino> I think yours is the same idea
2009 Jul 20 00:31:11  <aamar> okay, more or less similar -- you drew z as a separate environment
2009 Jul 20 00:31:31  <aamar> but we agree that it points back to the environment associated with x and y
2009 Jul 20 00:31:33  <inimino> the z has those ()() with only one line coming out which I take as indicating that they both point to x
2009 Jul 20 00:31:33  <aamar> ok
2009 Jul 20 00:31:35  <inimino> yeah
2009 Jul 20 00:32:27  <aamar> I think we should go to 3.21
2009 Jul 20 00:32:38  <mariorz> they point to (x,y) no?
2009 Jul 20 00:32:49  <mariorz> (x . y)
2009 Jul 20 00:33:32  <inimino> well, I think the line from x in the global environment is going to the wrong place
2009 Jul 20 00:33:50  <inimino> but yeah, I think we can move on to 3.21
2009 Jul 20 00:34:23  <aamar> I didn't have any y in this exercise, actually
2009 Jul 20 00:34:59  <mariorz> i got 3.21 wrong didnt noticed until i needed a print-deque
2009 Jul 20 00:35:01  <aamar> no wait, I did...
2009 Jul 20 00:35:02  <inimino> there's a y in E1
2009 Jul 20 00:35:13  <mariorz> bfixed print-deque but didnt get back to this one
2009 Jul 20 00:35:22  <inimino> (I think the line coming from x was meant to point to E1)
2009 Jul 20 00:35:44  <aamar> ok got it, thanks
2009 Jul 20 00:35:50  <aamar> okay 3.21
2009 Jul 20 00:35:58  <aamar> http://github.com/aalearn/aalearn-sicp/blob/6ce0edfd7bf669eaedd50d2249a4433457ee0453/ch3-outline.scm#L463
2009 Jul 20 00:36:14  <inimino> http://inimino.org/~inimino/projects/2009/SICP/chap_3/3.21
2009 Jul 20 00:36:20  <mariorz> oh cool
2009 Jul 20 00:36:32  <mariorz> thats what i have
2009 Jul 20 00:37:20  <mariorz> pushing to github
2009 Jul 20 00:37:25  <inimino> whoops, guess I could have used front-ptr there for consistency
2009 Jul 20 00:38:58  <mariorz> yeah but works jsut as well
2009 Jul 20 00:39:13  <aamar> I have trouble remembering to use the "interface" in some of these cases.
2009 Jul 20 00:39:29  <inimino> yeah, likewise
2009 Jul 20 00:39:37  <aamar> Seems good though
2009 Jul 20 00:39:41  <mariorz> http://github.com/mariorz/sicp/blob/ea99328e4f06518a4eee554f7e1f80212341c854/3.ss#L510
2009 Jul 20 00:39:52  <inimino> you end up needing to think about the actual representation to answer most of the exercises
2009 Jul 20 00:40:26  <aamar> ok, no explicit "display" in mariorz, but with my interpreter it would work fine
2009 Jul 20 00:40:51  <inimino> ok same answer
2009 Jul 20 00:41:00  <mariorz> t
2009 Jul 20 00:41:05  <inimino> oh yeah, display
2009 Jul 20 00:41:36  <inimino> this was the part of the book where I really start to miss algebraic datatypes :/
2009 Jul 20 00:42:16  <mariorz> how so?
2009 Jul 20 00:42:40  <inimino> well, using pairs for everything, and then, as aamar was saying trying to remember to use an interface
