I certainly understood the phrase as soon as it was mentioned on the list; and then started to recall things about "naked assignment slots that weren't supported. Can't remember too many more details about that --- basically could a naked assignment slot assign an immutable slot reached by delegation...
But then, I remembered the greatest legacy of the Self Project: Javascript! (argh!!)
OK: the second greatest legacy, Google.
"slot pair" is used in
- Chapter 4 of Craig's 1992 PhD thesis - the ECOOP '95 retrospective "Programming as an Experience" - the HOPLIII (2007) retrospective "Self" - the 2014 KORZ paper
so can we all agree to blame Craig, even if it's actually Dave's fault?
James
https://www.google.com/search?client=safari&rls=en&q=%22slot+pair%22...
On 18/04/2020, at 17:40PM, Randy randy.smith@gmail.com wrote:
I’m sure you’re right about that Dave .. for practical reasons, it was embedded in the code somewhere.
—Randy
On Apr 17, 2020, at 7:45 PM, David Ungar ungar@mac.com wrote:
At the very least, there was *something* in the mirror and likely VM code so that when you removed “a”, “a:” was also removed (assuming it was an assignment slot).
On Apr 17, 2020, at 2:53 PM, Randy randy.smith@gmail.com wrote:
Steve—
Hey maybe you could grep through the documentation for that term? But again, my memory is the concept is not formally defined in Self. Dave may have used it in the code for mirrors, but in discussing language semantics, or program design, or object structure, there was no set aside special name.
In the original Self group, if anyone had said “the slot pair for X” while chatting about some self object, there is a fair chance the listener may have understood the intent. But depending on the conversational context, the term could also be confusing I suppose. (It might mean the x slot in an object together with an x slot in a parent, if you were talking about inheritance and overriding.)
—Randy
On Apr 17, 2020, at 1:57 PM, Steve Dekorte steve@dekorte.com wrote:
Hi all,
Glad to see this list is still active. I have a question about slot terminology:
Self referred to a given property (object table key/value ) as a “slot”. So, a Point might have a “x" getter slot and an “x:" setter slot. Was there any name you for the set of slots related to a property (in the Point case, the set of “x” and “x:” slots)?
David wasn’t sure but thought this might have been called a “slot pair” and used in mirrors. Is that right the correct term?
Steve
Self-interest mailing list Self-interest@lists.selflanguage.org http://lists.selflanguage.org/mailman/listinfo/self-interest
Self-interest mailing list Self-interest@lists.selflanguage.org http://lists.selflanguage.org/mailman/listinfo/self-interest
Self-interest mailing list Self-interest@lists.selflanguage.org http://lists.selflanguage.org/mailman/listinfo/self-interest
Thank you, James! You’ve put a smile on my face. Hope you and your loved ones are doing well these days.
- David
On Apr 18, 2020, at 1:17 AM, kjx kjx@ecs.vuw.ac.nz wrote:
I certainly understood the phrase as soon as it was mentioned on the list; and then started to recall things about "naked assignment slots that weren't supported. Can't remember too many more details about that --- basically could a naked assignment slot assign an immutable slot reached by delegation...
But then, I remembered the greatest legacy of the Self Project: Javascript! (argh!!)
OK: the second greatest legacy, Google.
"slot pair" is used in
- Chapter 4 of Craig's 1992 PhD thesis
- the ECOOP '95 retrospective "Programming as an Experience"
- the HOPLIII (2007) retrospective "Self"
- the 2014 KORZ paper
so can we all agree to blame Craig, even if it's actually Dave's fault?
James
https://www.google.com/search?client=safari&rls=en&q=%22slot+pair%22...
On 18/04/2020, at 17:40PM, Randy randy.smith@gmail.com wrote:
I’m sure you’re right about that Dave .. for practical reasons, it was embedded in the code somewhere.
—Randy
On Apr 17, 2020, at 7:45 PM, David Ungar ungar@mac.com wrote:
At the very least, there was *something* in the mirror and likely VM code so that when you removed “a”, “a:” was also removed (assuming it was an assignment slot).
On Apr 17, 2020, at 2:53 PM, Randy randy.smith@gmail.com wrote:
Steve—
Hey maybe you could grep through the documentation for that term? But again, my memory is the concept is not formally defined in Self. Dave may have used it in the code for mirrors, but in discussing language semantics, or program design, or object structure, there was no set aside special name.
