- Pharo-dev - lists.pharo.org

Re: [Pharo-project] Interval protocol

by StÃ©phane Ducasse

On Dec 16, 2012, at 4:21 PM, Nicolas Cellier wrote: > Ah, i also forgot this feature: > > (2 to: 3) at: 1.5 -> 2.5 > > It's fun I'm not sure that this is the adjective that came to my mind.

Dec. 16, 2012

Re: [Pharo-project] Interval protocol

by StÃ©phane Ducasse

On Dec 16, 2012, at 4:11 PM, Sven Van Caekenberghe wrote: > I am probably stupid but this sounds crazy. > > Most people will have a pretty good idea of what an interval is from mathematics. > Stating that it is a finite, countable progression from a start to a stop with a step is more precise and makes sense. > > Now, [a,b] with b<a will seem wrong from this perspective. > > It could well be that Interval is (ab)used somewhere with b<a but is that a reason to allow this ? > > IMO, the specific cursor usage should have its own data structureâ¦ Yes :) I'm as stupid as you :)

Dec. 16, 2012

[Pharo-project] about isAbstractClass

by StÃ©phane Ducasse

Hi guys I will deprecate isAbstractClass and introduce defineAbstractMethods. why? because as soon as we have an abstract method in Object nearly all the classes becomes abstract and this does not make sense. Now I can understand that we want to know if a class defineAbstractMethods and sometimes taking into account inheritance. http://code.google.com/p/pharo/issues/detail?id=7123 Stef

Dec. 16, 2012

Re: [Pharo-project] Interval protocol

by Nicolas Cellier

Ah, i also forgot this feature: (2 to: 3) at: 1.5 -> 2.5 It's fun 2012/12/16 StÃ©phane Ducasse <stephane.ducasse(a)inria.fr>: > >>> For hash, I wonder if it would not be easier to just drop #(2 3) = (2 >>> to: 3) and let it answer false. > > + 1 > to me there are not equal! > > Stef > >>> We have hasSameElementsThan: or isSameSequenceThan: for that purpose. >>> >> or we haven't (? can't remember), but should have. >> >>> Nicolas >>> >>>> >>>> >> > >

Dec. 16, 2012

Re: [Pharo-project] Interval protocol

by StÃ©phane Ducasse

>> For hash, I wonder if it would not be easier to just drop #(2 3) = (2 >> to: 3) and let it answer false. + 1 to me there are not equal! Stef >> We have hasSameElementsThan: or isSameSequenceThan: for that purpose. >> > or we haven't (? can't remember), but should have. > >> Nicolas >> >>> >>> >

Dec. 16, 2012

Re: [Pharo-project] Interval protocol

by StÃ©phane Ducasse

O > > Currently: > > (3 to: 2) first -> 3. > (3 to: 2) last -> 2. > > But: > (2 to: 4 by: 3) last -> 2 > > Try to guess the logic behind thisâ¦ ;D > >> >> I see. But does it make sense to have (thinking aloud)? >> (3 to: 2) start -> 3 >> (3 to: 2) end -> 2 >> probably? > > Yes, I did that refactoring in Squeak 3.9, but I was sure it would be > rejected, too many methods were changed (TextMorph ParagraphEditor & > co). See http://bugs.squeak.org/view.php?id=6454 > > You also know that breaking the classes we use for browsing/debugging > is a dangerous sport ;) oh yes like CharacterScanner and Event :) > The simplest thing would be to add a few message to TextLineInterval > (first, last, ...) and move it somewhere else in hierarchy, because I > don't think it shares that much with Interval. > - I guess we don't use much of Collection protocol in TextLineInterval > - it spoils first/last uniformity > - it can't handle Interval step > - ... > >> >> So we should improve that. >> If what I said is correct and you confirm it I will open a bug entry >> Stef >> >> > > I found http://bugs.squeak.org/view.php?id=7002 indicating that most > weirdness of Interval were corrected, but I'm quite sure it still has > other reported or non reported bugs, except hash. > > I can still do an infinite loop with to:do: and Float, > 2 - Float epsilon to: 2 by: Float epsilon/4 do: [:i | Transcript cr; > show: i printString] > > but Interval>>do: is correct: (well I warned you, some rounding occurs...) > (2 - Float epsilon to: 2 by: Float epsilon/4) do: [:i | Transcript cr; > show: i printString] > > For hash, I wonder if it would not be easier to just drop #(2 3) = (2 > to: 3) and let it answer false. > We have hasSameElementsThan: or isSameSequenceThan: for that purpose. If you have changes that you want to propose for Pharo please go. I think that I will spend most of my charisma holidays fixing bugs :) Stef

