libera/#commonlisp - IRC Chatlog

6:47:40 Th30n` ** NICK Th30n

9:28:26 semarie ** NICK Guest7362

9:28:26 semarie_ ** NICK semarie

12:32:54 dbotton dlowe sorry, yesterday was busy day. So re-reading. "for unimported symbols, you have to specify :: instead of :" that is a violation of interface/protocol or perhaps internal "private" use of a symbol.

12:35:34 dbotton In the first attempts to demonstrate Ada's new 'class wide types and tagged-records (it's oo system) which like CL is "different" from smalltalk/java/c++ in that neither provides encapsulation, there was an attempt to teach using packages to provide the encapsulation of objects

12:36:55 dbotton granted packages in the two language are not exactly the same

12:37:54 dbotton It seems that the packages can (should?) have some place in OO on CL

12:38:46 beach They do.

12:38:47 dbotton but my initial question if anyone wrote any sort of style guide at all.

12:39:08 dbotton beach, I certainly think so

12:40:19 dbotton package exports seem outside of generic functions the some total (ok comments also) of CL's ability to write up a protocol/interface

12:41:28 beach I don't know what that means. Also I don't understand the grammar of your phrase.

12:42:28 beach Of course you can determine a protocol in Common Lisp. You just list the (often generic) functions and the types (often classes) they apply to, and the semantics of their use.

12:42:29 dbotton I think I need to better write up things then first, and I know still will lack clarity but at least can demonstrate practically my intentions

12:43:44 dbotton That is not enough for the protocol, you also have to export

12:43:52 beach dbotton: I usually turn my code into a collection of "modules" where each module consists of an ASDF system definition, one or more package definitions, and the code that implements the module protocol(s).

12:44:21 beach dbotton: The names of the functions and classes are symbols that are exported from some package.

12:45:01 beach dbotton: So when you write down those names, you also mention the package of the symbols in question.

12:45:29 dbotton I don't like leaving so much to comments

12:45:37 dbotton but seems no choice

12:45:49 beach Oh, so the complaint is that Common Lisp does not have first-class protocols?

12:46:07 dbotton That has always been an issue for me

12:46:52 dbotton No encapsulation the second for classes second

12:47:01 beach I see. Well, as I often point out, if Common Lisp were to include the preferences of everyone, then the standard would be closer to 10000 pages than to 1000 pages, and people already complain that it is too large.

12:47:10 dbotton right

12:47:31 dbotton What I am looking for is some best practice docs

12:47:39 dnhester26 Hi, a macro question: I have a loop which collects a list of forms, but it returns a list of forms so I get ((form1) (form2)) which gives me an error. How can I make it so that I get (form1)(form2)? Doing `,@(loop) didn't work for me... I could make some random `(let ((x 1)),@(loop)) but that's just a workaround...

12:47:45 beach Having classes do encapsulation is probably the reason why there are so many growing complaints about traditional object-oriented programming.

12:47:48 dbotton Which that site from dlowe sees to be working on

12:47:54 bike dnhester26: `(progn ,@(loop ...))

12:48:15 dnhester26 bike: ok thanks... is that normal or a workaround?

12:48:16 selwon there is at least one common lisp style guide

12:48:25 dbotton beach, agreed but packages can create an interesting encapsulation light

12:48:25 bike dnhester26: perfectly normal

12:48:42 dnhester26 bike: ok, thanks!

12:48:53 beach dbotton: But, that's precisely what we do.

12:48:54 dbotton selwon none of those guides deals with using packages + classes

12:49:05 dbotton which gets sticky

12:49:34 beach dbotton: And if you follow the CLIM II style, then the package that exports the symbols of the protocol contain only those symbols.

12:50:01 dbotton Is there a document on the CLIM II style?

12:50:25 beach There is the CLIM II specification. But it is mostly about the functionality of CLIM II.

12:50:25 dbotton or best example in it to look at?

12:50:53 dbotton Right, there is little to no docs on "how" or "tutorial like"

12:51:23 yitzi I think it is open for debate whether the "classes" or the generics are the protocol.

