No software product has ever been finished by your logic.
Windows 7 on its way :|
- Thread starter mito
- Start date
No software product has ever been finished by your logic.
Depends if the paying client is happy enough with it and if they consider the bugs to be show-stoppers or not.
nonsense, you've never written a program that works fine, doesnt crash doesnt produce errors and has all its features fully working ?
I have its been running in a local company since 1999, I have only recieved phone calls asking how to do things ive never recived a call telling me the program has done something wrong or produced an error or crashed
no doubt you will find a reason why my program doesnt count
If you write anything remotely complex it's going to have bugs.
Emacs is a very complex operating environment, even with people like RMS, you cannot kill every bug, but you can still make it usable to some decent or even good extent.
The only programs I've written that work without any known bugs are school assignments and lots of little things. When I tried writing my own version of Linux I found that bugs would creep up every so often, I'm currently at 80% completion in writing a Linux emulator for Linux. The important thing is to make note of bugs and prioritise.
Emacs is a very complex operating environment, even with people like RMS, you cannot kill every bug, but you can still make it usable to some decent or even good extent.
The only programs I've written that work without any known bugs are school assignments and lots of little things. When I tried writing my own version of Linux I found that bugs would creep up every so often, I'm currently at 80% completion in writing a Linux emulator for Linux. The important thing is to make note of bugs and prioritise.
Unknown Soldier
Veteran
Course you can if you spend the time
kiler they must absolutely love you at your local electrical shop
ladies and gentlemen i now introduce one of davros's world famous hypothetical scenarios
are you sitting confortably then i will begin.....
kiler visits his local electrical dealer :
kiler : "i'd like that sony hi-fi please"
assistant : " im sorry miss we dont have any in stock
kiler : "miss ?"
assistant : " im sorry i have a cold"
kiler : " i'll take the one on display then"
assistant : "ill have to check with the manager"
assistant : " a customer wants to buy a display model can we sell it"
manager : "no theres one in the loading bay give him that"
assistant : "but that one is broken it doesnt work properly"
manager : "its ok the customer is kiler he'll be fine with it"
assistant : "there you go sir enjoy your purchase"
The next day kiler returns with the hi-fi
kiler : "i'd like to see the manager"
manager : "yes how can i help you"
kiler : "this hi-fi you sold me doesnt work properly"
manager : "yes but you have to understand a hi-fi is a very complex peice of equipment you cant expect it to work or have all its features"
kiler : " ahh why did'nt you say so, thats fine then i am perfectly happy with it now"
Good job the overriding goal isn't to 'finish' under that reasoning, for any large-scale software project, or software would never ship.
EDIT: I see others have said much the same thing, sorry for thinking I was quoting the last post in the thread!
Davros, why don't you go ahead and elaborate on what your "program" actually does and the total use base.
I'd wager that A: your "program" doesn't even begin to show up on the radar (in terms of sheer scope and size) compared to something as "simple" as Microsoft Word, and B: that the number of people using your "program" don't even account for 1/10th of 1% of Microsoft's installed base in US state of Montana.
Saying that you have no bugs in a program that's 1000 lines of code and used by a population of ~5000 generally educated people is not a very big thing. Saying you have no bugs in an operating system that:
A: Has more than 10 million lines of code
B: Is still utterly dependant on drivers written by teams who are short on funds, pressured for timely delivery and don't mind taking shortcuts to get there
C: Is still utterly at the mercy of whatever program your client chooses to install
D: Is hated by the all the self-proclaimed "computer geeks" ofthe world because you *gasp* charge money for it and
E: Has an installed base bigger than the population of some third world countries
If you can say your operating system deals with all of the above and has zero bugs, then you're lying. Here's a newsflash: no version of Linux has ever held up to that standard. It doesn't work that way; it would be like building a car that never rusts, the engine never wears, the tires never go flat, it always gets 30mpg, the heater always works and the windshield wipers always wipe the rain away, the radio always has reception, the battery never dies, the clutch never wears out, the seals never dry up, the upholstery never gets thin and the rear view mirror never gets loose.
How could anyone make such a promise? There is a big common denominator for program bugs, and it's people. People ultimately write the code, and people ultimately test the code. People don't always write perfection on the first try, and people dont' always "test" in ways that you envisioned or that even make sense.
Hey Davros, here's a question:
Has anyone ever tried "attacking" your program with a buffer overflow? What about putting it on an external drive that has intermittent (but as-yet undetected) connectivity problems? How sure are you that there are no bugs, really?
