freenode/#clim - IRC Chatlog
Search
3:06:06
bmansurov
jackdaniel: I want to learn more about McCLIM internals and was wondering if you can recommend any beginner friendly bugs to fix? Thanks.
3:13:35
bmansurov
beach: I see. Hope he sees it when he's online. I'm open to suggestions if you have any.
3:14:52
bmansurov
I've seen some open issues, but haven't looked into the details. Maybe I should check them out.
3:15:27
beach
Again, I think most people working on McCLIM are in the EU, so it is still early for most of them.
3:18:56
beach
Some issues are associated with bounties, but they are probably the ones with a bit more work.
3:20:48
bmansurov
I hope to fix the low hanging fruits to start. Maybe help make the UI look a bit nicer, etc.
3:21:37
beach
I think jackdaniel is thinking about those issues, and in particular how to make it possible to use themes.
3:23:02
bmansurov
It'd be nice to tag the Github issues with difficulty levels like 'easy', 'medium', 'hard'. That way newcomers can pick up the right issues.
3:23:09
beach
Perfect. Like I said, I haven't kept up with the details. Now that there are so many good maintainers, it is best that I concentrate on my other project(s).
3:27:28
beach
I worked a bit on Second Climacs, the goal of which is to have excellent Common Lisp editing facilities.
3:28:06
beach
And Clordane will be the GUI debugger for Common Lisp, probably just SICL to start with.
3:28:48
beach
The thing is, many of those components depend on SICL, so that is highest priority right now.
3:30:41
beach
The debugger will use features that SBCL doesn't have, and is unlikely to acquire. See for instance http://metamodular.com/SICL/sicl-debugging.pdf
3:31:31
beach
The reason I started SICL is that I was convinced that the maintainers of existing Common Lisp implementations would not be willing to make their implementation evolve in the direction I need.
3:32:15
bmansurov
I see. At least you have the skills to start your own implementation. I wish I knew more ;)
3:33:19
beach
And it is well known in software engineering that it takes roughly the same amount of effort to understand existing code as it takes to implement it. Understanding an existing implementation and then modifying it to my needs would be way more work than starting from scratch.
3:34:13
beach
Most Common Lisp implementations started life way back, so the design decisions that are still with us are often no longer relevant.
3:34:57
bmansurov
I see. By your rough estimate, how much of SICL done? How long does it take to take it to a good spot?
3:36:11
beach
Hard to say. I recently finished a first version of x86-64 code generation. Now I am working on getting the code generator to work with the bootstrapping procedure. It feels like I am close, as in a few months.
3:37:05
beach
I am hoping that, once I have a working system, I will attract more people to help. But I am not counting on it.
3:37:38
beach
heisig has been doing a lot of work, and Bike is working on the Cleavir compiler framework.
3:38:43
bmansurov
It's getting late here. It was nice talking to you. Hope to learn from many things CL.
5:25:02
jackdaniel
bmansurov: maybe this? https://github.com/McCLIM/McCLIM/issues/802 to get only one feet wet
5:26:24
jackdaniel
as of adding difficulty gradation, more than just "first time issue" is too much effort - it is rarely obvious how tard the solution will be
5:27:17
jackdaniel
and we have the label "first time issue", we could go through the list to add this label where it is due
5:28:27
jackdaniel
no, I can go shopping! :) but we need to do the roundrobin with Justyna to take care of the little one
5:46:55
no-defun-allowed
What would I have to do, to implement a pane that is similar to a TABLE-PANE, but sharing column widths between every nth row instead of every row?
6:10:13
jackdaniel
you'd have to record each cell and run a similar layouting algorithm which formatting-table does
6:31:43
no-defun-allowed
It would look something like this artist-mode diagram: https://plaster.tymoon.eu/view/1725#1725, but I don't think it would be much effort to generalise.
6:34:15
jackdaniel
ach, table pane, I misunderstood. you talk about layouts. see how layout protocol is implemented for the table pane then
7:14:30
ecraven
a quick question (with probably a more involved answer): where are "object streams" described? are they only part of clim? or does CL itself support writing arbitrary objects to streams?
7:33:20
jackdaniel
gray streams define the protocol which permita non-character streams (i.e object streams)
8:02:42
ecraven
hm.. my return key does not work in the listener. this might be related to the fact that I run xcape, to have return double as shift. is there a way to see what key clim *thinks8