freenode/#lisp - IRC Chatlog

6:10:18 kamen beach: Yup, I'm new. Been hanging around some irc channels (emacs, etc.), but hadn't dared to venture in #lisp yet. No idea why.

6:16:42 pyc Nilby: yesterday, you said, "Everything that still works, works because they don't depend on external conditions. But for most useful things, the external conditions have changed enough that they don't run."

6:16:49 pyc Nilby: What does "external conditions" mean?

6:17:59 pyc kamen: how did you finally decide to join #lisp? for me the journey was opposite. I first decided to learn Lisp and that led me to Emacs.

6:22:53 Nilby pyc: Things like databases, web servers, some O/S things like file locations, and I was running on Solaris OS on 32 bit sparc hardware at the time.

6:23:29 Nilby But some things I still use to this day.

6:24:42 pyc Nilby: okay. I am still learning CL so not familiar with the term "external condition". Does "external condition" mean external package?

6:26:13 Nilby pyc: No. I just meant things like the O/S, network, and external programs, like database and web servers,

6:26:41 pyc Nilby: what does "condition" mean in that term?

6:26:55 Nilby pyc: The packages I used, like e.g. CLSQL still work the same.

6:27:47 Nilby Not Lisp conditions, just the software external to Lisp.

6:27:55 pyc Nilby: oh!

6:28:26 pyc Nilby: so if I understood you right, things like CLSQL still work the same. But things like O/S, network, etc. that you depended on in Solaris OS days, don't work the same and you had to rewrite some of that code?

6:31:58 Nilby Yes. Like for example, migrating from an Oracle database to Postgresql, and from Apache web server to hunchentoot or nginx

6:33:23 pyc Nilby: thanks

6:33:36 beach kamen: Great! Welcome!

6:34:36 beach kamen: What made you finally decide to join?

6:35:00 pyc Nilby: Can one create a web-based application in Common Lisp directly using CLHS functions and Nginx or do you need to use Hunchentoot for it?

6:35:20 Nilby pyc: In general, the CL code I've written is more portable, than any other language I've ever used.

6:36:24 pyc Nilby: that is great to know. what other languages are in your mind when you say that CL code is more portable than them? Python, Java? or something else?

6:37:41 Nilby pyc: Back then was before hunchentoot existed so I just used sockets code to talk to mod_lisp in an apache server. If the whole system was never touched, it could still run today, but people like to upgrade things

6:38:33 pyc Nilby: how about now? can I write a Common Lisp web application without Hunchentoot and serve it over the Internet via Nginx or something like Hunchentoot is necessary?

6:42:23 Nilby Yes. I mean you can make a HTTP 1 server from a few lines of CL code using implementation specific (or OS specific) socket code, but things like hunchentoot make things easier.

6:43:20 pyc okay, so I will learn hunchentoot once I become more comfortable with CL

6:44:43 pyc ah! looks like I have done a hello, world with Hunchentoot already. the Emacs4CL script I used to set up my Emacs introduced me to Hunchentoot in its docs. it had a Hunchentoot example to demo quicklisp: https://github.com/susam/emacs4cl#use-quicklisp

6:45:07 pyc Nilby: can you take a look at my other question above: what other languages are in your mind when you say that CL code is more portable than them?

6:45:22 pyc Nilby: just curious to know CL is more portable in comparison to what else?

6:46:06 Nilby The application I'm thinking of was in C++ before I converted it to Lisp. CLSQL gave me the cool ability to write mostly DB independent SQL, as well as immediately work on different DBs, but more importantly I could write SQL, HTML, and application logic using the same Lisp syntax.

6:46:48 Nilby C++ has more libraries and features now, but back then it was impossible.

6:51:34 kamen beach: Oh, nothing special. Since I had resurrected ERC, I thought I might as well hang around, lurk and read/listen to lispers. Also, I've stopped dabbling in CL and started taking it a bit seriously, so, as "ongoing education", it's a good idea to be here.

6:52:56 beach Sounds good. Don't hesitate to ask questions.

6:53:37 aggin is there a :folder like there is a :file in asdf:defsystem :components ?

6:53:48 aggin I want asdf to take files from a certain directory

6:54:04 aggin without defining a new module

7:01:00 pyc what does the term "sharplisper" mean? Found it in the channel logs.

7:01:28 beach It's a GitHub repository.

7:01:51 beach We gather abandoned code there and try to maintain it a bit.

