March 2015 - Pharo-users - lists.pharo.org

Re: [Pharo-users] Introduction to OOP for UPBE was: Some new book chapters :)

by stepharo

> I would like the opinion of the community on this. I have discussed > the addition of an introduction to OOP back when I was very actively > contributing to UPBE and the response I got was very positive. The > plan was similar to what I mentioned, just bare basics , making sure > that the reader understand the very basic fundamental concepts of OO, > like classes, inheritance, instances, class and instance variables, > class and instance methods etc. Essentially a chapter similar to my > video tutorials but going a bit deepers with more detailed example > still keeping it simple , clear and understandable even for non coders. If you write something I will review it and tell you if it is ok for the book. If this is more or less ok, I'm ready to do a pass. so I can help if what you write is ok. Now I do not know your writing skills so I do not want to tell you that this is ok without even getting a feel about what you can produce. > > I want to know whether such chapter would be included because I dont > like to waste my time or the precious time of the student that has > taken the task on her shoulders. > > "IMHO this is the exception to that obvious rule. OOP in particular is > misunderstood by almost everyone. If we as Smalltalkers describe it > from a forensic CS perspective that sounds to me like C++, I believe > it is better that we say nothing on the subject." > > Personally I dont care if people misunderstand OOP , I only care about > getting my code to work the way I want it. So far I had no problem > doing that even though I code using very simple concept some of them > may come even back when I was coding in C++. I also dont share the > sense of wonder that some of you may share about how Smalltalk is > doing OO, coming from python does not seem such a big deal for me. > > For example message passing suppose to be this super powerful feature > , that diffirentiate from method calls in the essence that the user > does not need to know the internals of the class. But I have not seen > a class that take advantage of this feature inside pharo. sorry but I do not get the last paragraph. To me method calls are just a different way to call message passing and I prefer message passing because it conveys that we do not directly refer to the method but that the method is dynamically selected. > In every single case if I had exchanged the message passing with > method calls I would not have even noticed the difference. How could you? in Smalltalk the only to implement what you call method calls is to access directly the method and use executeXXX on it. > In every case I have studied each message corresponds to a method with > exactly the same name. There are plenty of methods defined with the same name in the system so I do not get your point. > I am sure maybe one day will come handly , but I dont see why one > needs to explain a feature that rarely someone will use understand how > to code in a specific environment. ??? Stef

March 29, 2015

Re: [Pharo-users] Introduction to OOP for UPBE was: Some new book chapters :)

by kilon alios

first example is at page 5 . I don't overheat , I rarely get angry. So there is no reason to worry, all opinions are more than welcomed. If the styles of documentation that the community desires is on par with the link you provided then I know and this is not what I want to be working on. But since creating a chapter is quite an effort I rather spent efforts on things that are here to stay than things that get ignored. its ok , chapter cancelled, I hope an expert is found sometime in the near future to fill in our shoes. Have fun :) On Sun, Mar 29, 2015 at 5:59 PM, Hilaire <hilaire(a)drgeo.eu> wrote: > Hello, > > > [..] no need to list the chapters, I know the book. > > > Le 29/03/2015 16:43, kilon alios a Ã©crit : > > > The iink your provided is a reference documentation and I dont like > > this kind of documentation sinking inside theory with few practical > > examples. Quite the opposite of PBE and frankly a really bad way to > > teach coding especially to a newcomer and someone unfamiliar with OOP. > > > > We are probably not reading the same article then, the one I pointed to > you comes with a lot of examples. Nevertheless, only its first > paragraphs were of interest related to the point Sean raised. > Anyway I gave my opinion and will stop on this discussion before it > overheats. > > Best > > -- > Dr. Geo - http://drgeo.eu > iStoa - http://istoa.drgeo.eu > > > >

March 29, 2015

Re: [Pharo-users] MySQL framework for Pharo to analyze futures and options.