2009 Jul 20 00:42:59  <inimino> as opposed to the way you'd do this in Haskell
2009 Jul 20 00:43:31  <inimino> especially in the deque problem, I found using pairs really inconvenient and error-prone
2009 Jul 20 00:43:44  <inimino> but if you could say something like
2009 Jul 20 00:44:16  <inimino> well
2009 Jul 20 00:44:36  <inimino> I won't try to translate into Haskell, that would be too off-topic
2009 Jul 20 00:45:17  <mariorz> inimino: you do alot of haskell?
2009 Jul 20 00:45:20  <inimino> but I just really like using data constructors and pattern matching to say exactly what you mean
2009 Jul 20 00:45:49  <inimino> mariorz: not a lot, I used it a lot about a year ago writing an email spam filter
2009 Jul 20 00:46:14  <mariorz> i should look into it a bit
2009 Jul 20 00:46:46  <inimino> I really recommend learning it
2009 Jul 20 00:47:13  <inimino> it's got all the elegance of Scheme but without the icky mutable state stuff we are getting into now :-)
2009 Jul 20 00:47:53  <aamar> :)
2009 Jul 20 00:48:03  <inimino> anyway, </off-topic>
2009 Jul 20 00:48:10  <mariorz> k
2009 Jul 20 00:48:15  <aamar> where were we?
2009 Jul 20 00:48:38  <mariorz> 3.22 i think
2009 Jul 20 00:49:29  <inimino> yeah
2009 Jul 20 00:49:45  <inimino> http://inimino.org/~inimino/projects/2009/SICP/chap_3/3.22
2009 Jul 20 00:50:10  <aamar> http://github.com/aalearn/aalearn-sicp/blob/6ce0edfd7bf669eaedd50d2249a4433457ee0453/ch3-outline.scm#L472
2009 Jul 20 00:51:04  <inimino> hm, I think mine is broken
2009 Jul 20 00:51:46  <inimino> specifically it should fail to insert into an empty queue
2009 Jul 20 00:51:56  <mariorz> i was about to say mine is like inimino's but its literally the same, didnt do this one either
2009 Jul 20 00:52:06  <mariorz> er aamar's
2009 Jul 20 00:53:58  <mariorz> inimino:  yeah, an if in inert would take care of that
2009 Jul 20 00:55:01  <inimino> yeah
2009 Jul 20 00:56:26  <aamar> I don't see any other consequential differences
2009 Jul 20 00:56:32  <inimino> me neither
2009 Jul 20 00:56:39  <inimino> alright, 3.23?
2009 Jul 20 00:57:09  <mariorz> yaya this one i did
2009 Jul 20 00:57:13  <mariorz> was fun
2009 Jul 20 00:57:36  <inimino> yeah
2009 Jul 20 00:57:44  <mariorz> http://github.com/mariorz/sicp/blob/ea99328e4f06518a4eee554f7e1f80212341c854/3.ss#L510
2009 Jul 20 00:57:56  <inimino> I wasn't a huge fan of this one
2009 Jul 20 00:58:03  <aamar> http://github.com/aalearn/aalearn-sicp/blob/6ce0edfd7bf669eaedd50d2249a4433457ee0453/ch3-outline.scm#L512-591
2009 Jul 20 00:58:07  <aamar> Took a long time for me
2009 Jul 20 00:58:28  <mariorz> same here a couple fo hours yesterday
2009 Jul 20 00:58:38  <mariorz> print-deque took me the longest
2009 Jul 20 01:00:41  <inimino> yeah it took me a long time too
2009 Jul 20 01:01:01  <inimino> I didn't even write a print-deque
2009 Jul 20 01:01:59  <aamar> okay, scanning for differences...
2009 Jul 20 01:02:45  <aamar> Okay, so we're implementing doubly-linked lists.
2009 Jul 20 01:03:36  <aamar> mariorz: in front-insert-deque!, I think you also need to set the "previous" pointer to new_pair for whatever used to be at the front of the list.
