Where I contract for at the moment doesn't really *depend* on a closed source app - which is saying a lot considering we develop using the MS.Net toolchain. We could switch to SharpDevelop and move SVN and the devs onto *nix boxes - which wouldn't be overly difficult. We could probably target Mono - theres a few dependancies that could likely be replaced.
I'd imagine Infrastructure would have the worst time, exporting everything out of Exchange / AD. Say a $250k of peoples time for the switch?
Productivity would be down. Taking Resharper away from the junior developers would probably drop their output 30% straight up.
Bigger companies tend to understand risk mitigation - even if only via natural selection. We make software that is used in automation - people rely on this, and downtime is expensive. The code for any deployed binaries is in escrow.
It would be hard for Windows to end up "unmaintained" anyway. Windows as a property is worth probably around a hundred billion dollars. Any company that ends up owning it would have to be publically listed - and would be obligated by its shareholders and the massive investment in aquiring it to keep developing/selling.
But yeah theres a definite "free as in speech" fixation by some people. I'm not sure where it comes from either - I'm personally sick of reading other people's code - it usually drives me insane.;)
If you're not, you're not likely to understand how picking up maintenance of some complex software is a waste of your goddamn time, and its much easier to just do your day job and pay someone familiar with the code to fix it.
Why the hell do I need source for my OS? All the problems have been solved in some 70/80's textbook, and any modern algorithm is described in a research paper + patent doc.
Wait, so you are saying the Linux community needs to do exactly what Microsoft is doing and send in some sales guys to demonstrate the benefits of using their software?
Less than that of any other OS. Things you can do on Windows that you are unable to on other OSs without breaking the (stupid) law include playing BluRay movies, and PlaysForSureSometimes DRM tracks.
Other OS restrict what you can play by not supporting them.
You might find thats either an SP1 thing, or an ATI != Crappy Intel Onboard issue.
Your laptop is faster than mine (admittedly I have 4gig because I use it for development / VMs and RAM is cheap as hell) - Vista was crap until some update before SP1 (may have been an Intel driver update).
Seriously, you're not expecting the iTunes police to come crashing through the window with guns blazing are you?
As a member of the iTunes Police, I take strong exception to this. Firearms safety has always been a core tenet of iTP training. An iTP officer will only open fire if a copyright violation is in progress, or the officer has reasonable belief that lethal force is the only way to prevent a copyright violation.
iTunes Police would never "come crashing through a window with guns blazing". The very thought of it!
All that stuff is great. Like all tools, if you don't know how the damn things works, then you can't use it properly.
So many people downloading the latest "cool" tools and not knowing how it works. People love getting an IOC container and trying to use it without understanding the component lifecycle. Heck, they have the same problem with ASP.Net - I've worked with "senior" developers that don't understand the basic page lifecycle (and expecially that the page instance is _not_ reused between requests).
Drives me insane. So yeah, the GC might work most of the time, but you are still going to have devs getting screwed because they don't understand how finalization works.
If you don't have a rough idea on how to build the tool you are using (or how it is built), then you shouldn't be using it.
If you knew shit about your problem domain, then you would be able to determine when your model matches the problem domain, and can support the functional requirements.
A rollercoaster is a similar size to a hill - and is basically waterfall, with versions.
When you take the tiny tiny iterations of the Agile process, the bumps become so small that the closest analogy is sticking a vibrator up your ass - which describes the process quite exactly, IMHO.
...too small to have a reputation bad enough to drive people away...
I've worked at a place with one full time developer who used to hire one or two contractors. I say "used to" because after my predecessor and my time there, no contracting agency in the city will place staff there.
This was merely from a small company thinking because they all worked weekends and worked their ass off to make their business a "success", that we should too. Tip: Your business, your risks. Same goes for "I choose the risky option"... "What the hell, it all went wrong, this is your fault!!".
Look at it this way, if I cost £50 an hour, and a new fast-as-blazes CPU costs £100, then will two hours of my work make as much improvement as putting a faster chip in?
Depends. Are we deploying this to more than one node?;)
I think that defect rate being a function of QA effort and invese to developer noobitude is a fair assumption. Both of those are costs.
How much?
Hardware manufacturers can usually specify FIT with a handful of parts which can fail. Software has potential defects on every line. You could pick a naieve measurement, say defects per kloc, measure this per dollar spent on QA.
That would give you a nice diminishing returns graph. Pick a point where your defect rate is acceptable and that gives you your cost. If your manager asks in his PHB way how much it will cost for zero defects, then you can give him the one-in-a-million cost (which will be expensive) and ask if he wants to spend even more.
(Someone else can go google for the actual evidence, I'm going off memory and I really CBF searching for things).
$50k these days will get you a basic developer that has probably heard of unit testing, and can code according to spec, but can't be trusted with design.
$100k will get you someone that can handle complex algorithms and understand modelling, and can effectively code-review the juniors, kicking some arse where appropriate.
$200k will get you someone that won't fuck the domain model up either, and can make the seniors feel sufficiently inferior that they'll keep their dicks in their pants and not overengineer the hell out of things. (Keep them doing what they can do well, and not overextending themselves).
