00:00 (join) neilv
00:02 neilv: if i want to install the racket version that is on the path to becoming 5.1.2, do i get it from http://pre.racket-lang.org/installers/ or somewhere else?
00:03 neilv: > You are about to download: plt-5.1.2.3-src-unix.tgz (16M)
00:03 neilv: seems promising, but I'm just double-checking
00:04 offby1: couldn't tell ya
00:08 (join) kennyd
00:11 (join) asurai
00:11 (quit) asurai: Client Quit
00:14 neilv: the c part of the build seems a lot faster since the gui ffi stuff, unsurprisingly
00:14 (quit) bmp: Ping timeout: 260 seconds
00:21 dnolen: anyone here familiar with Racket's pattern matcher (implementation details) ?
00:32 (join) jonrafkind
00:40 (quit) neilv: Ping timeout: 255 seconds
00:54 (join) jeapostrophe
01:03 (quit) jeapostrophe: Quit: jeapostrophe
01:15 (join) neilv
01:30 (quit) dnolen: Quit: dnolen
01:34 (quit) jonrafkind: Ping timeout: 250 seconds
01:44 (join) hkBst
01:57 (quit) dherman: Quit: dherman
02:14 (quit) hussaibi: Ping timeout: 252 seconds
02:14 (quit) hussaibi_: Ping timeout: 252 seconds
03:05 (quit) sheikra: Ping timeout: 255 seconds
03:16 (quit) neilv: Ping timeout: 255 seconds
03:59 (quit) realitygrill: Quit: realitygrill
04:03 (join) tfb
04:08 (join) neilv
04:08 (quit) neilv: Changing host
04:08 (join) neilv
04:08 (quit) neilv: Client Quit
04:56 (join) masm
05:13 (quit) DGASAU: Ping timeout: 264 seconds
05:38 (quit) doc_who: Read error: No route to host
05:49 (quit) hkBst: Ping timeout: 255 seconds
06:02 (quit) gienah: Quit: leaving
06:02 (join) hkBst
06:20 (quit) hkBst: Ping timeout: 240 seconds
06:24 (join) hkBst
06:37 (join) MayDaniel
06:47 (quit) hkBst: Ping timeout: 240 seconds
07:03 (join) sheikra
07:05 (join) hkBst
07:09 (quit) DT``: Ping timeout: 264 seconds
07:10 (quit) hkBst: Ping timeout: 276 seconds
07:14 (join) hkBst
07:18 (quit) hkBst: Ping timeout: 258 seconds
07:20 (join) hkBst
07:22 (join) DT``
07:31 (quit) MayDaniel: Read error: Connection reset by peer
07:34 (quit) hkBst: Ping timeout: 252 seconds
07:36 (join) hkBst
07:50 (join) dnolen_
07:54 (quit) hkBst: Ping timeout: 255 seconds
07:56 (join) hkBst
08:24 (join) drdo
08:45 (join) jeapostrophe
08:50 (join) RacketCommitBot
08:50 RacketCommitBot: [racket] plt pushed 1 new commit to master: https://github.com/plt/racket/commit/fba1777b8ad6d84200e17c85896f9f6d210b0d1d
08:50 RacketCommitBot: [racket/master] fix contract - Matthew Flatt
08:50 (part) RacketCommitBot
09:15 (join) sheikra_
09:19 (join) sheikra__
09:19 (quit) sheikra: Ping timeout: 276 seconds
09:22 (quit) sheikra_: Ping timeout: 246 seconds
09:23 (quit) dnolen_: Quit: dnolen_
09:37 (join) MayDaniel
09:53 (quit) hkBst: Read error: Connection reset by peer
09:53 (join) ChibaPet
10:27 (join) realitygrill
10:34 (join) anRch
10:36 (join) Flatlander
10:36 (quit) Flatlander: Client Quit
10:47 (join) dnolen
10:49 (join) jao
10:54 (quit) sheikra__: Read error: Connection reset by peer
10:54 (join) sheikra__
10:57 (quit) jeapostrophe: Quit: jeapostrophe
11:02 (quit) realitygrill: Quit: realitygrill
11:02 (join) upwardindex
11:03 (join) realitygrill
11:06 (quit) realitygrill: Client Quit
11:08 (part) dnolen
11:08 (join) dnolen
11:11 (join) jeapostrophe
11:15 (quit) jeapostrophe: Client Quit
11:22 (join) jonrafkind
11:32 (join) jeapostrophe
11:37 (quit) jeapostrophe: Quit: jeapostrophe
11:41 (quit) anRch: Quit: anRch
11:51 (join) anRch_
12:01 (join) anRch
12:02 (quit) anRch_: Read error: Connection reset by peer
12:20 (quit) anRch: Quit: anRch
12:21 (quit) upwardindex: Quit: upwardindex
12:26 ChibaPet: Eli, are you about?
