February 2015 - Pharo-users

a successful success story

by Alain Busser

It is in French so I try to translate some parts. My goal was to teach as fast as possible the binomial distribution (number of successes when independently repeating a Bernoulli experiment) to students who, either have never seen an algorithm, or have seen some algorithmics in math lessons and don't like it at all. Both categories loved what I did with them. I precise that almost all of them are allergic to english language. The assignment was "throwing 10 dices, what is the probability of having at least 2 times "6" around the 10 results?" The whole job is related here: http://irem.univ-reunion.fr/IMG/pdf/binomiale.pdf (but I did only pages 1-5 in one hour) The first step was to show how one can simulate the throwing of 10 dices. I can do it without having to make a variable vary (explicitely at least), thanks to Seymour Papert: 10 timesRepeat: [ Transcript show: (6 atRandom)]. Then I spoke about the usefulness of the Bag object when there are statistics: Adding the 10 results (small integers) inside a Bag makes it an object where one can see and count the 6es: | urn | urn := #( ) asBag. 10 timesRepeat: [ urn add: (6 atRandom)]. Transcript show: urn. (Why "urn"? This is the word used by Bernoulli, he was thinking about balls concealed inside a vase for which the latin word was "urna"; the balls where called "billets" by Condorcet, so that they could instead be pieces of paper; whatever they were, a vase was use to conceal them) Then there remains only one difficulty: How can I automatize the counting of the 6es inside the bag? There is the most frightening challenge: Use a block with students who have never seen Pharo before. Very good surprise: They didn't seem to find this that difficult, I just had to say that 1. The pipe symbol means "such that" like in my lessons 2. The first time I use the letter "x" I have to tell Pharo that it is yet no variable to evaluate, hence the need to precede it with a colon. With that in mind, counting the 6es is the same as counting the x-es such x equals 6: | urn | urn := #( ) asBag. 10 timesRepeat: [ urn add: (6 atRandom)]. Transcript show: (urn count: [ :x | x=6]). Afterwards it remained only to repeat the whole experiment and, instead of showing the number of 6es, adding it to another bag called "stats", then displaying this bag. The end was made with a tool which exists only in mathsOntologie: The displaying of a bar chart in a translucent Morph, so that the most rapid of the students could compare several bar charts by superposition and guess the shape of the distribution. Once again, they all loved it, either because they found Pharo a better tool than whatever they knew before (and disliked), or because they knew nothing similar before (one exception: A girl who never learned algorithmics but just happens to love playing with a computer, here she is: http://reunion.la1ere.fr/2014/06/04/miss-reunion-muriel-roger-candidate-nde… ). I insist on the fact that they never saw anything like Smalltalk before, and did not seem to "rtfm": Is it because Smalltalk is a natural language or because MathsOntologie mimics the french language, I lack elements of comparison to answer to this. But I insist also to say that these students seem to dislike probability theory too. Finally, some stats about these stats: I can rapidly find the bugs when the text appears in red, and most of the time a point was missing at the end of a line, or a right bracket was missing or misplaced. Sometimes they add a space inside a word like asBag which becomes "as Bag". Alain Busser IREM La RÃ©union

Feb. 12, 2015

Re: [Pharo-users] SmaCC: First steps

by Trygve Reenskaug

The problem with reading class oriented code (C++, Java, C#, Ruby, ..., Smalltalk) is that the code does not reveal how the system will work at runtime (polymorphism). The essence of object orientation is that objects collaborate to achieve a goal. The DCI programming paradigm adds code for how the system works at runtime: * A class says everything about the inner construction of an object but nothing about how it is used in its interaction with other objects. * The new DCI context says everything about how objects are used when interacting with other objects but nothing about their insides. Class and context separates the code into two materially independent concerns: What the system IS (classes) and what the system DOES (Contexts). BabyIDE is a programming environment that supports the DCI paradigm. It is available in Squeak 4.5 and may be available in Pharo later. Cheers --Trygve On 11.02.2015 18:36, kilon alios wrote: > Well I tried in the past to use openqwaq , but what I learned from the > experience is that sometimes understanding others code can be even > more time consuming than remaking it yourself. I spent like an > afternoon trying to understand openqwaq before starting Ephestos from > complete scratch. Suffice to say that in another afternoon I had the > prototype of Ephestos working with the ability to send simple python > strings to the cptyhon interpreter to be execute by the time I was not > even able to understand the basic architecture of openqwaq. > > The module you recommended alone contains like 11.000 lines of code. > Thats a lot of work to just read that code and I cant find any class > named CModel. In any case I like it that I started from scratch, I > want to keep my project small, light, well documented and super easy > to hack. Even if that means less features :) -- /The essence of object orientation is that objects collaborateto achieve a goal. / Trygve Reenskaug mailto: trygver(a)ifi.uio.no <mailto:%20trygver@ifi.uio.no> Morgedalsvn. 5A http://folk.uio.no/trygver/ N-0378 Oslo http://fullOO.info Norway Tel: (+47) 22 49 57 27