7:02:21 beach I didn't create it, but I am convinced it's a pun, like #lisp-ers.

7:05:11 jackdaniel a sharp lisper is a lisper who is sharp

7:05:29 Nilby edgey sharplispers keeping those old blades from rusting :)

7:06:24 beach jackdaniel: As opposed to "flat"?

7:07:12 jackdaniel beach: I thought of "dull" or "blunt"

7:07:20 beach Just checking.

7:07:39 edgar-rft let's write lisp in a minor

7:08:24 jackdaniel (but a music reference is a nice touch too)

7:10:00 jackdaniel that said, I'm to make an owl from a play dough, see you \o

7:20:57 pyc what am I doing wrong in this statement: (uiop:with-current-directory "/tmp/" (uiop:getcwd)). I get: Error while parsing arguments to DEFMACRO UIOP/FILESYSTEM:WITH-CURRENT-DIRECTORY:

7:25:55 Nilby pyc: When a macro has an arg in parens, it has to be in parens when you use it. So in your case it's (with-current-directory ("/tmp/") ...)

7:28:04 beach pyc: Common Lisp has "statements" only inside TAGBODY. In general, an expression meant to be evaluated is called a "form".

7:29:36 pyc thanks Nilby beach

7:44:15 ldbeth good afternoon

7:46:14 beach Hello ldbeth.

7:47:09 ldbeth just come up to head, but are there any style guides mention when to use `&aux'?

7:48:41 beach Most people say "never", but I found a use for it, as described in my paper on method combinations.

7:50:18 ldbeth got it

7:51:22 jackdaniel it is useful when you want to introduce a binding that wraps whole body (i.e you want to rebind an argument)

7:51:53 jackdaniel (defun foo (string &aux (string (string string))) string)

7:52:01 jackdaniel there is only so many strings you can put in a string

7:53:10 beach That's beautiful.

7:53:46 ldbeth ah, it's some kind of coercesion

7:53:55 beach And that's precisely the use in the paper I mentioned. It needs to wrap an entire body of externally-supplied code.

7:54:38 jackdaniel ah, happy coincidence

7:55:49 Nilby please LET &aux die

7:56:18 beach Nilby: Then I couldn't implement my technique for method combinations.

7:57:30 Nilby I have total faith that you could

7:58:16 beach I don't think it has any great use in a function that is entirely written by the programmer. But in this case, I needed it because the body of the code could be arbitrary, and I had to wrap it as jackdaniel said.

7:58:43 beach http://metamodular.com/SICL/representing-method-combinations.pdf

8:11:15 Nilby someday I'll go back in time and get rid of &aux and nobody will miss it

8:13:35 beach Nilby: If so, I challenge you to come up with an alternative solution to the problem in that paper.

8:14:14 beach Nilby: It could be your ELS submission this year: "Method combinations without &aux".

8:15:39 Carkh aggin: maybe the :pathname key in a system definition does what you want ?

8:15:51 aggin ok thanks

9:27:21 scymtym beach: another use of &AUX is computing the value of a structure slot depending on constructor arguments: (defstruct (foo (:constructor make-foo (bar &aux (baz (1+ bar))))) bar baz)

9:29:42 beach Interesting. I didn't catch that one because I never think about using structures.

9:33:05 dim I must admit I like using structures a lot, and that might be because when using CCL it's possible to change their definitions online, much like classes

9:33:47 beach But then your code is implementation specific.

9:36:22 Carkh isn't there efficiency considerations to take into account ?

9:36:34 Carkh like generic accessors

9:36:52 Carkh vs static

9:37:47 pyc Is it true that 'loop' is discouraged and something like 'mapcar' is encouraged to adhere to more functional programming style?

9:38:02 beach Carkh: It depends on the generic dispatch algorithm that your implementation is using. But, yes, with current implementations, it is usually agreed that structures are faster than standard classes/objects.

9:38:12 beach pyc: No.

9:38:15 Carkh beach : thanks

9:38:19 pyc beach: thanks

9:38:20 gj hello!

9:38:25 beach Hello gj.

9:39:14 gj I am here again because I am still a little bit confused about using packages in a lisp program.

9:39:27 beach pyc: Common Lisp is a multi-paradigm language. You will find the functional style mostly in macro expanders. Of course, someone who wants to program in Common Lisp using a functional style may want to avoid imperative constructs.

9:40:20 beach Carkh: Now, if current implementations used my algorithm for generic dispatch and my technique for call-site optimization, the conclusions may be different. :)

9:40:23 gj Things like asdf and quicklisp seem to only work in a REPL? how does it work if I want to do things outside of a REPL?

