Ultimately, they concluded they needed to find a way to start all over with their OS. Microsoft will wind up doing the same, eventually.
I don't think that's a foregone conclusion, although it would be a welcome one. They're going to put out Vista first though, come hell or high water.
As the NYT article states, MS holds backwards compatibility sacred. And yes, they've broken it in small ways in the past (XP SP2 was probably one of the largest breaks in recent years), but nothing big. They're scared of breaking it in a "big" way because they know it could cause market fragmentation -- if I'm stuck on Windows because of some proprietary app that we lost the code for years ago and that's essential to my business, well if Windows 2010 breaks it, then why should I stick with Windows?
Frankly, I think they're a little over-concerned on this front. Yes, there are a number of apps out there like that. But that's solvable now -- they could spin up an entire virtualized copy of XP in a VM. It'd be slow, but it'd at least work. And most businesses would stick with Windows over the alternatives because that's what their IT knows, it has the widest range of apps available, and the widest hardware support.
Heck, how much crap could they be rid of if they simply ditched DOS and the entire 16-bit layer? What about for crap like WMF and other archaic data formats? Would ditching FAT32 as a bootable FS (or an FS for "special" files, like profiles and swap) buy anything?
Of course, I suspect that that's not the real sticking points when it comes to "compatibility". There's craptastic API calls all over the Windows API. There are entire layers of APIs that MS stopped promoting years ago, but are still used. And what about the craptastic IE5/6 renderer? Talk about a support and development nightmare.
Apple had the "fortune" circumstance of being a bit player. Microsoft doesn't. If Apple fragmented its market share, well, there wasn't all that much to lose in the first place. The same cannot be said for MS, and MS's entire business plan has revolved around a unified OS (supplied by it) for decades.
When you switch desktops, you see all the windows minimising and all the ones for the next desktop un-minimising
Then you're using some really crappy virtual desktop software (the PowerToy is known to be... well... toy; I've never tried the ones in the drivers, simply because they are feature poor).
I highly recommend either VirtuaWin or DeskWin. I've used both and prefer the former, although I haven't used the latest version of the latter.
Are they perfect? Nope. I've had apps get "lost" and become unrecoverable -- usually when they go off and become unresponsive to Windows messaging while on another desktop (the biggest offender on my work desktop is Outlook). But that's pretty rare with VirtuaWin (the main reason I prefer it).
especially when "taskbar grouping" is turned on, the interface is completely unsuitable for someone who has a large number of apps running.
So disable it. It's easily done. And, frankly, you're simply wrong. I prefer a virtual desktop method myself, but I have coworkers who regularly have 20+ programs open and have no issues with switching between apps efficiently. It's all a matter of what you're used to. I suspect they would find virtual desktops to be a pain; frankly the one thing that perpetually annoys me about any virtual desktop software is the redraw time on them compared to simply switching between apps on the same desktop. And yes, I've used virtual desktops dating all the way back to tvtwm.
Even if the engineers aren't citizens of the U.S. Got to love that.
Are you honestly that clueless?
The FBI, DoD, etc. often require US citizens for the support as well. In fact, for some projects you have to have security clearance to offer tech support, since you might have to be disclosed some sensitive infrastructure information.
because Nvidia/ATI/whoever won't and the distro makers can't
Nvidia and ATI can't either. The drivers contain technology licensed from other companies (such as S3). And in many cases they'd be damn near useless without that technology.
DON'T BUY HARDWARE THAT ONLY HAS PROPRIETARY DRIVERS
In other words -- don't buy 3D graphics cards and expect them to work. Yes, I'm aware of the OSS ATI driver. I also know it sucks. Poor speed, poor compatibility, and poor stability. None of which have been improving.
Sorry, but your "take home lesson" is utterly and completely devoid of use in the real world. The reality is that proprietary hardware and software must be supported sometimes. Who that support should fall to is the next question, and right now nobody is willing to do so. Making it fall to the users just means that the users are likely to say "fuck this" and go to a solution where it is supported -- namely Windows.
Am I the only one who does not like Google collecting surfing habits or using email to decide what ads to send my way.
Yes. You're the only one who has ever had any misgivings over it. I'm so glad you finally said something, because certainly nobody else has ever thought it before.
Seriously - is your tin foil covering your entire body now? Or are you just that much of an egotist?
