Showstopping Bug Delays Windows Vista RTM

[Descartes]

I'm also in software, and 499 bugs in any application as large as Vista is incredible. As pinion said, it's not as though these are all crucial bugs. They could be misplacement of buttons, misspellings, driver issues with some esoteric third party and everything in between.

I've worked on hundreds of projects, some with as many as 100 team members (still small by Vista standards), and if we got away with 499 bugs we'd be proud. It simply doesn't happen. There's simply no such thing as a defect free deliverable in software unless you reconsider your idea of a defect.

 

[Descartes]

Originally posted by: mugs

Originally posted by: pinion9
<------ Software Engineer (although not my day job...)

At any rate, 499 bugs for 50 million lines of code isn't that terrible. And you don't even know the nature of the bugs. For all you know one of the bugs could be that Windows Calculator crashes when you try to divide by zero twenty seven times in a row. There may be a memory leak with solitaire and that 2 MB stays resident even after closing the program. You get the idea.

There are software glitches, and there are show stoppers. It is important to know the difference between the two because minor glitches may only affect a minority of the population on occasion.

Please, if you have never written any software (I'm not just talking about "Hello, World!" or whatever you did in your introduction to C class) then don't comment on how terrible it is to have 500 bugs.

Exactly. :thumbsup:

:thumbsup:

 

[BD2003]

Originally posted by: loup garou
499 known bugs on release is pretty good for such an intricate piece of software.

Yep, and these bugs may be VERY specific to very unusual situations and hardware combinations, or require such massive overhauling to fix that it's not feasible. No software is perfect, especially a operating system measured in gigabytes.

 

[Drekce]

All of those here who are freaking out about 499 known bugs have probably never been involved in a large scale software project. It is often better to leave minor bugs alone than to fix them, with the possibility of causing even more, worse bugs. Known bugs are good, because they can be evaluated, prioritized and worked around.

Of course, no bugs is the optimal case, but that will never happen in a large scale software system.

 

[Injury]

Where has anyone said it's supposed to be out before the end of the year? The article plainly states that it has a release of 1.7.07.

"Nope not one bit can I see how computer sales, because of Vista, wont go up any at all this christmas over last"

It's funny that you were being sarcastic... but it's actually true. All of you "push for the holidays" people don't understand that "RTM" means "ready to manufacture"... Meaning they are sending it off to have large quanitites of CDs pressed at that point. RTM does not mean a release to the public. No hardware manufacturer could get OEM copies of XP ready and imaged on to hard drives fast enough to have them all over stores by december. Sorry, not happening.

This also doesn't mean that 499 bugs are going to go unfixed... I can almost promise that 300-400 of them have a windows update ready days before the release.

edit: And I'd also like to add that yes, it is a VERY good strategy to release before fixing the bugs because which people are going to be the first to buy it? Us. Geeks. The ones who can find problems better and figure out why they occur much better. NOT TO MENTION that it's going to have to be out a few months before the average person even considers switching. Why not release it now, so that by the time people start making the switch in mass, the bugs are fixed? There's NEVER going to be a bug free OS. They've been fixing bugs on XP since it's been out, and guess what? There are still bugs! So who gives a damn?

 

[Descartes]

Originally posted by: Drekce
All of those here who are freaking out about 499 known bugs have probably never been involved in a large scale software project. It is often better to leave minor bugs alone than to fix them, with the possibility of causing even more, worse bugs. Known bugs are good, because they can be evaluated, prioritized and worked around.

Of course, no bugs is the optimal case, but that will never happen in a large scale software system.

Excellent point and absolutely correct.

 

[FoBoT]

only people unfamiliar with software development lifecycle and "bug" tracking would be surprised with sending out an OS, which is a HUGE piece of software with "500 bugs"

a "bug" could be a typo, like a . where there should be a , in a help file

so what?

 

[waggy]

Originally posted by: pinion9
<------ Software Engineer (although not my day job...)

At any rate, 499 bugs for 50 million lines of code isn't that terrible. And you don't even know the nature of the bugs. For all you know one of the bugs could be that Windows Calculator crashes when you try to divide by zero twenty seven times in a row. There may be a memory leak with solitaire and that 2 MB stays resident even after closing the program. You get the idea.