In the original Self group, if anyone had said “the slot pair for X” while chatting about some self object, there is a fair chance the listener may have understood the intent. But depending on the conversational context, the term could also be confusing I suppose. (It might mean the x slot in an object together with an x slot in a parent, if you were talking about inheritance and overriding.)
—Randy
On Apr 17, 2020, at 1:57 PM, Steve Dekorte steve@dekorte.com wrote:
Hi all,
Glad to see this list is still active. I have a question about slot terminology:
Self referred to a given property (object table key/value ) as a “slot”. So, a Point might have a “x" getter slot and an “x:" setter slot. Was there any name you for the set of slots related to a property (in the Point case, the set of “x” and “x:” slots)?
David wasn’t sure but thought this might have been called a “slot pair” and used in mirrors. Is that right the correct term?
Steve
Self-interest mailing list Self-interest@lists.selflanguage.org http://lists.selflanguage.org/mailman/listinfo/self-interest
Self-interest mailing list Self-interest@lists.selflanguage.org http://lists.selflanguage.org/mailman/listinfo/self-interest
Self-interest mailing list Self-interest@lists.selflanguage.org http://lists.selflanguage.org/mailman/listinfo/self-interest
Nice to hear from you James! Thanks for the 4 references. I'm not sure there is much mileage left on what certainly is a small topic, but it's amusing to note that our two Self papers you cite mention "slot pair" at the end of a paragraph discussing assignment, remarking that the term could be defined and reified in a language, but in Self we don't do that!
Apparently Dave is saying that, well, actually, in Self mirror land (which he built so he should know) it *is* a named and reified concept! (I wonder how one might modify the UI to better represent it.)
The term appears only once in each of the other two references, with qualifying surrounding verbiage ( "...a data slot/assignment slot pair," and "a data slot pair (including an assignment slot).") despite substantial discussion of "assignment" and "assignment slot" in each.
So I think we should just blame James for muddying the waters, and let Dave and Craig off the hook.
On Apr 18, 2020, at 1:17 AM, kjx kjx@ecs.vuw.ac.nz wrote:
I certainly understood the phrase as soon as it was mentioned on the list; and then started to recall things about "naked assignment slots that weren't supported. Can't remember too many more details about that --- basically could a naked assignment slot assign an immutable slot reached by delegation...
But then, I remembered the greatest legacy of the Self Project: Javascript! (argh!!)
OK: the second greatest legacy, Google.
"slot pair" is used in
- Chapter 4 of Craig's 1992 PhD thesis
- the ECOOP '95 retrospective "Programming as an Experience"
- the HOPLIII (2007) retrospective "Self"
- the 2014 KORZ paper
so can we all agree to blame Craig, even if it's actually Dave's fault?
James
https://www.google.com/search?client=safari&rls=en&q=%22slot+pair%22...
On 18/04/2020, at 17:40PM, Randy randy.smith@gmail.com wrote:
I’m sure you’re right about that Dave .. for practical reasons, it was embedded in the code somewhere.
—Randy
On Apr 17, 2020, at 7:45 PM, David Ungar ungar@mac.com wrote:
At the very least, there was *something* in the mirror and likely VM code so that when you removed “a”, “a:” was also removed (assuming it was an assignment slot).
On Apr 17, 2020, at 2:53 PM, Randy randy.smith@gmail.com wrote:
Steve—
Hey maybe you could grep through the documentation for that term? But again, my memory is the concept is not formally defined in Self. Dave may have used it in the code for mirrors, but in discussing language semantics, or program design, or object structure, there was no set aside special name.
In the original Self group, if anyone had said “the slot pair for X” while chatting about some self object, there is a fair chance the listener may have understood the intent. But depending on the conversational context, the term could also be confusing I suppose. (It might mean the x slot in an object together with an x slot in a parent, if you were talking about inheritance and overriding.)
—Randy
On Apr 17, 2020, at 1:57 PM, Steve Dekorte steve@dekorte.com wrote:
Hi all,
Glad to see this list is still active. I have a question about slot terminology:
Self referred to a given property (object table key/value ) as a “slot”. So, a Point might have a “x" getter slot and an “x:" setter slot. Was there any name you for the set of slots related to a property (in the Point case, the set of “x” and “x:” slots)?