Will Google one day sell this information to employers?
Not without either violating their privacy policy or significantly changing it. But don't let that get in the way of your paranoia.
Will credit card companies and banks join a data mining company to share collected information?
Sorry, I was too lazy to link every single letter to a different data mining company, otherwise I could've included ones that operate predominantly outside of the US (although the big 3 all have non-US operations).
Your concerns on this issue are about 50 years out of date. And, somehow, I doubt that you know that much about the system as a whole either (and yes, I do).
Can people imagine if their bank, ISP, and employer joined forces to paint a complete profile of a person?
A rather large amount of that information, particularly the financial data, is already available. See above. If I pull a report on you from the credit bureaus then I can already tell a great deal about you -- where you live, how badly you are in debt (or if you're not), how much you're paid (roughly), possibly what kind of car you drive.
Can that data, when taken as a whole, be used to predict things like how much a person will cost in health insurance, and that data be used to not hire a person?
Not bloody likely. Even in Right to Work states you'd have a hard time pulling that one off. You might try, but if you were ever found out then you'd lose far, far more in legal bills than you'd ever gain in insurance savings. Not to mention that you'd get your ass sued off for invasion of privacy -- no matter what waivers you had employees (or potential employees) sign.
How do you know he doesn't have smart, zoned HVAC? It's not horribly uncommon anymore, and it's even DIY-able.
As for the insulation -- if you're building a new house, you have options. Spray foam insulation or (even better) prefabbed walls with integrated insulation are reasonable and likely to pay for themselves. If you're not building a new home, however, then they simply won't give you a return on your money. Close the cracks, add attic insulation, and insulate any outside walls that aren't already insulated. Beyond that, you're just wasting money.
super-efficient insulation and windows (say, even the LCD dimmable variety)
Even the most effectively insulated window may as well be treated as a hole in the wall -- they're just not that effective. And by "effectively insulated" I'm talking about double or triple paned with aerogel sandwiching. That's high tech. That's also an R value of about 5. Whoop de doo. If you have a huge solarium with HVAC then it might pay to go this way, but for your average window pane? You'll never recoup the costs over a "standard" high efficiency window (vinyl or wood frame, double or triple paned with good seals and a tint). Even that is going to take a long time to pay off over an average window -- really, think of a window as a hole in the house and you've got it.
ust a general attention to using advanced materials, design and techniques in the construction of the building itself
There are a lot of these, but they're not going to be done by your average builder. Take a look at any of the advanced, energy efficient showcase homes and you'll see this kind of stuff. And realize that most of the stuff in the showcase home is already available. Honestly, if you use spray foam insulation or the pre-fab insulated walls then you're not going to find much more in the way of energy savings (of course, the flipside is that you're also screwed if you want to run new cable/pipes through them -- plan ahead!). They're already hideously efficient. We're talking about a 5000 sq ft house in the Northeast US with heating bills of $100 in the middle of winter (IIRC).
And that's why libraries are distributed under LGPL, not GPL ?
Not all libraries are. Readline for instance, is GPL only. The Linux kernel, which could be considered a library of sorts, is GPL. There are many libraries distributed under the GPL. Are you seriously unaware of this?
And the LGPL is hardly a real fix. The wording in it is sufficiently unclear that many companies that write proprietary code will not consider linking to anything under it. The differentiation between interface and object code is particularly muddy when you consider functions implemented in the header files, functions that may be inlined by the compiler (comingling your code and the library code), and so forth.
This is why my company, which is a very large software company (top 5 software companies in the world, and in the Fortune 500) has such strict limitations on what OSS licenses are acceptable. Too many people don't understand the potential legal issues, and too many licenses are written too sloppily. You do not give up rights unless you explicitly disclaim them, and most licenses don't.
A few weeks ago some coworkers were pointing at some code they found online. The author appears to have posted it intending for others to use. But there's absolutely no license info anywhere -- no separate file, nothing in the code or header files. As such it's copyrighted material and cannot be used by anyone, in any project, even for personal, non-distrbution use (although the odds of you being sued on that are pretty much zero). But I had to explain this repeatedly... and I still don't think some of them understood the idea.
but in fact you need to provide source only if someone requests it
That's a bogus argument because you're only safe then if nobody ever requests it, and that's not a viable position to take as a company that sells proprietary code. You'd get your ass laughed out of the room if you were to propose that as a reason to use such code to a manager or lawyer.