I'd wager that A: your "program" doesn't even begin to show up on the radar (in terms of sheer scope and size) compared to something as "simple" as Microsoft Word, and B: that the number of people using your "program" don't even account for 1/10th of 1% of Microsoft's installed base in US state of Montana.
Saying that you have no bugs in a program that's 1000 lines of code and used by a population of ~5000 generally educated people is not a very big thing. Saying you have no bugs in an operating system that:
A: Has more than 10 million lines of code
B: Is still utterly dependant on drivers written by teams who are short on funds, pressured for timely delivery and don't mind taking shortcuts to get there
C: Is still utterly at the mercy of whatever program your client chooses to install
D: Is hated by the all the self-proclaimed "computer geeks" ofthe world because you *gasp* charge money for it and
E: Has an installed base bigger than the population of some third world countries
If you can say your operating system deals with all of the above and has zero bugs, then you're lying. Here's a newsflash: no version of Linux has ever held up to that standard. It doesn't work that way; it would be like building a car that never rusts, the engine never wears, the tires never go flat, it always gets 30mpg, the heater always works and the windshield wipers always wipe the rain away, the radio always has reception, the battery never dies, the clutch never wears out, the seals never dry up, the upholstery never gets thin and the rear view mirror never gets loose.
How could anyone make such a promise? There is a big common denominator for program bugs, and it's people. People ultimately write the code, and people ultimately test the code. People don't always write perfection on the first try, and people dont' always "test" in ways that you envisioned or that even make sense.
Hey Davros, here's a question:
Has anyone ever tried "attacking" your program with a buffer overflow? What about putting it on an external drive that has intermittent (but as-yet undetected) connectivity problems? How sure are you that there are no bugs, really?
"What about putting it on an external drive that has intermittent (but as-yet undetected) connectivity problems? How sure are you that there are no bugs, really?"
er you cant really blame the program for not running corectly if the drive its running on is broken
if a program fails because of a fault in a video driver for example then its the fault of the video driver not your program
"Saying you have no bugs in an operating system that:"
I didnt say there were no bugs in the operating system
"How sure are you that there are no bugs, really?"
all i can say is i found none during testing (obviously i did but i fixed them before it was released)
and in the past 8 years no-one has discovered a bug
microsoft cannot say the same they knew there was a huge number of bugs unfixed in xp when it was released
i can understand you accepting that programs have bugs we dont have much choice, but being happy about it and actively defending them for it i cant understand
would you like to buy some old pc hardware from me it doesnt work properly but thats ok with you
and its nice to know that if anyone criticised me for selling hardware that doesnt work properly you will jump to my defense
ps: would you like to buy my old 9800pro 2d runs fine but it ctd's if you try and run a 3d app it ctd's
after all gpu's are very complex and contain millions of transistors its a bit unreasonable to expect it to work properly
er you cant really blame the program for not running corectly if the drive its running on is broken
if a program fails because of a fault in a video driver for example then its the fault of the video driver not your program
"Saying you have no bugs in an operating system that:"
I didnt say there were no bugs in the operating system
"How sure are you that there are no bugs, really?"
all i can say is i found none during testing (obviously i did but i fixed them before it was released)
and in the past 8 years no-one has discovered a bug
microsoft cannot say the same they knew there was a huge number of bugs unfixed in xp when it was released
i can understand you accepting that programs have bugs we dont have much choice, but being happy about it and actively defending them for it i cant understand
would you like to buy some old pc hardware from me it doesnt work properly but thats ok with you
and its nice to know that if anyone criticised me for selling hardware that doesnt work properly you will jump to my defense
ps: would you like to buy my old 9800pro 2d runs fine but it ctd's if you try and run a 3d app it ctd's
after all gpu's are very complex and contain millions of transistors its a bit unreasonable to expect it to work properly
Davros? Are you serious?
People blame Microsoft OSes for faulty hardware, faulty drivers, faulty applications and faulty users all day every day.
And when you blast Microsoft for releasing new OSes when the old ones are "incomplete" by your definition without lobbing the same at FreeBSD, Suse, RedHat, Apple OSX, Solaris, Core and any other operating system out there -- then you are doing nothing more than flaming.
No piece of operating system software ever has met your standards. So get off your high programming horse and come back to reality.
When you can program a modern OS with support for every major hardware component on the planet that has a manufacture date anywhere in the last five years that has all the features that any "normal" person would expect (efficient local, remote and policy management, event logging, plug-n-play, resource allocation and virtaulization, backwards-and-forwards compatibility within reason, application compliance) then you can talk about how much better your programming skills are then everyone elses.
