urbit/shrub
1
1
Fork 0
mirror of https://github.com/urbit/shrub.git synced 2024-12-19 08:32:39 +03:00
Code Issues Packages Projects Releases Wiki Activity
34fd7f5adf
shrub/pkg/arvo/sys/vane/khan.hoon

190 lines
4.7 KiB
Plaintext
Raw Normal View History

khan: vane skeleton
2022-01-21 21:48:05 +03:00
:: %khan, thread runner
!:
!? 164
::
|= our=ship
=> |%
+$ move [p=duct q=(wite note gift:khan)]
+$ note :: out request $->
khan: wip fyrd
2022-01-26 02:37:48 +03:00
$~ [%g %deal *sock *term *deal:gall]
khan: closer to starting threads
2022-01-22 04:46:59 +03:00
$% $: %g :: to %gall
$>(%deal task:gall) :: full transmission
== == ::
khan: vane skeleton
2022-01-21 21:48:05 +03:00
+$ sign :: in response $<-
khan: wip threads v0 super ghetto first pass with a bunch of stuff hard-coded to prevent sending giant vases over the wire.
2022-01-26 04:43:48 +03:00
$% $: %gall
$>(%unto gift:gall)
==
$: %khan
khan: add %fard/%arow for in-arvo threads These take and produce vases, and assign random tids (rather than deducing them from the input duct.) Since %fard does not require mark conversion, we make the mark/beak on $thread-state optional (and use this to decide whether to send %avow or %arow.) Provide a state adapter since it's possible that people have been experimenting with this vane. This makes the negative case of %avow/%arow kind of clunky, since there is no content difference, but the following does not seem possible within the Hoon type system: =/ gif ?~ p.tad %arow %avow [hen %give gif %| p.cag tang]~
2022-02-08 01:46:02 +03:00
$> ?(%arow %avow)
gift:khan
khan: wip threads v0 super ghetto first pass with a bunch of stuff hard-coded to prevent sending giant vases over the wire.
2022-01-26 04:43:48 +03:00
== ==
khan: vane skeleton
2022-01-21 21:48:05 +03:00
+$ khan-state
khan: add %fard/%arow for in-arvo threads These take and produce vases, and assign random tids (rather than deducing them from the input duct.) Since %fard does not require mark conversion, we make the mark/beak on $thread-state optional (and use this to decide whether to send %avow or %arow.) Provide a state adapter since it's possible that people have been experimenting with this vane. This makes the negative case of %avow/%arow kind of clunky, since there is no content difference, but the following does not seem possible within the Hoon type system: =/ gif ?~ p.tad %arow %avow [hen %give gif %| p.cag tang]~
2022-02-08 01:46:02 +03:00
$: %1
khan: cleanup - only store metadata in the persistent map. just enough to support (eventual) thread cancellation and output mark lookup. - try to delete thread state at other failure points not covered by %kick. - reflect back the passed output mark rather than form.dais. not sure about this one yet.
2022-01-27 07:09:36 +03:00
hey=duct :: current unix duct
run=(map duct thread-state) :: running threads
khan: vane skeleton
2022-01-21 21:48:05 +03:00
==
khan: add %fard/%arow for in-arvo threads These take and produce vases, and assign random tids (rather than deducing them from the input duct.) Since %fard does not require mark conversion, we make the mark/beak on $thread-state optional (and use this to decide whether to send %avow or %arow.) Provide a state adapter since it's possible that people have been experimenting with this vane. This makes the negative case of %avow/%arow kind of clunky, since there is no content difference, but the following does not seem possible within the Hoon type system: =/ gif ?~ p.tad %arow %avow [hen %give gif %| p.cag tang]~
2022-02-08 01:46:02 +03:00
+$ thread-state
[tid=@ta p=(unit [mak=mark bek=beak])]
khan: vane skeleton
2022-01-21 21:48:05 +03:00
--
khan: closer to starting threads
2022-01-22 04:46:59 +03:00
=>
|%
khan: functionally done
2022-01-27 04:41:01 +03:00
++ get-dais
|= [=beak =mark rof=roof]
^- dais:clay
?~ ret=(rof ~ %cb beak /[mark])
~|(%mark-unknown !!)
