Closed mikolajpp closed 5 years ago
Geez, nice. What's the !,*twig of that? The != ?
On Friday, 6 April 2018, mikolajpp notifications@github.com wrote:
&33:~ bails out with out of loom error.
Calling &33:~[[%leaf p="syntax error"] [%leaf p="{1 11}"]] hangs urbit until CTRL+C whereupon it prints
/~sogdus-sibmep-ponmep-litsem/home/~2018.4.2..13.59.16..4850/sys/vane/ford <[431 9].[443 11]> /~sogdus-sibmep-ponmep-litsem/home/~2018.4.2..13.59.16..4850/sys/vane/ford <[431 13].[431 31]> /~sogdus-sibmep-ponmep-litsem/home/~2018.4.2..13.59.16..4850/sys/vane/ford <[431 17].[431 31]> /~sogdus-sibmep-ponmep-litsem/home/~2018.4.2..13.59.16..4850/sys/vane/ford <[475 7].[485 15]> /~sogdus-sibmep-ponmep-litsem/home/~2018.4.2..13.59.16..4850/sys/vane/ford <[478 13].[485 9]> /~sogdus-sibmep-ponmep-litsem/home/~2018.4.2..13.59.16..4850/sys/vane/ford <[478 17].[478 40]> /~sogdus-sibmep-ponmep-litsem/home/~2018.4.2..13.59.16..4850/sys/vane/ford <[478 21].[478 40]> /~sogdus-sibmep-ponmep-litsem/home/~2018.4.2..13.59.16..4850/sys/vane/ford <[1.290 7].[1.294 9]> /~sogdus-sibmep-ponmep-litsem/home/~2018.4.2..13.59.16..4850/sys/vane/ford <[1.290 11].[1.290 54]> /~sogdus-sibmep-ponmep-litsem/home/~2018.4.2..13.59.16..4850/sys/vane/ford <[1.290 15].[1.290 54]> /~sogdus-sibmep-ponmep-litsem/home/~2018.4.2..13.59.16..4850/sys/vane/ford <[1.290 24].[1.290 52]> -find.+17,179,869,182
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/urbit/arvo/issues/703, or mute the thread https://github.com/notifications/unsubscribe-auth/ABxXhstueYunyu7mmTqgTPbmUCY96iVhks5tl54CgaJpZM4TKYmm .
Oh right &33 as in +(32), somewhere the nock interpreter isn't checking for indirect atoms presumably.
On Friday, 6 April 2018, Anton Dyudin antechno777@gmail.com wrote:
Geez, nice. What's the !,*twig of that? The != ?
On Friday, 6 April 2018, mikolajpp notifications@github.com wrote:
&33:~ bails out with out of loom error.
Calling &33:~[[%leaf p="syntax error"] [%leaf p="{1 11}"]] hangs urbit until CTRL+C whereupon it prints
/~sogdus-sibmep-ponmep-litsem/home/~2018.4.2..13.59.16..4850/sys/vane/ford <[431 9].[443 11]> /~sogdus-sibmep-ponmep-litsem/home/~2018.4.2..13.59.16..4850/sys/vane/ford <[431 13].[431 31]> /~sogdus-sibmep-ponmep-litsem/home/~2018.4.2..13.59.16..4850/sys/vane/ford <[431 17].[431 31]> /~sogdus-sibmep-ponmep-litsem/home/~2018.4.2..13.59.16..4850/sys/vane/ford <[475 7].[485 15]> /~sogdus-sibmep-ponmep-litsem/home/~2018.4.2..13.59.16..4850/sys/vane/ford <[478 13].[485 9]> /~sogdus-sibmep-ponmep-litsem/home/~2018.4.2..13.59.16..4850/sys/vane/ford <[478 17].[478 40]> /~sogdus-sibmep-ponmep-litsem/home/~2018.4.2..13.59.16..4850/sys/vane/ford <[478 21].[478 40]> /~sogdus-sibmep-ponmep-litsem/home/~2018.4.2..13.59.16..4850/sys/vane/ford <[1.290 7].[1.294 9]> /~sogdus-sibmep-ponmep-litsem/home/~2018.4.2..13.59.16..4850/sys/vane/ford <[1.290 11].[1.290 54]> /~sogdus-sibmep-ponmep-litsem/home/~2018.4.2..13.59.16..4850/sys/vane/ford <[1.290 15].[1.290 54]> /~sogdus-sibmep-ponmep-litsem/home/~2018.4.2..13.59.16..4850/sys/vane/ford <[1.290 24].[1.290 52]> -find.+17,179,869,182
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/urbit/arvo/issues/703, or mute the thread https://github.com/notifications/unsubscribe-auth/ABxXhstueYunyu7mmTqgTPbmUCY96iVhks5tl54CgaJpZM4TKYmm .
Yeah, 33 is actually the smallest number which causes the bail out :-))
This can be a source for great Urbit inside jokes...
&34:~[~]
, &35:~[~ ~]
, etc also bail.
Can't repro the hang case in #704 though, on Linux or Mac; does the test code look accurate?https://github.com/urbit/arvo/blob/bc6826d3b7a74038e21a80abe5295118541aad36/tests/bug/gh-703.hoon#L6-L9
You are right. It seems you have to take actual output of parsing for hang to work. Try this:
=a (mule |.((ream '(bomb 2 2')))
&33:p.a
Guaranteed hang :-))
Ah, this gets compiled as &33.p.a
, so not surprised it's different.
And hangs trying to compile actually, increasing but immediately noticeable amounts as you go !=(&1.p.a)
,!=(&2.p.a)
,!=(&5.p.a)
etc. The problem here is the fork type, p.a
is either stack trace or a ++hoon
AST, grabbing an arbitrary axis from which gets a bit huge:
fixed in urbit/urbit#1063, regression test added in #889
&33:~ bails out with out of loom error.
Calling
&33:~[[%leaf p="syntax error"] [%leaf p="\{1 11}"]]
hangs urbit until CTRL+C wherein it prints