Brief Report - "Plain English Compiler" First Impressions
-
Jeremy Falcon wrote:
You can't offer a different paradigm while taking things away from an enviroment we're used to, which is what this guy did.
Sure you can. Look at MyXaml. Declarative programming without Intellisense, forms have to be created by editing XML instead of using a form designer, errors are detected at runtime instead of during the build, and so forth. Yet people use it, and quite successfully, depending on the problem domain they're trying to solve. Marc Pensieve Functional Entanglement vs. Code Entanglement Static Classes Make For Rigid Architectures
I meant in the language itself, not the tools. :) Jeremy Falcon
-
It's an interesting idea. I just played around with it too (so, hmm, the lounge post actually seemed to work). Floating point support is done by rational numbers, BTW. What I'm intrigued by is, for example, I have some really complex code that handles auto-generating SQL. It's complex enough that every single line of code has a comment describing what it does. There's a sort of redundancy there, and in fact, I often wish that I could code simply by writing the comments! This is sort of like that, I think. And oddly, it's reminiscent of something I tried a long time ago on a Commodore PET/64, when I was doing pretty much assembly language programming, and thinking, why can't I just express my thoughts in plain English? So yeah, variable definition, declaration, and assignment is ridiculously verbose. But the real power of something like this comes to play when expressing higher level concepts, that are built on lower level concepts. Whether this program actually succeeds at that, or is simply a glorified macro compiler, I don't know yet. [edit]And yes, the articles as they stand are inappropriate. But inappropriate to the point of deleting them? I'm not convinced of that, because I could argue the case that whatever the paradigm shift is, it IS interesting and has some worth.[/edit] Marc Pensieve Functional Entanglement vs. Code Entanglement Static Classes Make For Rigid Architectures -- modified at 0:53 Tuesday 7th March, 2006
Marc Clifton wrote:
Floating point support is done by rational numbers, BTW.
I disagree. AFAIK, you can't express sqrt(2) as a rational number. I need that square root to a decent precision! Don't take it away from me! :(
Marc Clifton wrote:
But the real power of something like this comes to play when expressing higher level concepts, that are built on lower level concepts.
Agreed - but I don't think they're quite there yet. - F
-
Marc Clifton wrote:
Whether this program actually succeeds at that, or is simply a glorified macro compiler, I don't know yet.
Well I didn't try the product, but so far it seems to me like it's one step forward and two steps back. You can't offer a different paradigm while taking things away from an enviroment we're used to, which is what this guy did. Or, so it seems. Jeremy Falcon
Jeremy Falcon wrote:
You can't offer a different paradigm while taking things away from an enviroment we're used to, which is what this guy did.
I think what you mean to say is that you cannot take stuff away just by saying, "It sucks". There's guys at MIT that are working on smaller paradigm shifts than this and have 1000's of pages of documentation to justify their assertions. I'm all for a paradigm shift taking us from A directly to Z and leaving no trace of A or anything in between so long as compelling arguments are given, a just amount of research has been done and credibility has been established. In his opening post he mentions 4 developers of significant stature that would be treated as kings if they even posted a picture of their unclothed butts here. Why? Because they have dedicated years and years of their lives to what they do. These guys have pumped out way more documentation than code. They've been on standards committees they've worked for leading companies. They've literally paved the roads of development, platforms and standards. Of course they would be treated differently they've earned it, proven they've earned it and do not in any way act like they deserve it (tongue in cheek on that last comment, most of them do it for the love of doing it). I welcome any well grounded, thoroughly researched and exhaustively documented paradigm shift. Documentation written by Adam Sandler just isn't going to wet my whistle... - Rex
I only read CP for the articles. Code-frog System Architects, Inc.
-
I meant in the language itself, not the tools. :) Jeremy Falcon
Jeremy Falcon wrote:
meant in the language itself, not the tools.
