[squeak-dev] GalaxyObject Scenario - remote wiring in promise pipelining during ephemeral 3rd party promise relocation and chaining.
rabbit
rabbit at callistohouse.org
Sun Oct 23 11:38:55 UTC 2022
I would note from yesterday that GC messages travel from the Inz to the Outz as a remote reference to a shared object is no longer held strongly.
—
Have a good one; keep it, light.
Kindly, langohr . .. … ‘…^,^ 🐇🐇🐇
Sent from Callisto House Mobile - Europa
:: decentralized mobile homeless solutions ::
> On Oct 23, 2022, at 07:31, rabbit <rabbit at callistohouse.org> wrote:
>
> What what happens when the little #whenMoreResolved: in Alice hits remote.4….
>
> —
> Have a good one; keep it, light.
> Kindly, langohr . .. … ‘…^,^ 🐇🐇🐇
>
> Sent from Callisto House Mobile - Europa
> :: decentralized mobile homeless solutions ::
>
>
>>> On Oct 23, 2022, at 07:28, rabbit <rabbit at callistohouse.org> wrote:
>>>
>>
>> For those still following along I’d say I appreciate you, I appreciate your interest, I appreciate your consideration, I appreciate your discernment!
>>
>> promise.1 now points to remote.2 and remote.1 is totally cleaned up in Carol while the 2nd message is all that is referencing remote.2 in Alice. Bob’s still waiting to be #greet:ed…
>> <image1.jpeg>
>>
>>
>> Now the 2nd message (what was it? #doYouLikeHer perhaps…) has been disgorged to Bob with the establishment of remote.4 in Alice. Shorting the promises…
>> <image2.jpeg>
>>
>>
>> —
>> Have a good one; keep it, light.
>> Kindly, langohr . .. … ‘…^,^ 🐇🐇🐇
>>
>> Sent from Callisto House Mobile - Europa
>> :: decentralized mobile homeless solutions ::
>>
>>>> On Oct 22, 2022, at 13:31, rabbit <rabbit at callistohouse.org> wrote:
>>>>
>>>
>>>
>>> Carrying on…
>>>
>>> In this slide we see the little #whenMoreResolved: in Carol has hit r1 and now p1 points to r2 to the message continuation in Bob. Due to r1 no longer being referenced a GCAnswer is sent to Alice to let her know r1 is no longer referenced.
>>>
>>> I’ve just realized that my table description may be wrong…sigh. As a GCAnswer is sent for r1, then the Burgundy/Pink tables must be Questions/Answers. Answers would be Pink and outgoing ref and GCAnswer cleans that entry. I’ll take a closer look in a minute…
>>>
>>> <image0.jpeg>
>>>
>>>
>>> Ok! As the resolver for r1 is no longer referenced in Alice, a GCExport is sent to Carol to cleanup the entry in Carol’s Exports table. So Exports must be Yellow.
>>>
>>> Yes I was all turned around. My apologies to those trying to follow. This makes much more sense semantically.
>>>
>>> Exports & Answers are Outz
>>> Imports & Questions are Inz
>>>
>>> ———- Scope tables ———-
>>> Outz:
>>> - Exports (Yellow)
>>> - Answers (Pink)
>>> Inz:
>>> - Imports (Orange)
>>> - Questions (Burgundy)
>>>
>>> <image1.jpeg>
>>>
>>> —
>>> Have a good one; keep it, light.
>>> Kindly, langohr . .. … ‘…^,^ 🐇🐇🐇
>>>
>>> Sent from Callisto House Mobile - Europa
>>> :: decentralized mobile homeless solutions ::
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.squeakfoundation.org/pipermail/squeak-dev/attachments/20221023/3491cc78/attachment.html>
More information about the Squeak-dev
mailing list
|