12:28 (quit) tfb: Ping timeout: 260 seconds
12:28 (join) carleastlund
12:36 (nick) elliottcable -> sad_elliottcable
12:38 (join) JuanDaugherty
12:38 (nick) sad_elliottcable -> elliottcable
12:41 (join) jeapostrophe
12:44 (quit) jeapostrophe: Client Quit
12:44 (join) lucian
12:51 (join) shofetim
12:55 (quit) shofetim: Remote host closed the connection
12:56 (join) shofetim
12:59 clklein: I was pleasantly surprised to see that `for-template' works in `provide' but now unpleasantly confused about how to use it.
13:01 drdo: net/imap doesn't support SEARCH or am i missing something?
13:01 jonrafkind: i think you can only provide things for-template that were required for-template
13:02 clklein: For a module to provide a binding at some level, does it need to already be defined at that level? I take it, for example ...
13:02 jonrafkind: yes
13:02 clklein: jonrafkind: that's what I was going to ask :)
13:03 clklein: I figured what I thought I found was too convenient to be true
13:03 (join) tunes
13:21 (join) sheikra
13:31 (quit) tunes: Read error: Operation timed out
13:31 sheikra: Is there a way to let the racket interpreter repl work in typed racket mode by default?
13:31 (join) littlebobby
13:34 (join) joshcough
13:36 (join) anRch
13:39 (join) viller
13:39 viller: is there a document that highlights the ways racket has improved on standard scheme?
13:41 jonrafkind: I haven't heard of such a thing.. but it sounds like a good addition to somewhere in the docs
13:46 (join) tunes
13:52 sheikra: Can I compile a racket program into a standalone executable?
13:52 (quit) viller: Quit: ChatZilla 0.9.87 [Firefox 5.0/20110615151330]
13:54 ohwow_: sheikra: yes
13:54 ohwow_: man raco
13:54 ohwow_: http://docs.racket-lang.org/guide/exe.html
13:54 sheikra: thanks. I'll have a look
13:57 clklein: sheikra: "racket -I typed/racket"
13:59 sheikra: clklein: it works. thank you.
14:00 clklein: np
14:00 (join) jeapostrophe
14:01 (quit) jeapostrophe: Client Quit
14:04 (quit) JuanDaugherty: Remote host closed the connection
14:16 (quit) lucian: Read error: Connection reset by peer
14:18 (join) lucian
14:24 sheikra: hmm.. here is a weird thing I'm fighting with the module system with ...
14:25 sheikra: I have a "virtual" function in a file, which I hope its redefinition would change the behavior of the code that depend on it. But racket would not let me redefine it...
14:26 sheikra: It feels like that this "virtual" function should become the module's "parameter"...
14:28 (quit) lucian: Quit: Ex-Chat
14:28 (quit) carleastlund: Quit: carleastlund
14:32 (join) lucian
14:40 clklein: sheikra: What do you mean by a "virtual" function?
14:40 sheikra: clklein: ah, I'm abusing the terms here. I mean a function whose redefinition can change the behavior of all the code that depend on it.
14:40 clklein: Are you after something that's like `require' but doesn't force you to refer to a particular module?
14:41 clklein: And what do you mean by redefinition?
14:42 sheikra: clklein: say the function already has a definition in the module A. Now from module B, I want to redefine this function, so that all the code in module A would be affected by the change
14:43 (quit) lucian: Read error: Connection reset by peer
14:43 clklein: Do you mean "edit the source code of module B" or "write B so that running it changes code that refers to module A"?
14:43 (quit) anRch: Quit: anRch
14:43 sheikra: the latter
14:43 clklein: Sorry, what I meant by "what do you mean by redefinition" probably wasn't clear :)
14:44 clklein: Anyway, I understand now.
14:44 sheikra: Is there a way, or do I need to change my design?