Ah. :) Just to put my foot in it, I remember when people howled that C# took away their beloved pointers and multiple inheritence. :-D (BTW, I'm off to bed. It's 1:11 AM here EST, and I wouldn't have been up except to work with a client on CA time, figuring out a corrupt database issue.) Marc Pensieve Functional Entanglement vs. Code Entanglement Static Classes Make For Rigid Architectures
-
Jeremy Falcon wrote:
meant in the language itself, not the tools.
Ah. :) Just to put my foot in it, I remember when people howled that C# took away their beloved pointers and multiple inheritence. :-D (BTW, I'm off to bed. It's 1:11 AM here EST, and I wouldn't have been up except to work with a client on CA time, figuring out a corrupt database issue.) Marc Pensieve Functional Entanglement vs. Code Entanglement Static Classes Make For Rigid Architectures
Marc Clifton wrote:
Just to put my foot in it, I remember when people howled that C# took away their beloved pointers and multiple inheritence.
Well, I still haven't completely moved to C#. Although, I've done VB for years (I repent, hey it's where the market is where I live). I still like my good ol' C, but like with anything else the higher level languages are a tool.
Marc Clifton wrote:
BTW, I'm off to bed.
Then I'm sure you'll have a splended morning when you arise and find my post in your inbox. Life couldn't get any better. :-D Jeremy Falcon
-
Jeremy Falcon wrote:
You can't offer a different paradigm while taking things away from an enviroment we're used to, which is what this guy did.
Sure you can. Look at MyXaml. Declarative programming without Intellisense, forms have to be created by editing XML instead of using a form designer, errors are detected at runtime instead of during the build, and so forth. Yet people use it, and quite successfully, depending on the problem domain they're trying to solve. Marc Pensieve Functional Entanglement vs. Code Entanglement Static Classes Make For Rigid Architectures
Marc Clifton wrote:
Sure you can. Look at MyXaml. Declarative programming without Intellisense, forms have to be created by editing XML instead of using a form designer, errors are detected at runtime instead of during the build, and so forth. Yet people use it, and quite successfully, depending on the problem domain they're trying to solve.
True, but you don't have to give up the power of the language for the ease of use of MyXaml (which I have looked at and passed on to others), it's "in addition" to using what you already have. So you either accept the improvement, or you don't. It has left me thinking of doing something similar in 3D world space, but time to write new tools is at a premium. (though I guess I could be doing so now... ;) ) _________________________ Asu no koto o ieba, tenjo de nezumi ga warau. Talk about things of tomorrow and the mice in the ceiling laugh. (Japanese Proverb)
-
Marc Clifton wrote:
Floating point support is done by rational numbers, BTW.
I disagree. AFAIK, you can't express sqrt(2) as a rational number. I need that square root to a decent precision! Don't take it away from me! :(
Marc Clifton wrote:
But the real power of something like this comes to play when expressing higher level concepts, that are built on lower level concepts.
Agreed - but I don't think they're quite there yet. - F
Fisticuffs wrote:
I need that square root to a decent precision!
10 digits? 20? 14142135623730950488016887242097 -------------------------------- 1000000000000000000000000000000 not reduced. ;P Marc Pensieve Functional Entanglement vs. Code Entanglement Static Classes Make For Rigid Architectures
-
Fisticuffs wrote:
I need that square root to a decent precision!