I can write a 10,000 line application that has no bugs so far as I know. But it isn't going to do a whole lot, and the first time you give it to a 25 year-old career hacker, it's going to fall apart.
People blame Microsoft OSes for faulty hardware, faulty drivers, faulty applications and faulty users all day every day.
And when you blast Microsoft for releasing new OSes when the old ones are "incomplete" by your definition without lobbing the same at FreeBSD, Suse, RedHat, Apple OSX, Solaris, Core and any other operating system out there -- then you are doing nothing more than flaming.
No piece of operating system software ever has met your standards. So get off your high programming horse and come back to reality.
When you can program a modern OS with support for every major hardware component on the planet that has a manufacture date anywhere in the last five years that has all the features that any "normal" person would expect (efficient local, remote and policy management, event logging, plug-n-play, resource allocation and virtaulization, backwards-and-forwards compatibility within reason, application compliance) then you can talk about how much better your programming skills are then everyone elses.
I can write a 10,000 line application that has no bugs so far as I know. But it isn't going to do a whole lot, and the first time you give it to a 25 year-old career hacker, it's going to fall apart.
Bouncing Zabaglione Bros.
Legend
I have to admit that I've found myself having to explain on several occasions recently why computers and OSes are really incredibly complex things that allow you to do very complicated tasks fairly easily. It's amazing that software and OSes are relatively easy for the layman to work to do very complex things. It's almost it's own worst enemy, as it actually hides so much stuff going on behind the scenes that people don't understand why they should ever have problems, completely forgetting all the time they spent talking to people on the other side of the world, editing their movies, touching up their digital photographs, booking their airline tickets and hotels, etc all things that were very difficult (if not impossible) and expensive to do at home just a few years back.
Oses, software and hardware for the home user are vastly more complex and capable than they've ever been, and while there are bugs (that IMHO are more related to the way we program these ever more complex systems), what we achieve with our PCs has vastly outpaced that.
Do XP and Vista have their issues? Yes. Are they perfect? No. Do they enable us to do more than ever before anyway? Yes.
Oses, software and hardware for the home user are vastly more complex and capable than they've ever been, and while there are bugs (that IMHO are more related to the way we program these ever more complex systems), what we achieve with our PCs has vastly outpaced that.
Do XP and Vista have their issues? Yes. Are they perfect? No. Do they enable us to do more than ever before anyway? Yes.
110% agreed with BZB on every point you made.
People get "numb" to what really has to happen underneath the OS to make our everyday computer tasks actually work. It's amazing how much effort you have to put in to a program when you start dealing with something as simple as remote administration of a PC...
What if you ask for a machine name and the person types in an IP address? Easy right? Except then you need to make sure they typed a valid IP address (four octets, all numbers, no spaces, no numbers above 254 and below 1).
What if you ask for a machine name and they type in a fully qualified domain name? Nothing wrong with that, but is it correct? No spaces, no special characters)
Is it pingable? Did the DNS resolution cache actually send you to the right machine (problems in our current environment -- DNS names don't "tombstone" fast enough, so pinging ABC1234 will return an IP, but that IP may actually belong to machine BCD2345)
You can do a WMI query to Win32_ComputerSystem to make sure you actually have the right computer, but what if it's the wrong OS? What if you dont' have sufficient rights to read it? What if you have access but that machine's WMI is broken and kicks back a bad return value?
Ok, so everything prior to now worked, now you need to read the event log. What if it's corrupt? What if it has seperate permissions that your account doesn't have rights? What if it's 500mb and you're connected to a machine on a slow link?
There is a GARGANTUAN amount of details you have to think of to create even a basic process like this; imagine the amount of trouble an OS creator goes through doing the same things....
People get "numb" to what really has to happen underneath the OS to make our everyday computer tasks actually work. It's amazing how much effort you have to put in to a program when you start dealing with something as simple as remote administration of a PC...
What if you ask for a machine name and the person types in an IP address? Easy right? Except then you need to make sure they typed a valid IP address (four octets, all numbers, no spaces, no numbers above 254 and below 1).
