Why I still use vc6
-
People often accuse those still using VC6 as holdouts, dinosaurs, or too lazy to embrace the new. Well in my case it's none of those. I'm still using VC 6 because Microsoft can't seem to produce a compiler that produces faster binaries!!! Same codebase, same machine: vc6 - Took 0.17346851 seconds or 173.4685 milliseconds vc80 binaries default optimization settings Took 0.54633341 seconds or 546.3334 milliseconds vc80, turned on favor speed optimization settings Took 0.32663722 seconds or 326.6372 milliseconds vc90 binaries default optimization settings Took 0.34099583 seconds or 340.9958 milliseconds vc90, turned on favor speed optimization settings Took 0.33104513 seconds or 331.0451 milliseconds So, after nearly a 10 year wait between the releases of VC 6 (1998) and VC 9 (2007) my program runs twice as slow. Sigh...
¡El diablo está en mis pantalones! ¡Mire, mire! Real Mentats use only 100% pure, unfooled around with Sapho Juice(tm)! SELECT * FROM User WHERE Clue > 0 0 rows returned Save an Orange - Use the VCF! VCF Blog
A nice thing about vc6 (among many others) is that it generates MFC code that only assumes mfc42.dll ... which has been part of the standard install since Win98-First-Edition (Win95 had mfc40.dll) You can link dynamically for really small executables and be confident that mfc42.dll will be present on 99% of the computers that will use your software (and many with Win95 will have picked up mfc42.dll somewere). I've also found that vc71, vc8, and vc9 generate somewhat slower code. However, I think the "search in files" and "tabbed property page per source file" work better in vc71. Also, "hover" to see definitions is handy. I tend to use vc71 for development and debugging, and generate the final executable for release with vc6.
-
I also use vc6, but I don't think vc6 is famous for its compiler. As everyone knowns, it is famous for its IDE.
AFAIK VC6 is infamous for code generation bugs. I know of one program in which a for(i=0;i<10;i++) would mysteriously bump i to 150 even though i was NEVER accessed inside the loop!
-
:mad: You beat me to the comment :)
Chris Austin wrote:
You beat me to the comment
A 5 to make you feel better.... obviously I upset a few people with my comment though. :)
_________________________ Asu no koto o ieba, tenjo de nezumi ga warau. Talk about things of tomorrow and the mice in the ceiling laugh. (Japanese Proverb)
-
AFAIK VC6 is infamous for code generation bugs. I know of one program in which a for(i=0;i<10;i++) would mysteriously bump i to 150 even though i was NEVER accessed inside the loop!
But since it was obviously defined outside the context of the for loop, it could have been incremented ANYWHERE. The bug is your fault, not the fault of the compiler.
"Why don't you tie a kerosene-soaked rag around your ankles so the ants won't climb up and eat your candy ass..." - Dale Earnhardt, 1997
-----
"...the staggering layers of obscenity in your statement make it a work of art on so many levels." - Jason Jystad, 10/26/2001 -
But since it was obviously defined outside the context of the for loop, it could have been incremented ANYWHERE. The bug is your fault, not the fault of the compiler.
"Why don't you tie a kerosene-soaked rag around your ankles so the ants won't climb up and eat your candy ass..." - Dale Earnhardt, 1997
-----
"...the staggering layers of obscenity in your statement make it a work of art on so many levels." - Jason Jystad, 10/26/2001Not really. It worked fine in VC8 - and I declared it one line before the loop.
-
Chris Austin wrote:
You beat me to the comment
A 5 to make you feel better.... obviously I upset a few people with my comment though. :)
_________________________ Asu no koto o ieba, tenjo de nezumi ga warau. Talk about things of tomorrow and the mice in the ceiling laugh. (Japanese Proverb)
Thank you very much :)
El Corazon wrote:
obviously I upset a few people with my comment though.
Now that I don't get. Ahhh people are odd.
-
People often accuse those still using VC6 as holdouts, dinosaurs, or too lazy to embrace the new. Well in my case it's none of those. I'm still using VC 6 because Microsoft can't seem to produce a compiler that produces faster binaries!!! Same codebase, same machine: vc6 - Took 0.17346851 seconds or 173.4685 milliseconds vc80 binaries default optimization settings Took 0.54633341 seconds or 546.3334 milliseconds vc80, turned on favor speed optimization settings Took 0.32663722 seconds or 326.6372 milliseconds vc90 binaries default optimization settings Took 0.34099583 seconds or 340.9958 milliseconds vc90, turned on favor speed optimization settings Took 0.33104513 seconds or 331.0451 milliseconds So, after nearly a 10 year wait between the releases of VC 6 (1998) and VC 9 (2007) my program runs twice as slow. Sigh...