Then if your project is big enough you may need someone that can take an overall technical lead role. Of course now we are well into fantasy land - because its extremely unlikely you can convince your managers to have more than one "senior" on 100k, supervising a whole bunch of juniors (add more men to get through those man-months!).
That doesn't sound very conducive to there being complex software avaliable. Can you imagine any modern OS being avaliable with full warranty? Can you imagine any complex software at all being released without serious defects?
Software is too complex to provide at $50 a license / "average contribution an hours work" per license if you want it to come with a warranty.
Hell you'd think BIND would be fairly bug free right now, being effectively "new Dictionary<string, IP>()".
Or was your "Full Warranty" remedy you linked to designed to only apply to Microsoft because of some butthurt about them being a monopoly?
I pull a 6-digit salary (even when you convert our shitty AUD to USD;)) - and its worse.
If anything I am more rushed than I was when I was a junior dev. Managers flat out don't care what you are doing when you are on 50k. You could whack off all day if you like.
As soon as you start earning more than your manager it all goes to shit. Possibly an ego thing - I don't really care about the reasons - but every place I've had managers over my shoulder monitoring my times, trying to cut schedules, etc. Where I'm consulting at the moment is a good case in point - every 3 weeks theres a new "final deadline". I think its to try and motivate us to finish up (theres a good several months of work left) - all they are getting is bandaid fixes.
Are you fucking retarded? There was two, count them, two years where anyone had free access to the Windows Driver Development Kit for Vista - where Vista was in "Beta" (and thus the DDK was roughly unchanged).
DX9 is really quite a good API level. Even with the D3D side of things, where I know I have a long history of people thinking I'm antagonistic against it. Microsoft has done a very, very good job of sensibly evolving it at each stepâ"they're not worried about breaking backwards compatibility - and it's a pretty clean API. I especially like the work I'm doing on the 360, and it's probably the best graphics API as far as a sensibly designed thing that I've worked with.
TBH I'd need an honest review. The original is a bit of a known beast - whereas with a foss (or not) clone, I'd be risking a fair bit of time checking it out. (I normally don't even bother with demos these days - just wait until theres a review from a mate, or a site I trust).
People playing your game is very little to do with the "OSS gaming revolution". You need to have a think about how you are going to attract developers.
Checking on Steam, theres a bunch of Tycoon games for ten bucks each. Most of the "free clone" foss games are clones of "good old games" which retail now for around the ten dollar mark, which swings the value quite lot in favour of just buying the damn thing.
Slashdot Mirror
And just what is the fucking point anyway, I doubt this works any better than slight randomisation...
Where I contract for at the moment doesn't really *depend* on a closed source app - which is saying a lot considering we develop using the MS .Net toolchain. We could switch to SharpDevelop and move SVN and the devs onto *nix boxes - which wouldn't be overly difficult. We could probably target Mono - theres a few dependancies that could likely be replaced.
I'd imagine Infrastructure would have the worst time, exporting everything out of Exchange / AD. Say a $250k of peoples time for the switch?
Productivity would be down. Taking Resharper away from the junior developers would probably drop their output 30% straight up.
Bigger companies tend to understand risk mitigation - even if only via natural selection. We make software that is used in automation - people rely on this, and downtime is expensive. The code for any deployed binaries is in escrow.
It would be hard for Windows to end up "unmaintained" anyway. Windows as a property is worth probably around a hundred billion dollars. Any company that ends up owning it would have to be publically listed - and would be obligated by its shareholders and the massive investment in aquiring it to keep developing/selling.
But yeah theres a definite "free as in speech" fixation by some people. I'm not sure where it comes from either - I'm personally sick of reading other people's code - it usually drives me insane. ;)
Are you a good programmer?
If you're not, you're not likely to understand how picking up maintenance of some complex software is a waste of your goddamn time, and its much easier to just do your day job and pay someone familiar with the code to fix it.
Why the hell do I need source for my OS? All the problems have been solved in some 70/80's textbook, and any modern algorithm is described in a research paper + patent doc.
That was his point. Windows works "better" for more people.
Hence Bill Gates needing double suspenders to prevent his heavy wallet from pantsing him in public.
"Is that the Conquistidor?" "No, its MS Dong"
Wait, so you are saying the Linux community needs to do exactly what Microsoft is doing and send in some sales guys to demonstrate the benefits of using their software?
I thought that was evil?
Less than that of any other OS. Things you can do on Windows that you are unable to on other OSs without breaking the (stupid) law include playing BluRay movies, and PlaysForSureSometimes DRM tracks.
Other OS restrict what you can play by not supporting them.
You might find thats either an SP1 thing, or an ATI != Crappy Intel Onboard issue.
Your laptop is faster than mine (admittedly I have 4gig because I use it for development / VMs and RAM is cheap as hell) - Vista was crap until some update before SP1 (may have been an Intel driver update).
As a member of the iTunes Police, I take strong exception to this. Firearms safety has always been a core tenet of iTP training. An iTP officer will only open fire if a copyright violation is in progress, or the officer has reasonable belief that lethal force is the only way to prevent a copyright violation.
iTunes Police would never "come crashing through a window with guns blazing". The very thought of it!
