freenode/#lisp - IRC Chatlog
Search
19:17:03
Harag
:beach ...thanx for side tracking me there for a couple of hours some interesting reading, there is a lot of projects out there but they all look seriously short of manpower :(
19:31:01
Harag
:beach ...just thinking out loud ...why not start with just a server version server version of a lispos? Something that can be put in a docker container...
19:44:04
Harag
:dlowe ...well yes minus the linux os that would normally used... would it be much smaller? maybe throw in a library for the docker containers to work well/easy together...I am no expert just wondering
19:46:05
White_Flame
as beach said, there's no real benefit at the user level or application developer level between running bare metal and running as a process on top of an existing OS
19:50:46
Harag
:White_Flame not sure what you mean...are you saying that there would be no benefits in application design, roll out or performance to a lispos?
19:50:47
dlowe
imagine being able to just pass a complex object and its referants to another program
19:52:52
White_Flame
Harag: since common lisp is an image based system, the running lisp *is* already an OS.
19:57:22
dlowe
The running lisp isn't an OS. wtf. There's no interrupt handling, no drivers, no DMA, no anything that qualifies
19:58:55
White_Flame
dlowe: right, the hardware tends to be abstracted away, but it still loosely counts in terms of managing your execution environment
19:59:22
dlowe
What's needed is an army of programmers to write hardware drivers and interfaces and code to coordinate between them all.
19:59:49
White_Flame
why not, at least in the interim, take advantage of all that which already is there in extant OSes
20:00:22
White_Flame
but the "comprehensive lisp-based user environment" still isn't built up, at least not to the level of lisp machines
20:03:40
Harag
:White_Flame what are the components that need to make up the lisp-based user environment? This stuff is far out of my scope of experience so excuse the obviously dumb questions
20:04:04
White_Flame
running multiple graphical applications in a full-screen environment in a single lisp image
20:10:17
Harag
:White_Flame ...ok so you are saying that some one should start working on VR user interface instead or even better a direct brain interface... that way it might be ready by the time the non lisp world has implemented there version *wink* *wink*
20:16:41
Harag
Jokes aside is "comprehensive lisp-based user environment" such a difficult thing to do or is their just not enough will power and manpower out there?
20:18:01
White_Flame
commercial ventures unified development both of teams creating lisp machines with full lisp OSes, as well as unifying Common Lisp from all the naturally fragmented lisps before it
20:21:22
Harag
:White_Flame ...yes I remember the days before I went full lisp... I never rolled my own
20:39:32
jack_rabbit
Hey, jackdaniel. I was looking at ECL, and it looks like it's actively developed. I don't see any releases since ~2016. (https://common-lisp.net/project/ecl/static/files/release/) Are there any plans to publish new releases of ECL?
21:29:54
jackdaniel
you may check out the milestone on our repository to see remaining two issues before we start testing
21:32:42
jackdaniel
green threads are not hard given you have delimited continuations and timers (for a scheduler)
21:35:58
jackdaniel
delimited continuations require changes of how runtime environment is handled, the actual context switch part could be implemented with longjmp
21:36:59
jackdaniel
timers are not difficult but also take time - I want to copy sbcl's api and functionality instead of providing simpler version
21:40:17
jackdaniel
sometims when I try to take a step out to see my plans from the broader perspective I see this: https://pics.awwmemes.com/you-know-what-this-place-needs-giant-fuckin-triangles-thats-58136426.png :) good night \o
21:41:10
jack_rabbit
hahaha! Well I wish you good luck. Looks like you've been working on it for a long time and it's getting close.