My opinion is that C++ will slowly decline *for new projects* over the next 10 years or so. It's certainly not going away, and it's by no means dead for new projects, but it will become harder and harder to justify as being the best option. For high performance apps C# and VB will become much more common as they offer good enough performance and are much easier to learn/write/debug than C++. The .NET runtime is putting people of for now but in a couple of years it will be as common as the C runtime library which nobody even thinks about how it gets installed any more. For many other windows application I think things like Python will become much more common. For speed and ease of development these kind of languages can't be beaten and it's starting to look to me like python might be the first langauge of this type to achieve enough critical mass of development to break into the mainstream for real applications. Once you've used a language with automatic memory management it's hard to go back. And there are fewer and fewer reasons to need to any more.

Frankly I find it totally sickening that so many people have gone out marching in support of Saddam Hussain being allowed to just ignore the rest of the world telling him to disarm. What the hell do they think they are doing?

Is this supposed to be a joke? It's just a political attack as far as I can tell.

Well the original sim city fitted on two floppy disks as I remember. And didn't seem to be significantly different either,

David Wulff wrote: As to the CV put everything down as experience that you have in a summary on the top of the CV. That way a future employer doesn't have to read all the boring details. But be careful. I've seen lots of CVs where someone with only a few years experience claims to be an expert at c++, sql, java, tcp/ip, web development, and half a dozen more things and you just think that there is no way they could be an expert in all of those things after a few years only. It might be better to put key skills (which should be exactly those you think the possible employer is looking for) and "others" just so they know you know something about a wide range of things without looking stupid because you've claimed to know everything about everything.

Or in one step :- inx x = 100; System.Console.WriteLine("Number is 0x{0:X4}", x);

Well I'm not quite sure what you mean but this defines a class called Foo.Bar.Class1 so maybe you can just have a namespace of Foo.Bar

You might want to look at this :- http://www.caida.org/tools/utilities/netgeo/[^]

If you like small pcs you might also like this site www.mini-itx.com[^]

Given that the windows API is probably the easiest API to base a Windows.Forms implementation on because naturally microsoft chose to make use of the windows API where is was convenient to do so, there would have to be a form of win32 emulation in mono. So why not make use of the work that has already been done in this area? I don't agree that this is backpedalling - I think it would have been easier for them to implement their own windowing base system if they only cared about 90% compatilibilty. This way it will be easier to achieve 100% compatilibility with Windows.Forms. It doesn't matter what fraction of the win32 api is supported as presumably the mono people will only be making use of the parts that are supported. It doesn't have to support all of win32 well, only the parts they choose to use to implement the .net functionallity.

Post almost *anything* into comp.lang.c++ and see how many people jump on your post. I have to say it's the most unhelpful group on usenet. The people there seem to take much more interest in putting people down than it trying to help them.

No, I'm in the UK and always treat monday as the first day of the week. And so does everyone else as far as I know.

I've added the tivonet adapter to mine and use tivoweb for the remote access. It works really well and was very easy to add I've not managed to get the video extraction tools to work at all though.

You can't. Well, not without keeping some extra information anywhere. Really this is a design problem, not an implementation problem - it's good practice to allocate and free memory at the same level of a program, so that level already knows how it allocated the memory. For example char* str = new char[100]; somefunction(str); delete[] str; char* str = "Hello world"; somefunction(str); // No need to delete anything The point is that "somefunction" didn't allocate the memory so it can't know how to free it and shouldn't try. Basically whenever you allocate some memory you need to think about how it's going to be freed, and ensure that it's done at the same level as it was allocated, which is probably the only place that knows how to do it.

You define an interface when you have two or more classes implementing common behavoir. For example "Customer" and "Supplier" might both implement a IContact interface with methods for setting their address, name etc. You only want to define an interface when there is more than one class that implements that interface, or if there might be in the future.

Until a few weeks ago I had BT Openworld in the UK who have no explicit bandwidth limits for £29.99 per month for 512K/256K ADSL. I've now cancelled that and am moving to A&A who have a Home Lite scheme for about the same price which does have a download limit of 5Mbytes per hour averaged over a month during 9AM-6PM weekdays. Outside of those times there are no specific limits although they do say that you should remember that it is a contended service at 20:1 and if they do have bandwidth problems they will ask people who greatly exceed that to decrease their usage. So why am I moving from an unrestricted service to a restricted one for the same price? Because they appear to be a small company who actually know what they are doing, and also will give me a block of 8 routed IP addresses so I properly run servers and so on using my ADSL line. And have no sillyness like blocked ports, or forcing you to use their servers. When I first had ADSL I downloaded movies and stuff to try it out, but I'm over that now, and find my average usage to be no more than one or two Gbytes per month.

And it would be *much* easier for the malicious application just to run the code itself. Unless you are expecting some other application to have enhanced privilages which you hope to get at this way...? In that case that's a huge bug in that application rather a vulnerability of sendmessage. There is an article somewhere by microsoft saying that code that creates windows should not be privilaged, and privilaged code should not create windows.

If only they'd change the name it might catch on, but it's not got a chance until they do.

They don't make this easy to find. Double click the message to get it displayed in its own window. Then choose View/Options from the menu of that window and in the dialog you will find the "internet headers". At least that's what works on the version of outlook I have here.

Run your own mail server instead of replying on the ISPS and use some of the blacklists that exist to refuse to accept mail from known spammers IP addresses. With some other rules you can filter at least 90% of the spam. Sadly, to do this properly you need a permanent internet connection, and to be running some kind of unix as the windows tools don't seem quite up to the job yet.