How can referencing someone else's library, regardless of their license, encumber my source code?
If that library is under the GPL and you distribute your executable then you are violating the GPL unless you distribute your source as well.
If the code is under some other license then you can still violate the license by not complying to it -- e.g., if it's under the BSD/MIT license and you don't include the original license text or try to represent the code as your own, original work. That doesn't encumber your source, but it's still illegal.
Yes, if you paste OSS code into your software project, you will need to follow their license.
If you link to a library you have to follow the license as well. And, sadly, this is where many (I'd even say most) developers just don't get it. They don't understand the differences between the varying licenses, or the difference between OSS licensing and commercial licensing. Most developers are familiar with commercial licenses where you buy the code once and you can now link to the library, using the documented API, and incur no further obligations. There are some commercial licenses that require you to pay additional money, but hey, that's handled by management.
OSS is different. People see the code out on the 'net and think they can just use it -- that it's free (as in beer) with no restrictions whatsoever. Excepting code that is in the public domain, that's just not true.
Hell, just yesterday I had to correct a fellow developer on the usage of the TinyXML library. No, it's not shareware. No, it's not freeware. In fact, when I looked at the license (png/zlib for the version we have) I realized that we need to go contact the lawyers. Because it's not one of the two licenses approved for corporate use (BSD and MIT are the only ones approved w/o special permission). I don't expect we'll have an issue here, but we were acquired last year, the new rules about licensing went into place immediately, and up until right now nobody even bothered to check the licensing of the OSS libraries we were using. And I'm not the only OSS aware developer here (there is, to my knowledge, precisely one other in our group of about two dozen developers). Who knows what some of the other projects are using (about another 75-100 developers).
The fact of the matter is that there's a disconnect between management and developers when it comes to OSS, and it's pretty easy for an ignorant programmer to include OSS code into a project without management knowledge or consent. Yes, that's a failure of both parties, but the fact remains that it's simply easier with OSS code than it is with proprietary code.
And to be clear -- I prefer to use OSS whenever possible. But I know the issues behind various licenses and will make sure we comply to them. All the article says is to do the same diligence on OSS that you would with a commercial license.
I believe the traditional solution to the egg problem is some sort of energy-absorbing shell- much like what was used in TFA.
Read the rules he posted -- the competition isn't so much to get an egg to fall w/o breaking (although that is required), but to take as much time as possible in doing so. The longest amount of time in the air wins.
Unsure how a parachute is "out of the box" thinking though.
The traditional egg drop contest went the way of the dodo when someone realized that a piece of bread and some creamy peanut butter could beat just about any other design out there. The bread provides a small, lightweight cushion and the peanut butter has rather stunning impact distribution and absorption capabilities.
when it comes to hiring, there are people out there who know app Y but not app X.
Exactly. And, more to the point, does it make sense to force them to learn a new app? Taking someone you pay $30-100k a year and forcing them to relearn the tools of their trade, when those tools cost $1000/year in their time as they relearn the toolset. And that's assuming that the new tool can replace the old one with no loss of functionality.
I'm all for switching to a non-Windows desktop when it makes sense... but the fact of the matter is that will not occur until the industry standard apps are available (either ported or via work-a-like OSS apps). There are some rare instances where the *nix tools are actually superior, but those instances have become more and more uncommon over the past decade. Particularly when you consider that you can actually get a lot of the *nix tools on the Windows platform as well.
is there anybody who has spent some time on X11 with a decent window manager who thinks that the Windows window manager is more useable?
More usable? No. Roughly equally usable? Yup. It takes some 3rd party software to get there, but if you're going to object to that then realize that pretty much any XWM you choose falls into "3rd party software" as well as most of the widgets and other software that you probably use on *nix.
And yes, I am a Unix C++ developer. But except for one position (my first, quite some time ago) all of my development work has been done on Windows workstations.
almost fake virtual desktops
Why are you "almost" faking virtual desktops? Why don't you get a program that does them for real? I have no experience w/ OS X, but for Windows I can recommend VirtuaWin and Deskwin. And they're both OSS too! I find VirtuaWin better -- more options, less incidences of lost apps, and less issues with apps that aren't responding well to desktop switching.
