I was trying to way that 1 of Microsofts problems[with the w3c] was that the box width =width = content width + padding + borders.
For instance I always you em's for content sizes, but often use px's for margins or borders. (usually about 1px-5px)
The Big problem is Width = 100% and I want a margin or x pixels, I screwed if I use width 100% margin 5pixels because my page will end up being 100% wide + 10pixels.
You're asking for elements which could be siblings, ancestors, descendants to all share a common value that, for each, has to be calculated independently.
How do tables work at the moment then?
Wrap a container around, give that the width and leave padding/border on the inner container.
No can do, I want a dynamic page, not a crappy one size fits none, 400 pixel wide, centred, white background page, how am I supposed to make anything accessible.
Just like table layout, you can use display:table and display:table-cell to do this easily. It was in CSS2.
Except I want my menu to be at the bottom of the page for text only, Braille or screen readers, and generally want it to customisable to the user, a table isn't a one size fits all. divs with box model layout are a lot more flexible.
1: the whole margins thing is odd (and this is one of M$'s main problem areas in IE)
width = width + margin, not content width = width - margin, as you would expect.
This makes layouts trickey.
Also, it's hard to properly layout dynamic content, say for instance: I want all members of class abc to be the same with, but that width is dynamic dependant on the content, or I want my page body to be the menu width away from the edge and I want the menu width to be the size of the largest entry + a 1em margin. No can do.
You should be able to use group sizes and reference other elements sizes in style sheets. Otherwise it's almost impossible to make a nice dynamically sizing website using CSS.
mmap, ramfs and stream encryption locally for really sensitive data blockdevices + stream encryption locally data and ssl tunnelling for network data
assuming that they run the encryption loop hardware so the whole system doesn't grind to a halt.
The question that needs asking is: Why would a user not want to be able to access there data and is that law full to prevent them?
DRM breaks current copyright legeslation if it doesn't use a key escrow to release it into the public domain. (yes things are currently released into the public domain when they fall out of copyright)
However, its method for creating arrays is different from C. Instead of having int[10][5] map into an int[50] space, each level of the array is merely a container for a child array. i.e. int[10][5] is equivalent to:
this is no good for raw data maps, like BMP, or pixelmaps or anything you could mmap in C, for that you have to use a ray block of char's and index into it.
using a raw block of char's is fairly trivial, either that or you game isn't going to be work shit if using a block of char's was your biggest problem.
an array of pointers to an array or, if you want it cstyle.
This inconvenient.... class 2darray{
int xsize;
type data[];
type getElement(x,y){
return data[y * xsize + x]
}
2darray(x,y){
data = new type[x*y];
} }
VB.NET is not VB, just like the west indies isn't India.
People you compare VB to VB.NET are foolish, they should be comparing a stable language like Java that has a huge developer and resource base instead of sucking on Microsoft marketing campaign.
MFC, if you know what you're doing, with the Visual Studio dialog editor and classwizard. Almost everyone hates MFC and I don't understand why - it's great.
That's like calling TCL/TK RAD, MFC is crap because well it just is a poor wrapper for winapi.
Why are you asking about OpenGL when you don't appear to know what the hell it is.
A: If it supports one extension then it support all extensions.
B: Pixel and vertex shaders are just entensions, see a.
C: Is supports OpenGL 1.5, are there any other extents, except OpenGL 2.0 and OpenGL 1.3, GLX is a different matter, because X windows only supports GLX 1.2 (no pbuffers) and ATI and NVidia put the extra functions in there libraries to hack around this limitiation.
D: what about CG for Nvidia or Render Monkey code for ATI. Java supports extrenal libraies too.
E: what about pointers that you need for A* pathfinding and artificial intelligence?
I prefer to use an irregular QTree or bounding boxes to perform two/n-tiered path finding. you don't need pointers to do this.
the benchmarks on this site were for all 800x600? huh?. Jesus, you must be a troll.
Most of you 'betters' are only a matter of opinion, and all can be worked arounbd reletivley easly in VB.