by Merwan Ouddane

On 29/03/2015 15:58, Mariano Martinez Peck wrote: > > > On Sun, Mar 29, 2015 at 6:07 AM, Sven Van Caekenberghe <sven(a)stfx.eu > <mailto:sven@stfx.eu>> wrote: > > > > On 29 Mar 2015, at 04:47, Pierce Ng <pierce(a)samadhiweb.com > <mailto:pierce@samadhiweb.com>> wrote: > > > > On Fri, Mar 27, 2015 at 02:21:01PM -0700, nacho wrote: > >> I'm using MySQL to hold some data on commodities prices for futures > >> valuation and want to - of course - use Smalltalk as a > front-end. Which > >> framework would you recommend? > > > > There is a pure Smalltalk MySQL driver on Squeaksource. Not sure > whether > > it has been moved to SS3 or STH. > > > > http://www.squeaksource.com/StdbMysqlProtocol > Here it is: http://smalltalkhub.com/#!/~DBXTalk/MysqlDriver > > > > Others have suggested Glorp. Glancing at the code, seems Glorp > currently only > > works with VisualWorks' MySQL driver. Based on my experience > integrating > > NBSQLite3 into Glorp, I'd say it will take about a day or two, > depending on the > > design of the driver. I've tested abovementioned MySQL driver > with some version > > of MySQL5 some months ago. > > It would be very good to have another Glorp solution with a pure > Smalltalk implemented DB driver, like the PostgresV2 one. IMHO > this is way less error prone than the whole library/driver/.. mess > on all these different platform. > > > Since our Glorp refactor back years ago...adding a Glorp adaptor > (DatabaseDriver subclasses) should be very very easy. > So yes, it would be nice to do it for the MySQL driver...but I thought > we already had that... > Me and some other students made an adaptator for the MySQL Driver written in smalltalk based on the one for PostgresV2, you can find it here: http://smalltalkhub.com/#!/~ThomasHeniart/GlorpDriverMySQL There is a method NativeMySQLDriver>>beGlorpDefaultDriver to use it as native driver for glorp. We posted it on the mailing list one and a half month ago ^^ > > -- > Mariano > http://marianopeck.wordpress.com

March 29, 2015

Re: [Pharo-users] MySQL framework for Pharo to analyze futures and options.

by nacho

Hi! @Cameron: Yes the data is in a MySQL database. I could do a dump and save it as CVS or plain text, should be no problem but I think I prefer to keep it as a SQL database as the new data gets updated automatically. @Mariano: I installed Glorp and as you say, the driver is there, altough in the class comments it says that is not complete yet. I would take a look these days to see if it works properly. Thanks for your support!! Cheers Nacho ----- Nacho Smalltalker apprentice. Buenos Aires, Argentina. -- View this message in context: http://forum.world.st/MySQL-framework-for-Pharo-to-analyze-futures-and-opti… Sent from the Pharo Smalltalk Users mailing list archive at Nabble.com.

March 29, 2015

Re: [Pharo-users] Introduction to OOP for UPBE was: Some new book chapters :)

by Hilaire

Hello, [..] no need to list the chapters, I know the book. Le 29/03/2015 16:43, kilon alios a Ã©crit : > The iink your provided is a reference documentation and I dont like > this kind of documentation sinking inside theory with few practical > examples. Quite the opposite of PBE and frankly a really bad way to > teach coding especially to a newcomer and someone unfamiliar with OOP. > We are probably not reading the same article then, the one I pointed to you comes with a lot of examples. Nevertheless, only its first paragraphs were of interest related to the point Sean raised. Anyway I gave my opinion and will stop on this discussion before it overheats. Best -- Dr. Geo - http://drgeo.eu iStoa - http://istoa.drgeo.eu