All that stuff is great. Like all tools, if you don't know how the damn things works, then you can't use it properly.
So many people downloading the latest "cool" tools and not knowing how it works. People love getting an IOC container and trying to use it without understanding the component lifecycle. Heck, they have the same problem with ASP.Net - I've worked with "senior" developers that don't understand the basic page lifecycle (and expecially that the page instance is _not_ reused between requests).
Drives me insane. So yeah, the GC might work most of the time, but you are still going to have devs getting screwed because they don't understand how finalization works.
If you don't have a rough idea on how to build the tool you are using (or how it is built), then you shouldn't be using it.
You finish Design, when it is Done.
If you knew shit about your problem domain, then you would be able to determine when your model matches the problem domain, and can support the functional requirements.
A rollercoaster is a similar size to a hill - and is basically waterfall, with versions.
When you take the tiny tiny iterations of the Agile process, the bumps become so small that the closest analogy is sticking a vibrator up your ass - which describes the process quite exactly, IMHO.
I've worked at a place with one full time developer who used to hire one or two contractors. I say "used to" because after my predecessor and my time there, no contracting agency in the city will place staff there.
This was merely from a small company thinking because they all worked weekends and worked their ass off to make their business a "success", that we should too. Tip: Your business, your risks. Same goes for "I choose the risky option" ... "What the hell, it all went wrong, this is your fault!!".
Depends. Are we deploying this to more than one node? ;)
I think that defect rate being a function of QA effort and invese to developer noobitude is a fair assumption. Both of those are costs.
How much?
Hardware manufacturers can usually specify FIT with a handful of parts which can fail. Software has potential defects on every line. You could pick a naieve measurement, say defects per kloc, measure this per dollar spent on QA.
That would give you a nice diminishing returns graph. Pick a point where your defect rate is acceptable and that gives you your cost. If your manager asks in his PHB way how much it will cost for zero defects, then you can give him the one-in-a-million cost (which will be expensive) and ask if he wants to spend even more.
(Someone else can go google for the actual evidence, I'm going off memory and I really CBF searching for things).
$50k these days will get you a basic developer that has probably heard of unit testing, and can code according to spec, but can't be trusted with design.
$100k will get you someone that can handle complex algorithms and understand modelling, and can effectively code-review the juniors, kicking some arse where appropriate.
$200k will get you someone that won't fuck the domain model up either, and can make the seniors feel sufficiently inferior that they'll keep their dicks in their pants and not overengineer the hell out of things. (Keep them doing what they can do well, and not overextending themselves).
Then if your project is big enough you may need someone that can take an overall technical lead role. Of course now we are well into fantasy land - because its extremely unlikely you can convince your managers to have more than one "senior" on 100k, supervising a whole bunch of juniors (add more men to get through those man-months!).
That doesn't sound very conducive to there being complex software avaliable. Can you imagine any modern OS being avaliable with full warranty? Can you imagine any complex software at all being released without serious defects?
Software is too complex to provide at $50 a license / "average contribution an hours work" per license if you want it to come with a warranty.
Hell you'd think BIND would be fairly bug free right now, being effectively "new Dictionary<string, IP>()".
Or was your "Full Warranty" remedy you linked to designed to only apply to Microsoft because of some butthurt about them being a monopoly?
Yeah, thought so.
I pull a 6-digit salary (even when you convert our shitty AUD to USD ;)) - and its worse.
If anything I am more rushed than I was when I was a junior dev. Managers flat out don't care what you are doing when you are on 50k. You could whack off all day if you like.
As soon as you start earning more than your manager it all goes to shit. Possibly an ego thing - I don't really care about the reasons - but every place I've had managers over my shoulder monitoring my times, trying to cut schedules, etc. Where I'm consulting at the moment is a good case in point - every 3 weeks theres a new "final deadline". I think its to try and motivate us to finish up (theres a good several months of work left) - all they are getting is bandaid fixes.
Newton was in on it too.
Are you fucking retarded? There was two, count them, two years where anyone had free access to the Windows Driver Development Kit for Vista - where Vista was in "Beta" (and thus the DDK was roughly unchanged).
Thats not even including any CTPs...
Carmack's 1996 rant has very little relevance.
Especially when in 2007 he said:
Huh? Change your shell to cmd.exe (or powershell if you want something actually useful).
Thats it, you aren't allowed to shit anymore.
Lots of things use Windows CE, which is fine.
The problem is with the Freescale Semiconductor's* RTC driver. So if you aren't using that specific chip and driver then CE is unaffected.
* No, this doesn't excuse MS from proper QA.
TBH I'd need an honest review. The original is a bit of a known beast - whereas with a foss (or not) clone, I'd be risking a fair bit of time checking it out. (I normally don't even bother with demos these days - just wait until theres a review from a mate, or a site I trust).
People playing your game is very little to do with the "OSS gaming revolution". You need to have a think about how you are going to attract developers.
Checking on Steam, theres a bunch of Tycoon games for ten bucks each. Most of the "free clone" foss games are clones of "good old games" which retail now for around the ten dollar mark, which swings the value quite lot in favour of just buying the damn thing.