1. VB has the option to enable or disable automatic integer overflows and array index bound checking. In some instances it might seem like a good thing to have these turned on, however, you don't always need to. Lets say for instance, it's all internal, meaning, you know the size of your array, pretty static environment, but it auto checks all this for you. That right there is 'OVERHEAD'! Because in a client for something, you might need to check these things, you might not, but better to check them manually than to needlessly do so automatically for instances when it's not necessary.
A: Your using VB you take some overhead for it's simplicity. B: YOU CAN TURN IT OFF, read your own post man.
2. VB forces a function to be Public (Program Wide) inorder to multi-thread it or even point to it (what limited pointer access VB does have). In C++ I can point to any function, sub routine, public/private/protected/virtual/static/extern, YOU NAME IT! Obviously you can't retrieve the location in RAM to something private from outside of a class without a property, but atleast I can point to a function within a class in C/C++!
a: Two points in one well done, No VB doesn't force a function to be public if you encapsulate it properly, you can have more than one module in a project you know.
b: In vb I can use CreateObject to create anonymous objects and easly wrap up anything I want to be dynaic as an object. It's closer to the sain mans use of function pointers.
I can also use functions in callbacks, which means all I have to do is write a few lines of C, wrap it up in a dll and include it in my VB app to get exactly the same kind of functionality in VB, a dynamic function call becomes an integer for the function pointer, and array of arguments and a call to a dll.
3. Along the lines of #2, since VB only supports 'AddressOf' pointing to a function in the rare chance an API might use it, you can't use 'AddressOf' in your own code to 'CALLBACK' to a Sub/Function of your own program (keeping in mind 'AddressOf' only works on 'Module Public' Subs/Functions). There's indirect ways to 'CALLBACK' to a VB sub/function (SetWindowLong) and filtering wMsg(s) sent to that window. However, that requires API, and C/C++ supports 'CALLBACK's natively!
a: You don't need this in VB, as you say there are simple work arounds, and you can use object wrappers for perform arbetrary callbacks, much sainer if you ask me.
4. VB forces all integers (however many bytes they are 1, 2, 4, possibly even 8 byte integers in.NET) all to be SIGNED. This doesn't matter if you're only using integers for (bitwise AND/OR/XOR/NOT) binary operators, but for alot of other instances, I'm sure you can think of several, UNSIGNED integers would be nice biggrin.gif.
a: can you code man? I can use a integer as a flag and perform bitwise operations on it, unlike some who must rely on the limitations of a language I could write code to perform operations on arbetrart sized integers or floating point numbers in VB or C.
5. As you say, you can use resources nativley why the hell would you HARD CODE a,b,c,d if you could soft code it into a resoure.
6. With C/C++ I can actually use REAL STRINGS!, why are you using byte arrays as strings? You can also use vb 5 that doesn't support unicode if you want to.
7. There's several APIs VB cannot use, because it'll crash the VB IDE, such as Create Thread.
a: you can use them, it will not crash the API at compile time, it's also easy to crate out of process objects in VB.
8. More 'OVERHEAD', native VB functions are slower than any equivalent native C/C++ or 'standard' library functions
a: your wrong, pascall calling convension is actually a few cycles faster than C calling convension.
9: I'd say I can be sure when I make an array in C/C++ that it's 8 bytes, no more, no le
Thre's a product called Gambas it's not very mature yet and has some odd quirks (like the object members not being available in the constructor or destructor), but it's getting there and it's FOSS, so it may not be to hard to make it understand VB6.
Powerbasic used to make an exelent product, maybe a bit too good for the VB crowed.
Failing that there must be enough VB programmers out there to to a re-write? in VB?
I don't know why but OSS doesn't seem to have the balls to get it together and write something truly innovative.
Things like ReiserFS were only possible when OSS moved from 'free' to commercially funded free, firefox also seems to have quite a bit of funding, and has been going for years.
Most of the other components are still in the dark ages of unix, X Windows has only just got clipping, it took them years to implement DRI, the kernel has only been thread safe for the past year or two etc....
Why can't we grab the bull by the horns and write tomorrows software instead of yesterdays?
So do I, but the arseholes at BT put my name in the phone book which I told expressly them not to. for one I don't even have a phone at home I only use the line for the Internet.
that's not quite correct, he[the company] owns any changes he made because he is the origional author, all other changes belong to there authors.