March 29, 2015

Re: [Pharo-users] Introduction to OOP for UPBE was: Some new book chapters :)

by kilon alios

Not really because those concepts are already explained in other chapters. Blocks and conditions and why conditions are the way they are are explained in Chapter 3. Chapter 4 explains messages in general including the use of parentheses , how messages are sent , the different kind of messages , when to use a block and when to use parentheses, cascades etc etc. Chapter 5 explains the model system of Pharo, method lookup , self sends, super sends and much more. So the chapter we are writing will go between Chapter 4 and 5 and offer a smoother introduction to OO. The focus is to offer a practical approach to the very fundamentals with as little overlap as possible with other chapters. Part II of PBE goes deeper also, chapter 13 covers classes and metclasses in great detail, Chapter 14 which is about Reflection describes method dictionaries , accessing run time context, intercepting MNUs etc. Stef once also published a very detailed chapter on Blocks with many advanced features . This used to be in Pharo for the Enterprise , now should be in PharoInProgress. And yes I understand how messages work. The iink your provided is a reference documentation and I dont like this kind of documentation sinking inside theory with few practical examples. Quite the opposite of PBE and frankly a really bad way to teach coding especially to a newcomer and someone unfamiliar with OOP. On Sun, Mar 29, 2015 at 5:17 PM, Hilaire <hilaire(a)drgeo.eu> wrote: > Le 29/03/2015 15:18, kilon alios a Ã©crit : > > For example message passing suppose to be this super powerful feature > > , that diffirentiate from method calls in the essence that the user > > does not need to know the internals of the class. But I have not seen > > a class that take advantage of this feature inside pharo. In every > > single case if I had exchanged the message passing with method calls I > > would not have even noticed the difference. In every case I have > > studied each message corresponds to a method with exactly the same > > name. I am sure maybe one day will come handly , but I dont see why > > one needs to explain a feature that rarely someone will use understand > > how to code in a specific environment. > > Hello, > > You may really take a look at the *first* paragraph of Alan Lovejoy's > article mentioned in my previous message, and pass it to the student as > well: > > > http://www.smalltalk.org/articles/article_20100320_a3_Getting_The_Message.h… > > > Without this understanding, one fails to understand why #ifTrue: > argument is a block, or why #and: message's argument is a block[] > surrounded by parenthesis(), then the beginner may have a hard time to > use correctly the class libraries. > > Best, > > -- > Dr. Geo - http://drgeo.eu > iStoa - http://istoa.drgeo.eu > > > >

March 29, 2015

Re: [Pharo-users] Introduction to OOP for UPBE was: Some new book chapters :)

by Hilaire

Le 29/03/2015 15:18, kilon alios a Ã©crit : > For example message passing suppose to be this super powerful feature > , that diffirentiate from method calls in the essence that the user > does not need to know the internals of the class. But I have not seen > a class that take advantage of this feature inside pharo. In every > single case if I had exchanged the message passing with method calls I > would not have even noticed the difference. In every case I have > studied each message corresponds to a method with exactly the same > name. I am sure maybe one day will come handly , but I dont see why > one needs to explain a feature that rarely someone will use understand > how to code in a specific environment. Hello, You may really take a look at the *first* paragraph of Alan Lovejoy's article mentioned in my previous message, and pass it to the student as well: http://www.smalltalk.org/articles/article_20100320_a3_Getting_The_Message.h… Without this understanding, one fails to understand why #ifTrue: argument is a block, or why #and: message's argument is a block[] surrounded by parenthesis(), then the beginner may have a hard time to use correctly the class libraries. Best, -- Dr. Geo - http://drgeo.eu iStoa - http://istoa.drgeo.eu

March 29, 2015

Re: [Pharo-users] MySQL framework for Pharo to analyze futures and options.