Feb. 12, 2015

Re: [Pharo-users] Getting notified when a morph moves

by Stephan Eggermont

On 12/02/15 10:20, Hilaire wrote: > Frankly I don't like the idea to subscribe to the HandMorph, because at > each mouse move event, you have to check for the morph A and B if they > move. It is a waste of CPU cycles. > Ideally you will want receive event from morph A and B whenever they > changed. I am surious to know if it is possible. Thanks Hilaire, This works for me. I can attach the listener when the dragging starts, and detach it when it stops. That should do. Stephan

Feb. 12, 2015

Re: [Pharo-users] recover lost changes

by Hartmut

Thanks to all of you ! > Le 11/2/15 18:58, Sean P. DeNigris a ?crit : >> Sean P. DeNigris wrote >>> I can not reproduce this... >> Okay, I see. It only happens when Nautilus (the browser) is on the >> class-side. Looking deeper, there is in general a flaw in Nautilus - at >> least in announcement handling - because it doesn't take class vs. metaclass >> into account. I started fixing it, but don't have time now. >> >> The basics for anyone that's interested: selected class could be a class, >> metaclass, or nil, so there needs to be a nil guard first, then something >> like selectedClass theNonMetaClass = aClass. Browsing (scoped to Nautilus) >> senders of #selectedClass is a good start. HTH 1. the <DoIt: ChangesLog reset; default> repairs the ChangesLog and 2. I can identify (and eventually remove) the empty class comments written by the meta class Hartmut -- Hartmut Krasemannâ¨ KÃ¶nigsberger Str. 41 c â¨D 22869 Schenefeldâ¨ Tel. 040.8307097â¨ Mobil 0171.6451283â¨ krasemann(a)acm.org

Feb. 12, 2015

Re: [Pharo-users] pharo and ui again

by Hilaire

Hi Joachim, I agree with your disagreement. Lack of developer support is really a problem. Just get me right: nobody is to blame. We are still a small community, although a growing one. For me one of the biggest lack is documentation (although the huge efforts done by several people). Polymorph for example is really nice to build UI, it offer a nice framework for common UI block but you are still free to hack it with Morph, but we do not have much documentation. You can also read my recent struggle with Magritte, the framework is marvelous but once you get on the trivial use case, you are really on your own to fit to your need. In all these situations your productivity is severely hit. What to do? Personally I try to discipline myself in community contribution: help in the mailing list when possible, even responding to myself when I ask for help and find myself an answer: it is likely it could be useful for other, and easily retrieved from the archive. Now I would like to do a bit more on writing notes on a kind-of knoweldge base for Pharo, in a wiki place or so. But I don't know where yet... Any idea, suggestion? Hilaire Le 12/02/2015 10:03, Joachim Tuchel a Ã©crit : > Hilaire. > > I completely disagree. The 2 areas you spend most of your time (and often waste it ) in my experience with business applications are > > a) The gui > b) persistency > > The tool support for both is pretty poor and both areas have been reinvented over and over again. Most of the reinventions were extremely poor in supporting the developer in getting things done, not to mention maintenance. Sick, isnt it? > > I can only underline what Sebastian said: if building a ui is hard, a system is probably not worth wasting time learning it. > > Just my 2 cents, > > Joachim > > Am 12.02.2015 09:46 schrieb Hilaire <hilaire(a)drgeo.eu>: >> Le 11/02/2015 23:28, Sebastian Heidbrink a Ã©crit : >>> I honestly try to bring Pharo/Smalltalk to the crowd, but somhow I >>> feel like the UI development removes all the productivity that one >>> usually has with Smalltalk. >> You are pretty right. For a small UI project, the UI development can be >> high. >> For a bigger project, where the UI part represent a small fraction of >> the whole project, it is a different story as most of the development >> will be out of the UI land and will benefit from the Pharo productivity. >> This is what I experienced with DrGeo. >> >> By the way, before Polymorph, the situation was even worst. >> Did you look at the Polymorph example in the WidgetExamples class ? >> There are good starter. >> >> Hilaire >> >> -- >> Dr. Geo - http://drgeo.eu >> iStoa - http://istoa.drgeo.eu >> >> >> >> -- Dr. Geo - http://drgeo.eu iStoa - http://istoa.drgeo.eu