9:40:35 gj sorry if there is a really obvious answer to this.

9:41:05 beach gj: You use those tools to build your program. And you would not usually build your program at run time.

9:41:15 Carkh beach : i was under the impression they were doing call site optimization ? or was it java ?

9:41:54 beach Carkh: Nothing like my draft paper suggests: http://metamodular.com/SICL/call-site-optimization.pdf

9:42:34 beach Carkh: Some very special cases are optimized in high-performance implementations, but nothing this general.

9:42:54 Carkh oh i had this paper in the queue

9:43:05 beach Wow! :)

9:43:18 Carkh probably grabbed it from here =)

9:43:35 beach Very likely.

9:43:40 Carkh yes you talked about it !

9:44:40 heisig We should start handing out DOIs for things posted on #lisp :)

9:45:29 beach I never did figure out who has the authority to create those.

10:40:17 pyc Found this post on Reddit: https://www.reddit.com/r/spxy/comments/luiq8k - "makes me wonder how it is webdev with Lisp. I know some old stories, like reddit itself"

10:40:24 pyc Was Reddit really written in Lisp? Which Lisp?

10:41:20 jackdaniel its initial version was, then it was rewritten in python

10:41:42 beach Oh, wow! Why would anyone want to do that?

10:41:55 jackdaniel here is an explanation: http://www.aaronsw.com/weblog/rewritingreddit

10:41:59 jackdaniel it has some merit to it

10:42:19 Gnuxie[m] The original is also on github now

10:42:21 jackdaniel mainly (afair): ecosystem

10:43:16 jackdaniel but I'm reading this right now: https://funcall.blogspot.com/2009/03/not-lisp-again.html

10:43:23 jackdaniel and it is far more encouraging :)

10:45:01 flip214 > The Python version had less code that ran faster

10:45:04 beach I need to put the first one on the queue. I can't figure out how Python would be faster. Maybe they were using CLISP initially?

10:45:14 flip214 hmmm, has that ever been the case?

10:46:01 jackdaniel maybe the lisp code was badly written code?

10:46:34 jackdaniel I mean - experienced python programmer can write faster code than inexperienced lisp programmer who takes into the heart, that "everything is a list"

10:46:39 kamen Ah, the reddit from-lisp-to-python case. What a shitstorm that ended up as.

10:49:00 flip214 jackdaniel: the article says "knows about lisp", though

10:49:12 flip214 well, never mind.

10:49:21 jackdaniel or: overly ambitious lisp programmer tries to do the right thing, while not exactly the right things gives faster code

10:49:33 jackdaniel (aka worse is better trivia ,)

10:49:37 Gnuxie[m] https://github.com/tamurashingo/reddit1.0

10:50:20 flip214 Gnuxie[m]: any hints what lisp was used?

10:50:40 jackdaniel reddit.asd seems to hint common lisp

10:50:43 jackdaniel same the readme

10:50:50 Nilby spez rewrite digg in lisp .. PR0F1T! .. me rewrite unix in lisp ... D3d x_x

10:50:55 jackdaniel ah, what impementation

10:53:37 Nilby (defparameter *secret* "blargo")

11:00:35 pyc flip214: which article says, "knows about lisp"?

11:05:11 flip214 pyc: the rewritingreddit one. > the others knew Lisp

12:11:26 Carkh if i set the fill-pointer of a vector to a lower value, should i first fill the upper array cells with nil in order to help the garbage collector ?

12:12:03 beach You might want to do that, yes.

12:12:11 Carkh sounds like this might be implementation dependant

12:12:13 beach Because there are operations that ignore the fill pointer.

12:12:23 beach No, I don't think so.

12:12:28 Carkh allright makes sense, thanks

12:12:41 beach AREF ignores the fill pointer, so the objects beyond the fill pointer can't be reclaimed.

12:12:59 Carkh i didn't know that, i was expecting an error

12:13:32 beach clhs aref

12:13:33 specbot http://www.lispworks.com/reference/HyperSpec/Body/f_aref.htm

12:14:04 Carkh ahh there it is =) thanks

12:16:36 Carkh in other news, after so many years doing clojure, i'm having the hardest time coming up with imperative algorithms ><

12:17:05 Carkh feels like it's even harder than the adaption i had to undergo the other way back then

12:17:13 Carkh adaptation*

12:17:58 beach Carkh: Try something with a huge graph, like adding an edge. If you do that functionally, it might very well be unbearably slow.

