| |
| |
| |
|
Page: 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Comments:
<0> we didnt want to read it on a paper
<1> Read: Abstract: Visual programming is a general label for non-textual programming and has the potential of making software programming efficient and easier. A software architecture gives a high-level view of a software system. Architecture Description Languages or ADLs are languages used to model software architectures.
<2> "modelling ADLs in UML" reminds me somewhat of http://discuss.joelonsoftware.com/default.asp?joel.3.219431.12
<1> Pi-ADL is one such language, formally grounded in the Pi method and capable of modeling both software architectures and expression their implementation details.
<1> This paper describes how a visual model of Pi-ADL, an Architecture Description Language, can address the open issues in visual programming, and provide a basis for a complete visual programming environment.
<3> _m_: too lazy to check, does it bash UML?
<2> MrAshe: no, it's not about UML per se
<2> MrAshe: It's about adding another layer around established concepts.
<3> IRR: now I understand why you think C++ is complicated, you want legos ;)
<0> IRR: explain how you use pi calculi for visual programming?
<2> And about repeating that until total uselessness.
<1> MrAshe ADLs are not the same thing as Components
<3> Ah yeah I've read tha tone
<3> IRR: I just mean you want something a kid can play with
<3> Or you
<1> The lego concept is popular
<1> And stupid
<0> IRR: first you need to bring down the psychic link me and ashe have
<3> Well, I know all the types of legos, and you don't cause you come from a piece of desert ;)
<3> So I'm not ignorant :D
<1> If software could be modelled like lego, we wouldn't need software
<0> IRR
<1> Everything could be accomplished in hardware
<0> explain how Pi-calculi can be used for visual programming
<4> god
<1> And you come from a country the size of a parking lot
<4> coding up a GUI is so BORING
<3> Yes, and we have parking lots
<3> You don't have cars
<3> :D
<1> Quantum`` do you know about the Deutsch Limit?
<3> It's the lowest your bank account can go if you have an account at Deutsche Bank
<0> yes
<0> MrAshe :)
<0> IRR: yea
<0> how does that fall back to Pi-calculi at all?
<1> Quantum`` ok, so Pi-ADL is presented as a candidate that has the language constructs that address that cognitive problem
<1> Through hierarchical modeling and step wise refinement semantics
<1> There are other approaches as well like aspect oriented programming
<3> At least it'll sound smart, even if it doesn't say much
<0> ok see
<0> I know about pi calculus and aop
<0> but they have nothing to do with visual programming
<0> thats like saying template metaprogramming and visual programming in the same line
<3> Even I know about pi calculus
<1> And here is my proposed work: The model presented sets the stage for subsequent research in applied ADL visual modeling. The model will be used as a teaching tool in software architectures coursework, and the feedback will be used to further improve the model. We intend to implement Pi-ADL as a .NET language on the Microsoft Common Language Infrastructure.
<3> But that might be due to the psychic link
<1> The visual model presented in this paper will then form the foundation of a true visual programming environment that generates compiled CLR code and leverages the immense power and versatility of the .NET framework.
<1> This work will give insights into both the design of Pi-ADL and that of its visual model. Our goal, eventually, is to develop a general purpose visual programming environment that addresses the efficiency, scalability and extensibility needs of modern software systems.
<1> Quantum`` there is a difference between pi calculus and pi adl
<3> "immense power and versatility of the .NET framework"
<3> hahah
<1> Its like saying lambda calc is the same thing as lisp
<0> IRR: I see...
<1> MrAshe and you disagree?
<0> the only way to go from pi to visual, is circles for processes and arcs with actions? :)
<3> There's nothing immense in .NET
<0> am I wrong IRR?
<2> MrAshe: there is. Marketing.
<1> Quantum`` yes
<3> _m_: yeah, and maybe the bloat
<0> IRR: so I am wrong or not?
<1> Quantum`` you are wrong
<0> ok so explain in short Pi ADL
<1> block diagrams are not the only option in visual programming
<3> Porn diagrams++
<0> well it must be visual right? so arcs and boxes? no?
<3> If you place the ***** next to the penis, it makes a sound
<1> Pi-ADL is an Architecture Description Language presented in [13]. The purpose of Architecture Description Languages is to model software architectures, and according to [17] they focus on the high-level structure of the overall application rather than the implementation details of any specific source module.
