That's (the snipped stuff) all way beyond my comprehension, DoN. That's why I use good ol' Bill's stuff. I get the distinct impression he intends to provide software for those that don't know much about computers. I lead the pack in that regard.
A coupe posts above I make mention that I run only one of each. I have never been infected, and I'm (heavily)bombarded daily. I also made mention that my ISP screens and deletes anything harmful, and not much gets past them. When it has, on rare occasion, NAV has trapped and disposed of it nicely. I'm firmly sold and convinced that there is a place for Bill's wares, especially for folks like me that don't prefer to spend as much time learning computers the way you understand them. I'd rather be working on a steam locomotive, to be quite honest. I see my using his software the same way an inexperienced person sees running a CNC when they lack proper training on manual machines. It relieves me of the learning curve, which, in both cases, can be a tremendous expenditure of time.
I'm of the opinion that there's more to it than that, though. If another system becomes as popular as Bill's, those that have no life (writers of a virus) will turn their attention towards that system. I can't imagine a system, ANY system, that can't be messed with. How could that be possible? If someone is smart enough to write it, there certainly must be someone that is smart enough to understand it and do harm if they so desire. The lesser popular systems, while perhaps more fool proof, are safe more from the standpoint of "who cares" than from their better design, in my opinion. (For what it's worth, considering my lack of knowledge, anyway! ) That is not to say they may not be better systems.
I'm sure I speak for the masses when I say we appreciate your knowledgeable contributions, DoN.
Did you try "File name with space.txt"? Dos required you to enclose strings with literal spaces in quotes to not get them confused as seperators. Windows still does--for filenames in shortcuts. --Glenn Lyford
It is possible, it just isn't likely because following the rules to make software provably correct, bug free, and secure imposes constraints on programmers which they do not like. Most programmers tend to think of what they do as an art instead of thinking of it as engineering with standardized Codes and practices.
The brightest programmers are almost always hackers in the old good sense of hot-rodder, ie they try to get the machine to do things faster and better than others can. Nearly every time a computer language or programming tool comes along which attempts to enforce the rules of secure and provably correct practices, programmers shun it.
There's also the legacy problem for Windows. Because it has its roots as a toy for hobbyists, ie CP/M->DOS->Windows, and has retained backwards compatibility with older applications which made undocumented system calls, it has become a horrible kludge. A thorough and total rewrite could resolve most of its issues (NT was a Nice Try), but Microsoft has *so much* invested in the existing code base, and users have *so much* invested in existing applications, that such a total rewrite is extremely unlikely at this point.
OTOH, unix, and particularly Linux, started with a clean programming model, and stuck to it. All of the source code is freely available, so there has been intensive outside peer review. Errors and kludges have been promptly found, the programmers responsible ridiculed, and the problems fixed. Knowing that people, probably smarter than you, are going to be reading your code has been a powerful motivator for programmers to use best practices when coding for Linux.
Consider that nearly every server on the internet is running some form of unix. Since servers, particularly those run by major ISPs and vendors, are used by *lots* of people for *lots* of things, they're a much more tempting target for hackers than desktop machines. In other words, hackers could create a lot more chaos by disabling a few of these machines than they could by disabling thousands of desktop machines.
So the *attempts* against them are many many times more likely and more numerous than the attacks against desktop machines. Yet few of those attacks succeed. The primary reason is that such systems are inherently more secure than Windows systems. Vendors and ISPs would be fools to run systems which are not robust and secure. Even Microsoft runs unix servers for this reason.
So while running unix on the desktop doesn't prevent attacks by hackers, the likelihood that they'll succeed is greatly reduced. Since Linux is *free*, there isn't a financial burden to bear by doing so either. There *is* a learning curve, but that's mostly involved with doing things most users don't do. For most users, Linux can be as point and click as any Windows system.
There are also thousands of free, or very inexpensive, application programs for unix. So if you're willing to do a little legwork, you can find a program to do most anything that a Windows program can. There are specialized exceptions. The big one is games.
Because the big game publishers almost invariably copy protect their products, and that depends to a large degree on user ignorance and operating system internals obscurity, it is difficult to do on a unix system where the user has full access and control, and the internals are available to anyone as source code. (There are ways to copy protect programs on a unix system, it's just harder, and a bit more clumsy, to do on a desktop machine.)
So you won't find many of the major games available for unix systems. (But there are games available for unix systems, most of them free, some of them very good.) A few other niche applications fall into this category. For example, there aren't any good individual income tax preparation programs for unix.
The major reason Windows has the lock on the desktop that it does is the licensing agreements Microsoft has signed with the major computer vendors. Since Windows comes *pre-installed* on nearly every PC sold, people use it. They often don't even realize that something else is available. It hasn't won by being *popular* with people. It has won mainly because people aren't even offered an alternative.
Programmers tend to write for it too, because that way they're assured of the largest potential customer base. It has become a self-perpetuating cycle. But it has also become an example of Gresham's Law, where the bad overwhelms the good.
All that said, you may notice that I'm writing this on a Windows machine. The irony of that isn't lost on me. Familiarity with Windows systems and applications has become a business necessity. Using the same sort of system at home that I use at work is just easier than switching back and forth each day. I don't have to remember which machine I'm on, and which keystrokes or mouse clicks do what.