¡El diablo está en mis pantalones! ¡Mire, mire! Real Mentats use only 100% pure, unfooled around with Sapho Juice(tm)! SELECT * FROM User WHERE Clue > 0 0 rows returned Save an Orange - Use the VCF! VCF Blog
You are all crazy. I can't stand the vc6 ide, but most importantly the compiler doesn't even try to comply with c++ standards. As far as the ide goes, not having tabbed windows is enough to make me never want to use it. Plus it just looks old.
-
You are all crazy. I can't stand the vc6 ide, but most importantly the compiler doesn't even try to comply with c++ standards. As far as the ide goes, not having tabbed windows is enough to make me never want to use it. Plus it just looks old.
Oh yeah, I forgot to mention the fact the the intellisense just stops working for some random reason like 50% of the time I use it.
-
GCC outshines the newer VC's too, especially in C mode.
xacc.ide - now with IronScheme support
IronScheme - 1.0 alpha 2 out nowGCC rocks! In spite of its flaws, it's the only one that lets me get good performance out of code that runs on all major platforms, without using say a Java VM. Since both LLVM (awesome) and many university projects in optimization use GCC, I can also reap those benefits without buying a new compiler. For instance, the PEAK optimizer uses GCC by default: http://cobweb.ecn.purdue.edu/~zpan/PEAK/[^] Of course, I would like to see it cleaned up a bit, and fully ported to Windows. No sense having GCC and MingGW when you can have just one, with a few separate modules handling OS-specific stuff. That's how I did my apps. Later, yall!
-
ori kovacsi wrote:
a VC6.0 developer has x3 times more job oportunities then C# / VB / .NET programmers.
That's a bunch of crap [paraphrasing here]. I have not found this to be true anywhere. Post some links if you find otherwise...
Gary
-
People often accuse those still using VC6 as holdouts, dinosaurs, or too lazy to embrace the new. Well in my case it's none of those. I'm still using VC 6 because Microsoft can't seem to produce a compiler that produces faster binaries!!! Same codebase, same machine: vc6 - Took 0.17346851 seconds or 173.4685 milliseconds vc80 binaries default optimization settings Took 0.54633341 seconds or 546.3334 milliseconds vc80, turned on favor speed optimization settings Took 0.32663722 seconds or 326.6372 milliseconds vc90 binaries default optimization settings Took 0.34099583 seconds or 340.9958 milliseconds vc90, turned on favor speed optimization settings Took 0.33104513 seconds or 331.0451 milliseconds So, after nearly a 10 year wait between the releases of VC 6 (1998) and VC 9 (2007) my program runs twice as slow. Sigh...
¡El diablo está en mis pantalones! ¡Mire, mire! Real Mentats use only 100% pure, unfooled around with Sapho Juice(tm)! SELECT * FROM User WHERE Clue > 0 0 rows returned Save an Orange - Use the VCF! VCF Blog
Does your function have stack-based buffers? If so, it's probably due to the /GS switch ('Buffer Security Check') which is enabled by default on both the new compilers. If you don't think you're likely to be affected, use /GS- to turn it off. Generally I have better results with Favor Code Size Over Speed - smaller code fits into caches better and generates less paging.
DoEvents: Generating unexpected recursion since 1991
-
I also use vc6, but I don't think vc6 is famous for its compiler. As everyone knowns, it is famous for its IDE.
I use both VC6 and VS2008. VC6 had the best help and easiest to follow. The examples were much better than VS2002-2008. The help on VS2008 is the most painful. Startup, VC6 is instant. VS2008, even without the startup page is still slow. (If you add the addons like Resharper, its even slower.) VS2008 does have a better IDE in that I can use tabbed windows. Also, these tab windows can be split into sets of tab windows. Line number are useful, for me, now a necessity. It seems that the more you get from a product (like Visual Studio, more pain you have to endure). Something that I have found useful for C++ development in VS2008 is the find usage popup menu command. It opens a window showing all the calls for a function. (This can use some work, since it works much better for C#). That is my two cents for this discussion.
