Closed aranega closed 3 years ago
It should be the same, except for the case of denormalized numbers (gradual underflow) - that is when exponentBits
is zero.
Pick the exponentFromBitPattern
from Squeak.
In Cuis, use the partBits:
method, it will be much nicer.
Thanks Folks. This is fixed now (in Cuis code style) at #4530
Hi I see that there are some packages that fail to load due to the fact that "SqueakCompatibility" was deleted from the repo. [image: Capture2.PNG]
For instance, when I try to load Erudite I get: [image: Capture.PNG]
Fortunately I had a copy of that package and now Erudite works. cheers
Ignacio Sniechowski
On Tue, Feb 9, 2021 at 2:33 PM Juan Vuletich @.***> wrote:
Closed #182 https://github.com/Cuis-Smalltalk/Cuis-Smalltalk-Dev/issues/182.
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/Cuis-Smalltalk/Cuis-Smalltalk-Dev/issues/182#event-4310470610, or unsubscribe https://github.com/notifications/unsubscribe-auth/AA23RWQUKGBG3TY57MBW7MTS6FWWNANCNFSM4UOS22FA .
Hi,
While reading and decoding/executing bytecodes of the image without some primitives, I ran into an infinite loop for
SmallFloat64
with theexponent
method. Where there is not the primitive number 553, theSmallFloat64>>exponent
method fallback on this codeWhen running
10.0 exponent
, theFloat>>floorLog: 2
method is executed:askind again for the
exponent
to10.0
. I know the primitive is not supposed to fail, but just in case. I think I fixed it changingFloat>>floorLog:
for:but I'm unsure it will produce the right result, on some examples it seems it does (I didn't find tests).
EDIT> Nope, the modification I made does not solve nothing, the behavior is different