There are software glitches, and there are show stoppers. It is important to know the difference between the two because minor glitches may only affect a minority of the population on occasion.

Please, if you have never written any software (I'm not just talking about "Hello, World!" or whatever you did in your introduction to C class) then don't comment on how terrible it is to have 500 bugs.

yeap yeap :thumbsup:

 

[pinion9]

My brother works for a government contractor. He works with highly specialized software. He often gets bored at work and does some additional testing to take his findings to the software engineers. He finds the STUPIDEST bugs ever, that no one should ever happen upon.

For example, in one application you are supposed to click a button to spawn a different screen. He wanted to know what happened if you kept clicking back to that same button without closing the other view. He kept clicking until the application crashed. Wow. A bug. Yes, the developer should have a method to detect if that particular window was open. But they didn't, and yes, if you try to open it 50 times, it crashes. It will get fixed eventually, but in this case, the message was "QUIT CLICKING IT THEN!" He found another bug where if you hold down certain (randomly found) keys (like ctrl-alt-f7-tab) and clicked a certain button you got an error. Why would anyone EVER do that? Who knows. But it is now a documented "bug."

Similarly, I bet you could open up enough IE processes that it would eventually crash. It may take hundreds, but you could do it. That would technically be a bug. But why bother fixing it? If you notice you have 500 IE windows, maybe close one before opening another.

 

[Baked]

Originally posted by: pinion9
<------ Software Engineer (although not my day job...)

At any rate, 499 bugs for 50 million lines of code isn't that terrible. And you don't even know the nature of the bugs. For all you know one of the bugs could be that Windows Calculator crashes when you try to divide by zero twenty seven times in a row. There may be a memory leak with solitaire and that 2 MB stays resident even after closing the program. You get the idea.

There are software glitches, and there are show stoppers. It is important to know the difference between the two because minor glitches may only affect a minority of the population on occasion.

Please, if you have never written any software (I'm not just talking about "Hello, World!" or whatever you did in your introduction to C class) then don't comment on how terrible it is to have 500 bugs.

That shut the OP up. Good job. :thumbsup:

 

[pulse8]

Originally posted by: franguinho

Originally posted by: pulse8

Originally posted by: franguinho

Originally posted by: loup garou
499 known bugs on release is pretty good for such an intricate piece of software.

i beg to differ...
499 bugs is pretty good for BETA software

i mean really... NOBODY is gonna rush out and buy this.... maybe a handful but definitely not the large companies that MS is aiming for... its been in the works for like what 6 years?

and now suddenly it HAS to be out before the end of the year??? WHY????!!!!!!

man this is so retarded. can't they just wait a couple more months until its more stable, optimized, has more 3rd party software developed for it and actually has 0 or close to 0 known bugs?

software doesn't have to be perfect and that's pretty tough to accomplish specially with such an intricate and complex piece of software... but really that's just pitiful in my opinion...

You're welcome to create your own bug free OS. In fact, I think you should give it a shot and report back.

you're also welcome to remove your head from inside your ass. although i think you're prolly better off like this anyway.

Good one, douche.

 

[JulesMaximus]

Originally posted by: loup garou
499 known bugs on release is pretty good for such an intricate piece of software.

I think I just figured out how Microsoft determines the pricing of it's products! Let's charge $.50/bug!!!

 

[Alex]

Originally posted by: Descartes

Originally posted by: Drekce
All of those here who are freaking out about 499 known bugs have probably never been involved in a large scale software project. It is often better to leave minor bugs alone than to fix them, with the possibility of causing even more, worse bugs. Known bugs are good, because they can be evaluated, prioritized and worked around.

Of course, no bugs is the optimal case, but that will never happen in a large scale software system.

Excellent point and absolutely correct.

fair enough... i stand corrected!

i'm a 1st year comp sci student so i have no experience whatsoever with large-scale projects like that so if you guys say that 499 bugs is ok then so be it! i'm just surprised that's all but its all good!
i for one will definitely not be buying it for a long time as i'm pretty happy with XP!

Originally posted by: JulesMaximus

Originally posted by: loup garou
499 known bugs on release is pretty good for such an intricate piece of software.