?~ u.ret
~|(%mark-invalid !!)
?> =(%dais p.u.u.ret)
!<(dais:clay q.u.u.ret)
khan: closer to starting threads
2022-01-22 04:46:59 +03:00
++ start-spider
khan: add %fard/%arow for in-arvo threads These take and produce vases, and assign random tids (rather than deducing them from the input duct.) Since %fard does not require mark conversion, we make the mark/beak on $thread-state optional (and use this to decide whether to send %avow or %arow.) Provide a state adapter since it's possible that people have been experimenting with this vane. This makes the negative case of %avow/%arow kind of clunky, since there is no content difference, but the following does not seem possible within the Hoon type system: =/ gif ?~ p.tad %arow %avow [hen %give gif %| p.cag tang]~
2022-02-08 01:46:02 +03:00
|= [our=@p =vase]
khan: wip fyrd
2022-01-26 02:37:48 +03:00
^- note
khan: add %fard/%arow for in-arvo threads These take and produce vases, and assign random tids (rather than deducing them from the input duct.) Since %fard does not require mark conversion, we make the mark/beak on $thread-state optional (and use this to decide whether to send %avow or %arow.) Provide a state adapter since it's possible that people have been experimenting with this vane. This makes the negative case of %avow/%arow kind of clunky, since there is no content difference, but the following does not seem possible within the Hoon type system: =/ gif ?~ p.tad %arow %avow [hen %give gif %| p.cag tang]~
2022-02-08 01:46:02 +03:00
[%g %deal [our our] %spider %poke %spider-start vase]
khan: closer to starting threads
2022-01-22 04:46:59 +03:00
++ watch-spider
khan: wip fyrd
2022-01-26 02:37:48 +03:00
|= [our=@p =path]
^- note
[%g %deal [our our] %spider %watch path]
khan: closer to starting threads
2022-01-22 04:46:59 +03:00
--
khan: vane skeleton
2022-01-21 21:48:05 +03:00
=| khan-state
=* state -
|= [now=@da eny=@uvJ rof=roof]
=* khan-gate .
^?
|%
:: +call: handle a +task:khan request
::
++ call
|= $: hen=duct
dud=(unit goof)
wrapped-task=(hobo task:khan)
==
^- [(list move) _khan-gate]
::
=/ =task:khan ((harden task:khan) wrapped-task)
khan: cleanup
2022-01-26 08:28:45 +03:00
?+ -.task [~ khan-gate]
khan: basic +call, types
2022-01-22 01:09:01 +03:00
%born
khan: functionally done
2022-01-27 04:41:01 +03:00
:- ~
%= khan-gate
khan: cleanup - only store metadata in the persistent map. just enough to support (eventual) thread cancellation and output mark lookup. - try to delete thread state at other failure points not covered by %kick. - reflect back the passed output mark rather than form.dais. not sure about this one yet.
2022-01-27 07:09:36 +03:00
hey hen
run *(map duct thread-state)
khan: functionally done
2022-01-27 04:41:01 +03:00
==
khan: add %fard/%arow for in-arvo threads These take and produce vases, and assign random tids (rather than deducing them from the input duct.) Since %fard does not require mark conversion, we make the mark/beak on $thread-state optional (and use this to decide whether to send %avow or %arow.) Provide a state adapter since it's possible that people have been experimenting with this vane. This makes the negative case of %avow/%arow kind of clunky, since there is no content difference, but the following does not seem possible within the Hoon type system: =/ gif ?~ p.tad %arow %avow [hen %give gif %| p.cag tang]~
2022-02-08 01:46:02 +03:00
::
%fard
=* fad p.task
=/ tid=@ta
(cat 3 'khan-fard--' (scot %uv (sham eny)))
=/ =beak
?@(bear.fad [our bear.fad %da now] bear.fad)
=/ args
[~ `tid beak name.fad data.fad]
=/ start-moves=(list move)
%+ turn
:~ (watch-spider our /thread-result/[tid])
(start-spider our !>(args))
==
|=(=note ^-(move [hen %pass //g note]))
=. run (~(put by run) hen tid ~)
[start-moves khan-gate]
khan: cleanup - only store metadata in the persistent map. just enough to support (eventual) thread cancellation and output mark lookup. - try to delete thread state at other failure points not covered by %kick. - reflect back the passed output mark rather than form.dais. not sure about this one yet.