Because the company has chosen to distribute the changes that they own along with GPL'd code they must release the changes under GPL (or distrpbute them as a standalone product) the origional author does not own their changes.
This is why the FSF ask you to hand over copyright to them, and it is why the work I've been doing on DirectX often starts off
Copyright 2002-2004 Jason Copyright 2002-2004 Rapheal Copyright 2005 Oliver
more than none, there are several time keeping packages used in large munti-national companies that require IE, that's a hell of a lot of desktops dependant on IE for an ActiveX contol on a web server.
Every company that migrates there servers from Windows to Linux or starups that start with Linux chip away at the use of windows tied desktop products.
Microsoft Exchange is another clasic example, I'm tempted to wrap up MAPI with Wine so that Linux clients can access MAPI and Exchange.
And regardless, the vast majority of Windows apps DO look the same and use native widget.
Which two versions of Microsoft Office would this be? do Realplayer, Windows Media player and Quck dime look the same?
etc..... Most windows applications, even different office products in the same suite have different menus with different structures and try using ctrl+insert to copy and you'll soon find out how many non-native widgets are used in windows.
If you think the majority of Windows applications use standard toolkits and interfaces then I can only assume your application base consists of hello world examples written in VB6.
Slashdot Mirror
Gawd...the irony of me saying that. :-)
Just think, even Microsoft agrees with me and it seems a lot better.
I have no idea why Microsoft implemented it incorrectly, since w3c usually give formula for calculating everything in the standards.
ok, more verbosely.
char a [3] = {1,2,3};
char b [3] = {1,2,3};
char c [3] = {1,2,3};
char foo*[3] {a , b, c};
is not the same as
char bar[3][3] {1,2,3,1,2,3, 1,2,3};
One is whole bytes the other is pointers and bytes, I can do *(bar+10) = 123; I can't do *(foo+10) = 123;
Java may be good at pixelmaps, but the method you suggested for multi-dimensional arrays isn't.
I was trying to way that 1 of Microsofts problems[with the w3c] was that the box width =width = content width + padding + borders.
For instance I always you em's for content sizes, but often use px's for margins or borders. (usually about 1px-5px)
The Big problem is Width = 100% and I want a margin or x pixels, I screwed if I use width 100% margin 5pixels because my page will end up being 100% wide + 10pixels.
You're asking for elements which could be siblings, ancestors, descendants to all share a common value that, for each, has to be calculated independently.
How do tables work at the moment then?
Wrap a container around, give that the width and leave padding/border on the inner container.
No can do, I want a dynamic page, not a crappy one size fits none, 400 pixel wide, centred, white background page, how am I supposed to make anything accessible.
Just like table layout, you can use display:table and display:table-cell to do this easily. It was in CSS2.
Except I want my menu to be at the bottom of the page for text only, Braille or screen readers, and generally want it to customisable to the user, a table isn't a one size fits all. divs with box model layout are a lot more flexible.
Anyhow CSS3 has columns now.
columns are easy, dynamic columns are a right bastard. CSS needs groups and referncing.
myclass{
width = grouped
}
myclass2{
left = myclass.right
height = id.height
}
etc...
1: the whole margins thing is odd (and this is one of M$'s main problem areas in IE)
width = width + margin, not content width = width - margin, as you would expect.
This makes layouts trickey.
Also, it's hard to properly layout dynamic content, say for instance: I want all members of class abc to be the same with, but that width is dynamic dependant on the content, or I want my page body to be the menu width away from the edge and I want the menu width to be the size of the largest entry + a 1em margin. No can do.
You should be able to use group sizes and reference other elements sizes in style sheets. Otherwise it's almost impossible to make a nice dynamically sizing website using CSS.
So you could probably get away with patches to your own copy of the source
That's often good enough.
I would expect that Borland would consent to you releasing binaries too.
Except you can get the source code for Java, and most of the source code for Delphi (well all Borlands components)
So if Borland or SUN drop Java or Delphi tomorrow you can still support the language.
I suppose C# has mono now, I don't know if that's a bad thing or a good thing.
The editor never reads it.
Plausible Deniability.
mmap, ramfs and stream encryption locally for really sensitive data
blockdevices + stream encryption locally data
and ssl tunnelling for network data
assuming that they run the encryption loop hardware so the whole system doesn't grind to a halt.