12:51:24 dbotton People coming to CL have to read CLIM code to best practice how to use CL is bad policty

12:51:45 dbotton yitzi GF for sure are imo

12:52:09 beach dbotton: Yes, there are too few people using Common Lisp. Do you have suggestions on how to fix that?

12:52:15 dbotton My issue is this is a very poorly documented area of CL

12:52:28 beach Both types (often classes) and functions (often generic) are part of a protocol.

12:52:55 dbotton beach I am working on it :) I am seeing new people all the time with CLOG and want to pull more

12:53:01 beach dbotton: Again, yes, there are too few people using Common Lisp. Do you have suggestions on how to fix that? Or are you saying we should dissuade people from using Common Lisp?

12:53:26 dbotton I am saying I want to make it easier for people to use CL

12:53:38 dbotton and that means showing the how to use it

12:53:38 beach Great!

12:54:03 beach dbotton: I will be happy to read and comment on your style guide.

12:54:03 dbotton So I write code and docs

12:54:45 dbotton What I am looking for is a picture of what you are doing and share your protocols

12:55:06 dbotton what you are calling a protocol and what would and would not be part of it

12:55:07 dbotton etc

12:55:44 beach http://metamodular.com/protocol.pdf

12:55:45 dbotton As of now the only towards that is code snippets

12:56:45 dbotton will read through - and is there a step by step how to set up a software project so that it revolves around the protocols?

12:57:40 beach dbotton: Like I said, I usually turn my code into a collection of "modules" where each module consists of an ASDF system definition, one or more package definitions, and the code that implements the module protocol(s). Each module lives in an OS directory.

12:58:25 dbotton can you send me a link to the most representative of that, so perhaps I can use that as part of a tutorial

12:58:46 beach dbotton: So the step by step is: 1. Create the directory. 2. Create the ASDF system definition. 3. Create the package(s) file(s). 4. Write the code.

12:59:47 beach dbotton: This one: https://github.com/robert-strandh/Cluffer comes close.

13:00:21 beach It does not use the CLIM II style, but maybe I'll fix that at some point.

13:01:02 dbotton is there a link tot he CLIM II style?

13:01:32 beach http://bauhh.dyndns.org:8000/clim-spec/index.html

13:02:11 beach Specifically, I was referring to: http://bauhh.dyndns.org:8000/clim-spec/2-2.html#_8

13:03:04 beach "CLIM is the package where the symbols specified in this specification live. It contains only exported symbols and is locked in those implementations that allow package locking."

13:03:48 dbotton we have a link to see it in code?

13:04:02 beach *sigh*

13:04:43 dbotton (sorry for being frustrating, but I see this undocumented area of CL being a frequent wall to people new to CL)

13:04:44 beach https://codeberg.org/McCLIM/McCLIM

13:06:08 beach dbotton: The frustration is because all the things I have showed you are discussed here frequently, so I am a bit surprised you haven't heard of them.

13:07:33 dbotton The issue is that the parts are not a practical guide and the theory is what is discussed

13:08:29 dbotton A non-cs programmer is not going to glean from what is said or pointed out (even today) a practical system to use

13:09:06 dbotton I am being frustrating today so that I can make sure I have enough to write it myself :)

13:10:12 dbotton At this point I can get just about anyone with basic skills quickly capable of using and reading CL code

13:10:54 dbotton what I need to do is get them to a point they can design a system and then program it in CL

13:11:39 dbotton one that does not turn out like the majority of what I see, tons of random bits thrown in a package or two

13:12:04 dbotton just "collections" of utilities

13:14:37 dbotton CLOG has now a solid IDE and a solid GUI builder, so so templates, I need a solid set of guide lines and tools to keep a project organized and reusable

14:00:04 dnhester26 a question about restarts and conditions: My use case is like this: I would like to check some input data, if it has a problem signal a condition. Now on the handler, when receiving that condition, I would like to check the input data programmatically and see if I can change it in a way that solves the problem, and try again. If the data can't be changed, then do whatever the handler-case would do. is this possible? When getting to a