12:18:32 Carkh it's slower, but the trickj is to flatten the data structur

12:18:33 Carkh e

12:18:57 Carkh but yeah, there's a reason i'm back to CL

12:19:13 beach Sure, if we don't care about performance, anything can be done functionally.

12:20:06 Carkh though CL is not always that friendly towards non-consing imperative stuff... sometimes feels like fighting it

12:20:49 beach Oh?

12:21:05 beach Ah, yes, sorry. Misread. Sure.

12:21:44 beach The, consing is not a problem in general. Only in Common Lisp implementations with bad garbage collectors.

12:23:57 Carkh i'd say yes and no

12:24:41 beach That remark kind of requires some elaboration.

12:24:51 Carkh wasn't it ITA or whatever is the name of this trip calculating application that had to keep its data in a C application ?

12:25:23 beach Could be. Why was that?

12:25:37 Carkh ok so here it is : i'm trying to make a game, 60fps not required, and i'm glad it isn't because i don't think i could maintain it with CL

12:26:26 Carkh sure a small toy thing will work nicely

12:26:43 Carkh but even for the small thing you have to be VERY carefull about consing

12:27:00 beach Why is that?

12:27:10 Carkh frame drops

12:27:24 beach Because of the GC?

12:27:44 Carkh that's my guess anyways

12:27:52 beach That just confirms what I said. It is a problem only with Common Lisp implementations with bad garbage collectors.

12:28:03 Carkh sbcl has a bad garbage collector ?

12:28:08 beach Absolutely.

12:28:15 Carkh which one is good ?

12:28:31 beach SBCL started life many decades ago, and we didn't know then what we do now.

12:28:47 Carkh i mean, i'm pretty sure SBCL has good throughput

12:28:59 Carkh but this isn't especially suited to games i guess

12:29:10 beach Pretty much all Common Lisp implementation currently used, at least the free ones, started life a very long time ago, and nobody has implemented a good concurrent and parallel GC for them.

12:30:01 beach Last time I looked, SBCL had a stop-the-world GC.

12:30:24 Carkh will you dive into concurrent GC for SICL ?

12:30:34 beach That's the plan, yes.

12:30:49 Carkh wew dear i'm glad you're doing it rather than me =)

12:30:58 beach See the relevant chapter in http://metamodular.com/SICL/sicl-specification.pdf

12:31:24 beach It hasn't been tested, but the basic design is based on fairly recent research.

12:32:17 Carkh guess i'll have to come back in about 10 years =)

12:32:49 beach I totally hope it won't take us that long to get SICL going.

12:34:24 beach The applications I had in mind for this design are "analysis, processing, and synthesis of sound"

12:34:40 beach And sound is even more sensitive to pauses than video.

12:34:44 Carkh yes that's soft realtime right thre

12:34:54 beach Exactly.

12:36:02 Carkh though with sound you can schedule stuff a bit in abvance depending on buffer size

12:36:49 Carkh do you intend to treat windows as first class environment with SICL ?

12:37:29 beach You have to make sure that the application responds adequately fast to the user pressing some key on the MIDI keyboard.

12:37:44 beach "windows"?

12:38:17 Carkh microsoft windows... I notice it's nothing but an afterthought for most implementations

12:38:53 beach I have no time or energy for Microsoft products, and very little for Unix as well, so I think the answer is "no".

12:39:12 beach But I am sure there will be someone else doing it.

12:39:34 beach Just like someone will probably make SICL support CFFI or something like that.

12:39:36 Carkh that's something i'll never understand, the users are on windows, most of them anyways

12:39:54 beach I am a researcher. I don't care if I am the only user of SICL.

12:40:18 Carkh no common lisp advocacy on your part then ?

12:40:46 beach I gave up trying to convince people of just about anything many decades ago.

12:41:09 beach I do my job and try to show good results. If people don't care, then that's their problem.

12:41:36 Carkh that was a loaded question anyway

12:41:52 Carkh i mean there are advantages to having a larger community

12:42:07 Carkh maybe more hands to help you with SICL for instance

12:42:43 beach Yes, but the kind of people that would be attracted by "advocacy" are not the kind of people I would like help from.

12:42:55 jackdaniel beach succesfully attracted quite a few people without marketing deparatment :)

12:43:04 beach There is already a handful of people doing real good work for me. That's enough.