2022-01-27 07:09:36 +03:00
::
khan: basic +call, types
2022-01-22 01:09:01 +03:00
%fyrd
khan: cleanup
2022-01-26 08:28:45 +03:00
=/ wir=wire (head hen)
=/ rid=@ta (rear wir)
khan: accepts thread arguments with marks
2022-01-27 03:26:30 +03:00
=* fyd p.task
khan: cleanup - only store metadata in the persistent map. just enough to support (eventual) thread cancellation and output mark lookup. - try to delete thread state at other failure points not covered by %kick. - reflect back the passed output mark rather than form.dais. not sure about this one yet.
2022-01-27 07:09:36 +03:00
?: (~(has by run) hen)
khan: functionally done
2022-01-27 04:41:01 +03:00
~|(%fyrd-duplicate-rid !!)
khan: closer to starting threads
2022-01-22 04:46:59 +03:00
=/ tid=@ta
(cat 3 'khan-fyrd--' rid)
khan: accepts thread arguments with marks
2022-01-27 03:26:30 +03:00
=/ =beak
khan: +bear allows a full +beak
2022-01-28 09:28:30 +03:00
?@(bear.fyd [our bear.fyd %da now] bear.fyd)
khan: functionally done
2022-01-27 04:41:01 +03:00
=/ =dais:clay (get-dais beak p.data.fyd rof)
khan: wip fyrd
2022-01-26 02:37:48 +03:00
=/ args
khan: accepts thread arguments with marks
2022-01-27 03:26:30 +03:00
:* ~ `tid beak name.fyd
(vale.dais q.data.fyd)
khan: closer to starting threads
2022-01-22 04:46:59 +03:00
==
khan: wip fyrd
2022-01-26 02:37:48 +03:00
=/ start-moves=(list move)
%+ turn
khan: cleanup
2022-01-26 08:28:45 +03:00
:~ (watch-spider our /thread-result/[tid])
khan: add %fard/%arow for in-arvo threads These take and produce vases, and assign random tids (rather than deducing them from the input duct.) Since %fard does not require mark conversion, we make the mark/beak on $thread-state optional (and use this to decide whether to send %avow or %arow.) Provide a state adapter since it's possible that people have been experimenting with this vane. This makes the negative case of %avow/%arow kind of clunky, since there is no content difference, but the following does not seem possible within the Hoon type system: =/ gif ?~ p.tad %arow %avow [hen %give gif %| p.cag tang]~
2022-02-08 01:46:02 +03:00
(start-spider our !>(args))
khan: wip fyrd
2022-01-26 02:37:48 +03:00
==
|=(=note ^-(move [hen %pass //g note]))
khan: add %fard/%arow for in-arvo threads These take and produce vases, and assign random tids (rather than deducing them from the input duct.) Since %fard does not require mark conversion, we make the mark/beak on $thread-state optional (and use this to decide whether to send %avow or %arow.) Provide a state adapter since it's possible that people have been experimenting with this vane. This makes the negative case of %avow/%arow kind of clunky, since there is no content difference, but the following does not seem possible within the Hoon type system: =/ gif ?~ p.tad %arow %avow [hen %give gif %| p.cag tang]~
2022-02-08 01:46:02 +03:00
=. run (~(put by run) hen tid `[out-mark.fyd beak])
khan: closer to starting threads
2022-01-22 04:46:59 +03:00
[start-moves khan-gate]
khan: basic +call, types
2022-01-22 01:09:01 +03:00
==
khan: vane skeleton
2022-01-21 21:48:05 +03:00
:: +load: migrate an old state to a new khan version
::
++ load
khan: add %fard/%arow for in-arvo threads These take and produce vases, and assign random tids (rather than deducing them from the input duct.) Since %fard does not require mark conversion, we make the mark/beak on $thread-state optional (and use this to decide whether to send %avow or %arow.) Provide a state adapter since it's possible that people have been experimenting with this vane. This makes the negative case of %avow/%arow kind of clunky, since there is no content difference, but the following does not seem possible within the Hoon type system: =/ gif ?~ p.tad %arow %avow [hen %give gif %| p.cag tang]~