Want a decent shell? Cygwin gives it to you, in any flavor you could want. There's also MKS (we use it here), but I really recommend Cygwin if you can work around its wonkiness with the Windows drive structure (if we could freaking disable that we could ditch MKS... sigh). Want perl/python/etc? Nearly any you want are also available on Windows, both via Cygwin and in natively compiled versions.
Want other window manager nice-ities? Nearly anything you want for X is also available in Windows, for free, via 3rd party programs. All it takes is some Google searches.
By necessity, the center of mass (radially from the surface of the Earth) must be at or near geosynchronous orbit, so it naturally remains centered over its ground anchor
For the simple case, yes. But (IIRC) Robert Forward proposed a modified concept that utilized solar sails to stabalize the orbit and allow for them to be in other orbits. Or it may have just allowed for non-equatorial placement, or both -- I don't recall exactly and I'm certainly not a rocket scientist/orbital mechanics expert.
I've not installed any extension (unless you count Flash as an extension).
Flash is a plugin, and plugins do not get unloaded when the page using them is no longer displayed. Furthermore it's widely believed that the Flash plugin is a huge memory leak in and of itself.
Uninstall the Flash plugin and I bet your memory usage will remain stable.
Note that Adobe Reader and Java are also plugins, and they can (and do) use large amounts of memory.
I only have adblock and bugmenot, so it's not extensions causing the problem.
No, it is extensions causing the problem. Specifically adblock. Adblock appears to be a horribly written piece of shit -- it leaks memory all over the damn place. Use Adblock Plus instead -- I think it still leaks memory a bit, but I can surf for several days without reloading Firefox and be at <200M usage, while I'd hit 400M with Adblock in a day.
And I've actually added a couple extensions since switching to Adblock Plus, so if anything my memory should be up, not down. Adblock Plus still works with FilterSet.G, and it also adds whitelists. There are a few esoteric features it doesn't have compared to Adblock, but I certainly haven't missed them.
Extensions are the really powerful bit of Firefox, and something matched by no other browser in ease of development and capability. Unfortunately they seem to also be prone to memory leaks. Firefox, in and of itself, doesn't leak memory (much), but a lot of extensions and plugins (Flash) do. If you want to test this, go for it -- start Firefox in Safe Mode and watch its memory usage over time. Note that plugins (fucking Flash) are still enabled, so if that's what's leaking memory (yes) then you'll still see memory usage increase over time. Surf to sites that don't require plugins and you shouldn't see much of a memory leak though. Remove the Flash plugin (or maybe use NoScript or Flashblock -- although I've had many issues with the latter, including conflicts with GreaseMonkey scripts) and you'll eliminate what's the second biggest memory leak.
But, really, ditch Adblock and replace it with Adblock Plus. You'll solve most of your problems right there. It's the biggest memory leak I've seen in a long, long time.
q[One more positive side is that open sourcing GSX may trigger few separate public projects based on it (depends on what license GSX sources will be provided under).]q
It's provided under the "here are the binaries; you may not reverse engineer them" license.
Read the topic again -- free as in beer, not free as in speech. Just because I give you the beer for free doesn't mean I have to provide you the recipe.
Re:Do people actually log-in when searching Google
on
Google Toolbar v.4
·
· Score: 4, Informative
does anyone actually use these features and log-in to Google.com
If you use gmail or Google Talk then you're logged in. Some people also use Personal Google which requires you to login to be of any use.
I'm unsure if Google News customization uses cookies or just login. I know it can use login -- if you're logged in then your customizations will be shared to any browser that you login under. I think it used to remember customization by cookies though (and may still if you're not logged in).
I'm not willing to make it any easier for big brother than it already is.
And yet you logged into/. instead of posting as an AC... your tin foil hat is slipping.
I haven't bothered looking at the new toolbar since it's IE only, but one think I've always looked for in "shareable bookmarks" is the ability to partially share -- to either mark some bookmarks as "do not share" or (better) different groupings that can be shared independantly. I know that there's some bookmarks on my home machines that I simply don't want coming anywhere near my work system. Ahem.
To my mind, there's a significant difference between buying a pair of Chinese-made shoelaces and offering a search engine that blocks links about Tibet and Taiwan.
You're splitting hairs.
In either case you're supporting a company that's willing to do business with a totalitarian regime that surpresses free speech. And I rather doubt that their native employees are given wide lattitude either.