14:45 clklein: You can have A export a box containing the function that B replaces.
14:46 sheikra: sound like it should work...
14:46 sheikra: are there better ways? This sounds a little hacky ;-)
14:47 clklein: Yes, it's not a good thing :)
14:47 clklein: I would change my design.
14:47 sheikra: but maybe my design is bad. it's not very modular
14:49 (join) dherman
14:50 (join) lucian
14:51 sheikra: Let me make it concrete, I'm making a cross-language "semantic diff" tool.  Each language has its own module. From the generic "diff" module there are some functions which will get the name of a node using a function named "get-name". But this function should be "overridden" by the modules for different languages.
14:52 stamourv: sheikra: It sounds like either the class system or the unit system would do what you want.
14:52 sheikra: stamourv: what's the unit system?
14:52 stamourv: Units are first class modules.
14:52 (join) jeapostrophe
14:53 clklein: sheikra: http://docs.racket-lang.org/guide/units.html
14:53 stamourv: You basically specify the dependencies of a unit, and hook it up to any other unit that exports the dependencies.
14:53 (quit) jeapostrophe: Client Quit
14:53 clklein: or http://docs.racket-lang.org/guide/classes.html
14:54 stamourv: So you could have your `diff' unit import a signature that includes `get-name' and others.
14:54 stamourv: And the unit for each language would export the signature, providing concrete bindings for `get-name' and the rest.
14:56 sheikra: stamourv: I guess that's it. I'll figure out how to use it. thank you
14:57 stamourv: If you have any questions, you're welcome to ask.
15:01 (quit) shofetim: Remote host closed the connection
15:05 (quit) sheikra: Ping timeout: 250 seconds
15:05 (quit) sheikra__: Ping timeout: 255 seconds
15:15 (join) RacketCommitBot
15:15 RacketCommitBot: [racket] plt pushed 1 new commit to master: https://github.com/plt/racket/commit/3f4074296882a40d25be10c88327535484456335
15:15 RacketCommitBot: [racket/master] Remove bash completion for commands we don't provide. - Vincent St-Amour
15:15 (part) RacketCommitBot
15:17 (join) JuanDaugherty
15:24 (join) jeapostrophe
15:31 (quit) dnolen: Quit: Page closed
15:41 eli: ChibaPet: ?
15:44 ChibaPet: Eli, I was going to ask about your desktop issue, but someone posted what I was going to suggest to the mailing list.
15:45 ChibaPet: Did you resolve it?
15:46 clklein: or http://docs.racket-lang.org/guide/classes.html
15:46 clklein: oops, sorry... "up-enter" in wrong window
15:47 (join) dnolen
15:49 (quit) MayDaniel: Read error: Connection reset by peer
15:49 (join) MayDaniel
15:50 eli: ChibaPet: It wasn't actually solved, but the main thing that bugged me went away.
15:50 eli: I'll post a reply in a few minutes...
15:50 ChibaPet: ok
15:55 eli: ChibaPet: Um...  what I apparently *meant* to say is that Carl would reply...
15:55 ChibaPet: Just read that. That works. I've seen Macs behave poorly with unusual monitor configurations in the past, so what he describes seems reasonable as an explanation.
15:56 eli: Yeah, that was our best guess, though it's still weird.
15:56 (join) jules2
15:57 eli: I also tried looking at various /var/log/stuffs but there was no error message or anything similar.
15:57 ChibaPet: When in doubt, blow away parts of your library, where those preferences are stored.
15:57 ChibaPet: (per-user library, of course)
15:58 eli: I was trying to find these too, but that didn't help either.
15:58 eli: BTW, where is that "Detect Displays" button supposed to be?
15:58 ChibaPet: Monitor config utility, as I remember.
15:58 ChibaPet: (I'm not in front of a Mac right now.)
16:00 eli: Where would I find that?
16:00 ChibaPet: Apple menu, system settings (forgetting actual name), display settings.
16:01 ChibaPet: It should be on the hardware row, if I remember correctly.
16:01 ChibaPet: For your in-home-directory prefs, I believe you can get a completely fresh set my moving aside your Library folder.
16:02 (quit) JuanDaugherty: Quit: JuanDaugherty
16:02 jules2: how would you write something like a CPS transform macro, that transforms a tree recursively and does different things depending on the kind of node it's at (a constant, a variable, a function application)?
