new Smalltalk programmer's thoughts
Locked
12
12
new Smalltalk programmer's thoughts
 
|
Hi list,
at http://people.squeakfoundation.org/article/57.html I wrote down my thoughts while learning how to programm in Smalltalk. Greatings, Wolfgang |
Re: new Smalltalk programmer's thoughts
|
|
Welcome! I've posted a reply BTW.
On Apr 30, 2006, at 10:19 PM, Wolfgang Helbig wrote: > Hi list, > at > http://people.squeakfoundation.org/article/57.html > > I wrote down my thoughts while learning how to programm in Smalltalk. > > Greatings, > Wolfgang > > |
Re: new Smalltalk programmer's thoughts
|
|
And I like your reply :)
even if the point of wolfgang was valid but from a C perspective. > Welcome! I've posted a reply BTW. > > On Apr 30, 2006, at 10:19 PM, Wolfgang Helbig wrote: > >> Hi list, >> at >> http://people.squeakfoundation.org/article/57.html >> >> I wrote down my thoughts while learning how to programm in Smalltalk. >> >> Greatings, >> Wolfgang >> >> > > |
Re: new Smalltalk programmer's thoughts
|
|
In reply to this post by Wolfgang Helbig-2
Hi Stev,
you wrote: >even if the point of wolfgang was valid but from a C perspective. and I must frankly admit, that I can't parse that sentence. Ascribe this to my mother tongue not being English. So in case you meant my article is valid only from a C programmer's perspective, let me stress, that my perspective is from a programmer who aims to convince himself that the program at hand is correct. If you want to attribute this frame of mind to a profession, I think a mathematician trying to prove a theorem is more accurate than a C programmer. Proving a program being correct while writing it is much easier, if the first natural number is zero. This convention leads to more elegant and "natural" statements about the state of instance variables. This is not confined to indexes, but file stream positions as well. The position is zero after initialization and thus does not address a byte. It needs to be incremented before it is a valid pointer. This made it much harder to write and prove the AltoFileStream class correct. Unix and C let zero be the first natural number. As you might know, the Unix V6 kernel does not contain a single bug. Ironically, the only V6 program that I consider buggy is the debugger. Maybe this is because Ken Thompson and Dennis Ritchie didn't need the debugger. And maybe this is because they were trying to prove their programs to be correct while writing. And maybe this is the reason they decided to let zero be the first natural number. The last sentences are purely speculative and Dennis might shed some light on them. But it is not speculative that Smalltalk is the only programming language that allows indexed variables and forces the programmer to use one as the first index. If you find another one let me know. Greetings, Wolfgang -- Weniger, aber besser. |
Re: new Smalltalk programmer's thoughts
|
|
Wolfgang Helbig wrote:
> >The last sentences are purely speculative and Dennis might shed some light on >them. But it is not speculative that Smalltalk is the only programming language >that allows indexed variables and forces the programmer to use one as the first >index. If you find another one let me know. > > http://en.wikipedia.org/wiki/Array lists BASIC and Fortran as having 1-based array indexing, and it even discusses some of the relative merits of and arguments for 0-based and 1-based array indexing. Which one is more natural is a matter of personal taste and opinion. In most everyday programming situations I consider the 1-based approach much more natural, but the low-level stuff where you're addressing hardware is definitely an exception which can be handled pretty well by implementing a specialized arrayed class which does 0-based indexing. Having learned BASIC as a first language in school, I found C's convention of 0-based indexing pretty awkward, until I realized that C is mostly a kind of portable assembly language which lets you get close to the hardware. It's not a language I would consider well suited for expression my thoughts in a program... The question is: do we want to let hardware characteristics dominate our way of thinking? I at least don't. When I've got 5 elements in a collection, I want the first at index 1 and the fifth at index 5. When I'm using Integers, I don't want to think about 16 vs 32 bits, signed vs unsigned etc. The argument about loop bounds is somewhat related. I've seen quite some beginner's C code which did "for (i=0; i<=10; i++)" to address a 10-element array :-) Seasoned C programmers don't do that anymore, but it tells me a bit about what is "natural" :-) Cheers, Hans-Martin |
Re: new Smalltalk programmer's thoughts
|
|
In reply to this post by Wolfgang Helbig-2
Hi Hans-Martin,
you mentioned: >http://en.wikipedia.org/wiki/Array lists BASIC and Fortran as having >1-based array indexing not any more :-) In Fortran 77 you specify a Real Array with indexes from 0 to 10 as REAL(0:10) This leaves BASIC and Smalltalk on the wikipedia list. Greetings, Wolfgang -- Weniger, aber besser. |