Dec. 16, 2012

Re: [Pharo-project] Interval protocol

by Sven Van Caekenberghe

I am probably stupid but this sounds crazy. Most people will have a pretty good idea of what an interval is from mathematics. Stating that it is a finite, countable progression from a start to a stop with a step is more precise and makes sense. Now, [a,b] with b<a will seem wrong from this perspective. It could well be that Interval is (ab)used somewhere with b<a but is that a reason to allow this ? IMO, the specific cursor usage should have its own data structureâ¦ On 16 Dec 2012, at 11:04, Nicolas Cellier <nicolas.cellier.aka.nice(a)gmail.com> wrote: > What Frank means is that Smalltalk Interval are countable (and most are finite). > > However, the original usage of Interval was for Text editing, and an > Interval from 5 to: 4 means that the cursor is before (after?) > position 5. That's why Interval is a strange beast, not exactly > polymorphic to an Array of Number. Those two views have introduced a > number of unconsistencies in the library... > > Last thing, I don't recommend using Interval of Float, (1.5 to: 1.6 > by: 0.01), rather use Integer or Fraction. You never know where the > Float Interval really ends, it's progression is generally not > constant, and to:do: used to behaved differently than do: > > Nicolas > > 2012/12/16 Igor Stasenko <siguctua(a)gmail.com>: >> On 16 December 2012 09:26, Frank Shearar <frank.shearar(a)gmail.com> wrote: >>> On 16 December 2012 01:06, Igor Stasenko <siguctua(a)gmail.com> wrote: >>>> On 15 December 2012 23:45, Chris Muller <asqueaker(a)gmail.com> wrote: >>>>>> and i always hitting the wall with my head again and again.. because >>>>>> there's no such protocol >>>>>> and instead (after opening browser and looking at class, i figuring >>>>>> that i should use #first/#last, but it is completely unintuitive to >>>>>> me).. >>>>> >>>>> Me too, when I was using what I knew would only ever need to be, >>>>> simply, an Interval. >>>>> >>>>> Later the complexity increased and it would be either a Interval or >>>>> Array of numbers. It was a transparent / painless improvement. After >>>>> gaining that appreciation, now I think of Intervals as just "efficient >>>>> Arrays" and think naturally to use #first / #last. >>>>> >>>> >>>> Well, my math background prevents me from thinking this way. >>>> Since in math, intervals are defined on sets of numbers. >>>> (like Real set).. you cannot treat it as array , simply because there >>>> is an infinite number >>>> of values inside any non-empty interval, and you cannot enumerate them. >>> >>> But it's a subset of the integers, not the reals! >>> >> what? >> then how you explain this: >> >> (1.5 to: 1.6 by: 0.01) collect: [:i | i ] >> >>> frank >>> >>>> -- >>>> Best regards, >>>> Igor Stasenko. >>>> >>> >> >> >> >> -- >> Best regards, >> Igor Stasenko. >> >