The question that needs asking is:
Why would a user not want to be able to access there data and is that law full to prevent them?
DRM breaks current copyright legeslation if it doesn't use a key escrow to release it into the public domain.
(yes things are currently released into the public domain when they fall out of copyright)
However, its method for creating arrays is different from C. Instead of having int[10][5] map into an int[50] space, each level of the array is merely a container for a child array. i.e. int[10][5] is equivalent to:
this is no good for raw data maps, like BMP, or pixelmaps or anything you could mmap in C, for that you have to use a ray block of char's and index into it.
using a raw block of char's is fairly trivial, either that or you game isn't going to be work shit if using a block of char's was your biggest problem.
no pointers and nothing above 1D arrays.
an array of pointers to an array or, if you want it cstyle.
This inconvenient....
class 2darray{
int xsize;
type data[];
type getElement(x,y){
return data[y * xsize + x]
}
2darray(x,y){
data = new type[x*y];
}
}
that's ok, I'm going to troll on the spelling and grandma forum next week.
VB.NET is not VB, just like the west indies isn't India.
People you compare VB to VB.NET are foolish, they should be comparing a stable language like Java that has a huge developer and resource base instead of sucking on Microsoft marketing campaign.
MFC, if you know what you're doing, with the Visual Studio dialog editor and classwizard. Almost everyone hates MFC and I don't understand why - it's great.
That's like calling TCL/TK RAD, MFC is crap because well it just is a poor wrapper for winapi.
Look at the Borland toolset for a good wrapper.
Any new products where VB was considered, should be using C# instead.
Why not Java or Delphi? It seems logical if Microsoft killed VB that it's users should move to a laugnage with better support.
Why are you asking about OpenGL when you don't appear to know what the hell it is.
A: If it supports one extension then it support all extensions.
B: Pixel and vertex shaders are just entensions, see a.
C: Is supports OpenGL 1.5, are there any other extents, except OpenGL 2.0 and OpenGL 1.3, GLX is a different matter, because X windows only supports GLX 1.2 (no pbuffers) and ATI and NVidia put the extra functions in there libraries to hack around this limitiation.
D: what about CG for Nvidia or Render Monkey code for ATI.
Java supports extrenal libraies too.
E: what about pointers that you need for A* pathfinding and artificial intelligence?
I prefer to use an irregular QTree or bounding boxes to perform two/n-tiered path finding. you don't need pointers to do this.
the benchmarks on this site were for all 800x600? huh?. Jesus, you must be a troll.
Most of you 'betters' are only a matter of opinion, and all can be worked arounbd reletivley easly in VB.
.NET) all to be SIGNED. This doesn't matter if you're only using integers for (bitwise AND/OR/XOR/NOT) binary operators, but for alot of other instances, I'm sure you can think of several, UNSIGNED integers would be nice biggrin.gif.
1. VB has the option to enable or disable automatic integer overflows and array index bound checking. In some instances it might seem like a good thing to have these turned on, however, you don't always need to. Lets say for instance, it's all internal, meaning, you know the size of your array, pretty static environment, but it auto checks all this for you. That right there is 'OVERHEAD'! Because in a client for something, you might need to check these things, you might not, but better to check them manually than to needlessly do so automatically for instances when it's not necessary.
A: Your using VB you take some overhead for it's simplicity.
B: YOU CAN TURN IT OFF, read your own post man.
2. VB forces a function to be Public (Program Wide) inorder to multi-thread it or even point to it (what limited pointer access VB does have). In C++ I can point to any function, sub routine, public/private/protected/virtual/static/extern, YOU NAME IT! Obviously you can't retrieve the location in RAM to something private from outside of a class without a property, but atleast I can point to a function within a class in C/C++!
a: Two points in one well done, No VB doesn't force a function to be public if you encapsulate it properly, you can have more than one module in a project you know.
b: In vb I can use CreateObject to create anonymous objects and easly wrap up anything I want to be dynaic as an object. It's closer to the sain mans use of function pointers.
I can also use functions in callbacks, which means all I have to do is write a few lines of C, wrap it up in a dll and include it in my VB app to get exactly the same kind of functionality in VB, a dynamic function call becomes an integer for the function pointer, and array of arguments and a call to a dll.