by Mariano Martinez Peck

On Sun, Mar 29, 2015 at 6:07 AM, Sven Van Caekenberghe <sven(a)stfx.eu> wrote: > > > On 29 Mar 2015, at 04:47, Pierce Ng <pierce(a)samadhiweb.com> wrote: > > > > On Fri, Mar 27, 2015 at 02:21:01PM -0700, nacho wrote: > >> I'm using MySQL to hold some data on commodities prices for futures > >> valuation and want to - of course - use Smalltalk as a front-end. Which > >> framework would you recommend? > > > > There is a pure Smalltalk MySQL driver on Squeaksource. Not sure whether > > it has been moved to SS3 or STH. > > > > http://www.squeaksource.com/StdbMysqlProtocol > > > > Others have suggested Glorp. Glancing at the code, seems Glorp currently > only > > works with VisualWorks' MySQL driver. Based on my experience integrating > > NBSQLite3 into Glorp, I'd say it will take about a day or two, depending > on the > > design of the driver. I've tested abovementioned MySQL driver with some > version > > of MySQL5 some months ago. > > It would be very good to have another Glorp solution with a pure Smalltalk > implemented DB driver, like the PostgresV2 one. IMHO this is way less error > prone than the whole library/driver/.. mess on all these different platform. > > Since our Glorp refactor back years ago...adding a Glorp adaptor (DatabaseDriver subclasses) should be very very easy. So yes, it would be nice to do it for the MySQL driver...but I thought we already had that... -- Mariano http://marianopeck.wordpress.com

March 29, 2015

Introduction to OOP for UPBE was: Some new book chapters :)

by kilon alios

I would like the opinion of the community on this. I have discussed the addition of an introduction to OOP back when I was very actively contributing to UPBE and the response I got was very positive. The plan was similar to what I mentioned, just bare basics , making sure that the reader understand the very basic fundamental concepts of OO, like classes, inheritance, instances, class and instance variables, class and instance methods etc. Essentially a chapter similar to my video tutorials but going a bit deepers with more detailed example still keeping it simple , clear and understandable even for non coders. I want to know whether such chapter would be included because I dont like to waste my time or the precious time of the student that has taken the task on her shoulders. "IMHO this is the exception to that obvious rule. OOP in particular is misunderstood by almost everyone. If we as Smalltalkers describe it from a forensic CS perspective that sounds to me like C++, I believe it is better that we say nothing on the subject." Personally I dont care if people misunderstand OOP , I only care about getting my code to work the way I want it. So far I had no problem doing that even though I code using very simple concept some of them may come even back when I was coding in C++. I also dont share the sense of wonder that some of you may share about how Smalltalk is doing OO, coming from python does not seem such a big deal for me. For example message passing suppose to be this super powerful feature , that diffirentiate from method calls in the essence that the user does not need to know the internals of the class. But I have not seen a class that take advantage of this feature inside pharo. In every single case if I had exchanged the message passing with method calls I would not have even noticed the difference. In every case I have studied each message corresponds to a method with exactly the same name. I am sure maybe one day will come handly , but I dont see why one needs to explain a feature that rarely someone will use understand how to code in a specific environment.

March 29, 2015

Re: [Pharo-users] Some new book chapters :)

by Hilaire

Le 29/03/2015 13:19, kilon alios a Ã©crit : > "It does not emphasize the key features of OOP - > message sending and late binding - or the tremendous benefits." > > I am glad it does not , because if it did , it would be not a > Introductory chapter to OOP. Hello, I don't know the article Sean is talking about so I am only discussing the point about message and late binding. What Sean explains is that message sending and late binding is at the essence of Smalltalk (and therefore Pharo). More over it is an *easy* concept to understand, and finally and probably more important: a learner getting it right at the first time will learn and understand *more easily* the way the Class libraries are built and practised. Hopefully there are already valuable resources on the topic one can reuse: http://www.smalltalk.org/articles/article_20100320_a3_Getting_The_Message.h… Probably more: https://duckduckgo.com/?t=lm&q=smalltalk+essence+to+OOP And the points from Smalltalk creator: http://www.c2.com/cgi/wiki?AlanKaysDefinitionOfObjectOriented Thanks -- Dr. Geo - http://drgeo.eu iStoa - http://istoa.drgeo.eu

March 29, 2015