<1> However Pi-ADL does provide programmatic constructs which allow it to model implementation details so it can be seen as both a modeling and an implementation language. Further, it is complemented with Pi-ARL, an Architecture Refinement Language for the stepwise refinement of software architectures [12].
<1> Quantum`` close: in our case its components and connectors
<1> And ports and connections
<3> You just said it was not components
<1> And Composites and abstractions
<1> MrAshe component research is different from architectural components
<1> It'll be helpful if you get a little educated
<3> And architectural components are not components? :D
<1> architectural components == design abstractions
<3> So a component is not a component?
<1> components == COM, EJB etc.
<5> "a little education is a dangerous thing"
<1> Certain languages refer to cl***es as templates
<1> so C++ templates are different from those templates which are different from word document templates
<2> I have difficulties to see the bridge between pi calculus and a programming language.
<1> So a template is not a template?
<3> Nope, a template is a template
<3> An idiot is an idiot
<3> And an IRR is an IRR
<1> Precisely
<3> So a component is a component
<3> You lose
<3> (again)
<5> geezus, MrAshe you've never encountered technical terms that mean diffferent things in different contexts?
<0> _m_, so do I, but I know what he is doing I "think" just show a process as a circle, ports are dots on the circle and arcs are the connections between the ports
<6> it is quite pointless for you to ask me to show you my code when i'm asking you how i'm supposed to write the code
<3> vawjr: I'm just making fun of him cause he believes he's smart
<1> _m_ you write a language and formally verify its syntax against a calculus. Thats what "formally founded" means
<5> ViPr no, we ask to see what you're tried
<3> ViPr asked a question?
<0> but Pi Calculus is to reason about Parallel processes, NO???
<5> no, he was bitching about our instructions in the topic
<2> Quantum``: Doesn't the pi calculus describe a soup of potentially executable code snippets waiting for their respective start condition?
<0> and about communications
<1> Quantum`` yes
<1> yes
<2> Being a soup, I don't see the arcs.
<0> _m_, communicating processes, the channels can be arcs
<7> a soup ?
<5> Quantum`` yes, I liked Kevlar
<5> still do actually
<0> :) thanks vawjr :)
<0> they brought down the site I just looked it up
<0> prolly maintainance that screwed up
<5> <sigh> I sure hope I got everything copied off it
<6> ok wait i think i'm understanding something. there's a difference between template parameter and function parameter
<3> Heheh
<6> but then how do i call a function that has template parameters
<3> Like any other function
<5> ViPr yup template <typename templateparameter> void myfunc(int functionparameter)
<3> And you specify the template arguments if they can't be deduced
<3> template <typename T> void func( T const & ) // can be deduced, so func( somevar ); works
<3> template <typename T> void func( void ) // can't be deduced, so func<int>(); works
<0> anyway IRR, I dont get it very well, just guesses... but once you got it in a more substantial format ill be interested in looking into it
<6> so to use the function i must writesomething like: function<x,y,z>(a,b);
<3> Yeah
<5> ViPr NO
<4> haha
<5> it depends on how you'e got the arguments deducable
<5> you can always write it as function<x,y,z>(a,b) if you wish
<1> Quantum`` sure
<6> or function <4,7,2>(a,b);
<3> Why not
<0> the language should be changed, so that types are deduced as much as possible
<3> They're already deduced as as much as possible
<5> Quantum`` you've voting FOR auto then, I presume
<0> the way of the future, the way of the future, the way of the future
<0> yes vawjr
<0> I am :)
<4> yea I'm with Quantum``
<6> so within the <> i write the static constant template parameter data and within the () i write the dynamic variables?
<0> MrAshe, not just templates
<3> template <typename T> T func() { return T(); }
<0> MrAshe, anything
<5> ViPr that depends
<3> How do you deduce it ;)
<4> werd
<5> you haven't given us an example of what your function does
<6> what difference does it make what it does
<0> IRR: How the hell is Aspect Oriented Programming even remotely visual???
<5> well, it MAY be that the types of a and b suffice to deduce the template args
Return to
#c++
or
Go to some related
logs:
#linux
#linux
#linux
unix + torrents
#chatzone
philosopher warts frogs
#apache
Preprocessor Guards + c++
#AllNiteCafe
#linux
|
|