I was thinking way back like 8-10 years ago I always had problems with it.

It's nothing anyone would have to deal with at work, but the first several releases of Battlecruiser 3000AD were basically inoperable, with menu interfaces that took multiple clicks to open anything, random crashes all over, a manual that was full of incorrect information and typos, and basically nothing resembling gameplay; only complicated statistical optimization methods could ensure living through the tutorial.

XCode is a steaming pile of ****.

I loved Xcode 3. 4 regressed in functionality and disappoints me terribly.

What is so bad about the Windows command prompt?

Javascript was never broken. People's understanding of how to use Javascript was broken.

+1 on Xcode. Apple is nothing but a hardware company in disguise and 99% of the software they take upon themselves to write turns out being absolute **** with the exception of iOS (Please don't give them credit for their computer OS, most of what they done is put a nice UI on top of BSD).

90% of all websites?

I was expecting someone to mention Dreamweaver. Can't nominate it because I don't know anything about it, but I thought this one was pretty well reviled in this forum.

The latest FF (10?) is much better than 3 or 4. I wish Opera was as good as Opera Mini, though.

I can make it as wide as I want. I'm sure it's possible on your system as well.

Right click on the title bar go to Layout and increase "Screen Buffer Size's" width-value. Then you should be able to increase the width with the mouse as expected.
Alternatively you can just increase the window size in the same tab(Layout).

Totally agree with myNameIsInga.
I once tried to increase the performance of a piece of software written in C# by porting the most time consuming part of it to C++ and compiled it with all thinkable performance settings for execution speed including compiling specifically for the very CPU on which the program was supposed to be running and even in the "benchmark environment" (this is where you will actually get bigger performance gains than in practice) the performance gain for only this time consuming part was not even 5% and for the overall application it would have been even less.

Given the development time, porting was absolutely not worth it even though speed was important and it didn't even take that long to port it, still getting more than 5% faster hardware would have been way cheaper.

Overall I was extremely impressed with the performance of .NET.

Cliffs:
Ported C# software to C++ for performance reasons -> was absolutely not worth it.
.NET's performance is excellent.

Thanks! didn't know about this. but then again I don't use it much

Just copying C# over to C++ and recompiling is not likely to realize much of the potential performance gain that does exist with C/C++.

C++ gives you access to things you just cant do in C# (or Java)... stuff like intrinsic functions (or inline assembly), cache control and memory alignment, SSE instructions, etc...

By taking advantage of these things there can be huge performance gains for most non-trivial code sections. I'm talking speedups of 2x, 5x... or more...

Just take the SSE instructions for starters... C# code will not be vectorized to use SSE (actually, until VS11 shows up or you use the Intel compiler, neither will C++ be vectorized by the compiler... you have to use intrinsic funcs). By not using SSE, you're guaranteed that 75% of the register file is wasted.

C# has a lot going for it, but maximizing performance is not one of them.


And to bring it back to the topic... SourceSafe is the worst.

Two words: Cross-site scripting.

And there's actually more to Mac OS than you seem to think. So many unix bigots open up a shell, see a FreeBSD environment with some extra GNU bits, and think oh, this is just unix.

But there's a lot more to it.

Javascript is plenty broken. There's an amazing little language hiding in there, but there's also a bunch of garbage, which you can't just blame on developers.

Well, how many programs are optimized in that way? I can see games making heavy use of those types of optimizations, but I don't think most software do. The ability to inline functions or make heavy use of makros C-style, can obviously increase performance. I guess thats one of the weaknesses of the JIT solution.

I have seen some benchmarks that suggest that the JIT can't optimize the way a c/c++ compiler can or does. Part of that is probably a choice done by the designers of the .Net stuff. Having a JIT that optimizes as aggressively as a C++ compiler would risk giving very slow start up times.

Sorry for derail and for sounding like a .Net shill, but I often get the sense that much of the critisism of .Net has more to do with hating Microsoft than it has to do with the quality of .Net.

Inga

@sng_jason
Great avatar, btw!

C++ has its place that was not my point.
I was replying to someone who said .NET performs awfully. Now what does the fact that a straightforward porting of code comes with almost no performance gain tell you?

When a language performs awfully a straight-forward porting to C++ is the exact thing that comes with a huge performance gain. I personally expected that the way .NET works would come with a bigger overhead compared to C++.
Back then I researched benchmarks (I'm aware of the problems with benchmarks) for C# compared to other programming languages and including my own experience with it I was quite impressed with its performance, especially in relation to developer productivity. I expected it to be slower and it's a common misconception that people expect C# to be slow.

Jon Skeet gave a great answer in regards to that.

The example I was mentioning was a situation where going to great lengths like inline assembly would have been absolutely not worth the development time.
I didn't say you should use C# when all you want is maximizing performance I was replying to the common misconception that it's slow. The performance you get out of the box with C# especially in relation to developer productivity is very good.