 |
|
|
I have a site up and running, although not being used at the moment using DNN. Nothing special at all, pretty basic and not all that pretty atm, just a very slight tweak of one of the default themes. Clickety[^]
He who makes a beast out of himself gets rid of the pain of being a man
|
|
|
|
|
|
|
I am specifically talking about extending, and the pain that goes with that. Once it runs, it runs fine, but I cant wait 20 minutes for the site to compile every time I make a tiny change.
|
|
|
|
|
|
|
Ah ok I'm with you, can't be of any help though, good luck! 
He who makes a beast out of himself gets rid of the pain of being a man
|
|
|
|
|
|
|
The way we had it implemented, it would not have to compile the core more than once. Ever. The things that did compile were the DNN website, which references the DNN libraries, and our custom modules, which we wrote in C#. We had one VS solution that was the core, and one VS solution that had the website and custom modules in it. We worked really hard to not modify the core, or at least keep our changes documented and short so that when the next version came out (which it did about every 6-8 weeks), we could upgrade without any problems.
I would suggest trying our solution to your problem: splitting up your VS solutions into core DNN libraries and the website with custom modules.
Success is the happy feeling you get between the time you do something and the time you tell a woman what you did.
--Dilbert
My left name is Tremendous Savings, Ms. America – Señor Cardgage
|
|
|
|
|
|
|
This is the way a large project I'm working on is structured. The only time we need a whole recompile is when interfaces or web service method signatures change between projects.
Todd Smith
|
|
|
|
|
|
|
Not sure if this[^] helps.
There are ll kinds of people in the world, those who understand binary and those who understand Roman numerals. Web - Blog - RSS - Math
|
|
|
|
|
|
|
http://www.sciencedaily.com/releases/2008/07/080716154355.htm[^]
"The key lies in an esoteric reformulation of mathematics called ‘type theory’ based on the notion of computation. In this approach, the specification for a computational task is stated as a mathematical theorem. The program that performs the computation is equivalent to the proof of the theorem. By proving the theorem the program is guaranteed to be correct."
Sounds like TDD to me, just with math terms. Write the test then write the code to pass the test.
only two letters away from being an asset
|
|
|
|
|
|
|
Mark Nischalke wrote: Sounds like TDD to me
Sounds like another formal specification language to me. A bit like Z[^].
|
|
|
|
|
|
|
Yeah, doesn't seem like they are saying anything new. Makes me think of Spark[^].
Simon
|
|
|
|
|
|
|
I dreaded that module at uni, was hoping I'd never hear of it again, bleh 
He who makes a beast out of himself gets rid of the pain of being a man
|
|
|
|
|
|
|
But then some architect would write a framework that removes all types because it is more enterprisey.
Need a C# Consultant? I'm available.
Happiness in intelligent people is the rarest thing I know. -- Ernest Hemingway
|
|
|
|
|
|
|
We know how to write proofs for software. That's not the problem. The problem is that it costs way too much, in both money and time, to do it for any but the most safety-critical software.
We also know how to perform enough inspection and testing to catch very large proportions of errors. Again, it's cost-prohibitive.
The fact is, our customers don't want to pay for fault-free software. Or, we haven't explained to them just how much it will cost and allowed them to choose the value of this parameter.
Proof isn't really compatible with evolutionary development as you really have to have a thorough design to attempt to prove.
By the way, your compiler already does some type-theory proving - well, unless you're using a weakly-typed language (like VB with Option Strict off). Most of us don't use a very strong language like Ada because again, defining your own integer types and converting between them as required is time-consuming and hence costly. C++ and C# are strongly-typed for custom types, but there's no concept of different types of integer. I suppose you could design some structures with overloaded operators (e.g. Velocity, Distance, Time: summing any quantity would give the same type, multiplying Velocity and Time gives Distance, etc) but you still have to have access to the raw value to interoperate with any other code.
I'd like to see anyone prove a user interface. You're so dependent on your host platform working correctly, which of course you don't have a proof for.
DoEvents: Generating unexpected recursion since 1991
|
|
|
|
|
|
|
Like many of the comments thus far, this is only applicable to safety-critical software that can be formally specified. The problem with 'software' in general is that most of it cannot be formally defined, or, if it were, the errors would still be there -- just abstracted to the formal spec. Also mentioned prevoiously -- try to formally prove a UI design.
In my opinion, there is no single software development strategy that applies to all software. For example, the waterfall method requires the specification, design, implementation, all be done completely and correctly (or at least in sequence). This is a greatly inefficient way to develop software that has loosely-stated requirements (and you're never going to get the requirements right anyway)... Agile works only if there is a rapid prototyping framework in place. If you can't quickly build and release components to a target in an embedded environment, this approach becomes ineffective (not to mention the potential CM issues that will arise from any project of more than trivial scope). TDD? Same fate as Agile.
So in essence the crux of the problem (i.e. why 'software' hasn't been pinned down like other engineering disciplines) is that the term "software" is all-encompassing and no single method is good for all applications. Don't get me wrong, technologies like Formal Methods has its place, but the idea here is that there is no 'silver bullet' or single process. It comes down to properly scoping a project, selecting the tools and techniques that best suit the project, and establishing adequatre process control such that the effort can be managed -- the continuous balance of cost, schedule, and risk.
|
|
|
|
|
|
|
Also we overestimate how precise engineering is. I used to work in the oil and gas industry. There are still lots of mistakes but "reliability" is achieved by redundancy.
