<bipt>waressearcher2, yes, though probably not that well at this point <bipt>basically. i think someone is working on it but it wasn't actively worked on for a while <bipt>oh, it's not included in guile yet. we only have scheme, elisp, ecmascript and brainfuck in the main distribution <bipt>waressearcher2, it's just a toy language. maybe someone should implement unlambda so guile has a toy language with a less controversial name (-: <bipt>waressearcher2, the toy language is provided precisely because it is very simple, as an example of how to use the compiler-building infrastructure <bipt>guile does a lot of the heavy lifting, as high level languages can basically be compiled to scheme, but writing complete and fully compatible support for a real language takes months of work in most cases *cky wants to implement GolfScript as a Guile language. <cky>Maybe it'll run faster than the reference implementation. <waressearcher2>so it is possible to embed guile in any software to make that software upgradable with plugins ? <nalaginrut>alright, after read Appel's paper, I realized SSA is not hard to be implemented with Scheme too... <taylanub>bipt: hoi :) any plans on rebasing your emacs branch onto the new canonical master branch, since Emacs moved to git? is it difficult?.. <ArneBab>civodul: I now looked into SRFI-42, and adding futures doesn’t look as easy as I had hoped (but actually not really expected) <ArneBab>I’m sure that someone who really understands the implementation of SRFI-42 can add futures easily, but that’s not the case for me… <davexunit>are there any good naming conventions for functional setters? <davexunit>I don't really like the 'set-object-field' naming that the guile manual uses for the (srfi srfi-9 gnu) examples. <taylanub>well it also uses "update!" for non-functional updates. <davexunit>I'm writing what I hope will be a declarative API for something, so I want the language I use to reflect that. ***dsmith-w` is now known as dsmith-work
<civodul>davexunit: another option is something like 'inherit' in Guix <davexunit>civodul: yeah, I was thinking about doing something like that. <davexunit>but then I'd probably need that record macro of yours <davexunit>I'm hacking on sly, and I have a data type that describes the state for a single rendering operation that I am calling a 'model'. <davexunit>I want to write a series of functions that do things like change the color, texture, transformation matrix, etc. <civodul>there may be performance considerations as well <davexunit>and it would be nice if the code was declarative. I guess an 'inherit' concept would work. <civodul>like, do you want to reallocate a complete model every time you change a single parameter <davexunit>it just contains references to the expensive to create stuff like the vertex array <ArneBab>davexunit: could you use keyword arguments, changing only the parameters you request changed and inheriting all others? <davexunit>(set-model-color (make-model ...) white) works of course, but it just doesn't use the language that I would like. <davexunit>ArneBab: I thought of that, but not sure how to do that well. <ArneBab>(I’ve got some lists of immutable datastructures in my python stuff which I update like that) <davexunit>yeah that could work, but I couldn't use lambda* to specify the accepted parameters. <davexunit>I would need some special default value that I can check for to know that the user didn't want that field changed. <davexunit>instead, I could take a 'rest' arg and map over it <davexunit>converting keywords to symbols for field names. <davexunit>is there an easy way to assert that two objects are of the same type? <jmd>configure complains: <jmd>checking whether libunistring was built with iconv support... no <jmd>configure: error: No iconv support. Please recompile libunistring with iconv enabled. <jmd>But I have done that. and it still complains. <jmd>Is it possible to build guile without threading? <jmd>civodul: So how does one do it? --without-threads is ignored. <civodul>--without-threads shouldn't be ignored <civodul>false alarm, --without-threads does the right thing <jmd>Yeah on closer look, its gc that is expecting the threads. ***dje is now known as xdje
<dsmith-work>Hey. All you U.S. guile-heads have a great Thankgiving weekend <amirouche>I'm doing some bindings (wiredtiger) and it's not that difficult