David wasn’t sure but thought this might have been called a “slot pair” and used in mirrors. Is that right the correct term?
Steve
Self-interest mailing list Self-interest@lists.selflanguage.org http://lists.selflanguage.org/mailman/listinfo/self-interest
Self-interest mailing list Self-interest@lists.selflanguage.org http://lists.selflanguage.org/mailman/listinfo/self-interest
Self-interest mailing list Self-interest@lists.selflanguage.org http://lists.selflanguage.org/mailman/listinfo/self-interest
Thanks to everyone for the responses and James for the references. I guess that’s where I should have started :)
It looks like “slot pair” was used in Self and made sense for it, but I’m still looking for a more general term as it doesn’t fit as well for other dynamic languages (such as Javascript or Python) where it's more of a slot triple of getter, setter, and value properties. Also, I’d like it to describe an abstraction around representing/managing a larger (and extensible) number of closely inter-related slots, such as slots which get called on mutation or access of the related value.
Steve
On Apr 18, 2020, at 12:17 PM, Randy Smith randy.smith@gmail.com wrote:
Nice to hear from you James! Thanks for the 4 references. I'm not sure there is much mileage left on what certainly is a small topic, but it's amusing to note that our two Self papers you cite mention "slot pair" at the end of a paragraph discussing assignment, remarking that the term could be defined and reified in a language, but in Self we don't do that!
Apparently Dave is saying that, well, actually, in Self mirror land (which he built so he should know) it *is* a named and reified concept! (I wonder how one might modify the UI to better represent it.)
The term appears only once in each of the other two references, with qualifying surrounding verbiage ( "...a data slot/assignment slot pair," and "a data slot pair (including an assignment slot).") despite substantial discussion of "assignment" and "assignment slot" in each.
So I think we should just blame James for muddying the waters, and let Dave and Craig off the hook.
On Apr 18, 2020, at 1:17 AM, kjx kjx@ecs.vuw.ac.nz wrote:
I certainly understood the phrase as soon as it was mentioned on the list; and then started to recall things about "naked assignment slots that weren't supported. Can't remember too many more details about that --- basically could a naked assignment slot assign an immutable slot reached by delegation...
But then, I remembered the greatest legacy of the Self Project: Javascript! (argh!!)
OK: the second greatest legacy, Google.
"slot pair" is used in
- Chapter 4 of Craig's 1992 PhD thesis
- the ECOOP '95 retrospective "Programming as an Experience"
- the HOPLIII (2007) retrospective "Self"
- the 2014 KORZ paper
so can we all agree to blame Craig, even if it's actually Dave's fault?
James
https://www.google.com/search?client=safari&rls=en&q=%22slot+pair%22...
On 18/04/2020, at 17:40PM, Randy randy.smith@gmail.com wrote:
I’m sure you’re right about that Dave .. for practical reasons, it was embedded in the code somewhere.
—Randy
On Apr 17, 2020, at 7:45 PM, David Ungar ungar@mac.com wrote:
At the very least, there was *something* in the mirror and likely VM code so that when you removed “a”, “a:” was also removed (assuming it was an assignment slot).
On Apr 17, 2020, at 2:53 PM, Randy randy.smith@gmail.com wrote:
Steve—
Hey maybe you could grep through the documentation for that term? But again, my memory is the concept is not formally defined in Self. Dave may have used it in the code for mirrors, but in discussing language semantics, or program design, or object structure, there was no set aside special name.
In the original Self group, if anyone had said “the slot pair for X” while chatting about some self object, there is a fair chance the listener may have understood the intent. But depending on the conversational context, the term could also be confusing I suppose. (It might mean the x slot in an object together with an x slot in a parent, if you were talking about inheritance and overriding.)
—Randy
> On Apr 17, 2020, at 1:57 PM, Steve Dekorte steve@dekorte.com wrote:
Hi all,
Glad to see this list is still active. I have a question about slot terminology:
Self referred to a given property (object table key/value ) as a “slot”. So, a Point might have a “x" getter slot and an “x:" setter slot. Was there any name you for the set of slots related to a property (in the Point case, the set of “x” and “x:” slots)?