12:43:29 beach Carkh: What jackdaniel said.

12:43:34 Carkh that's only an example, there is also more brains to come up with ideas

12:43:50 jackdaniel another thing is that effort of managing low quality contributions sometimes is bigger than not accepting them

12:44:07 jackdaniel (or persuading)

12:44:17 beach I seem to come up with more ideas per time unit than I have time to implement. So the backlog is already considerable.

12:44:23 jackdaniel heck, even a consensus with knowledgeble people is hard in quantity of 2-3 devs!

12:44:29 Carkh sur the newly "advocated" person is useless, but he/she might grow

12:44:48 Carkh sure*

12:45:10 Carkh community gardening =)

12:45:14 jackdaniel might or might not, the free time you put in growing someone you don't know doesn't grow on trees

12:45:37 jackdaniel well, I'm sure that there are positive examples, but there is some merit in not making a community bubble :)

12:45:38 beach Carkh: It is part of my job to help people make progress, but I don't do it hoping that they will contribute to SICL.

12:46:21 Carkh i shouldn't have mentioned SICL, it was just an example

12:47:33 beach Carkh: For the past 40 years or so, I have "attracted" people by showing what I can do, without talking much about it, and certainly without trying to "market" it. I think I have a good track record of attracting people that have understood the point of what I do.

12:48:41 Carkh so you're saying there's no value in advocacy

12:49:29 jackdaniel I think that you extrapolate too much

12:49:32 Carkh i disagree on that, but to each their own

12:49:36 beach Almost. It is certainly not worth *my* time, but that can be because I am not very good at it.

12:50:18 jackdaniel i.e fact that I'm not going to the cinema does not equal to saying that I don't see value in going to the cinema

12:50:29 Carkh ok i understand

12:50:59 beach Carkh: I also think it is risky. I frequently see people asking questions here in order to "sell" Common Lisp to their manager, but those people would have no clue whatsoever when they then are asked to handle problems with their proposed solutions.

12:51:35 beach ... and such a situation would be very bad for the image of Common Lisp indeed.

12:51:55 beach There are exceptions, of course. Like flip214, who knows what he is talking about.

12:52:00 Carkh again thinking about clojure, rich hickey spent a lot of time doing advocacy, and this resulted in a vibrant community and some tools and solutions that are objectively the best for their purpose (that i know of)

12:52:17 beach Good for him.

12:52:45 aggin anyone familiar with the maintainer of https://github.com/naryl/cl-tui/commits/master ?

12:52:50 beach I would be OK with Kent Pitman doing some advocacy for Common Lisp.

12:52:58 aggin * https://github.com/naryl/cl-tui

12:54:11 Carkh he's more like the historian of CL =)

12:56:10 Carkh anyways back to work, thanks for the chat

12:56:16 ldbeth I think I see "MIDI" in the conversation

12:56:21 beach Carkh: Also, this is probably not a good time to push Common Lisp to a wider audience. Our development tools are not that great, and as you pointed out, the garbage collectors of existing implementations are not that great either. We are working on improving it though.

12:56:50 ldbeth Is there any material I can reference for how to do "real time" signal processing in CL?

12:57:16 jackdaniel n.b abcl relies on jvm gc, so it is rather good given how much effort has been put into that technology

12:57:23 Carkh beach: i sure feels like there is more activity these days than 10 years ago

12:57:29 beach jackdaniel: Good point.

12:57:30 Carkh it*

12:58:30 Carkh though it still firmly smells funny =)

12:59:12 ldbeth to be more specific, I'd like to know how to keep video and sound in sync, but unfortunatly I have no experience in game programming etc

13:00:49 Carkh i'm yet to tackle the sound issue, looks like there are a few libs to be found with borodust

13:01:02 Carkh calling into C libraries

13:03:53 beach ldbeth: I know of no such material. Even with more traditional languages, the OS is going to get in your way a lot. I can't even get a real-time synthesizer program to give adequate response times on Linux. I am told I need to recompile the kernel for that.

13:04:01 Carkh namely bodge-openal

13:04:44 Carkh bodge-sndfile and its sndfile-blob

13:04:46 Lycurgus yeah that's standard

13:05:00 Lycurgus the normal linux kernel isn't real time

13:05:16 Lycurgus the build is pretty straightforward though

13:05:41 beach So why is it not real time by default?

13:06:01 Lycurgus not the default use case?