So even though I know it is a better OS, I've relegated my Linux machine to specialized tasks where the available free applications make it a clear winner, ie where I'd have to spend real money to do the same thing on a Windows box.
Thanks, Gary. Very enlightening, and encouraging as well. In a sense, you've added credibility to my uneducated guess. Windows, good or bad, is the choice by a large margin, and for various reasons. The learning curve you spoke of certainly will keep me from ever changing, barring some drastic reason to do otherwise.
By the way, games are not a consideration for me. I don't play them, none of them. I prefer to waste my time in other ways. Just a personal thing, YMMV. I realize how big that part of the industry is.
I think if some software engineers would just use PURE software to check open file handles and the like it would be much cleaner. Also it wouldn't eat up so much memory that seems to be the problem with each and every new version.
Yes -- they will. And experimental virii have been written which will infect unix computers and spread through infected files. (These were harmless ones -- but proved the point.)
*But* -- it is much more difficult to damage a unix system with a virus run by a normal user. To do *serious* damage, requires running with higher privileges. (Windows NT, and Windows 2000 have that idea, though they can install (and often are installed) with only the administrator account (and the guest account -- which *can't* be eliminated in Windows 2000), and set up with *no* password, so it becomes easier for anyone who has physical access to the system. (Given physical access -- there is *always* a way to break into a system, but for some systems, it requires a lot more work. :-)
There have been unix versions certified by the government for handling highly classified data -- in multi-user mode -- but they are very difficult to use and to administer, thanks to all of that security. Normal unix is a compromise giving a lot more security than Windows of any version, but not all that is possible.
There was even one try which was provably A-1 secure (SunOs used to ship with an option which could tighten it to C-2 level only as an example.) That A-1 system was so secure that *nobody* could use it, so it was scrub off the OS, and install the next trial version. :-)
But mostly, unix does not do the stupid things that Windows does (such as Outlook Express being easy to fool into extracting and running an attachment without user intervention.) It is trying too hard to be user friendly.
No -- normal unix is not intrusion proof -- nor is any computer which does anything useful -- but, it eliminates a lot of the *gaping* holes which Windows leaves there -- hidden inside an OS with no peer review of the code possible. But those holes *can* still be found, and
*are* found by crackers (who want to be called "hackers", but who dirty the term.)
Note, also -- linux is not normally as secure as one would like. My own preference in the free unix world is OpenBSD -- which comes with a *lot* of things made a lot tighter as freshly installed from the CD-ROM than most other unicies.
[ ... ]
I have several computers connected with direct access to the outside net -- but *none* of those will be a Windows box. These days, in some networks (such as colleges), in the time needed to download the security patches from Microsoft, the system has been exposed to four or five infection attempts -- so it is already too late to patch it once it is connected to the net.
And somewhere in the snipped data was a comment by Harold that Windows was made to make things easier for the user untutored in the use of computers. It might be worth considering that the Mac was so designed long before Windows started reaching for that target, and I believe that it is still ahead in real ease-of-use. But you won't find
*me* using a Mac -- because that ease of use comes at the cost of flexibility to think up and use other ways to do tasks. It is either the way that the programmer who wrote the program thought of, or no way. (Unless you become a programmer, yourself, at which point you've just blown away the "ease of use" part.
Overall, Gary, that was an excellent summary of the situation.
Yes -- because I already knew some variants of Unix, as well as OS-9. That did not work, in the version of MS-DOS which I was dealing with -- it may have been 2.0 -- or perhaps 3.0, but certainly no later.
O.K. But it did not work in the *early* versions of MS-DOS, based on my own experience.
Another nasty is the way wildcarding works. I was accustomed to unix, and in a test run, I created a bunch of files with an x in the name, and to get rid of them, I tried "DEL *X*.*", and discovered that it removed *everything*. (Again -- back around MS-DOS 3.1 or so.) In unix, that would have required an explicit 'X' *somewhere* in the file name. Unfortunately, MS-DOS wildcarding took the first '*' to translate to "????????." -- so there was no place for the 'X' to match.
Even worse, if you typed "DEL *.*" it would say "ARE YOU SURE?", but because what I typed was not precisely "DEL *.*" (even though it had the same meaning to MS-DOS) it did not bother to ask me "ARE YOU SURE?", so there went everything on the disk (a floppy, with no subdirectories, so it all was there to get zapped. :-(
I'm not sure whether they *ever* fixed this properly. I seem to remember encountering it as late as MS-DOS 6.22.
And -- of course, -- wildcarding in MS-DOS was all in the application programs, and could differ from one program to the other. In unix, the shell does the wildcard expansion, so if you want to see what it will match, you can type "echo *x*.*" and *see what would have been deleted had you used "rm" (unix equivalent of "DEL") instead of "echo". Since it is the shell doing the wildcard expansion -- each program behaves identically.
Don Nichols said it all when he wrote, "Overall, Gary, that was an excellent summary of the situation."
Second that! Dittos and kudos! Yours was absolutely the best explanation I have ever seen on servers and hackers. It took on particular relevance in light of the fact I have been fighting PC problems for several days. For one to make such an observation and then openly admit he too is a "user" is credible writing of grand style.
PolyTech Forum website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.