David wasn’t sure but thought this might have been called a “slot pair” and used in mirrors. Is that right the correct term?
Steve
Self-interest mailing list Self-interest@lists.selflanguage.org http://lists.selflanguage.org/mailman/listinfo/self-interest
Self-interest mailing list Self-interest@lists.selflanguage.org http://lists.selflanguage.org/mailman/listinfo/self-interest
Self-interest mailing list Self-interest@lists.selflanguage.org http://lists.selflanguage.org/mailman/listinfo/self-interest
Self-interest mailing list Self-interest@lists.selflanguage.org http://lists.selflanguage.org/mailman/listinfo/self-interest
You could say that Self regarded a slot-pair as a dirty little secret elephant hiding in the closet. (I wonder how Ole Agesen’s type inferencer (see: Sifting Out the Gold) dealt with them.) An implementation detail, a mere bag of shells.
On Apr 18, 2020, at 12:17 PM, Randy Smith randy.smith@gmail.com wrote:
Nice to hear from you James! Thanks for the 4 references. I'm not sure there is much mileage left on what certainly is a small topic, but it's amusing to note that our two Self papers you cite mention "slot pair" at the end of a paragraph discussing assignment, remarking that the term could be defined and reified in a language, but in Self we don't do that!
Apparently Dave is saying that, well, actually, in Self mirror land (which he built so he should know) it *is* a named and reified concept! (I wonder how one might modify the UI to better represent it.)
The term appears only once in each of the other two references, with qualifying surrounding verbiage ( "...a data slot/assignment slot pair," and "a data slot pair (including an assignment slot).") despite substantial discussion of "assignment" and "assignment slot" in each.
So I think we should just blame James for muddying the waters, and let Dave and Craig off the hook.
On Apr 18, 2020, at 1:17 AM, kjx kjx@ecs.vuw.ac.nz wrote:
I certainly understood the phrase as soon as it was mentioned on the list; and then started to recall things about "naked assignment slots that weren't supported. Can't remember too many more details about that --- basically could a naked assignment slot assign an immutable slot reached by delegation...
But then, I remembered the greatest legacy of the Self Project: Javascript! (argh!!)
OK: the second greatest legacy, Google.
"slot pair" is used in
- Chapter 4 of Craig's 1992 PhD thesis
- the ECOOP '95 retrospective "Programming as an Experience"
- the HOPLIII (2007) retrospective "Self"
- the 2014 KORZ paper
so can we all agree to blame Craig, even if it's actually Dave's fault?
James
https://www.google.com/search?client=safari&rls=en&q=%22slot+pair%22...
On 18/04/2020, at 17:40PM, Randy randy.smith@gmail.com wrote:
I’m sure you’re right about that Dave .. for practical reasons, it was embedded in the code somewhere.
—Randy
On Apr 17, 2020, at 7:45 PM, David Ungar ungar@mac.com wrote:
At the very least, there was *something* in the mirror and likely VM code so that when you removed “a”, “a:” was also removed (assuming it was an assignment slot).
On Apr 17, 2020, at 2:53 PM, Randy randy.smith@gmail.com wrote:
Steve—
Hey maybe you could grep through the documentation for that term? But again, my memory is the concept is not formally defined in Self. Dave may have used it in the code for mirrors, but in discussing language semantics, or program design, or object structure, there was no set aside special name.
In the original Self group, if anyone had said “the slot pair for X” while chatting about some self object, there is a fair chance the listener may have understood the intent. But depending on the conversational context, the term could also be confusing I suppose. (It might mean the x slot in an object together with an x slot in a parent, if you were talking about inheritance and overriding.)
—Randy
> On Apr 17, 2020, at 1:57 PM, Steve Dekorte steve@dekorte.com wrote:
Hi all,
Glad to see this list is still active. I have a question about slot terminology:
Self referred to a given property (object table key/value ) as a “slot”. So, a Point might have a “x" getter slot and an “x:" setter slot. Was there any name you for the set of slots related to a property (in the Point case, the set of “x” and “x:” slots)?
David wasn’t sure but thought this might have been called a “slot pair” and used in mirrors. Is that right the correct term?