10 digits? 20? 14142135623730950488016887242097 -------------------------------- 1000000000000000000000000000000 not reduced. ;P Marc Pensieve Functional Entanglement vs. Code Entanglement Static Classes Make For Rigid Architectures
-
In response to the ongoing debate regarding the appropriateness of the Plain English Compiler articles, I've requested and received a free copy of the compiler/IDE, available for purchase at the respective website for $100. As such, I'm going to report my first impressions (have to be brief, I actually have to study for my Friday exam at some point :P). The demo was sent immediately, and checked clean for malware/spyware (as some people were suspicious of). It included documentation and some sample programs, including the "plain english" source code for the compiler itself, which compiled and ran successfully. Unfortunately, I question the value of this product for professional programmers. Everything about it indicates this is a "proof-of-concept" project, and not a polished final product. The documentation and the IDE are simply not up to professional standards, and the language structure is questionable. The language syntax certainly appears to be English, loosely wrapped around Win32. The documentation uses cutsey, unintimidating language geared to a beginner, and frankly, the IDE reflects certain assumptions about the incompetence of the user: there is no stand-alone command-line compiler option as far as I could tell, and (among other things) the IDE runs full-screen and AFAIK, cannot be windowed, which makes multitasking a chore and just pisses me right off. The font they use looks like Comic Sans MS. Unintimidating, sure. Annoying, definitely. More importantly, several notable features draw the strength of the language into question: From the documentation:
- I don't do nested IFs - I don't do nested LOOPs - I don't do OBJECTS - I don't do REAL NUMBERS - I don't do EQUATIONSNow, some of this isn't necessarily a direct weakness of the language, particularly if we're trying to facilitate a paradigm shift. They claim that "Objects suck - we have limited record support, but the page editor should do just fine." Okay: but no equations? No real numbers? You CANNOT market a professional programming product without floating point support. They even go so far as to vaguely imply that people should stop being "math-heads." I'm not kidding.- No case sensitivityAside: I generally find this more frustrating than helpful.- "I don't care where you put your definitions"This may seem like a feature, but IMO, it's a weakness, because in large, complex programs, you're going to be hunting down your definitions when you gExcellent review... Do you suppose that this might be an appropriate language for a beginning course on programming concepts? I teach at the local college, and sadly our Intro to Programming is badly lacking a decent language for beginners. "...a photo album is like Life, but flat and stuck to pages." - Shog9
-
Okay, you just started the same thing I was going to but in his blog if I could contrive it. I agree with pretty much everything you just said. However and in addition to I was really struck at the similarities between this (Osmosian Order) and Visual Basic. The similarities are just two strong. Abstract away the technical as much as possible. Remove the ability to do truly powerful programming by ham-stringing the language to make it safe. Involve wordy but still technical (at times) phrases that don't accomplish a whole lot less or more than int i = 0; as an example. There is so much I have to say on this but really don't have time. I don't belive there's an audience/market that exists for this product as it is. Simply put it's a mildly different twist on Visual Basic and it does far less. When I first read what was in the articles I thought there was some real potential here. But honestly there isn't. Something like this (in order to work) must sit on an operating system built just for it. Which seems a bit backwards to me. Build an OS to support the compiler??? If I had the time I could shoot some different approaches at the concepts involved but honestly it's just another VB to me. Finally the real downside to me is that the author promotes/documents/emails in the same manner and tone. I got an email. That had one thing really sticking out. "If you like it go buy it." About 2 other sentences dressed that statement and I immediately deflated. I don't think there's much here other than a price tag. I don't think the author is nearly as interested in advancing his work/science as he is in advancing his account balance. I dunno. I'm probably done on this one. There's a bun but no meat. - Rex
I only read CP for the articles. Code-frog System Architects, Inc.
code-frog wrote:
There's a bun but no meat.
As the old saying goes, "You just can't beat good meat." :rolleyes: Jeremy Falcon
-
Excellent review... Do you suppose that this might be an appropriate language for a beginning course on programming concepts? I teach at the local college, and sadly our Intro to Programming is badly lacking a decent language for beginners. "...a photo album is like Life, but flat and stuck to pages." - Shog9
NO! Not at all. My gawd it would only make Americans stupider. As a country we need to understand bitwise shifting, direct memory access and violation. I've observed that no matter how gradually you start in programming, no matter how gentle the introduction. You will eventually find yourself contemplating something incredibly difficult. If that's database theory or compilers in school or brute force linear programming over a timeline to predict yields of a build process you are going to have to bite the turd sooner or later. Might as well be sooner I think.
I only read CP for the articles. Code-frog System Architects, Inc.
-
Excellent review... Do you suppose that this might be an appropriate language for a beginning course on programming concepts? I teach at the local college, and sadly our Intro to Programming is badly lacking a decent language for beginners. "...a photo album is like Life, but flat and stuck to pages." - Shog9
A bit like code frog: NO, not at all! For different reasons. What about teaching a language used in a real world? Of course I'm think C#, but there is also Java and VB noteworthy. All these 3 remove useless complexity while still providing full access to the concepts and programing power. I didn't mention C++ here because while it provide a bit more power, its complexity outweight its benefit, IMHO. (That is, unless you need its low-level easy and full access to the API).