16:02 eli: There's a "displays" the hardware row, but no detection button there.
16:02 ChibaPet: But then, I've never encountered precisely that issue, so I'm only guessing it's there. What I've encountered in the past involves windows living on no-longer-present displays.
16:02 ChibaPet: Checking, co-worker has a Mac.
16:04 eli: jules2: A macro should be able to do that fine...
16:04 jules2: can you do it with syntax-rules?
16:04 (quit) littlebobby: Quit: Ex-Chat
16:04 ChibaPet: Yeah, it ought to be there. System Settings -> Displays, and "Detect Displays" in the bottom right corner. There's also evidently a "detect displays" available in the "monitors" menu item you can add. Co-worker's Mac is running Lion, but this isn't a new thing.
16:05 eli: jules2: No, you'll need `syntax-case' for that.
16:07 eli: ChibaPet: I don't have that...  http://tmp.barzilay.org/x.png
16:07 stamourv: eli: If you're doing this on kauai/weatherwax, they're running pretty old versions of OS X.
16:08 stamourv: Apple doesn't ship updates for PPC machines anymore, so they're stuck with 10.4, or something.
16:08 eli: stamourv: But I suspect that something like detecting which monitor you have would have been pretty ancient.
16:08 ChibaPet: That's interesting. I suspect it might know you're on virtual hardware. Do you also not have the item in the menu bar that notes?
16:08 eli: It's 10.5.8
16:09 eli: ChibaPet: Yeah -- there's just a few resolution choices, and then "Number of Recent Items" and "Displays Preferences...".
16:10 stamourv: eli: If the lack of monitors is really a problem (i.e., it's not just that you don't like the desktop picture), there's a spare monitor lying around, and I can hook it up temporarily to either of the machines.
16:10 ChibaPet: Strange. If it were me I'd move aside my library and see if that helped, and if it didn't I'd chock it up to opaque proprietary badness.
16:10 stamourv: But I'm not leaving it there long term, my desk is already crowded.
16:12 (part) joshcough
16:13 eli: stamourv: I don't think it'd matter.  Just replied.
16:21 (quit) jeapostrophe: Quit: jeapostrophe
16:22 (quit) jules2: Ping timeout: 258 seconds
16:25 eli: ChibaPet: No, blowing out ~/Library didn't work...  Got me back to the annoying Aurora bg...
16:35 (join) tfb
16:44 (quit) DT``: Read error: Connection reset by peer
16:45 (quit) lucian: Ping timeout: 240 seconds
16:57 (join) RacketCommitBot
16:57 RacketCommitBot: [racket] plt pushed 5 new commits to master: https://github.com/plt/racket/compare/3f40742...63af93f
16:57 RacketCommitBot: [racket/master] Much improved `wrapped-output'. - Eli Barzilay
16:57 RacketCommitBot: [racket/master] Revert "Have DrDr keep ignoring the random seeds." - Eli Barzilay
16:57 RacketCommitBot: [racket/master] Fix `unstable/time' & xrepl use. - Eli Barzilay
16:57 (part) RacketCommitBot
17:09 (join) lucian
17:25 (join) DT``
17:26 (quit) lucian: Quit: Ex-Chat
17:30 (join) RacketCommitBot
17:30 RacketCommitBot: [racket] plt pushed 3 new commits to master: https://github.com/plt/racket/compare/63af93f...b972a09
17:30 RacketCommitBot: [racket/master] added a #:security-guard argument to: - Robby Findler
17:30 RacketCommitBot: [racket/master] add test case for PR 12084, clean up exception predicates in test suite a little - Robby Findler
17:30 RacketCommitBot: [racket/master] Added online expansion and compilation of files - Robby Findler
17:30 (part) RacketCommitBot
17:31 jonrafkind: yay parallel check syntax
17:44 (quit) kennyd: Ping timeout: 276 seconds
17:47 (quit) dherman: Quit: dherman
17:49 (join) kennyd
17:50 (join) dherman
17:50 (quit) dherman: Client Quit
17:51 (quit) MayDaniel: Read error: Connection reset by peer
18:00 (join) dherman
18:12 (quit) ChibaPet: Quit: Leaving.