Re: new Smalltalk programmer's thoughts
|
|
In reply to this post by Wolfgang Helbig-2
Hi Hans-Martin,
you posed the question >The question is: do we want to let hardware characteristics dominate our >way of thinking? And the answer is no! And I deeply appreciate Smalltalk for the #// and #\\ operators, for Integers that are unlimeted, and for things like garbage collection that no real machine can achieve! Smalltalk leverages on being run by a virtual machine. And it does very well so. Furthermore, it lets you adopt things to your likes much easier than any other language I know. (I don't know Lisp) So, in this respect, I couldn't agree more with you. >collection, I want the first at index 1 and the fifth at index 5. When >I'm using Integers, I don't want to think about 16 vs 32 bits, signed vs >unsigned etc. >The argument about loop bounds is somewhat related. I've seen quite some >beginner's C code which did "for (i=0; i<=10; i++)" to address a >10-element array :-) Seasoned C programmers don't do that anymore, but >it tells me a bit about what is "natural" :-) And that is my point: Different challenges need different numbers for the first index. When I am implementing a Gaussian algorithm to invert matrices, I'd never use 0 as the first index. But whenever I am computing indexes, I'd never use one as the first index. And this is not supported by BASIC and Smalltalk. Greetings, Wolfgang > >Cheers, >Hans-Martin -- Weniger, aber besser. |
Re: new Smalltalk programmer's thoughts
|
|
In reply to this post by Wolfgang Helbig-2
2006/5/1, Wolfgang Helbig <[hidden email]>:
> Hi Hans-Martin, > you mentioned: > > >http://en.wikipedia.org/wiki/Array lists BASIC and Fortran as having > >1-based array indexing > > not any more :-) > > In Fortran 77 you specify a Real Array with indexes from 0 to 10 as > REAL(0:10) > This leaves BASIC and Smalltalk on the wikipedia list. IIRC XPath uses 1 based indices too. |
Re: new Smalltalk programmer's thoughts
|
|
In reply to this post by Wolfgang Helbig-2
Hi Wolfgang,
Smalltalk, as an ambience, can be used for doing other activities than writing "programs" (aProgram, aComponent, or aModule). When you compare Smalltalk with languages you are comparing syntax (the only thing comparable from a language perspective). Newbies use to compare smalltalk from language perspective, or from OO way (through design, abstractions, patters) to the tools they used in the past. Only with time, opportiunities of real use in production and contact with people that use smalltalk in another way (not simply like the best OOL); new ways of doing informatics emerges. Smalltalk as a open system, let you do non-formal development (e.g. pass the limits imposed by formalization, reductionism & encapsulation). As a non-formal device/medium, it is not the opposite of formal design; it is a convenient complement (the only one competitive complement to formal design used in software industry today) best, Ale. ----- Original Message ----- From: "Wolfgang Helbig" <[hidden email]> To: <[hidden email]> Sent: Monday, May 01, 2006 9:02 AM Subject: Re: new Smalltalk programmer's thoughts > Hi Hans-Martin, > you posed the question > >The question is: do we want to let hardware characteristics dominate our > >way of thinking? > > And the answer is no! > And I deeply appreciate Smalltalk for the #// and #\\ operators, for Integers > that are unlimeted, and for things like garbage collection that no real machine > can achieve! Smalltalk leverages on being run by a virtual machine. And it does > very well so. Furthermore, it lets you adopt things to your likes much easier > than any other language I know. (I don't know Lisp) So, in this respect, I > couldn't agree more with you. > > >collection, I want the first at index 1 and the fifth at index 5. When > >I'm using Integers, I don't want to think about 16 vs 32 bits, signed vs > >unsigned etc. > >The argument about loop bounds is somewhat related. I've seen quite some > >beginner's C code which did "for (i=0; i<=10; i++)" to address a > >10-element array :-) Seasoned C programmers don't do that anymore, but > >it tells me a bit about what is "natural" :-) > > And that is my point: Different challenges need different numbers for the > index. When I am implementing a Gaussian algorithm to invert matrices, I'd never > use 0 as the first index. But whenever I am computing indexes, I'd never use > one as the first index. And this is not supported by BASIC and Smalltalk. > > Greetings, > Wolfgang > > > >Cheers, > >Hans-Martin > > -- > Weniger, aber besser. > > |
Re: new Smalltalk programmer's thoughts
|
|
In reply to this post by Wolfgang Helbig-2
