libera/#commonlisp - IRC Chatlog
Search
3:55:39
beach
We are making great progress with SICL code generation, so I am quite pleased with that. I can see the light at the end of the tunnel.
3:56:45
Michal
Well that is great :) I read about it from time to time, so looking forward to seeing it. Well let me make it on-topic :)
3:57:06
Michal
I was curious on if there was a way to search for existing functions loaded in my image by their name or docstring
7:18:25
beach
pjb: I think I have taken steps to make myself as replaceable as possible. Most of the techniques are documented in the form of published papers or in the specification. Much of the code has been extracted to independent repositories, usually well documented and including tests. Remaining code is often highly modular and often well commented.
7:20:43
beach
pjb: Bike and karlosz know the compiler fairly well. heisig knows a lot about HIR because he wrote the HIR evaluator. The register allocator was largely written by hayley. The reader is Eclector, so scymtym knows about it. heisig also wrote the entire implementation fo the sequence functions.
7:21:08
beach
pjb: That leaves the details of the bootstrapping technique that I am currently the only one who understands.
7:21:36
beach
Plus, the bootstrapping procedure not being entirely finished, the documentation is not accurate at this point.
7:30:06
beach
Gnuxie knows the assembler and the disassembler. lonjil did a large part of the printer.
7:31:41
beach
I would love to have someone else work on bootstrapping, because that aspect is essential for a final system.
7:33:33
mfiano
Make sure you are using the latest release of SBCL and the Quicklisp dist, and then probably a question for #sbcl
7:43:19
flip214
CodeBitCookie[m]: you need to have some libraries available - in this case, OpenSSL.
14:13:38
akater[m]
IIUC, it means, the implementation must define a method with specializers (standard-generic-function t). SBCL only has (STANDARD-GENERIC-FUNCTION METHOD). Is it my misunderstanding of CLHS, or is it SBCL?
14:16:48
akater[m]
Also CLHS: > remove-method must not signal an error if the method is not one of the
14:17:53
akater[m]
I mean, it could be interpreted as “when the second arg is not a method at all, there should be no error”.
14:19:43
specbot
The ``Arguments and Values'' Section of a Dictionary Entry: http://www.lispworks.com/reference/HyperSpec/Body/01_ddc.htm
14:20:06
beach
Which means that the consequences are undefined if a method is not given to REMOVE-METHOD.