18:34 (quit) dnolen: Ping timeout: 252 seconds
18:37 (part) drdo: "ERC Version 5.3 (IRC client for Emacs)"
18:37 (join) misterm
18:38 (quit) jonrafkind: Ping timeout: 260 seconds
18:44 (quit) tfb: Ping timeout: 276 seconds
19:16 (join) dnolen
19:17 (join) tfb
19:35 (join) ChibaPet
19:42 (join) JuanDaugherty
19:48 (quit) Daemmerung: Quit: Smoove out.
20:12 (quit) tfb: Quit: sleeping
20:14 (quit) dherman: Quit: dherman
20:22 ChibaPet: Is there any point to reporting a bug against the last release? My wife downloaded a nightly build on her Mac, and in the nightly build, she can paste pictures from HTDPv2 into DrRacket. With the release (MacOS Intel) the paste didn't work.
20:22 ChibaPet: (I've got her transforming animate values to get her rocket to move in different directions. It's fun watching her command of the variables developing.)
20:27 (join) sheikra
20:28 (join) sheikra__
20:32 ChibaPet: posted this to the list
20:36 (quit) JuanDaugherty: Remote host closed the connection
20:43 (quit) masm: Ping timeout: 276 seconds
20:44 (quit) misterm: Read error: No route to host
20:45 offby1: I'd report it
20:57 (join) kevin01123
20:59 (quit) kevin01123: Remote host closed the connection
21:23 eli: ChibaPet: Is that the last release, as in 5.1.1?
21:24 eli: Or the one that will be published very soon -- using the pre-release candidate builds?  (Currently 5.1.1.900)
21:24 sheikra: hey, there is a minor bug in http://docs.racket-lang.org/guide/Linking_Units.html
21:25 sheikra: There should be a line that says (require "toy-store-sig.rkt")
21:25 eli: sheikra: Where on the page?
21:26 sheikra: eli: beginning of the first program
21:26 sheikra: "store-specific-factory-unit.rkt"
21:26 sheikra: it only required one signature
21:27 sheikra: it needs both signature files
21:28 sheikra: The module system is very interesting. I'm trying to use it to solve some of the headaches in my design.
21:29 eli: You mean the *unit* system?
21:29 sheikra: eli: oh yes. unit system
21:29 eli: In any case, it sounds questionable, because it looks like the code is actually evaluated.
21:30 eli: But there are other weird things there -- the last example has "(list (toy )(toy ))"
21:30 sheikra: eli: really? I got an error when I load it
21:30 sheikra: oh I didn't notice that
21:31 sheikra: yep. that's something weird.
21:31 eli: It might be a typo -- that example might not be evaluated, or maybe evaluated but not displayed.
21:32 eli: The missing `require', that is.
21:32 eli: That last thing might be some problematic rendering of `toy' values and the constructor-style printouts.
21:32 eli: I'll forward it to FF.
21:36 eli: Ha, yeah, it's faking it -- it only evaluates the `define-unit' form.
21:36 sheikra: There seems to be a problem even if I add another (require ..) in the file
21:37 eli: Yeah?
21:37 sheikra: I had to require both signature files in the repl in order to make the last examples evaluate
21:37 sheikra: I'll have breakfast and try again..
21:38 eli: OK.  Do you want me to CC you so you can provide more details when you get there?
21:42 sheikra: sure. yinwang0@gmail.com
21:45 eli: sheikra: Um, are you not Eric Dobson?
21:45 eli: Ugh, I think I'm getting senile.
21:45 sheikra: eli: heheh :)
21:46 eli: Unmnemonic usernames confuse me.
21:47 eli: sheikra: Aha!  I'm not *that* senile.
21:48 eli: Eric uses `mithos28' in here, and his git username is `shekari'.
21:49 eli: sheikra: Even your avatar things look similar...
21:49 eli: (on github.)
21:50 sheikra: eli: really!
21:50 eli: He's has a star that I always thought is an eye...
21:50 eli: (of some animal)
21:50 sheikra: eli: The two signature files need to be required before the linking
21:51 sheikra: eli: what's his github user name? I'd like to have a look
21:52 eli: But one is required -- "toy-factory-sig.rkt" no?
21:52 sheikra: eli: oh, you just mentioned that. I'm getting senile this time
21:52 eli: sheikra: https://github.com/shekari
21:52 eli: Yeah...  It's infectious.
21:53 sheikra: eli: the signatures are never required at top-level
21:53 eli: Yeah, I just got there..