Steve
Self-interest mailing list Self-interest@lists.selflanguage.org http://lists.selflanguage.org/mailman/listinfo/self-interest
Self-interest mailing list Self-interest@lists.selflanguage.org http://lists.selflanguage.org/mailman/listinfo/self-interest
Self-interest mailing list Self-interest@lists.selflanguage.org http://lists.selflanguage.org/mailman/listinfo/self-interest
Self-interest mailing list Self-interest@lists.selflanguage.org http://lists.selflanguage.org/mailman/listinfo/self-interest
Nice to hear from you James! Thanks for the 4 references. I'm not sure there is much mileage left on what certainly is a small topic, but it's amusing to note that our two Self papers you cite mention "slot pair" at the end of a paragraph discussing assignment, remarking that the term could be defined and reified in a language, but in Self we don't do that!
Apparently Dave is saying that, well, actually, in Self mirror land (which he built so he should know) it *is* a named and reified concept! (I wonder how one might modify the UI to better represent it.)
The term appears only once in each of the other two references, with qualifying surrounding verbiage ( "...a data slot/assignment slot pair," and "a data slot pair (including an assignment slot).") despite substantial discussion of "assignment" and "assignment slot" in each.
So I think we should just blame James for muddying the waters, and let Dave and Craig off the hook.
On Apr 18, 2020, at 1:17 AM, kjx kjx@ecs.vuw.ac.nz wrote:
I certainly understood the phrase as soon as it was mentioned on the list; and then started to recall things about "naked assignment slots that weren't supported. Can't remember too many more details about that --- basically could a naked assignment slot assign an immutable slot reached by delegation...
But then, I remembered the greatest legacy of the Self Project: Javascript! (argh!!)
OK: the second greatest legacy, Google.
"slot pair" is used in
- Chapter 4 of Craig's 1992 PhD thesis
- the ECOOP '95 retrospective "Programming as an Experience"
- the HOPLIII (2007) retrospective "Self"
- the 2014 KORZ paper
so can we all agree to blame Craig, even if it's actually Dave's fault?
James
https://www.google.com/search?client=safari&rls=en&q=%22slot+pair%22...
On 18/04/2020, at 17:40PM, Randy randy.smith@gmail.com wrote:
I’m sure you’re right about that Dave .. for practical reasons, it was embedded in the code somewhere.
—Randy
On Apr 17, 2020, at 7:45 PM, David Ungar ungar@mac.com wrote:
At the very least, there was *something* in the mirror and likely VM code so that when you removed “a”, “a:” was also removed (assuming it was an assignment slot).
On Apr 17, 2020, at 2:53 PM, Randy randy.smith@gmail.com wrote:
Steve—
Hey maybe you could grep through the documentation for that term? But again, my memory is the concept is not formally defined in Self. Dave may have used it in the code for mirrors, but in discussing language semantics, or program design, or object structure, there was no set aside special name.
In the original Self group, if anyone had said “the slot pair for X” while chatting about some self object, there is a fair chance the listener may have understood the intent. But depending on the conversational context, the term could also be confusing I suppose. (It might mean the x slot in an object together with an x slot in a parent, if you were talking about inheritance and overriding.)
—Randy
On Apr 17, 2020, at 1:57 PM, Steve Dekorte steve@dekorte.com wrote:
Hi all,
Glad to see this list is still active. I have a question about slot terminology:
Self referred to a given property (object table key/value ) as a “slot”. So, a Point might have a “x" getter slot and an “x:" setter slot. Was there any name you for the set of slots related to a property (in the Point case, the set of “x” and “x:” slots)?
David wasn’t sure but thought this might have been called a “slot pair” and used in mirrors. Is that right the correct term?
Steve
Self-interest mailing list Self-interest@lists.selflanguage.org http://lists.selflanguage.org/mailman/listinfo/self-interest
Self-interest mailing list Self-interest@lists.selflanguage.org http://lists.selflanguage.org/mailman/listinfo/self-interest
Self-interest mailing list Self-interest@lists.selflanguage.org http://lists.selflanguage.org/mailman/listinfo/self-interest
_______________________________________________ Self-interest mailing list Self-interest@lists.selflanguage.org http://lists.selflanguage.org/mailman/listinfo/self-interest
self-interest@lists.selflanguage.org
-
David Ungar -
Douglas Simons -
kjx -
Randy Smith -
Steve Dekorte