ANSI is dead and this is good. Now this is interesting to learn from the analysis and the APIs. > On 30 Jun 2025, at 14:23, Richard O'Keefe <raoknz@gmail.com> wrote: > > I have the final ANSI Smalltalk standard as published. I obviously > cannot give it to you, but I *can* tell you that painfully few of the > obvious errors in the 1.9 draft were corrected. For example, the > DateAndTime class. The basic idea is sound, but we are told that the > range of zone offsets is -12 hours to +12 hours. Well, I live in a > country where the local time is UTC+13 hours for half the year and > part of the country gets up to UTC+13 hours 45 minutes. Then there's > the Line Islands that get up to +14 hours... Worse than that, the > operations on DateAndTime literally *cannot* be implemented as > specified, as they require you to know the leap second rules for > thousands of years in the future. And the standard appears to assume > that all places where the offset from UTC is such and such at one time > has the same offset at all times. It sort of muddles up time zones > with zone OFFSETS and then forgets about zones. The standard also > suffers from being pre-Unicode, so it effectively assumes that random > access by bytes makes sense in external text files. > > There's another problem with the standard, which is that nobody much > seems to have care about it once they'd finished with it. > Case in point: the way to open a file in ANSI Smalltalk is to send a > message to FileStream. Note only did Squeak and Pharo never bother to > implement the standard messages for this purpose, Pharo has even > deleted the class. > > I'll get back to this tomorrow but my battery is about to say bye-bye. > > On Sun, 29 Jun 2025 at 23:29, Alok <zhalok24@gmail.com> wrote: >> >> Thnx a lot Richard for such great feedback. We'll definitely update the repo with all the necessary changes. >> >> Thanks Stephane for sharing all the resources. I'll definitely go through ANSI :) >> >> >> On Sun, 29 Jun, 2025, 13:40 Sebastian Jordan, <sebastian.jordan@inria.fr> wrote: >>> >>> Hello Richard, >>> >>> Thanks for your feedback. Just for context we are doing this project for the Google Summer of Code. I agree that the buffer should be polymorphic with the OrderedCollection. >>> Yes, we should signal empty collection when accesing an empty buffer (7) >>> >>> We will rok on that and come back to you >>> Sebastian >>> >>> Le 27 juin 2025 à 12:23, Richard O'Keefe via Pharo-users <pharo-users@lists.pharo.org> a écrit : >>> >>> My own Smalltalk library (and yes, I've tried to put it on github, I >>> don't know what I'm doing wrong) >>> includes Deque and BoundedDeque, both descendants of Collection. >>> Using addLast/removeLast gives you a stack (use #last for peeking) >>> Using addLast/removeFirst gives you a queue. (use #first for peeking) >>> >>> (1) I am puzzled why there are separate FIFO and LIFO classes rather >>> than a single BoundedDeque. >>> -- This has implications for performance. >>> (2) I am puzzled why #withCapacity: is used rather than #new:, >>> familiar from OrderedCollection. >>> -- These two points together make it hard to just swap >>> OrderedCollection and ?IFOBuffer. >>> (3) I am puzzled why #clear is used instead of the familiar #removeAll. >>> -- See note on question 2. >>> (4) I am extremely puzzled why ALL, like ALL, of the internals of the >>> data structure are exposed. >>> Did encapsulation fall out of favour and I didn't get the memo? >>> (5) It looks as though calling #capacity: at the wrong time can >>> destroy the integrity of one of >>> these containers, but there is nothing sayiing "don't do that". >>> (6) I am puzzled that they are not Collections. >>> (7) I am puzzled why trying to access an element in an empty buffer >>> does not signal >>> a CollectionIsEmpty exception >>> -- Is this related to (6)? >>> >>> The structure, with two separate classes and key performance-essential >>> methods being >>> template methods, hurts performance by about a factor of two in my tests. >>> >>> Now Pharo has a design MOOC and if Stephane Ducasse says "this is >>> great", these things >>> that puzzle me must be good design. I would like to improve my >>> skills, so *why* is this good design? >>> (As it happens, I *have* had occasion to 'push' from both ends of a >>> single Deque.) >>> >>> None of this is meant as criticism of the generosity or competence of >>> the authors. It expresses >>> genuine puzzlement. Like when I implemented deques I could not >>> imagine not making them >>> Collections. Principle of Least Surprise and all that. Maybe I >>> should be thinking differently. >>> >>> >>> On Fri, 27 Jun 2025 at 20:10, stephane ducasse via Pharo-users >>> <pharo-users@lists.pharo.org> wrote: >>> >>> >>> Thanks this is great! >>> >>> On 18 Jun 2025, at 12:13, Alok via Pharo-users <pharo-users@lists.pharo.org> wrote: >>> >>> Hello Everyone, >>> We're excited to share a new addition to the pharo-containers. An efficient Circular Buffer implementation, developed as part of Google Summer of Code 2025 project under the mentorship of Gordana Rakic and Sebastian Jordan Montaño. >>> >>> This package provides fixed-size buffers supporting both FIFO (queue-style) and LIFO (stack-style) behavior. It’s designed for use cases such as streaming data, undo/redo functionality, chat or browser history & more. >>> >>> You can find the repo here: Containers-Buffer >>> The README includes usage examples, installation steps etc. >>> >>> Feedback, suggestions, and contributions are very welcome ! >>> ThankYou ! >>> Alok Pathak >>> GSoC'25 Contributor >>> >>> >>> Stéphane Ducasse >>> http://stephane.ducasse.free.fr >>> 06 30 93 66 73 >>> >>> "If you knew today was your last day on earth, what would you do differently? ....ESPECIALLY if, by doing something different, today might not be your last day on earth.” Calvin & Hobbes >>> >>> >>> >>> >>> >>> Stéphane Ducasse http://stephane.ducasse.free.fr 06 30 93 66 73 "If you knew today was your last day on earth, what would you do differently? ....ESPECIALLY if, by doing something different, today might not be your last day on earth.” Calvin & Hobbes