21:54 eli: And that code *is* evaluated -- but the previous things are all fakes, all of the relevant definitions are basically evaluated directly, which makes the examples run even with these problems.
21:55 sheikra: eli: the (list (toy) (toy)) is because build-toys in store-specific-factory-unit.rkt doesn't not specify a color
21:55 sheikra: eli: just change the line in build-toys to (make-toy store-color)
21:55 eli: I didn't actually follow the code, but the current printout is definitely bogus.
21:55 eli: The spacing is wrong -- that ")(" should never happen.
21:56 eli: sheikra: OK, sent.
21:56 sheikra: eli: thanks
21:58 sheikra: eli: I saw racket-llvm on shekari's repositories. Does racket use llvm?
22:00 eli: No, we had an llvm experiment a good while ago, but it wasn't successful and eventually we went with gnu lightening.
22:00 eli: But that code looks like some different llvm interface.
22:01 (quit) sheikra__: Quit: Leaving
22:02 (join) RacketCommitBot
22:02 RacketCommitBot: [racket] plt pushed 4 new commits to master: https://github.com/plt/racket/compare/b972a09...37c43c2
22:02 RacketCommitBot: [racket/master] fix `namespace-attach-module{-declaration}' problem - Matthew Flatt
22:02 RacketCommitBot: [racket/master] fix docs - Matthew Flatt
22:02 RacketCommitBot: [racket/master] doc typo - Matthew Flatt
22:02 (part) RacketCommitBot
22:03 sheikra: got it. why would you need gnu lightening or llvm? schemers are not supposed depend on those things ;-)
22:03 sheikra: oh, let me try to my understanding of unit system and see if there are bugs in my mind...
22:04 sheikra: try to explain...
22:05 sheikra: ok, basically, the (import ...) and (export ...) pair is a signature of a function. So (import toy-store^)
22:05 sheikra: (export toy-factory^) is something similar to (toy-store^ -> toy-factory^), but at a module level ...
22:05 eli: sheikra: Strictly speaking you don't need them -- but in practice it's easier to use an established project to generate machine code rather than doing it from scratch.
22:07 sheikra: I wonder if I can understand it that way
22:20 sheikra: eli: typed racket seems to have a really powerful type system. are you familiar with that?
22:29 ChibaPet: Hi, sorry, was AFK. Yes, 5.1.1.
22:50 (join) dherman
22:53 eli: ChibaPet: In that case there's not much to worry about, since 5.1.2 will come out very soon.
22:53 eli: As in "a few hours away", provided that I don't crash on the way.
22:54 eli: sheikra: Yes, I know much more about the type system than the unit one.
22:55 ChibaPet: heh
22:56 ChibaPet: I hadn't realized it was a point release. Makes sense for that to be the correct fix then.
22:57 eli: Well, we increment the third number for most "normal" releases, and the second for major changes -- like the gui rewrite that took us from 5.0.2 to 5.1, so that's not specific for these kind of bugs.
23:02 ChibaPet: It gives the impression that there's not support for the older third-number release, with a suggestion to upgrade.
23:02 ChibaPet: I tend to think of, for instance, x.0.1 to x.0.2 to mean "absolutely safe" generally, as most people do it.
23:07 (join) jeapostrophe
23:10 (quit) jeapostrophe: Client Quit
23:13 (quit) jao: Ping timeout: 252 seconds
23:13 ChibaPet: g'night all
23:13 (quit) ChibaPet: Quit: Leaving.
23:26 (join) RacketCommitBot
23:26 RacketCommitBot: [racket] plt pushed 4 new commits to release: https://github.com/plt/racket/compare/f4cab3e...e75fc54
23:26 RacketCommitBot: [racket/release] Updated DrRacket images - Eli Barzilay
23:26 RacketCommitBot: [racket/release] Fix `unstable/time' & xrepl use. - Eli Barzilay
23:26 RacketCommitBot: [racket/release] Update version number for the v5.1.2 release - Eli Barzilay
23:26 (part) RacketCommitBot
23:36 (quit) sheikra: Read error: Connection reset by peer
23:37 (join) sheikra
23:38 (join) jonrafkind
23:57 sheikra: eli: cool! I sent a detailed fix to the documentation for units. I'll talk to you later about types when I get back (if you didn't crash :-))