Dec. 16, 2012

Re: [Pharo-project] Interval protocol

by Nicolas Cellier

2012/12/16 Nicolas Cellier <nicolas.cellier.aka.nice(a)gmail.com>: > 2012/12/16 StÃ©phane Ducasse <stephane.ducasse(a)inria.fr>: >> >> On Dec 16, 2012, at 11:44 AM, Nicolas Cellier wrote: >> >>> 2012/12/16 StÃ©phane Ducasse <stephane.ducasse(a)inria.fr>: >>>> polymorphically they are a collection with a first and last :) >>>> >>> >>> No. >>> >>> (5 to: 4) is different than (3 to: 2) because they specify a different >>> cursor position. >> >> Ok that I understand >> >>> Viewed as collection, they have no first nor last, they're just empty >>> >>> Nicolas > > Currently: > > (3 to: 2) first -> 3. > (3 to: 2) last -> 2. > > But: > (2 to: 4 by: 3) last -> 2 > > Try to guess the logic behind this... > >> >> I see. But does it make sense to have (thinking aloud)? >> (3 to: 2) start -> 3 >> (3 to: 2) end -> 2 >> probably? > > Yes, I did that refactoring in Squeak 3.9, but I was sure it would be > rejected, too many methods were changed (TextMorph ParagraphEditor & > co). See http://bugs.squeak.org/view.php?id=6454 > > You also know that breaking the classes we use for browsing/debugging > is a dangerous sport ;) > > The simplest thing would be to add a few message to TextLineInterval > (first, last, ...) and move it somewhere else in hierarchy, because I > don't think it shares that much with Interval. > - I guess we don't use much of Collection protocol in TextLineInterval > - it spoils first/last uniformity > - it can't handle Interval step > - ... > >> >> So we should improve that. >> If what I said is correct and you confirm it I will open a bug entry >> Stef >> >> > > I found http://bugs.squeak.org/view.php?id=7002 indicating that most > weirdness of Interval were corrected, but I'm quite sure it still has > other reported or non reported bugs, except hash. > > I can still do an infinite loop with to:do: and Float, > 2 - Float epsilon to: 2 by: Float epsilon/4 do: [:i | Transcript cr; > show: i printString] > > but Interval>>do: is correct: (well I warned you, some rounding occurs...) > (2 - Float epsilon to: 2 by: Float epsilon/4) do: [:i | Transcript cr; > show: i printString] > > For hash, I wonder if it would not be easier to just drop #(2 3) = (2 > to: 3) and let it answer false. > We have hasSameElementsThan: or isSameSequenceThan: for that purpose. > or we haven't (? can't remember), but should have. > Nicolas > >> >>

Dec. 16, 2012

Re: [Pharo-project] CI down?

by StÃ©phane Ducasse

Try https://ci.inria.fr/pharo/job/Pharo-2.0/lastSuccessfulBuild/artifact/ > I wanted to grab the latest build to check for a possible bug and the CI seems to be downâ¦ > > Max

Dec. 16, 2012

Re: [Pharo-project] Interval protocol

by Nicolas Cellier

2012/12/16 StÃ©phane Ducasse <stephane.ducasse(a)inria.fr>: > > On Dec 16, 2012, at 11:44 AM, Nicolas Cellier wrote: > >> 2012/12/16 StÃ©phane Ducasse <stephane.ducasse(a)inria.fr>: >>> polymorphically they are a collection with a first and last :) >>> >> >> No. >> >> (5 to: 4) is different than (3 to: 2) because they specify a different >> cursor position. > > Ok that I understand > >> Viewed as collection, they have no first nor last, they're just empty >> >> Nicolas Currently: (3 to: 2) first -> 3. (3 to: 2) last -> 2. But: (2 to: 4 by: 3) last -> 2 Try to guess the logic behind this... > > I see. But does it make sense to have (thinking aloud)? > (3 to: 2) start -> 3 > (3 to: 2) end -> 2 > probably? Yes, I did that refactoring in Squeak 3.9, but I was sure it would be rejected, too many methods were changed (TextMorph ParagraphEditor & co). See http://bugs.squeak.org/view.php?id=6454 You also know that breaking the classes we use for browsing/debugging is a dangerous sport ;) The simplest thing would be to add a few message to TextLineInterval (first, last, ...) and move it somewhere else in hierarchy, because I don't think it shares that much with Interval. - I guess we don't use much of Collection protocol in TextLineInterval - it spoils first/last uniformity - it can't handle Interval step - ... > > So we should improve that. > If what I said is correct and you confirm it I will open a bug entry > Stef > > I found http://bugs.squeak.org/view.php?id=7002 indicating that most weirdness of Interval were corrected, but I'm quite sure it still has other reported or non reported bugs, except hash. I can still do an infinite loop with to:do: and Float, 2 - Float epsilon to: 2 by: Float epsilon/4 do: [:i | Transcript cr; show: i printString] but Interval>>do: is correct: (well I warned you, some rounding occurs...) (2 - Float epsilon to: 2 by: Float epsilon/4) do: [:i | Transcript cr; show: i printString] For hash, I wonder if it would not be easier to just drop #(2 3) = (2 to: 3) and let it answer false. We have hasSameElementsThan: or isSameSequenceThan: for that purpose. Nicolas > >

Dec. 16, 2012