As promised, here is a further issue from the standard, where I think the standard has a good idea but nobody paid any attention. Smalltalk has a practice of providing conversions between classes when that makes sense. They are named #asResult where Result is the result of the conversion. The "as" prefix is used for no other purpose. In the caseut of things like numbers, we expect that the result will be the closest representable value to the original, and since numbers are theoretically immutable, it doesn't matter whether the receiver or a copy of it is returned when the receiver is already what's wanted. But what about when it *does* matter? What about things like #asArray, #asBag, #asBytearray, #asOrderedCollection, #asSortedCollection, #asSortedCollection:, and #asString? You see, looking at CTFIFOBuffer I noticed the absence of any conversion methods to or from that class, and remember that I didn't have conversion methods to Deque or BoundedDeque either (though those classes inherited a full set of conversion methods *to* them from Collection). So I wondered what guidance the standard provided. Should I just have Collection>>asDeque ^Deque withAll: self or should I override this with Deque>>asDeque ^self On reading the standard closely, I realised two things. First, it took great care to avoid ruling on this. That was a warning sign to me. It indicated a lot of variation between existing Smalltalks of the day with nobody willing to change to a common behaviour. Second, the standard *did* specify an overriding for #asSortedcollection. It says that the result should have the *same* sortBlock as the receiver. SortedCollection>>asSortedCollection ^self copy So I investigated running versions of Smalltalk-80, Squeak, Pharo, GNU Smalltalk, Dolphin [%], VisualWorks, VisualAge, and Smalltalk/X. I found a complete lack of consistency both between Smalltalks and within each Smalltalk (with the single exception of GNU Smalltalk, which always returns a new collection, never the receiver). I also found that nobody had SortedCollection>>asSortedCollection doing the "standard" thing (with the single exception of astc, which does, NOW).. [%] I cannot get Dolphin 8 running on any of my Windows boxes because installing some M$ DLL it depends on doesn't seem to do what it expects. So I had to read the sources. Since I started astc before Pharo existed, I've had a policy of "follow the standard if it's definite and possible, otherwise Squeak". So some of my conversion methods return the receiver when I kind of wish they didn't, but SortedCollection>>asSortedCollection preserves the sortBlock. Now there's a reason for that. In fact there are two. (0) The standard may be intending to allow SortedCollection>>asSortedCollection ^self (1) Preserving the sortBlock means that a *new* SortedCollection can be constructed in linear time without sorting (2) What I call The Principle of Least Crashing: an operation shouldn't fail without good reason. Given a SortedCollection, we have no reason to believe that the default sort block would not crash on its elements. If you supply an explicit sort block, you are taking responsibility for it not crashing. I think it would have been a very good thing had the Smalltalk standard been taken seriously by the people who put it together and by the implementors of the day. We are now in a situation where there is no portable way to open a file, where ScaledDecimal doesn't even *exist* in some Smalltalk systems and is implemented inconsistently in others (and the fact that the ANSI Standard defers definition of ScaledDecimal to the Language Independent Arithmetic standard, which explicitly has nothing to say about the subject, shows me just how seriously the standardisers took their task), oh, I could go on. The ANSI standard left everyone free to do there own user interface, networking, data base, version management, &c &c. But it *does*, with a bit of charitable reading, define a useful base. Trying to implement it has taught me a lot. On Tue, 1 Jul 2025 at 00:23, Richard O'Keefe <raoknz@gmail.com> wrote: > > I have the final ANSI Smalltalk standard as published. I obviously > cannot give it to you, but I *can* tell you that painfully few of the > obvious errors in the 1.9 draft were corrected. For example, the > DateAndTime class. The basic idea is sound, but we are told that the > range of zone offsets is -12 hours to +12 hours. Well, I live in a > country where the local time is UTC+13 hours for half the year and > part of the country gets up to UTC+13 hours 45 