-
Jeremy Falcon wrote:
You can't offer a different paradigm while taking things away from an enviroment we're used to, which is what this guy did.
I think what you mean to say is that you cannot take stuff away just by saying, "It sucks". There's guys at MIT that are working on smaller paradigm shifts than this and have 1000's of pages of documentation to justify their assertions. I'm all for a paradigm shift taking us from A directly to Z and leaving no trace of A or anything in between so long as compelling arguments are given, a just amount of research has been done and credibility has been established. In his opening post he mentions 4 developers of significant stature that would be treated as kings if they even posted a picture of their unclothed butts here. Why? Because they have dedicated years and years of their lives to what they do. These guys have pumped out way more documentation than code. They've been on standards committees they've worked for leading companies. They've literally paved the roads of development, platforms and standards. Of course they would be treated differently they've earned it, proven they've earned it and do not in any way act like they deserve it (tongue in cheek on that last comment, most of them do it for the love of doing it). I welcome any well grounded, thoroughly researched and exhaustively documented paradigm shift. Documentation written by Adam Sandler just isn't going to wet my whistle... - Rex
I only read CP for the articles. Code-frog System Architects, Inc.
code-frog wrote:
I think what you mean to say is that you cannot take stuff away just by saying, "It sucks".
Well that and there has to be an alternative way to acheive simple mechanics to be able to perform relative tasks even in that new paradigm.
code-frog wrote:
Documentation written by Adam Sandler just isn't going to wet my whistle...
:laugh: Jeremy Falcon
-
Excellent review... Do you suppose that this might be an appropriate language for a beginning course on programming concepts? I teach at the local college, and sadly our Intro to Programming is badly lacking a decent language for beginners. "...a photo album is like Life, but flat and stuck to pages." - Shog9
Roger Wright wrote:
Do you suppose that this might be an appropriate language for a beginning course on programming concepts?
It would be like teaching someone to drive by giving them a Tonka truck...
Ryan
"Punctuality is only a virtue for those who aren't smart enough to think of good excuses for being late" John Nichol "Point Of Impact"
-
In response to the ongoing debate regarding the appropriateness of the Plain English Compiler articles, I've requested and received a free copy of the compiler/IDE, available for purchase at the respective website for $100. As such, I'm going to report my first impressions (have to be brief, I actually have to study for my Friday exam at some point :P). The demo was sent immediately, and checked clean for malware/spyware (as some people were suspicious of). It included documentation and some sample programs, including the "plain english" source code for the compiler itself, which compiled and ran successfully. Unfortunately, I question the value of this product for professional programmers. Everything about it indicates this is a "proof-of-concept" project, and not a polished final product. The documentation and the IDE are simply not up to professional standards, and the language structure is questionable. The language syntax certainly appears to be English, loosely wrapped around Win32. The documentation uses cutsey, unintimidating language geared to a beginner, and frankly, the IDE reflects certain assumptions about the incompetence of the user: there is no stand-alone command-line compiler option as far as I could tell, and (among other things) the IDE runs full-screen and AFAIK, cannot be windowed, which makes multitasking a chore and just pisses me right off. The font they use looks like Comic Sans MS. Unintimidating, sure. Annoying, definitely. More importantly, several notable features draw the strength of the language into question: From the documentation:
- I don't do nested IFs - I don't do nested LOOPs - I don't do OBJECTS - I don't do REAL NUMBERS - I don't do EQUATIONSNow, some of this isn't necessarily a direct weakness of the language, particularly if we're trying to facilitate a paradigm shift. They claim that "Objects suck - we have limited record support, but the page editor should do just fine." Okay: but no equations? No real numbers? You CANNOT market a professional programming product without floating point support. They even go so far as to vaguely imply that people should stop being "math-heads." I'm not kidding.- No case sensitivityAside: I generally find this more frustrating than helpful.- "I don't care where you put your definitions"This may seem like a feature, but IMO, it's a weakness, because in large, complex programs, you're going to be hunting down your definitions when you gI had this idea way back in university. I scrapped the idea, because it's just making programming more verbose. Who the hell wants to write a frigging 500 page novel, just to perform some simple tasks? And moreover, who'd want to read that 500 page (utterly boring) novel? The day when I will embrace natural programming languages is the day when I can tell the compiler to "Do my work, or fuck off!" :rolleyes:
-
I had this idea way back in university. I scrapped the idea, because it's just making programming more verbose. Who the hell wants to write a frigging 500 page novel, just to perform some simple tasks? And moreover, who'd want to read that 500 page (utterly boring) novel? The day when I will embrace natural programming languages is the day when I can tell the compiler to "Do my work, or fuck off!" :rolleyes:
Jörgen Sigvardsson wrote:
"Do my work, or f*** off!"
