|
I've actually since left the company & didn't end up using it much after all. Pity tho, as it does seem to have potential. The basic "gist" of it (for want of a better word) is that it'd allow you independence from any particular RDBMS (which is particularly useful to small ISV's I'd say). That's the upside - the downside is that it does require a significant time investment to get a decent, reusable DAL API that you could plug in to the various incarnations of one's project.
The reason I didn't get to use it as much as I would have liked to is that the project I was working on was way behind schedule and my team wasn't afforded enough time to get it (an iBatis DAL) up & running. Besides that, the company decided that they would be working exclusively with SQL Server in the future, and as such, the abstraction provided by using iBatis didn't provide any significant benefits.
I ended up building a lightweight DAL that could be applied to pretty much any SQL Server database with only a little extra code involved (setting up SQL parameters for specific objects). Other than that it does the job pretty well. Good article fodder were I eloquent enough 
But fortunately we have the nanny-state politicians who can step in to protect us poor stupid consumers, most of whom would not know a JVM from a frozen chicken. Bruce Pierson
Because programming is an art, not a science. Marc Clifton
|
|
|
|
|
I have a simple application that searches on terms like surname or phone number, and then displays results in a grid. Double clicking a row then opens a details edit form for the owner of the phone number. Editing owner details is easy, but I would like some suggestions on editing phone numbers for the owner. If it's a person, there probably is only one phone number, so I have a phone number field on the owner details form. It is populated from the first item in the PhoneNumbers collection for the owner.
Now, when a company is the owner of a whole range of numbers, I need to open a separate editor. I would like some commentary on my initial design choice: I have two listboxes, and above them two textboxes. I have three buttons, Add, which adds the number in textbox1 to the left listbox, which shows all numbers for the owner. AddRange, which adds all the numbers in the range from textbox1 to textbox2, and Remove, which removes the selected number from the list on the left to the list on the right. When the user clicks OK, changes are saved - new numbers are added, changed numbers are updated, and removed numbers are deleted.
How else could I do this?
|
|
|
|
|
Brady Kelly wrote: If it's a person, there probably is only one phone number
 ummmm I have 3
|
|
|
|
|
Of course, I have three as well. One is work, and two cell, oh, plus one fax.
I meant that, in the scenario of a person with one number, the first number in their numbers collection is exposed for editing without having to open the numbers collection editor I am trying to design. The first number in the collection will always appear in the single Phone Number field on the Person Details form.
|
|
|
|
|
I don't care much for the "Details Form" design. I prefer the Visual Studio editing design where you have methods of navigating your domain model, like trees and lists etc., and once you selected the item of interest you use the PropertyGrid for editing the values. This allows you to present a UI where the user can see everything at the same time, no popping up new windows (forms) on top of the previous one. This style of UI gives the user the impression that they are in control rather than the application "controlling the user".
|
|
|
|
|
That's way too involved for this little app. It's a 12 hour quote, and confined to only the functions described.
|
|
|
|
|
My mistake. I didn't realize it was a toy program.
|
|
|
|
|
Yes, the editing is essentially a toy. The app is used by a tracing agency primarily to do searches, on name, adress, or phone number, based on a database provided by a third party. Editing the records hardly makes sense, but it was a requirement that I still want to deliver and get paid for.
|
|
|
|
|
How can I use Runge Kutta algorithm to programme a fractal?
Noha.Y
|
|
|
|
|
Are you still working on this?
"Real programmers just throw a bunch of 1s and 0s at the computer to see what sticks" - Pete O'Hanlon
|
|
|
|
|
Hi, Can you help me to get a Neuronic Computation simple program?
Noha.Y
|
|
|
|
|
|
what is the advantages of functional dependencies ?
<nill>
|
|
|
|
|
Have a read of this[^] and forumlate your own answer based on it.