2022-02-08 01:46:02 +03:00
|= $= old
$% $: %0
hey=duct
run=(map duct [tid=@ta mak=mark bek=beak])
==
khan-state
==
khan: vane skeleton
2022-01-21 21:48:05 +03:00
^+ khan-gate
khan: add %fard/%arow for in-arvo threads These take and produce vases, and assign random tids (rather than deducing them from the input duct.) Since %fard does not require mark conversion, we make the mark/beak on $thread-state optional (and use this to decide whether to send %avow or %arow.) Provide a state adapter since it's possible that people have been experimenting with this vane. This makes the negative case of %avow/%arow kind of clunky, since there is no content difference, but the following does not seem possible within the Hoon type system: =/ gif ?~ p.tad %arow %avow [hen %give gif %| p.cag tang]~
2022-02-08 01:46:02 +03:00
=/ new=khan-state
?: ?=(%0 -.old)
:+ %1
hey.old
%- ~(urn by run.old)
|= [a=duct b=[tid=@ta mak=mark bek=beak]]
[tid.b `[mak.b bek.b]]
old
khan-gate(state new)
khan: cleanup
2022-01-26 08:28:45 +03:00
:: +scry: nothing to see as yet
khan: vane skeleton
2022-01-21 21:48:05 +03:00
::
++ scry
^- roon
|= [lyc=gang car=term bem=beam]
^- (unit (unit cage))
~
++ stay state
khan: cleanup
2022-01-26 08:28:45 +03:00
:: +take: handle responses.
::
khan: vane skeleton
2022-01-21 21:48:05 +03:00
++ take
|= [tea=wire hen=duct dud=(unit goof) hin=sign]
^- [(list move) _khan-gate]
?^ dud
khan: cleanup - only store metadata in the persistent map. just enough to support (eventual) thread cancellation and output mark lookup. - try to delete thread state at other failure points not covered by %kick. - reflect back the passed output mark rather than form.dais. not sure about this one yet.
2022-01-27 07:09:36 +03:00
=. run (~(del by run) hen)
khan: vane skeleton
2022-01-21 21:48:05 +03:00
~|(%khan-take-dud (mean tang.u.dud))
khan: functionally done
2022-01-27 04:41:01 +03:00
?. ?=(%gall -.hin)
[~ khan-gate]
?+ -.p.hin [~ khan-gate]
?(%poke-ack %watch-ack)
?~ p.p.hin [~ khan-gate]
khan: cleanup - only store metadata in the persistent map. just enough to support (eventual) thread cancellation and output mark lookup. - try to delete thread state at other failure points not covered by %kick. - reflect back the passed output mark rather than form.dais. not sure about this one yet.
2022-01-27 07:09:36 +03:00
=. run (~(del by run) hen)
khan: PR feedback, wip tests
2022-01-28 09:20:31 +03:00
%- (slog 'khan-ack' u.p.p.hin)
khan: functionally done
2022-01-27 04:41:01 +03:00
:_ khan-gate
khan: add %fard/%arow for in-arvo threads These take and produce vases, and assign random tids (rather than deducing them from the input duct.) Since %fard does not require mark conversion, we make the mark/beak on $thread-state optional (and use this to decide whether to send %avow or %arow.) Provide a state adapter since it's possible that people have been experimenting with this vane. This makes the negative case of %avow/%arow kind of clunky, since there is no content difference, but the following does not seem possible within the Hoon type system: =/ gif ?~ p.tad %arow %avow [hen %give gif %| p.cag tang]~
2022-02-08 01:46:02 +03:00
=/ tad (~(got by run) hen)
?~ p.tad
[hen %give %arow %| -.p.hin u.p.p.hin]~
khan: functionally done
2022-01-27 04:41:01 +03:00
[hen %give %avow %| -.p.hin u.p.p.hin]~
::
%fact
=* cag cage.p.hin
?+ p.cag ~&(bad-fact+p.cag !!)
