>I take it your a software person then.....
Yes, I am, and hardware too.
>bit touchy aren't we?....
Not really, but when people say something unfair and I can point it out, I do. Do you feel threatened by facts ?
>sarcy eh! :)
Have problem with sarcasms ?
>why not write programs to available memory
Oh, so you want to limit software programming to algorithms that do not use dynamic memory allocation ? That's going to be very funny.
By the way, "out of memory" was only one example of an error that should not happen but still might, there are plenty of others, such as hardware failures etc. In fact that particular one is an error that we are already able to circumvent fairly well (that is as long as the hard drive does not fail). Out of memory errors do not occur anymore unless you do have a serious bug in your software.
>or maybe if software was actually written in machine code
Writing in machine code would do very little for this, since it is the dynamic allocation that is the problem, and not the size of the compiled program. A program written in assembly can dynamically allocate just as much (ie: too much) as a program written in C, Java, whatever.
>or assembly like it used to be done and like some software engineers still do!
Yes, we still do, in fact most software engineers still have at least notions of assembly, and a good number are still "fluent" in it, but they are *all* aware that writing directly in assembly (as opposed to writing in an intermediate language like C and letting the compiler transform this in assembly) does not solve much at all beside making a binary that is smaller, and slightly faster (actually it happens that the compilers optimizations end up being better than your own assembly). They also know how many inconveniences it introduces (low readability, low evolutivity and...... *error handling which is harder to manage*!) There are times you must program in assembly (ie most microcode for microcontrollers is still written in assembly, and if you write a compiler you must know assembly to be able to translate other languages to it), but even coding for microcode is done more and more in C nowadays (there are even native java microcontrolers), and clearly writing in assembly is nothing like the solution you seem to think it is.
>the amount of blue screen deaths or page faults is ridiculous, concidering the amount of money software writers get and the cost of the final program..
Oh wait a minute, do you mean "windows" when you say "software" ?? Now that explains it.
I doubt VERY seriously that the landers use anything remotely related to windows. Most probably they are using some highly redundant in-house flavour of UNIX, perhaps based on some existing derivative, together with their own redundant hardware.
>I learnt computer language when it came out of a ticker tape machine.....
I wrote my first program in machine code in the mid 70's, then in the mid 80's wrote basic on a commodore 64...
>64k for heavans sake...
Sure, I started with 512 bytes. What does that say ? Nothing at all.
>before upgradeing to a wow IBM 8087 desktop running at a dizzy 8Mhz..
>today its all modular....get a mistake in one, that doesn't *always* appear, and that module is used over and over in larger programs...and *crunch*...
That's how computing has evolved since the beginning, since before your IBM8087. Once people got transistors to work, they decided to use them over and over in large CPUs. When C was created (much before your 8087), the Clib was written, and was since them reused over and over in larger programs.
Are you saying people should rewrite everything from scratch (possibly down to hardware) every time they write a software ? Wow, that's... "quite a revolutionary idea"....
>whether we like it or not windows is the dominant OS around and therefore it is for this system that most other software is written for....
Uh ? No. It's dominant on the desktop, but it's not dominant. Software has a broader meaning that you seem to think, it can be found in toys, watches, cell phones, mars landers etc, and lets not forget network servers (where unix is still largely dominant). If you have something against Windows, don't make it against the software development community.
>If software engineering was aeronautics we would never get off the
>ground and pushed from a cliff we could glide but never land safely.
>If you could the streamlined efficiency of programming directly into the
>computer in the early 80's combined with todays processiong power,
>we'd be light years ahead of where we are today in software
This completely ignores the combinatorial explosion that happens when software complexifies, it certainly is not by writing directly in assembly (rather than let the compiler translate to assembly for you) that you are going to handle it better, on the contrary. (btw, these quotes are here to answer bill gates quotes that aeronautics should be much further than it is had it evolved at the same rate as the computing industry, so let's not start the ping pong game).
>this is not me talking but silicon valley people....
I'm a "silicon valley people"' too, so I guess it's their opinion against mine, so I suggest that the fact that much more people in silicon valley are writing in C (or even "high level" languages like Java) than in assembly these days vindicates my position.
>so you don't need to be so sarcy with me now, do you?
I'm not. Are you one of those people who take facts and sarcasms as personal attacks ?
>because to be honest most modern programing is shoddy at best
That is true (and is due in great part to the combinatorial explosion I mentionned) but it has nothing to do with the fact that people introduce contigencies for errors which they know MIGHT happen, even if they also know they SHOULD NOT happen. You can't avoid this, because you do not control the way in which those errors occur (they may be outside your software, they may be due to the user, they may be due to cosmic rays, whatever).
This was what I was responding to, not to say that software programming was safe, secure, or anything close to perfect, but that pretending that errors can be avoided merely because they can be detected is ludicrous. It would be a demonstration of your integrity if you didn't try to make it look like I was answering something else, and I would appreciate that a lot.
[snip snip stuff about unreliability of modern day softwares as it has nothing to do with the point I was replying to]
>>" It seems endemic of computer users that they constantly think they know better than both hardware manufacturers and software developpers.
>depends which computer users your talking about.....
I'm talking about people like you, to be precise.