Deja View - the feeling that you've seen this post before.
|
|
|
|
|
Basically functional dependency focuses in database decomposition in tables. I had seen this ordinary link very earlier, so i had incept from it, i just wanted to verify that in bullets.
|
|
|
|
|
This is probably the first game I've ever worked on. I want to use it as a testing platform for concepts which will be used in another much larger project. To give you an idea of what "this game" is, I will describe it to you below:
It is a game with one or more characters, and some artificial intelligence opponents. It is a 3rd-person shooting game. The game is somewhat 3-dimensional, and uses Windows Presentation Foundation to draw the 3D scenery, but the maps are flat and tile-based (however, movement resolution is finer than a tile—They can have different positions/locations within the tile). Most calculations will be using two dimensions. The game has 3 major parts: the terrain, the terrain objects, and characters which move on the terrain and interact with other characters and objects on it. They are all contain in a level/stage/map.
The terrain is tile-based. A tile can accomodate any number of terrain object, and one character. Each tile is customizable by the designer (the image painted onto its surface). The entire collection of tiles may or may not form a rectangular terrain. The tiles are static and do not change during runtime.
Terrain objects are in-game objects which the characters can interact with. They may be walls, bushes, a roof over the head, explosive barrels, etc. Terrain objects have some properties: whether a character can walk through it, whether it is indestructible, how high it it from the tile it is on, etc. Several terrain objects can be on a tile. They can be placed on the terrain during design time or dynamically during runtime (such as setting explosives by a character, and detonating by another).
Characters are in-game entities that are mobile on the terrain. They are basically there to eliminate other [non-friendly/neutral] entities. Characters are also the most dynamic of them all. They can have any appearance. Their appearance changes when they are damaged, dead, or switching weapons. They can rotate anywhere from 1 to 360 degrees in very fine intervals. (The character on screen always faces forward. The camera is swung around the character to keep it that way.) Although character movements are not restricted to individual tiles (meaning they can move left a quarter of a tile rather than having to move to an entirely new tile) each tile can contain only one character. A character is a tile wide, which makes it impossible for two characters to be on the same tile. When a character uses a weapon (like a grenade, a gun, or setting mines) it always throws/shoots straight ahead towards whatever it is facing or in case of setting mines it drops it on the tile it is currently on.
With the above points in mind, I designed my game like this: (also keeping in mind it has to be easily extendable)
- Map data consists of a folder. The folder has an xml file describing the terrain, what objects are set on it, and and custom characters (or it can link to predefines ones). All images are external files link to in the xml file.
- When the map is chosen and the game begins loading, it loads the xml file first. The file is checked against a schema.
- The data is read and converted to in-game data structures. Any dependencies are loaded (like images)
- The data is passed to an instance of the "terrain management" class. The terrain management class creates an instance of a terrain object class for each terrain object and a character class for each instance of a character. (Those classes will be derived from for specific kinds of objects and characters.) The terrain manager simply acts as the communication medium which delivers status notificatiosn to all the proper instances. If an instance of a TripMine class will not be affected by a character moving from one tile to another, then it will not be notified of the event.
Example:
- TerrainObject
- Character
- AutonomousCharacter (Uses artificial intelligence)
- MainCharacter (Only one instance allowed, controlled using the keyboard and mouse)
- ExternalCharacter (Controlled via some other influence like over a network for multiplayer scenarios)
- Each character class may have a separate thread which controls its actions.
I would like some feedback on my preliminary design decisions, and any alternative designs that you may have come up with that will meet my requirements. (It has taken me a few nights to think it over.)
ROFLOLMFAO
|
|
|
|
|
How is your game coming along?
"Real programmers just throw a bunch of 1s and 0s at the computer to see what sticks" - Pete O'Hanlon
|
|
|
|
|
In the C++ community there is certain unhappy feeling towards the use of MACROs.
In a post someone was given the advice to use a inline function instead of a MACRO.
This is certainly good advice.
Now my question is the following:
Are there any reasons/issues or other situations where a MACRO can be preferred over another solution.
Or should one 'try' to find a better C++ approach.
One thing could be to decrease the amount of code, so it becomes more readable and less complex.