2009 Jul 20 01:04:19  <inimino> mariorz: yours looks very similar to mine
2009 Jul 20 01:04:39  <mariorz> aamar: i do that
2009 Jul 20 01:04:46  <mariorz> (set-cdr! (cdr new-pair) (front-ptr deque))
2009 Jul 20 01:04:57  <mariorz> oh wait
2009 Jul 20 01:05:16  <inimino> I missed that, but it looks like he's right
2009 Jul 20 01:05:55  <mariorz> eyah
2009 Jul 20 01:06:00  <mariorz> should add
2009 Jul 20 01:06:00  <mariorz> (set-cdr! (cdr (front-ptr deque) new-pair))
2009 Jul 20 01:06:10  <mariorz> but weird becuase my tests worked
2009 Jul 20 01:06:14  <inimino> you have a '() where you want (front-ptr deque) in the new-pair I think
2009 Jul 20 01:06:44  <mariorz> true better to do it from the let
2009 Jul 20 01:07:08  <mariorz> but still my tests worked
2009 Jul 20 01:07:22  <mariorz> maybe some other bug compensates? :P
2009 Jul 20 01:07:25  <inimino> wait scratch that
2009 Jul 20 01:07:35  <inimino> on line 596 you do set it
2009 Jul 20 01:08:39  <inimino> I think both rear-insert and front-insert might be correct
2009 Jul 20 01:08:49  <inimino> but they're asymmetrical
2009 Jul 20 01:09:54  <mariorz> im trying to remember what i did at, some issue presented when writing print-deque made me do that fiz
2009 Jul 20 01:09:57  <mariorz> x
2009 Jul 20 01:10:31  <aamar> In the case where the queue is not empty, you need to set 3 pointers, by my count.
2009 Jul 20 01:10:34  <aamar> 1. the front-queue
2009 Jul 20 01:10:45  <aamar> 2. "next" on the new-pair
2009 Jul 20 01:11:08  <inimino> no, I think aamar was right all along and your front-insert is wrong
2009 Jul 20 01:11:18  <aamar> 3. "prev" on the previous front-queue (or alternately on new-pair's next)
2009 Jul 20 01:11:26  <inimino> agreed
2009 Jul 20 01:11:42  <mariorz> inimino: 596 sets the enx pointer on the new_pair
2009 Jul 20 01:11:58  <mariorz> next
2009 Jul 20 01:11:59  <inimino> yes, that takes care of ptr 2 in aamar's list
2009 Jul 20 01:12:32  <inimino> but nothing sets his #3
2009 Jul 20 01:12:56  <mariorz> yes, my tests would not catch it unless y rear-delete all the way to the seocnd pointer
2009 Jul 20 01:13:09  <inimino> exactly
2009 Jul 20 01:13:15  <aamar> I think I can construct a test that breaks it -- basically do two front inserts, then two rear-deletes