-
I use both VC6 and VS2008. VC6 had the best help and easiest to follow. The examples were much better than VS2002-2008. The help on VS2008 is the most painful. Startup, VC6 is instant. VS2008, even without the startup page is still slow. (If you add the addons like Resharper, its even slower.) VS2008 does have a better IDE in that I can use tabbed windows. Also, these tab windows can be split into sets of tab windows. Line number are useful, for me, now a necessity. It seems that the more you get from a product (like Visual Studio, more pain you have to endure). Something that I have found useful for C++ development in VS2008 is the find usage popup menu command. It opens a window showing all the calls for a function. (This can use some work, since it works much better for C#). That is my two cents for this discussion.
-
People often accuse those still using VC6 as holdouts, dinosaurs, or too lazy to embrace the new. Well in my case it's none of those. I'm still using VC 6 because Microsoft can't seem to produce a compiler that produces faster binaries!!! Same codebase, same machine: vc6 - Took 0.17346851 seconds or 173.4685 milliseconds vc80 binaries default optimization settings Took 0.54633341 seconds or 546.3334 milliseconds vc80, turned on favor speed optimization settings Took 0.32663722 seconds or 326.6372 milliseconds vc90 binaries default optimization settings Took 0.34099583 seconds or 340.9958 milliseconds vc90, turned on favor speed optimization settings Took 0.33104513 seconds or 331.0451 milliseconds So, after nearly a 10 year wait between the releases of VC 6 (1998) and VC 9 (2007) my program runs twice as slow. Sigh...
¡El diablo está en mis pantalones! ¡Mire, mire! Real Mentats use only 100% pure, unfooled around with Sapho Juice(tm)! SELECT * FROM User WHERE Clue > 0 0 rows returned Save an Orange - Use the VCF! VCF Blog
Remember that the VC++ dev team are walking around in T-shirts saying: "10 is the new 6." :)
Kevin
-
You are all crazy. I can't stand the vc6 ide, but most importantly the compiler doesn't even try to comply with c++ standards. As far as the ide goes, not having tabbed windows is enough to make me never want to use it. Plus it just looks old.
chawkins1 wrote:
not having tabbed windows
There was the free and excellent WndTabs add-in for that though. Even today the existing tabs aren't a patch on WndTabs. But anyway, I don't do C++ these days.
Kevin
-
GCC rocks! In spite of its flaws, it's the only one that lets me get good performance out of code that runs on all major platforms, without using say a Java VM. Since both LLVM (awesome) and many university projects in optimization use GCC, I can also reap those benefits without buying a new compiler. For instance, the PEAK optimizer uses GCC by default: http://cobweb.ecn.purdue.edu/~zpan/PEAK/[^] Of course, I would like to see it cleaned up a bit, and fully ported to Windows. No sense having GCC and MingGW when you can have just one, with a few separate modules handling OS-specific stuff. That's how I did my apps. Later, yall!
-
People often accuse those still using VC6 as holdouts, dinosaurs, or too lazy to embrace the new. Well in my case it's none of those. I'm still using VC 6 because Microsoft can't seem to produce a compiler that produces faster binaries!!! Same codebase, same machine: vc6 - Took 0.17346851 seconds or 173.4685 milliseconds vc80 binaries default optimization settings Took 0.54633341 seconds or 546.3334 milliseconds vc80, turned on favor speed optimization settings Took 0.32663722 seconds or 326.6372 milliseconds vc90 binaries default optimization settings Took 0.34099583 seconds or 340.9958 milliseconds vc90, turned on favor speed optimization settings Took 0.33104513 seconds or 331.0451 milliseconds So, after nearly a 10 year wait between the releases of VC 6 (1998) and VC 9 (2007) my program runs twice as slow. Sigh...
¡El diablo está en mis pantalones! ¡Mire, mire! Real Mentats use only 100% pure, unfooled around with Sapho Juice(tm)! SELECT * FROM User WHERE Clue > 0 0 rows returned Save an Orange - Use the VCF! VCF Blog
Wow, I thought I was the only one doing that! The .NET-based compilers are so sluggish, and include a lot of .NET crap that I don't need in my product. The only disadvantage to still using VC6 is that the choice of target processor is limited to what was available 10 years ago. It won't utilize the latest SSE processor registers.