Feb. 12, 2015

Re: [Pharo-users] Getting notified when a morph moves

by Hilaire

Hi, Here is an implementation example. >From a workspace, do: | morphA morphB follower | morphA := Morph new position: 10@10. morphB := Morph new position: 100@100. follower := FollowerLineMoprh from: morphA to: morphB. morphA openInWorld. morphB openInWorld. follower openInWorld. Frankly I don't like the idea to subscribe to the HandMorph, because at each mouse move event, you have to check for the morph A and B if they move. It is a waste of CPU cycles. Ideally you will want receive event from morph A and B whenever they changed. I am surious to know if it is possible. Thanks Hilaire Le 11/02/2015 21:19, Stephan Eggermont a Ã©crit : > I'm trying to keep a line morph connected between two morphs > while moving one of them. > When I add the morph to the hand and start dragging the morph > no longer receives mouseMove:. I can think of several ways to get > position updates. Is there a best/easy way? > > Stephan > > -- Dr. Geo - http://drgeo.eu iStoa - http://istoa.drgeo.eu

Feb. 12, 2015

Re: [Pharo-users] pharo and ui again

by kilon alios

One note on the GUI situation of Pharo. The situation of Pharo is pretty much the same situation for most programming languages. I am coming from python and the situation is pretty much the same but scaled up to the size of the python community which is huge. So instead like pharo of having 6 or 7 GUIs frameworks , Python has hundreds. I will say 4 are the most popular, pyQT which wraps QT (and its alternative pyside) , wxPython (wraps wxWidgets) , pyGTK (wraps GTK) and tkinter (wraps tk gui api) . All of the above frameworks have tons of users so one thing that this shows us is that fragmentation is really beneficial to a community. The reason is simple, GUI is so complex subject but there will never be one ring to rule them all. Many have tried, all of them have failed. Pharo will follow and does follow a similar route. I am actually thinking of making my own GUI API since none seems to fit my needs or the way I like to work. In the end no generic solution will be able to beat a customized one. This is why we need fragmentation. On Thu, Feb 12, 2015 at 10:46 AM, Hilaire <hilaire(a)drgeo.eu> wrote: > Le 11/02/2015 23:28, Sebastian Heidbrink a Ã©crit : > > I honestly try to bring Pharo/Smalltalk to the crowd, but somhow I > > feel like the UI development removes all the productivity that one > > usually has with Smalltalk. > > You are pretty right. For a small UI project, the UI development can be > high. > For a bigger project, where the UI part represent a small fraction of > the whole project, it is a different story as most of the development > will be out of the UI land and will benefit from the Pharo productivity. > This is what I experienced with DrGeo. > > By the way, before Polymorph, the situation was even worst. > Did you look at the Polymorph example in the WidgetExamples class ? > There are good starter. > > Hilaire > > -- > Dr. Geo - http://drgeo.eu > iStoa - http://istoa.drgeo.eu > > > >