Oh, and your comment about "shoelaces" is a rather odd attempt to devalue how much stuff is made in China now. Odds are a rather large percentage of the stuff you buy (excluding foods) was made in China. Electronics, cars, appliances, clothing, tools, you name it -- many are now made in whole or part in China. Or an equally repressive (but not yet as technologically advanced) country.
I'm not particularly happy with Google's decision, but given the circumstances I think it was the best decision they could make. If you really want to be pissed off at someone, be pissed off at the Chinese government, not at Google. Sure, it would've been nice if they could make a stand on principles, but I seriously question what good would've come of such a thing (unlike their stand on principles regarding search query results vs the US Government -- at least there there's some hope of setting precedant).
Then the question comes from the Microsoft employee that basically says these processes are a joke, and that "in the trenches" they are just going through the motions. Why didn't he answer the question?
What proof do you have that the AC was actually a MS employee? Their statement. That's nice... and I'm Steve Jobs. If I post AC (which I won't) then you can neither prove nor disprove that statement, much less a more ambiguous statement like "I work for company X" -- heck, I doubt you could prove or disprove that even if I don't post AC. And no, buzzwords aren't proof -- there's too much leaking of buzzwords onto the net, not to mention to friends.
If they are an MS employee then he provided some very reasonable methods for correcting the issue in the group the guy works in.
Honestly, recent MS software releases are better about security than the older ones were. Do they have a long way to go? Sure. But most of the security issues (like, oh say, the WMF bug) are in legacy code that's merely being brought forward. That's going to happen for awhile -- there are things you can do to alleviate it (such as enabling NX bits on CPUs that support them, which is an option in Server 2003), but without completely throwing out the code and rewriting it all you simply can't prevent it.
Hopefully Microsoft will eventually devote developers exclusively to security, and in nontrivial numbers.
No. No, no, no, no. That's completely ass-backwards and doesn't work. By doing that you presume that certain parts of the code are inherently immune to security threats, so you don't need a security-devoted developer working on them. That's bogus and you know it. He even touched on that idea in the interview. Instead you must train everyone (developers and testers) to think about security from the start, because otherwise things will slip through the gaps.
You can still have a "tiger team", but that's best done as a test group that bangs on products in the QA cycle. And even then you still need your regular testers to be doing basic security testing, because that should be part of the cycle now.
He's a VP at Microsoft, and treats his family like the BOFH!
No, the BOFH would tell his "friends" to keep using the unstable, insecure, and crappy products that were Windows95, Windows98, and WindowsME rather than moving to XP.
Or are you honestly going to try and say that XP isn't both more stable and more secure than those three products? It's not perfect, but it's a helluva lot better.
Ah yes, the secret undocumented Microsoft super-API
It's not undocumented. There were several books (not by MS of course) that documented the undocumented API calls in Windows 3.0, 3.1, Windows95, and Windows98. The better of them even listed what advantages they had over the documentd calls and which Microsoft programs used them. If something like Word or Excel used them then you could damn well count on them being available in later versions as well, even though they were not officially supported.
Were they super? Generally no... but they did have advantages in ease of use (mostly) or speed (less often). There were a few cases where a single API call replaced a hundred lines or more of convoluted logic to acheive the same thing.
Note that I don't have an issue with undocumented APIs -- they're common in software. And there are reasons you don't expose them, particularly if the API in question is actually a clever hack that may break later on because it takes advantage of unexpected behavior, or a "feature" that you should really get rid of at some point. The issue is that Microsoft's own application division appears to have had access to these calls (and their documentation) when nobody else did. At that point in time Microsoft had a de facto monopoly on the OS (which, realistically, was done in a fair marketplace), but not on application software (they were, at best, in distant second). By leveraging the OS monopoly they were able to create an application software monopoly -- and that's illegal.
Go ahead. Google for "Undocumented Windows API". There's plenty of hard evidence for this "meme".
You are not forced to use their products due to that lack of any others because of the actions of Microsoft
Yeah, and back in the early 1900s you didn't have to buy Standard Oil either. Sure, the only non-Standard Oil store was 50 miles away, but you had that choice!