What if you ask for a machine name and they type in a fully qualified domain name? Nothing wrong with that, but is it correct? No spaces, no special characters)
Is it pingable? Did the DNS resolution cache actually send you to the right machine (problems in our current environment -- DNS names don't "tombstone" fast enough, so pinging ABC1234 will return an IP, but that IP may actually belong to machine BCD2345)
You can do a WMI query to Win32_ComputerSystem to make sure you actually have the right computer, but what if it's the wrong OS? What if you dont' have sufficient rights to read it? What if you have access but that machine's WMI is broken and kicks back a bad return value?
Ok, so everything prior to now worked, now you need to read the event log. What if it's corrupt? What if it has seperate permissions that your account doesn't have rights? What if it's 500mb and you're connected to a machine on a slow link?
There is a GARGANTUAN amount of details you have to think of to create even a basic process like this; imagine the amount of trouble an OS creator goes through doing the same things....
Your actually comparing a program you wrote for a local company in 1999 to Vista?
No of course im not comparing my program to vista, im just destroying the premise that many people beleive that it just isnt possible to create a program with no bugs and no missing features
Im a mediocre programmer at best, the people at these major companies are coding gods in comparison they can write bug free code if they wanted they just wont spend the time and resources on doing so
oh + ps: i do blame all companies that release buggy software not just ms
like i said i can understand people accepting it
but actively defending the practice of selling incomplete goods and your asking me am i serious
you wouldnt stand for it if it was hardware if ati or nvidia accidently left out the part of the gpu responsible for fsaa you'd be outraged (remember the uproar over the pentium fdiv bug)
yet because its software it perfectly fine by you that they can sell you it in any condition they see fit..
People still bought tons of 6800s even though though the vast majority of that serious had screwed up PureVideo acceleration. People still buy Nvidia and ATi cards even though their drivers are certainly not bug free. There are many "bugs" as you say yet you continue to buy the product.
That's your problem, right there. You somehow expect someone to write "perfect" code on a hardware platform that continually evolves, a software platform that continually evolves, and with a userbase that continually evolves.
What worked on yesterday's hardware may not work on today's. What people did to your code yesterday they may do differently today. If you are in some sort of technological vacuum where nothing EVER changes, then you may have a point. But when you're in real life where both hardware, software and user needs evolve at a break-neck pace, you can't just expect something to work.
Does your code work on Windows 64? Vista 64? Does it work on Windows 3.11? Does it work on IA64 processors? Does it work with Chinese localization?
Why not? Haven't you tested it? Again, how are you so sure that your code has no bugs?
and is it ok that purevideo doesnt work ?
id say no
albuquerque thinks its pefectly fine
/Davros waits for albuquerque to attack Skyring
"how dare you complain,
6800's are very complex,
untill you can create your own 6800 you have no right to critise "
No, that's not what I'm saying.
What you are saying is that since you can write "perfect code" with no bugs and you're just a mediocre programmer, that you find fault in Microsoft for writing an "unfinished" OS. I'm pointing out that your code is no more stable than theirs, and for you to make such an assertion is ridiculous.
What I am saying is that you are wrong, and to put yourself on that kind of pedestal while setting some unrealistic expectation of "perfect code" on Microsoft is both illogical and wrong.
and is it ok that purevideo doesnt work ?
id say no
albuquerque thinks its pefectly fine
/Davros waits for albuquerque to attack Skyring
"how dare you complain,
6800's are very complex,
untill you can create your own 6800 you have no right to critise "
That's not at all what Albuquerque has been saying to be honest. You must realize that in something so complex as an operating system you are going to have bugs, it comes down to a matter if those bugs are majors and if your client finds those bugs acceptable in relation to the overall product. No operating system ever created is bug free, no large scale application is bug free, these are human creations that are human tested and therefore errors will occur. Your application might be bug free (but then again, you probably don't have people making a living off of finding those bugs) but your application is on a very small scale. You must realize that what I said particularly is a hint at applications that are in mass use, in a open (meaning vulnerable) environment, and often pushed to its limits. I'm not talking about a small app for a single purpose. I'm talking large ones, OSes, word processors, games, content creation, web browsers.
To judge these applications as being complete based on "bugs" is rather silly. While if it has a bug that prevents the application from functioning I can see your position, from my knowledge (and probably the largest scale test I've ever seen) Windows XP does not have such a bug.
I would also like to state that you saying something is bad does little, especially when it comes to something that costs money. For example, you might complain but you still bought the graphics card and therefore continue to support the issue.
Correct i havnt tested it do you know why ?
they are not in the list of supported sytems
you know and this is just a wild guess mind you so dont hold me to it but it probably doesnt work on risc/os machines either
Similar threads