14:00:04 dnhester26 handler-case, change the data, and try running the same form again with the new data instead? Is that a restart? In the spec it refers to interactive which I understood to mean that a user is interacting with the REPL, though I may have mistunderstood

14:02:26 bike that's a restart

14:02:36 bike you need to use handler-bind rather than handler-case for that

14:02:49 bike "interactive" in this context does mean user interaction, but you can use restarts programmatically as well

14:05:09 dnhester26 bike: thanks. So handler-bind. I'm not sure if it's better to just make a more complicated function which deals with the data instead of using handler-bind which may be overly complicated because of the macro. What do you think? Is it better to keep it "simpler" and just use a longer form or to use a handler-bind which may be more advanced but probably a cleaner solution?

14:05:17 yitzi Did you look at the restart-case/restart-bind docs?

14:05:58 dnhester26 yitzi: restar-bind had no examples, I was just reading with-simple-restart, ahven't gotten to restart-case

14:13:58 dnhester26 yitzi: restart-case is just providing options for the user to choose in the REPL on how to deal with an error, not sure if that really helps me, but thanks, it's good to learn something new! :D

14:14:37 bike no, restart-case also provides restarts for the program to use.

14:15:02 bike the program can do (invoke-restart ...) to invoke one of the options in restart-case.

14:18:00 dnhester26 I don't get in the example: http://www.ai.mit.edu/projects/iiip/doc/CommonLISP/HyperSpec/Body/mac_restart-case.html

14:18:12 dnhester26 Why is the restart case starting from calling this (all-start-with-same-letter ice-cream sauce topping)?

14:18:38 dnhester26 bike: where would the invoke-start be called, inside the handler-case?

14:20:02 yitzi As bike said, handler-bind is actually used since handler-case unwinds.

14:20:26 dlowe dnhester26: ERROR is called which invokes the debugger, which gives a menu of available restarts

14:20:43 dlowe then the debugger invokes a restart once the user makes a selection

14:21:10 dlowe the all-start-* function isn't called in the body of the restart-case

14:21:21 dnhester26 right, sorry, handler-bind. I'm reading the spec trying to make sense of it

14:21:56 dlowe basically, the example loops until all-start-with-same-letter returns true, and it uses the debugger to fix the various ice cream problems

14:21:59 dnhester26 dlowe: oh! thank you! now I just got the example. So why did they add that function call? It seems completely unrelated to the example

14:22:56 dnhester26 you answered just before I sent that

14:22:59 dnhester26 thanks

14:24:02 dlowe probably the most common restart-case I've used is when decoding utf-8 from a socket, and the decoder will throw an error on invalid utf-8 but with restarts that allow the program to decide how to handle it

14:34:08 dnhester26 dlowe: pardon my ignorance for I have never used do. what is the last statement doing there? when is it called? it's after the test form, form the spec I don't understand

14:35:54 dnhester26 is it evaluating only if the test-form fails? That's what it looks like to me, but from the spec I couldn't understand

14:38:09 thuna` dnhester26: That DO evaluates its body - which is the RESTART-CASE - until the result of (ALL-START-WITH-SAME-LETTER ICE-CREAM SAUCE TOPPING) is non-nil.

14:38:22 dnhester26 In this line in the spec "At the beginning of each iteration, after processing the variables, the end-test-form is evaluated. If the result is false, execution proceeds with the body of the do (or do*) form. If the result is true, the result-forms are evaluated in order as an implicit progn, and then do or do* returns." I don't see in the sytnax neither the body of the do nor the result forms

14:38:36 dnhester26 thuna`: thanks

14:59:38 chkhd ** NICK chkhd`

16:37:14 hayley jmercouris: Consider writing --quiet with two hyphens before you smartass about wHaTeVeR rEaSoN.

17:09:54 edgar-rft_ ** NICK edgar-rft