minutes. Then there's > the Line Islands that get up to +14 hours... Worse than that, the > operations on DateAndTime literally *cannot* be implemented as > specified, as they require you to know the leap second rules for > thousands of years in the future. And the standard appears to assume > that all places where the offset from UTC is such and such at one time > has the same offset at all times. It sort of muddles up time zones > with zone OFFSETS and then forgets about zones. The standard also > suffers from being pre-Unicode, so it effectively assumes that random > access by bytes makes sense in external text files. > > There's another problem with the standard, which is that nobody much > seems to have care about it once they'd finished with it. > Case in point: the way to open a file in ANSI Smalltalk is to send a > message to FileStream. Note only did Squeak and Pharo never bother to > implement the standard messages for this purpose, Pharo has even > deleted the class. > > I'll get back to this tomorrow but my battery is about to say bye-bye. > > On Sun, 29 Jun 2025 at 23:29, Alok <zhalok24@gmail.com> wrote: > > > > Thnx a lot Richard for such great feedback. We'll definitely update the repo with all the necessary changes. > > > > Thanks Stephane for sharing all the resources. I'll definitely go through ANSI :) > > > > > > On Sun, 29 Jun, 2025, 13:40 Sebastian Jordan, <sebastian.jordan@inria.fr> wrote: > >> > >> Hello Richard, > >> > >> Thanks for your feedback. Just for context we are doing this project for the Google Summer of Code. I agree that the buffer should be polymorphic with the OrderedCollection. > >> Yes, we should signal empty collection when accesing an empty buffer (7) > >> > >> We will rok on that and come back to you > >> Sebastian > >> > >> Le 27 juin 2025 à 12:23, Richard O'Keefe via Pharo-users <pharo-users@lists.pharo.org> a écrit : > >> > >> My own Smalltalk library (and yes, I've tried to put it on github, I > >> don't know what I'm doing wrong) > >> includes Deque and BoundedDeque, both descendants of Collection. > >> Using addLast/removeLast gives you a stack (use #last for peeking) > >> Using addLast/removeFirst gives you a queue. (use #first for peeking) > >> > >> (1) I am puzzled why there are separate FIFO and LIFO classes rather > >> than a single BoundedDeque. > >> -- This has implications for performance. > >> (2) I am puzzled why #withCapacity: is used rather than #new:, > >> familiar from OrderedCollection. > >> -- These two points together make it hard to just swap > >> OrderedCollection and ?IFOBuffer. > >> (3) I am puzzled why #clear is used instead of the familiar #removeAll. > >> -- See note on question 2. > >> (4) I am extremely puzzled why ALL, like ALL, of the internals of the > >> data structure are exposed. > >> Did encapsulation fall out of favour and I didn't get the memo? > >> (5) It looks as though calling #capacity: at the wrong time can > >> destroy the integrity of one of > >> these containers, but there is nothing sayiing "don't do that". > >> (6) I am puzzled that they are not Collections. > >> (7) I am puzzled why trying to access an element in an empty buffer > >> does not signal > >> a CollectionIsEmpty exception > >> -- Is this related to (6)? > >> > >> The structure, with two separate classes and key performance-essential > >> methods being > >> template methods, hurts performance by about a factor of two in my tests. > >> > >> Now Pharo has a design MOOC and if Stephane Ducasse says "this is > >> great", these things > >> that puzzle me must be good design. I would like to improve my > >> skills, so *why* is this good design? > >> (As it happens, I *have* had occasion to 'push' from both ends of a > >> single Deque.) > >> > >> None of this is meant as criticism of the generosity or competence of > >> the authors. It expresses > >> genuine puzzlement. Like when I implemented deques I could not > >> imagine not making them > >> Collections. Principle of Least Surprise and all that. Maybe I > >> should be thinking differently. > >> > >> > >> On Fri, 27 Jun 2025 at 20:10, stephane ducasse via Pharo-users > >> <pharo-users@lists.pharo.org> wrote: > >> > >> > >> Thanks this is great! > >> > >> On 18 Jun 2025, at 12:13, Alok via Pharo-users <pharo-users@lists.pharo.org> wrote: > >> > >> Hello Everyone, > >> We're excited to share a new addition to the pharo-containers. An efficient Circular Buffer implementation, developed as part of Google Summer of Code 2025 project under the mentorship of Gordana Rakic and Sebastian Jordan Montaño. > >> > >> This package provides fixed-size buffers supporting both FIFO (queue-style) and LIFO (stack-style) behavior. It’s designed for use cases such as streaming data, undo/redo functionality, chat or browser history & more. > >> > >> You can find the repo here: Containers-Buffer > >> The README includes usage examples, installation steps etc. > >> > >> Feedback, suggestions, and contributions are very welcome ! > >> ThankYou ! > >> Alok Pathak > >> GSoC'25 Contributor > >> > >> > >> Stéphane Ducasse > >> http://stephane.ducasse.free.fr > >> 06 30 93 66 73 > >> > >> "If you knew today was your last day on earth, what would you do differently? ....ESPECIALLY if, by doing something different, today might not be your last day on earth.” Calvin & Hobbes > >> > >> > >> > >> > >> > >>