13:06:05 Carkh probably because there is a tension between throughput and realtimeness

13:06:25 Lycurgus also it's pretty easy on most distros

13:06:27 ldbeth Carkh: ah, so I accept that it is not easy to accomplish this with trivial programs.

13:06:28 beach Lycurgus: What use case would be disadvantaged if it were real time by default?

13:06:37 Lycurgus debian derived ones anyway

13:07:19 Lycurgus all my hosts in this LAN for example and this one I'm typing on where I expect interactive resposivity

13:07:44 beach I think I need a competent system administrator.

13:08:06 Lycurgus also switching will be easy from grub

13:08:16 beach I can't even get an adequate desktop environment without specific adaptations.

13:08:35 Lycurgus but this host hasn't been rebooted in a couple months

13:09:04 Lycurgus did have to restart X/vnc a few days ago loosing my desktop

13:10:08 beach I think I could convince someone to be my system administrator, but I have no idea how that would work remotely.

13:10:11 Lycurgus *responsivity

13:10:23 Lycurgus i'd do it on retainer :)

13:10:44 beach What does that mean?

13:11:17 Lycurgus what does retainer mean?

13:11:56 beach You mean you would get paid in advance?

13:11:59 beach I could do that.

13:12:23 beach But I still don't see how it solves the remote problem.

13:12:25 Lycurgus k, juan@acm.org ; ( to get back on topic )

13:14:05 beach I guess I'll wait until I have replaced my desktop. It is time for that.

13:14:19 beach My desktop computer I mean.

13:14:32 Lycurgus remote isn't a reall issue these days; was gonna sey, dedicate a box to rt if it's important

13:15:11 beach That's an idea.

13:17:36 Lycurgus in terms of defaultness order: server, desktop, rt and other specialty

13:21:04 beach Thanks for the advice. When I have time, I'll give it some more thought. It is a very attractive idea.

13:21:23 Lycurgus np

15:27:09 pyc Given a relative path "foo" what is the best way to convert it to an absolute pathname, example: #p"/home/pyc/project/foo/" ?

15:28:52 jackdaniel (truename "foo")

15:28:56 jackdaniel if the file exists

15:29:17 pyc thanks jackdaniel

15:29:28 jackdaniel or (make-pathname :name "foo" :defaults *default-pathname-defaults*) if it doesn't

15:30:01 jackdaniel (given that your default pathname is /home/pyc/project/)

15:32:36 pyc jackdaniel: yes, I will use the second option because it evaluates what I need when the directory name is a symbolic link to another path.

15:33:52 pyc just to ensure that I am not falling in the trap of X-Y problem, here is my problem X: given a relative source path "foo" and relative target path "bar", I want to copy all files and folders recursively from "foo" and copy them to "bar".

15:34:26 pyc so when I start walking through "foo", I get absolute paths for all files and directories found recursively, example, "/home/pyc/project/foo/dir1/file1.txt" and so on.

15:35:21 pyc so what I am going to do is resolve "foo/" to "/home/pyc/project/foo/" (the prefix), remove that from the source file path to get "dir1/file1.txt" and then append that to "bar" to get the destination path, that is, "bar/dir1/file1.txt"

15:36:02 Xach pyc: enough-namestring may help there

15:46:49 pyc What am I doing wrong here: (enough-namestring "/home/pyc/foo/bar" :defaults "/home/pyc/foo/"). Error: invalid number of arguments: 3

15:48:10 Bike enough-namestring takes an optional parameter for the defaults, not a keyword parameter

15:51:39 pyc oops! thanks for the help Bike! I am still misinterpreting the docs even though Practical Common Lisp has taught me the difference between keyword and optional.

16:06:29 aeth_ ** NICK aeth

16:13:17 semz I'm starting to think ITERATE is largely pointless. It's idiosyncratic with lots of gotchas, hard to extend (beats impossible I guess, but still), and purging it from my code showed that there were maybe two features I used that didn't have a direct LOOP equivalent.

16:13:55 beach Yeah, LOOP works for most cases.

16:14:32 lisp-machine hehe i've been trying to use map(c)(ar) exclusively lately as a challenge

16:14:36 lisp-machine loopless code

16:14:56 beach I am sorry to hear that. :)

16:15:42 beach I find mapping works fine when I have the function already made, but not so practical if I have to make up an anonymous function just for the call.

16:18:52 beach Also, if you want to collect resulting elements conditionally, mapping doesn't work.

16:19:22 beach You get some twisted code with mapcan and listifying intermediate results.