By that time the computer will being telling us that. :-D:rolleyes: Jeremy Falcon
-
I had this idea way back in university. I scrapped the idea, because it's just making programming more verbose. Who the hell wants to write a frigging 500 page novel, just to perform some simple tasks? And moreover, who'd want to read that 500 page (utterly boring) novel? The day when I will embrace natural programming languages is the day when I can tell the compiler to "Do my work, or fuck off!" :rolleyes:
Jörgen Sigvardsson wrote:
The day when I will embrace natural programming languages is the day when I can tell the compiler to "Do my work, or f*** off!"
I'd be too worried about which choice it would make... :~
Ryan
"Punctuality is only a virtue for those who aren't smart enough to think of good excuses for being late" John Nichol "Point Of Impact"
-
In response to the ongoing debate regarding the appropriateness of the Plain English Compiler articles, I've requested and received a free copy of the compiler/IDE, available for purchase at the respective website for $100. As such, I'm going to report my first impressions (have to be brief, I actually have to study for my Friday exam at some point :P). The demo was sent immediately, and checked clean for malware/spyware (as some people were suspicious of). It included documentation and some sample programs, including the "plain english" source code for the compiler itself, which compiled and ran successfully. Unfortunately, I question the value of this product for professional programmers. Everything about it indicates this is a "proof-of-concept" project, and not a polished final product. The documentation and the IDE are simply not up to professional standards, and the language structure is questionable. The language syntax certainly appears to be English, loosely wrapped around Win32. The documentation uses cutsey, unintimidating language geared to a beginner, and frankly, the IDE reflects certain assumptions about the incompetence of the user: there is no stand-alone command-line compiler option as far as I could tell, and (among other things) the IDE runs full-screen and AFAIK, cannot be windowed, which makes multitasking a chore and just pisses me right off. The font they use looks like Comic Sans MS. Unintimidating, sure. Annoying, definitely. More importantly, several notable features draw the strength of the language into question: From the documentation:
- I don't do nested IFs - I don't do nested LOOPs - I don't do OBJECTS - I don't do REAL NUMBERS - I don't do EQUATIONSNow, some of this isn't necessarily a direct weakness of the language, particularly if we're trying to facilitate a paradigm shift. They claim that "Objects suck - we have limited record support, but the page editor should do just fine." Okay: but no equations? No real numbers? You CANNOT market a professional programming product without floating point support. They even go so far as to vaguely imply that people should stop being "math-heads." I'm not kidding.- No case sensitivityAside: I generally find this more frustrating than helpful.- "I don't care where you put your definitions"This may seem like a feature, but IMO, it's a weakness, because in large, complex programs, you're going to be hunting down your definitions when you gVB I miss you! The tigress is here :-D
-
In response to the ongoing debate regarding the appropriateness of the Plain English Compiler articles, I've requested and received a free copy of the compiler/IDE, available for purchase at the respective website for $100. As such, I'm going to report my first impressions (have to be brief, I actually have to study for my Friday exam at some point :P). The demo was sent immediately, and checked clean for malware/spyware (as some people were suspicious of). It included documentation and some sample programs, including the "plain english" source code for the compiler itself, which compiled and ran successfully. Unfortunately, I question the value of this product for professional programmers. Everything about it indicates this is a "proof-of-concept" project, and not a polished final product. The documentation and the IDE are simply not up to professional standards, and the language structure is questionable. The language syntax certainly appears to be English, loosely wrapped around Win32. The documentation uses cutsey, unintimidating language geared to a beginner, and frankly, the IDE reflects certain assumptions about the incompetence of the user: there is no stand-alone command-line compiler option as far as I could tell, and (among other things) the IDE runs full-screen and AFAIK, cannot be windowed, which makes multitasking a chore and just pisses me right off. The font they use looks like Comic Sans MS. Unintimidating, sure. Annoying, definitely. More importantly, several notable features draw the strength of the language into question: From the documentation:
- I don't do nested IFs - I don't do nested LOOPs - I don't do OBJECTS - I don't do REAL NUMBERS - I don't do EQUATIONSNow, some of this isn't necessarily a direct weakness of the language, particularly if we're trying to facilitate a paradigm shift. They claim that "Objects suck - we have limited record support, but the page editor should do just fine." Okay: but no equations? No real numbers? You CANNOT market a professional programming product without floating point support. They even go so far as to vaguely imply that people should stop being "math-heads." I'm not kidding.- No case sensitivityAside: I generally find this more frustrating than helpful.- "I don't care where you put your definitions"This may seem like a feature, but IMO, it's a weakness, because in large, complex programs, you're going to be hunting down your definitions when you gGreat post! Means I do not have to look at the product ;) Back in the day, I always wanted programming to move to point-and-click drag-and-drop style on a graphic screen and drop all the text. Oh well, one of these days ;) Rocky <>< Latest Post: SQL2005 Server Managemnet Studio timeouts! Blog: www.RockyMoore.com/TheCoder/[^]
-
In response to the ongoing debate regarding the appropriateness of the Plain English Compiler articles, I've requested and received a free copy of the compiler/IDE, available for purchase at the respective website for $100. As such, I'm going to report my first impressions (have to be brief, I actually have to study for my Friday exam at some point :P). The demo was sent immediately, and checked clean for malware/spyware (as some people were suspicious of). It included documentation and some sample programs, including the "plain english" source code for the compiler itself, which compiled and ran successfully. Unfortunately, I question the value of this product for professional programmers. Everything about it indicates this is a "proof-of-concept" project, and not a polished final product. The documentation and the IDE are simply not up to professional standards, and the language structure is questionable. The language syntax certainly appears to be English, loosely wrapped around Win32. The documentation uses cutsey, unintimidating language geared to a beginner, and frankly, the IDE reflects certain assumptions about the incompetence of the user: there is no stand-alone command-line compiler option as far as I could tell, and (among other things) the IDE runs full-screen and AFAIK, cannot be windowed, which makes multitasking a chore and just pisses me right off. The font they use looks like Comic Sans MS. Unintimidating, sure. Annoying, definitely. More importantly, several notable features draw the strength of the language into question: From the documentation:
- I don't do nested IFs - I don't do nested LOOPs - I don't do OBJECTS - I don't do REAL NUMBERS - I don't do EQUATIONSNow, some of this isn't necessarily a direct weakness of the language, particularly if we're trying to facilitate a paradigm shift. They claim that "Objects suck - we have limited record support, but the page editor should do just fine." Okay: but no equations? No real numbers? You CANNOT market a professional programming product without floating point support. They even go so far as to vaguely imply that people should stop being "math-heads." I'm not kidding.- No case sensitivityAside: I generally find this more frustrating than helpful.- "I don't care where you put your definitions"This may seem like a feature, but IMO, it's a weakness, because in large, complex programs, you're going to be hunting down your definitions when you gFisticuffs wrote:
From the documentation: - I don't do nested IFs - I don't do nested LOOPs - I don't do OBJECTS - I don't do REAL NUMBERS - I don't do EQUATIONS
Quack quack THUD! (It's a dead duck then) The tigress is here :-D