3. Along the lines of #2, since VB only supports 'AddressOf' pointing to a function in the rare chance an API might use it, you can't use 'AddressOf' in your own code to 'CALLBACK' to a Sub/Function of your own program (keeping in mind 'AddressOf' only works on 'Module Public' Subs/Functions). There's indirect ways to 'CALLBACK' to a VB sub/function (SetWindowLong) and filtering wMsg(s) sent to that window. However, that requires API, and C/C++ supports 'CALLBACK's natively!
a: You don't need this in VB, as you say there are simple work arounds, and you can use object wrappers for perform arbetrary callbacks, much sainer if you ask me.
4. VB forces all integers (however many bytes they are 1, 2, 4, possibly even 8 byte integers in
a: can you code man? I can use a integer as a flag and perform bitwise operations on it, unlike some who must rely on the limitations of a language I could write code to perform operations on arbetrart sized integers or floating point numbers in VB or C.
5. As you say, you can use resources nativley why the hell would you HARD CODE a,b,c,d if you could soft code it into a resoure.
6. With C/C++ I can actually use REAL STRINGS!, why are you using byte arrays as strings?
You can also use vb 5 that doesn't support unicode if you want to.
7. There's several APIs VB cannot use, because it'll crash the VB IDE, such as Create Thread.
a: you can use them, it will not crash the API at compile time, it's also easy to crate out of process objects in VB.
8. More 'OVERHEAD', native VB functions are slower than any equivalent native C/C++ or 'standard' library functions
a: your wrong, pascall calling convension is actually a few cycles faster than C calling convension.
9: I'd say I can be sure when I make an array in C/C++ that it's 8 bytes, no more, no le
Thre's a product called Gambas it's not very mature yet and has some odd quirks (like the object members not being available in the constructor or destructor), but it's getting there and it's FOSS, so it may not be to hard to make it understand VB6.
Powerbasic used to make an exelent product, maybe a bit too good for the VB crowed.
Failing that there must be enough VB programmers out there to to a re-write? in VB?
I don't know why but OSS doesn't seem to have the balls to get it together and write something truly innovative.
Things like ReiserFS were only possible when OSS moved from 'free' to commercially funded free, firefox also seems to have quite a bit of funding, and has been going for years.
Most of the other components are still in the dark ages of unix, X Windows has only just got clipping, it took them years to implement DRI, the kernel has only been thread safe for the past year or two etc....
Why can't we grab the bull by the horns and write tomorrows software instead of yesterdays?
So do I, but the arseholes at BT put my name in the phone book which I told expressly them not to. for one I don't even have a phone at home I only use the line for the Internet.
I hope I can get them fined.
Make eeryone use IE5 and write the whole lot in flash.
Why not do conformance testing against W3C standards like everyone else.
that's not quite correct, he[the company] owns any changes he made because he is the origional author, all other changes belong to there authors.
Because the company has chosen to distribute the changes that they own along with GPL'd code they must release the changes under GPL (or distrpbute them as a standalone product) the origional author does not own their changes.
This is why the FSF ask you to hand over copyright to them, and it is why the work I've been doing on DirectX often starts off
Copyright 2002-2004 Jason
Copyright 2002-2004 Rapheal
Copyright 2005 Oliver
more than none, there are several time keeping packages used in large munti-national companies that require IE, that's a hell of a lot of desktops dependant on IE for an ActiveX contol on a web server.
Every company that migrates there servers from Windows to Linux or starups that start with Linux chip away at the use of windows tied desktop products.
Microsoft Exchange is another clasic example, I'm tempted to wrap up MAPI with Wine so that Linux clients can access MAPI and Exchange.
And regardless, the vast majority of Windows apps DO look the same and use native widget.
Which two versions of Microsoft Office would this be?
do Realplayer, Windows Media player and Quck dime look the same?
etc.....
Most windows applications, even different office products in the same suite have different menus with different structures and try using ctrl+insert to copy and you'll soon find out how many non-native widgets are used in windows.
If you think the majority of Windows applications use standard toolkits and interfaces then I can only assume your application base consists of hello world examples written in VB6.