Kevin
|
|
|
|
|
|
|
I thought this was going to be about MS dropping support for VB. *sigh*.
|
|
|
|
|
|
|
The core part of Scade[^] is written like that - except it uses ML as the implementation language, as it has an almost one-to-one mapping with the specification language. Oh - and don't forget the many, many pages of specification  .
As Mike said we already know how to do this (see CICS[^] - a classic poster boy for formal specification) - it's just that for most systems, the cost completely outweighs the benefit in terms of residual defects.
You're probably better off using a language with a higher level of abstraction (Ruby, Python?) so you write less code, on the basis that the less code you write, the less likelihood there is of introducing errors!
|
|
|
|
|
|
|
This sounds like utter BS to me (it's more likely a way for idiot clever college professors to get grants.)
Anyone who thinks he has a better idea of what's good for people than people do is a swine.
- P.J. O'Rourke
|
|
|
|
|
|
|
"A Journey of a Thousand Rest Stops Begins with a Single Movement"
|
|
|
|
|
|
|
Say no more........
Actually I am looking for a decent book on photography.
My daughter (14) is doing an additional course at school over the next two years (some sort of arty farty creative course) and she wants to do the majority of it via photos and manipulated images.
Is there a good book that takes someone from a point and clicker to a decent level of understanding and techniques?
I know several people here are well into photography so I am, unashamedly, trying to pick your brains.
Thanks.
I still remember having to write your own code in FORTRAN rather than be a cut and paste merchant being pampered by colour coded Intellisense - ahh proper programming - those were the days
|
|
|
|
|
|
|
i read a "Photography For Dummies" way back when i made the transition to SLRs. that's more than enough to get started. the basics of exposure and aperture aren't too complicated*, and should be easy to pick up - especially with a DSLR, where the results are instant and taking a picture is essentially free.
composition, technique, etc. are more art than mechanics, of course.
* - though they can be, if you want to get into the physics of what lenses do. but that's not necessary. you don't need to know why a 50mm lens shows a wider field of view than a 300mm lens to know that it does.
|
|
|
|
|
|
|
Books? You're so TwenCen! 
"It's so simple to be wise. Just think of something stupid to say and then don't say it."
-Sam Levenson
|
|
|
|
|
|
|
I've just bought these[^] two[^] (after getting one of these[^] ) - at first inspection, they look to be pretty good - for example, the article that says "Use aperture priority to control depth-of-field" - a clear, concise explanation of how changing aperture affects DoF.
|
|
|
|
|
|
|
I was just paroosing the CP forums and articles, and a couple of the developers (particularily VB ones) get shot down for their choice of language: I am a C# developer myself, but I can't seem to fathom the reasons why people keep squabbling over programming languages.
Firstly, you get the typical guy who says: "your language suckzzoorrzzz." Honestly he is showing himself up to be the real idiot because he is demonstrating that he [most probably] has experience only in one programming language. This means that if he has to get a job some day in a real company with real legacy systems he is going to be stuck in the middle of nowhere with his 1337 Foo# programming skillz.
Then you get the guy who simply says "my language is better than yours/best there is," basing his argument on statistics and the amount of people who use the language etc. Just because a lot of people use (or understand) the language doesn't make it the bees knees. Take a web scripting language for example, I am sure thousands of people use any specific one (and do amazing things with it): but it sure isn't possible to write a IMAP server in one. Just because it's popular doesn't make it universally applicable. Then take a language like x86 assembler, definitely not that many developers, but they must get paid bucket loads and have sooo much more skill than the rest of us.
The typical CP article message: "I need a version of this in [insert language here]." I really wonder how many copy'n'paste programmers are writing software that runs our lives today and it scares me. It seems as though the art and joy of programming has been lost to the ability to copy'n'paste (although a DirectShow interop library always helps ). Apart from that, most of these guys are asking for the stuff in C#: there are a couple VB-to-C# converters out there, so they are also showing a lack of Google skills.
Finally (although OT) you get the guys who say, "your GUI sucks." Honestly, I think the better programmers would be the ones who are useless at GUIs. If you want a fancy GUI find an article on a Office 2007-style ribbon and not on a complex network protocol. Infact, with the WPF trend, programmers need only make the bare essential GUIs and the designers are left to make them look pretty.
I think developers who want to say any of the above should seriously consider learning a new language: and they should take the language used by the author for a start.
I joined the Java IRC chat room a while back claiming that I wanted to write a Java Bytecode - IL translator in C# (with J2ME on Windows Mobile in mind). I was given a pretty poor reception by all except one guy. With the help from that one person I was able to write a .class reader in C# that should be able to benefit both communities one day.
Honestly, where has the patience gone? Any thoughts?
He who asks a question is a fool for five minutes. He who does not ask a question remains a fool forever. [Chineese Proverb]
Jonathan C Dickinson (C# Software Engineer)
|
|
|
|
|
|
|
I think you're pretty much on the mark. The most employed developers/programmers are the ones that pick the right tool for the job at hand, whatever that tool may be. And although my current preference is c# I really don't mind too much what I use though in the environment in which I presently work it is predominantly c# or java and mssql server or oracle so it makes commercial sense to stick with the tools that make the most money and/or keep me employed.
|
|
|
|
|
|
|
I think you're right on the money here.
Jonathan C Dickinson wrote: If you want a fancy GUI find an article on a Office 2007-style ribbon
Careful. The development team for Windows 7 are the same ones that did the Office 2007 Ribbon
Don't take any wooden nickels.
|
|
|
|
|
General 
News 
Question 
Answer 
Joke 
Rant 
Admin 