%thread-fail
=/ =tang !<(tang q.cag)
khan: PR feedback, wip tests
2022-01-28 09:20:31 +03:00
%- (slog 'khan-fact' tang)
khan: cleanup
2022-01-26 08:28:45 +03:00
:_ khan-gate
khan: add %fard/%arow for in-arvo threads These take and produce vases, and assign random tids (rather than deducing them from the input duct.) Since %fard does not require mark conversion, we make the mark/beak on $thread-state optional (and use this to decide whether to send %avow or %arow.) Provide a state adapter since it's possible that people have been experimenting with this vane. This makes the negative case of %avow/%arow kind of clunky, since there is no content difference, but the following does not seem possible within the Hoon type system: =/ gif ?~ p.tad %arow %avow [hen %give gif %| p.cag tang]~
2022-02-08 01:46:02 +03:00
=/ tad (~(got by run) hen)
?~ p.tad
[hen %give %arow %| p.cag tang]~
khan: functionally done
2022-01-27 04:41:01 +03:00
[hen %give %avow %| p.cag tang]~
khan: wip threads v0 super ghetto first pass with a bunch of stuff hard-coded to prevent sending giant vases over the wire.
2022-01-26 04:43:48 +03:00
::
khan: functionally done
2022-01-27 04:41:01 +03:00
%thread-done
:_ khan-gate
khan: add %fard/%arow for in-arvo threads These take and produce vases, and assign random tids (rather than deducing them from the input duct.) Since %fard does not require mark conversion, we make the mark/beak on $thread-state optional (and use this to decide whether to send %avow or %arow.) Provide a state adapter since it's possible that people have been experimenting with this vane. This makes the negative case of %avow/%arow kind of clunky, since there is no content difference, but the following does not seem possible within the Hoon type system: =/ gif ?~ p.tad %arow %avow [hen %give gif %| p.cag tang]~
2022-02-08 01:46:02 +03:00
=/ tad (~(got by run) hen)
?~ p.tad
[hen %give %arow %& q.cag]~
=/ mak mak.u.p.tad
=/ bek bek.u.p.tad
=/ =dais:clay (get-dais bek mak rof)
khan: cleanup - only store metadata in the persistent map. just enough to support (eventual) thread cancellation and output mark lookup. - try to delete thread state at other failure points not covered by %kick. - reflect back the passed output mark rather than form.dais. not sure about this one yet.
2022-01-27 07:09:36 +03:00
=/ =vase (vale:dais q.q.cag)
khan: add %fard/%arow for in-arvo threads These take and produce vases, and assign random tids (rather than deducing them from the input duct.) Since %fard does not require mark conversion, we make the mark/beak on $thread-state optional (and use this to decide whether to send %avow or %arow.) Provide a state adapter since it's possible that people have been experimenting with this vane. This makes the negative case of %avow/%arow kind of clunky, since there is no content difference, but the following does not seem possible within the Hoon type system: =/ gif ?~ p.tad %arow %avow [hen %give gif %| p.cag tang]~
2022-02-08 01:46:02 +03:00
[hen %give %avow %& mak q.vase]~
khan: wip threads v0 super ghetto first pass with a bunch of stuff hard-coded to prevent sending giant vases over the wire.
2022-01-26 04:43:48 +03:00
==
khan: del in %kick
2022-01-27 04:45:25 +03:00
%kick
khan: cleanup - only store metadata in the persistent map. just enough to support (eventual) thread cancellation and output mark lookup. - try to delete thread state at other failure points not covered by %kick. - reflect back the passed output mark rather than form.dais. not sure about this one yet.
2022-01-27 07:09:36 +03:00
[~ khan-gate(run (~(del by run) hen))]
khan: wip threads v0 super ghetto first pass with a bunch of stuff hard-coded to prevent sending giant vases over the wire.
2022-01-26 04:43:48 +03:00
==
khan: vane skeleton
2022-01-21 21:48:05 +03:00
--
Reference in New Issue Copy Permalink