Feb. 12, 2015

Re: [Pharo-users] pharo and ui again

by Joachim Tuchel

Let me try to be clearer.... Your observation may be right for some Applications, where there is some heavy lifting in computation and logic. So it is unfair to say you are wrong. There are, however, many areas where its all about searching some data, modifying and saving it. Think crm, erp, accounting, most web apps, and whatnot. These are the ones I had in mind. And in these, gui is the most important area, and the one that eats lots of time and effort. Joachim Am 12.02.2015 10:03 schrieb Joachim Tuchel <jtuchel(a)objektfabrik.de>: > > Hilaire. > > I completely disagree. The 2 areas you spend most of your time (and often waste it ) in my experience with business applications are > > a) The gui > b) persistency > > The tool support for both is pretty poor and both areas have been reinvented over and over again. Most of the reinventions were extremely poor in supporting the developer in getting things done, not to mention maintenance. Sick, isnt it? > > I can only underline what Sebastian said: if building a ui is hard, a system is probably not worth wasting time learning it. > > Just my 2 cents, > > Joachim > > Am 12.02.2015 09:46 schrieb Hilaire <hilaire(a)drgeo.eu>: > > > > Le 11/02/2015 23:28, Sebastian Heidbrink a Ã©crit : > > > I honestly try to bring Pharo/Smalltalk to the crowd, but somhow I > > > feel like the UI development removes all the productivity that one > > > usually has with Smalltalk. > > > > You are pretty right. For a small UI project, the UI development can be > > high. > > For a bigger project, where the UI part represent a small fraction of > > the whole project, it is a different story as most of the development > > will be out of the UI land and will benefit from the Pharo productivity. > > This is what I experienced with DrGeo. > > > > By the way, before Polymorph, the situation was even worst. > > Did you look at the Polymorph example in the WidgetExamples class ? > > There are good starter. > > > > Hilaire > > > > -- > > Dr. Geo - http://drgeo.eu > > iStoa - http://istoa.drgeo.eu > > > > > > > >

Feb. 12, 2015

Re: [Pharo-users] pharo and ui again

by Joachim Tuchel

Hilaire. I completely disagree. The 2 areas you spend most of your time (and often waste it ) in my experience with business applications are a) The gui b) persistency The tool support for both is pretty poor and both areas have been reinvented over and over again. Most of the reinventions were extremely poor in supporting the developer in getting things done, not to mention maintenance. Sick, isnt it? I can only underline what Sebastian said: if building a ui is hard, a system is probably not worth wasting time learning it. Just my 2 cents, Joachim Am 12.02.2015 09:46 schrieb Hilaire <hilaire(a)drgeo.eu>: > > Le 11/02/2015 23:28, Sebastian Heidbrink a Ã©crit : > > I honestly try to bring Pharo/Smalltalk to the crowd, but somhow I > > feel like the UI development removes all the productivity that one > > usually has with Smalltalk. > > You are pretty right. For a small UI project, the UI development can be > high. > For a bigger project, where the UI part represent a small fraction of > the whole project, it is a different story as most of the development > will be out of the UI land and will benefit from the Pharo productivity. > This is what I experienced with DrGeo. > > By the way, before Polymorph, the situation was even worst. > Did you look at the Polymorph example in the WidgetExamples class ? > There are good starter. > > Hilaire > > -- > Dr. Geo - http://drgeo.eu > iStoa - http://istoa.drgeo.eu > > > >

Feb. 12, 2015

Re: [Pharo-users] Nautilus groups

by kilon alios

I don't use groups because already pharo implement a fair share of grouping mechanism. For my needs package tags serve me very well and I have not so far a practical need to group things from different packages. When I want to browser diffrent classes at the same time I use window groups and I am very happy with it. I also use the history mechanism of the system browser recently which I find useful for going back and forth. The next step for me would be to become more familiar with keyboard shortcuts to further optimise my workflow. But I work so slow because I am still very unfamiliar with pharo that I have no immediate need for it On Thu, Feb 12, 2015 at 5:19 AM, Martin BÃ¤hr < mbaehr(a)email.archlab.tuwien.ac.at> wrote: > Excerpts from Christophe Demarey's message of 2015-02-10 14:28:19 +0100: > > dynamic groups: 'Last modified packages', 'Most viewed classes', 'Work' > > i just discovered a problem, in that when a group is selected, some > important > items are missing from the content menu. in particular, "add package..." > > it is entirely non-obvious that i need to have a package selected in order > to > add a new package which may be quite unrelated to the package i selected. > > what happened next was, that i thought "add group" is actually "add > package" so > i added a group and then wondered why, when creating a class, the template > didn't have the package pre-set. > > it would probably better if all generic menu items, that is, those that > don't > act on a specific group are present in all menus in that panel. > > greetings, martin. > > -- > eKita - the online platform for your entire academic > life > -- > chief engineer > eKita.co > pike programmer pike.lysator.liu.se caudium.net > societyserver.org > secretary > beijinglug.org > mentor > fossasia.org > foresight developer foresightlinux.org > realss.com > unix sysadmin > Martin BÃ¤hr working in china > http://societyserver.org/mbaehr/ > >

Feb. 12, 2015