I think I just figured out how Microsoft determines the pricing of it's products! Let's charge $.50/bug!!!

hahha nice... i guess that's why MS software is so expensive

 

[Miramonti]

I would bet my life (well, atleast my neighbor's life) there are not 500 bugs in Vista.

Only 500, that is.

 

[L1FE]

Along the same lines as what other people have said, most software developers usually categorize their bugs/TARs. We just went live with an important application for a huge energy provider and there was specific client sign off during the release that some known bugs were acceptable and would be fixed at a future date. Some of the signed off problems were actually pretty important for certain areas of functionality, but would only have an impact on say 1% of all customers ever. At some point in time, going live with the product made more business sense than sitting there fixing bugs while not generating revenue.

 

[purbeast0]

dude 500 bugs is almost NOTHING in software.

hell i don't even wanna tell you the total of bugs listed in the bugtrack for the software i develop here at work. it makes 500 look tiny.

 

[Alex]

Originally posted by: purbeast0
dude 500 bugs is almost NOTHING in software.

hell i don't even wanna tell you the total of bugs listed in the bugtrack for the software i develop here at work. it makes 500 look tiny.

yeah i feel kinda dumb now... i had no idea!

 

[Drakkon]

Originally posted by: purbeast0
dude 500 bugs is almost NOTHING in software.

hell i don't even wanna tell you the total of bugs listed in the bugtrack for the software i develop here at work. it makes 500 look tiny.

I'll back you up on this one....for a piece of software to have only 500 bugs at the stage windows vista is actually pretty impressive. I havent scene one application that didnt come back with a fair amount of bugs, and for an OS to come back so far with only 500 is amazing considering it as part of a whole.

 

[purbeast0]

Originally posted by: Drakkon

Originally posted by: purbeast0
dude 500 bugs is almost NOTHING in software.

hell i don't even wanna tell you the total of bugs listed in the bugtrack for the software i develop here at work. it makes 500 look tiny.

I'll back you up on this one....for a piece of software to have only 500 bugs at the stage windows vista is actually pretty impressive. I havent scene one application that didnt come back with a fair amount of bugs, and for an OS to come back so far with only 500 is amazing considering it as part of a whole.

Here I'll take the plunge . Here is a quote when accessing all active bugs in our bugtrack ...

Now showing Issues 1 - 20 of 8548

Keep in mind though, this is over 13 versions with .5's in between, and some of the bugs are for versions that were obsolete years and years ago. also, our code base is over 200 million lines.

so yah, 500, for an OS as intriciate as windows, is not bad at ALL.

 

[Injury]

Originally posted by: jjsole
I would bet my life (well, atleast my neighbor's life) there are not 500 bugs in Vista.

Only 500, that is.

Oh, most definately. I WOULD bet my life on that.

It's just that they haven't been made apparently even after public beta tests, so apparently the conditions for them aren't really frequent under normal usage.

 

[Pabster]

ROFL at all the n00bs and clueless in this thread.

Vista has over 50 million lines of code. It is by far the most complex Windows ever undertaken. If all that remains is 500 bugs, they have done a GREAT job.

It is fairly evident most of you have no experience with software design and development.

 

[sao123]

Just for a standard comparison...

Citation: Dennis R Jenkins Book:
Space Shuttle: the Histroy of the National Space Transportation Systems First 100 Flights.


The statistical average of errors per 1000 lines of code in critical systems (flight control, air traffic control, nuclear launch) is approximately 10-12.

The space shuttle PASS (primary avionics systems software) & BFS (backup flight systems)... which to date is the most thoroughly tested software ever coded... (over 3,000,000 man hours of software testing)
Both have been certified to approximately .11 errors per 1000 lines of code. the PASS & BFS are each about half a million lines of code. At that time, the average line of code cost approximately $50 (written, documented, & tested), while NASA paid over $1000 per line of code.

 

[MaxDepth]

According to DigiTimes, Vista build version 5824 reduced the bug count from from 1450 to around 500 since the second release candidate and was set for RTM (release to manufacturing) on October 25 before the bug was uncovered.

I thought RTM was something nasty like ATM.