Hi Wolfgang, Your point of view is correct, but sorry, it is a C point of view. That means a point of view of how can i better insure the low level implementation. And in fact, i trust you, it is more convenient to have 0-based if your are dealing with pointers and offset. But the aim of Smalltalk was not to ease such implementation. Smalltalk ignore the notion of pointer, and assume all this low level goes into the VM. Smalltalk goal is to ease the upper level programming. And at upper level, sorry, if i want the object of rank 100 in a collection, i naturally write (collec at: 100), and would find it very irritating to write (collec at: 99), having to think of how the processor will translate this into an offset. Looking at the history of numbers, you will discover that the zero is a relatively new invention. Initially, it just meant empty, none or nihil (our Smalltalk nil). And looking closer at the english words, you know very well that fourth must be associated to number four, fifth to number five, etc... If you associate first to 0, then you associate fourth to 3, something that does not sound natural to me. That is why we index starting at 1 in Smalltalk, and why we are reluctant to change that (though it is not at all in the syntax, just a convention choice in the library). Beside, it does not prevent us to use 0-based algorithm like with Stream position, when what we need is an offset. So i do not really see that as an obstacle. It is just a programming habit, and i can tell you i also made a lot of index-errors when switching between FORTRAN and C code, so the lack of errors is not an objective criterium... Nicolas |
0 or 1 index (was: new Smalltalk programmer's thoughts)
|
|
In reply to this post by Wolfgang Helbig-2
Wolfgang Helbig wrote on Mon, 1 May 2006 14:02:19 +0200 (MEST)
> And that is my point: Different challenges need different numbers for the first > index. When I am implementing a Gaussian algorithm to invert matrices, I'd never > use 0 as the first index. But whenever I am computing indexes, I'd never use > one as the first index. And this is not supported by BASIC and Smalltalk. To be fair, I have used more than one BASIC that allowed you to set either 0 or 1 as the base index for all arrays in a program. And the Self dialect of Smalltalk (http://www.merlintec.com:8080/Self/) does borrow several C conventions including 0 based arrays. The current program I am writing in Squeak simulates hardware and 1 based arrays were a bit inconvenient, so the first thing I did was define a ZArray class as a subclass of ArrayedCollection with at: index ^ super at: index + 1 at: index put: obj ^ super at: index + 1 put: obj do: block 0 to: self size - 1 do: [ :index | block value: (self at: index) ] and then all I had to do was write the rest of my application naturally and just remember to use "ZArray new: x" instead of "Array new: x". Inspecting ZArrays still shows the contents as having indexes starting with 1, but other than that small confusion (which I could easily fix if I wanted to) it works just great. There is probably a significant performance hit but I am not worried about that here. -- Jecel |
Re: new Smalltalk programmer's thoughts
|
|
In reply to this post by Hans-Martin Mosner
On 5/1/06, Hans-Martin Mosner <[hidden email]> wrote: Wolfgang Helbig wrote: So true. The admonition "remember arrays start at zero" is repeated countless times in most literature. You see this in just about every article or book on most modern languages. If it were so "natural" to count groups of things from 0 people wouldn't need to be reminded constantly. Regards, Laurence Cheers, |
Re: new Smalltalk programmer's thoughts
|
|
In reply to this post by Philippe Marschall
On 5/1/06, Philippe Marschall <[hidden email]> wrote: 2006/5/1, Wolfgang Helbig <[hidden email]>: Applescript and COBOL do as well. |
Re: new Smalltalk programmer's thoughts
|
|
On May 01, 2006, at 19:25, Laurence Rozier wrote: > On 5/1/06, Philippe Marschall <[hidden email]> wrote: >> >> 2006/5/1, Wolfgang Helbig <[hidden email]>: >> > This leaves BASIC and Smalltalk on the wikipedia list. >> >> IIRC XPath uses 1 based indices too. > > > Applescript and COBOL do as well. Lua to! http://lua-users.org/wiki/CountingFromOne Cheers -- PA, Onnay Equitursay http://alt.textdrive.com/ |
Re: 0 or 1 index (was: new Smalltalk programmer's thoughts)
|
|
In reply to this post by Wolfgang Helbig-2
On 5/1/06, Jecel Assumpcao Jr <[hidden email]> wrote: Wolfgang Helbig wrote on Mon, 1 May 2006 14:02:19 +0200 (MEST) This seems to me to be a good approach. I think Alan Kay said something like "hardware is just software crystallized early". If one accepts that notion then the special needs of hw should properly be subsets of the typical case. Perhaps, when the world is running on Plurion processors, that will come to pass ;-) Laurence -- Jecel |
Re[2]: new Smalltalk programmer's thoughts
|
|
In reply to this post by Wolfgang Helbig-2
Hello Wolfgang,
WH> index. If you find another one let me know. Matlab! And there's a big user group using it for dsp (digital signal processing) which are deeply frustrated about it. So if you want to look into this discussion search the comp.dsp archives. For this discussion: As a LISPer I'm annoyed by (nth 1 ofMyList) being the second element, as a DSPer I'm annoyed whenever I read some 1-based matlab indexing while zero-based indexing feels so natural in DSP. As a child of two worlds I bloody my nose with off by one errors more frequent than others but by now I have a rugged nose (TM), so I just laugh and change the indexing. Cheers Herbert mailto:[hidden email] |