Any thoughts on this?
codito ergo sum
|
|
|
|
|
BadKarma wrote: One thing could be to decrease the amount of code, so it becomes more readable and less complex.
No, that is hiding complexity, not decreasing it. MACROS are a bad substitute for redesigning bad, unnecessarily complex code. They also make debugging more difficult. I can't think of any situation where a MACRO would be the preferred solution.
|
|
|
|
|
The only time that it's acceptable is if you are deliberately attempting to create bad code where you are the only person who can maintain it. Hmmm - oh wait - that's the sign of a really bad app.
Deja View - the feeling that you've seen this post before.
|
|
|
|
|
Actually MACROs are widely used in the ATL library, that is no such bad design example.
MACROs are also used in MFC, and, for instance, to make more readable operations on ListView, ecc..., DirectX often hides COM complexity using macros.
MACROs are hazardous of course and there are arguments against their use. But by no means MACROs are the origin or the signature of a bad design.
If the Lord God Almighty had consulted me before embarking upon the Creation, I would have recommended something simpler.
-- Alfonso the Wise, 13th Century King of Castile.
|
|
|
|
|
ASSERT & HIWORD/LOWORD are just two examples that came to mind.
I certainly wouldn't use a inline function for either of those.
I'm finding the only constant in software development is change it self.
|
|
|
|
|
The only time I would see as an advantage of using macros is when programming in C (I haven't used C++), an instance of this was some code I wrote for a company who shall remain anonymous where the biggest emphasis was on speed because it was performing rainflow analysis of vast quantities of data, all on a micro. God forbid but most functions were eliminated in favour of copy and paste and gotos were used because the original idea was then to use an FPGA to do the analysis, i.e. perform it in hardware so because they don't have a concept of functions etc these were copied in and gotos were used because they're nice and fast, especially in hardware.
In the end the other company decided to use a micro and refactored back my optimisations, then hit a problem, it was taking 4 times too long. I was called in and noticed they were doing stupid things like calling these tiny little functions, if it the micro had been able to manage it it would have had to call this routine 1,000 / second (note this was only a tiny part of the entire, complex algorithm) so I changed the function into a macro so it "looked" like a function to them but actually wasn't.
Of course I guess this is exactly what inline functions are for in C++ but most if not all C compilers don't support them. So no I can't think of any good reason 
|
|
|
|
|
In general macros are bad and should be avoided. I don’t believe this “macros are bad full stop” line however; macros have their uses. And I'm not alone: The Boost libraries are highly acclaimed C++ libraries and they use macros when needed; Andrei Alexandrescu says he’s not afraid to use macros here[^].
Steve
|
|
|
|
|
Macros are processed at pre-compile time, so they serve a purpose of course. Think the ASSERT in MFC. Can't do that with inlined functions. Well you could, but what a waste.
MACROS some might argue, are bad practice. Personally, there have been many times while programming in JS or PHP where I have considered running my code through a pre-processor to gain the benefits of macro expansion.
Ask any interpreted language developer with experience in C/C++ and they'd likely agree.
When you program OOP, everything is an object, at least in a purists eyes. Using inline functions would seem messy and pointless. Are you going to throw all generic functions inside a class named CGeneric and call them statically?
That would be hack, IMHO.
In those cases where you need some simple code, that is common in more places than a single class, Macros are a good thing. Most Macros that I have used, such as LOWORD, etc...make sense as a macro, not an inlined function...
So to answer your question, no Macros are not evil, the developers who use them inappropriately are.
There are times when I write code, that is clean and effective but could be slightly more elegant or efficient if I could use a GOTO (no longer the case in most languages). I think macros, globals, goto's and even multiple inheritence can still be useful, but their use has to be completely understood and justified, otherwise they promote bad coding practice and you end up with crap.
Then again, I've seen purist OO code with-out macros, globals, goto, or MI and it was crap too.
Cheers
I'm finding the only constant in software development is change it self.
|
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