2009 Jul 20 01:13:18  <aamar> ok agreed
2009 Jul 20 01:13:20  <inimino> yeah
2009 Jul 20 01:13:35  <mariorz> right
2009 Jul 20 01:14:19  <mariorz> i catched the opposite bug which was fixed in rear-insert
2009 Jul 20 01:14:27  <mariorz> missed this one
2009 Jul 20 01:14:36  <inimino> yeah
2009 Jul 20 01:14:47  <inimino> alright, anything more on 3.23?
2009 Jul 20 01:15:06  <inimino> it's interesting that we all had basically the same implementation
2009 Jul 20 01:15:38  <aamar> one more thing...
2009 Jul 20 01:16:13  <aamar> mariorz on rear-delete-queue -- you should also be setting front-ptr to null in some cases
2009 Jul 20 01:16:18  <aamar> my version is actually a mess
2009 Jul 20 01:16:35  <aamar> but inimino has a better solution
2009 Jul 20 01:17:39  <aamar> same thing with front-delete-queue! -- if empty, set rear-ptr back to '()
2009 Jul 20 01:17:41  <inimino> oh I got distracted and never finished reading aamar's
2009 Jul 20 01:18:16  <inimino> yeah I see you added a check for singletons
2009 Jul 20 01:18:51  <wastrel> hi
2009 Jul 20 01:19:51  <inimino> hello
2009 Jul 20 01:20:04  <mariorz> ok, noted
2009 Jul 20 01:20:20  <aamar> well, I think we had more similar than different, overall
2009 Jul 20 01:20:22  <aamar> Ok
2009 Jul 20 01:20:24  <aamar> 3.24
2009 Jul 20 01:20:27  <mariorz> if not the front-ptr keeps point to the prev last
2009 Jul 20 01:20:35  <mariorz> pointing
2009 Jul 20 01:20:47  <aamar> http://github.com/aalearn/aalearn-sicp/blob/6ce0edfd7bf669eaedd50d2249a4433457ee0453/ch3-outline.scm#L593-618
2009 Jul 20 01:20:48  <mariorz> in the case of only one element
2009 Jul 20 01:20:58  <mariorz> man i was sure i had it right
2009 Jul 20 01:21:40  <inimino> aamar: I finished reading yours, it does look very similar to mine
2009 Jul 20 01:21:59  <mariorz> im working on this one
2009 Jul 20 01:22:30  <aamar> ok, cool
2009 Jul 20 01:22:32  <inimino> I didn't implement it, just the change
2009 Jul 20 01:22:34  <inimino> http://inimino.org/~inimino/projects/2009/SICP/chap_3/3.24
2009 Jul 20 01:23:27  <mariorz> i guess interface foe next/prev on the deques is def the way to go
2009 Jul 20 01:23:50  <inimino> aamar yours looks like what I had in mind there
2009 Jul 20 01:24:07  <inimino> (that's as far as I got also)
2009 Jul 20 01:24:09  <aamar> inimino: we have identical solutions
2009 Jul 20 01:24:11  <aamar> ok,
2009 Jul 20 01:24:14  <aamar> let's hold here...
2009 Jul 20 01:24:15  <inimino> yes
2009 Jul 20 01:24:57  <inimino> alright
2009 Jul 20 01:25:11  <mariorz> i feel cool when using cddr and such though
2009 Jul 20 01:25:18  <mariorz> makes me want to grow a beard
2009 Jul 20 01:25:22  <inimino> haha
2009 Jul 20 01:26:35  <inimino> studies have shown that consistent use of car and cdr stimulates the facial hair follicles
2009 Jul 20 01:27:40  <mariorz> o i know
2009 Jul 20 01:27:54  <mariorz> and everyone knows beards increase skill
2009 Jul 20 01:27:56  <aamar> brb
2009 Jul 20 01:28:41  <inimino> it's a virtuous cycle
2009 Jul 20 01:30:36  <aamar> back
2009 Jul 20 01:30:55  <aamar> checked for a beard, definitely coming in a bit thicker lately
2009 Jul 20 01:31:13  <inimino> hehe
2009 Jul 20 01:31:22  <aamar> ok, how far should we aim for next time?
2009 Jul 20 01:31:39  <inimino> I was just going to ask that :-)
2009 Jul 20 01:31:44  <aamar> 3.37 so we can be pretty sure to make it?
2009 Jul 20 01:31:50  <inimino> maybe finally finish up?
2009 Jul 20 01:31:51  <inimino> sure
2009 Jul 20 01:31:57  <aamar> sounds good
2009 Jul 20 01:32:04  <mariorz> agreed
2009 Jul 20 01:32:14  <inimino> I note that scroll bar still has a ways to go down in my browser window so I think that's about right
2009 Jul 20 01:32:50  <aamar> heh
2009 Jul 20 01:33:03  <aamar> lots of reading in the last few sections
2009 Jul 20 01:33:25  <inimino> yeah
2009 Jul 20 01:33:29  <inimino> and long exercises
2009 Jul 20 01:35:11  <inimino> I guess they are getting us ready for "Exer. 4.x.  Now write a Scheme interpreter mwahahaha"
2009 Jul 20 01:35:26  <inimino> so next meeting Aug. 2?
2009 Jul 20 01:35:31  <mariorz> lol
2009 Jul 20 01:35:55  <aamar> ok, good meeting
2009 Jul 20 01:35:59  <wastrel> i haven't shaved for a week
2009 Jul 20 01:35:59  <aamar> see you in a week or 2
2009 Jul 20 01:36:18  <mariorz> cool
2009 Jul 20 01:36:19  <inimino> alright, Aug. 2 it is :-)
2009 Jul 20 01:36:22  <inimino> see you guys
2009 Jul 20 01:36:24  <mariorz> see you guys