Re: new Smalltalk programmer's thoughts
|
|
In reply to this post by Wolfgang Helbig-2
Wolfgang,
> Unix and C let zero be the first natural number. As you might know, the Unix V6 > kernel does not contain a single bug. Ironically, the only V6 program > that I consider buggy is the debugger. Maybe this is because Ken Thompson and > Dennis Ritchie didn't need the debugger. And maybe this is because they were > trying to prove their programs to be correct while writing. And maybe this > is the reason they decided to let zero be the first natural number. The people's favorite language Perl has a special variable "$[" that let you specify the index of the first element in an array. As everyone wrote: it is about either to think about the n-th index , or the offset or distance and I'd say that you can't simply say which is more natural. Also, in Smalltalk, Streams are 0-based as nicolas and others mentioned. a _ ReadStream on: 'ABC'. a next. " => $A" a position. "=> 1" a position: 0. a next. "=> $A" a position: 3. a next. "=> nil" The AltoFileStream should've follow this convension, instead of allowing [1..size] indices. (The coordinates in a Form are 0-based, by the way...) Just my $0.02, -- Yoshiki |
Re: new Smalltalk programmer's thoughts
|
|
In reply to this post by Wolfgang Helbig-2
Hi Yoshiki,
as you and others pointed out: > Also, in Smalltalk, Streams are 0-based as nicolas and others >mentioned. > The AltoFileStream should've follow this convension, instead of >allowing [1..size] indices. I agree! It not only *should* follow this convention, but it *must* follow it. After all, I programmed AltoFileStream to be used by the Hobbes Emulator, which expects a FileStream to read the snapshot. So I subclassed FileStream and tried to implement its protocol. Later I found out, that ReadStream suffices for this purpose. Anyway, I had to implement the semantics of #postion and #position: as expected by the Hobbes Emulator. Since I didn't find a specification of that protocol, I naively assumed that position is supposed to be in [1..size]. After all, this is the natural range of indexes in Smalltalk, isn't it? But I was utterly wrong, which I found out when I tried to make Hobbes read the snapshot from an AltoDisk. The position is supposed to be in range [0..size]. Which looks very unnatural to me. With this range, position looses the nice property of being a valid index. So you have to make sure you never access the byte indexed by the current position. Instead you always use current position + 1. Unix files are more elegant in this respect. Their position is in the range [0..size), which happens to be the range of valid "Unix indexes". And I feel much more comfortable with [0..size) than with [0..size] or [1..size]. Thanks to everyone on this list for their answers! I've learned a lot! Greetings Wolfgang, a humble programmer -- Weniger, aber besser. |
Re: new Smalltalk programmer's thoughts
|
|
In reply to this post by Wolfgang Helbig-2
Dear Wolfgang,
> But it is not speculative that Smalltalk is the only programming > language that allows indexed variables and forces the programmer to > use one as the first index. If you find another one let me know. > To add to the list, there's Icon. Icon indexes in an interesting way. It indexes not the elements, but before, after, and between elements. It also indexes from both ends of a list or array. E. g., (1,-3) A (2,-2) B (3,-1) C (4,0) There are three elements, A, B, and C, and 8 indices, -3 to 4. 1 and -3 refer to the point before A, 2 and -2 to the point before B, and 3 and -1 to the point before C. 4 and 0 refer to the end. To access an element you use one of the indices that comes before it. Best regards, Bill |
Re: new Smalltalk programmer's thoughts
|
|
>
> >> But it is not speculative that Smalltalk is the only programming >> language that allows indexed variables and forces the programmer to >> use one as the first index. If you find another one let me know. Even if it were true that no other language used 1 origin array indexing, so what? Popularity is no indication of correctness. If we cared about being like all the other languages we wouldn't have Smalltalk. If we cared about being like all the other kids we wouldn't have geeks like us to create stuff like Smalltalk. And then where would the world be? Ultimately, Smalltalk, like all other languages, systems, cars, planes, boats, foods, political parties, comics, games etc is what it is. If you like it - Enjoy! If you don't - find one you do like and Enjoy! tim -- tim Rowledge; [hidden email]; http://www.rowledge.org/tim Asking whether machines can think is like asking whether submarines can swim. |
«
Return to Squeak - Dev
|
1 view|%1 views
| Free forum by Nabble | Edit this page |