The history of Windows (and MS products on it) is remarkably similar. Sure, you can use DR DOS. Oops, for some reason Windows 3.11 won't work with it! We're sure it's their problem though. Use MS-DOS instead. And you can use something other than Excel or Word too... except that those applications don't have the undocumented-yet-supported APIs that let Word/Excel run faster and more stable. Oh, and by the way -- Dell, Gateway? You can sell any operating systems you want. But you'll pay for Windows no matter what. Oh, sure, you don't have to do that -- but then we'll just charge you twice as much per license. That's fair, right?
Methinks you have a short memory for the "actions of Microsoft".
Question: If, as you say, Microsoft is a monopoly, are you than admitting that Linux and BSD are not viable operating systems? Of course that can't be true can it? We know in fact that Linux and BSD are viable operating systems.
Thanks for the sarcasm! Now go read up on the legal definition of monopoly. Pay particular attention to the concept of "de facto monopoly". In order to be a monopoly you do not need to be the only provider of a service -- merely an overwhelming majority. The other key part is that you must attain that status not through level competition, but by doing illegal things to your competitors (sabotaging products, not properly sharing information even handedly, giving preferential treatment to resellers based on whether or not they sell a competitor's product, etc).
And yes, MS was found guilty of being a de facto monopoly. That has no bearing on whether or not Linux, BSD, and Apple are "viable operating systems" or not.
If you're going to start talking about the facts, then it's useful to actually know them first.
Slashdot Mirror
Ultimately, they concluded they needed to find a way to start all over with their OS. Microsoft will wind up doing the same, eventually.
I don't think that's a foregone conclusion, although it would be a welcome one. They're going to put out Vista first though, come hell or high water.
As the NYT article states, MS holds backwards compatibility sacred. And yes, they've broken it in small ways in the past (XP SP2 was probably one of the largest breaks in recent years), but nothing big. They're scared of breaking it in a "big" way because they know it could cause market fragmentation -- if I'm stuck on Windows because of some proprietary app that we lost the code for years ago and that's essential to my business, well if Windows 2010 breaks it, then why should I stick with Windows?
Frankly, I think they're a little over-concerned on this front. Yes, there are a number of apps out there like that. But that's solvable now -- they could spin up an entire virtualized copy of XP in a VM. It'd be slow, but it'd at least work. And most businesses would stick with Windows over the alternatives because that's what their IT knows, it has the widest range of apps available, and the widest hardware support.
Heck, how much crap could they be rid of if they simply ditched DOS and the entire 16-bit layer? What about for crap like WMF and other archaic data formats? Would ditching FAT32 as a bootable FS (or an FS for "special" files, like profiles and swap) buy anything?
Of course, I suspect that that's not the real sticking points when it comes to "compatibility". There's craptastic API calls all over the Windows API. There are entire layers of APIs that MS stopped promoting years ago, but are still used. And what about the craptastic IE5/6 renderer? Talk about a support and development nightmare.
Apple had the "fortune" circumstance of being a bit player. Microsoft doesn't. If Apple fragmented its market share, well, there wasn't all that much to lose in the first place. The same cannot be said for MS, and MS's entire business plan has revolved around a unified OS (supplied by it) for decades.
When you switch desktops, you see all the windows minimising and all the ones for the next desktop un-minimising
Then you're using some really crappy virtual desktop software (the PowerToy is known to be... well... toy; I've never tried the ones in the drivers, simply because they are feature poor).
I highly recommend either VirtuaWin or DeskWin. I've used both and prefer the former, although I haven't used the latest version of the latter.
Are they perfect? Nope. I've had apps get "lost" and become unrecoverable -- usually when they go off and become unresponsive to Windows messaging while on another desktop (the biggest offender on my work desktop is Outlook). But that's pretty rare with VirtuaWin (the main reason I prefer it).
especially when "taskbar grouping" is turned on, the interface is completely unsuitable for someone who has a large number of apps running.
So disable it. It's easily done. And, frankly, you're simply wrong. I prefer a virtual desktop method myself, but I have coworkers who regularly have 20+ programs open and have no issues with switching between apps efficiently. It's all a matter of what you're used to. I suspect they would find virtual desktops to be a pain; frankly the one thing that perpetually annoys me about any virtual desktop software is the redraw time on them compared to simply switching between apps on the same desktop. And yes, I've used virtual desktops dating all the way back to tvtwm.
Even if the engineers aren't citizens of the U.S. Got to love that.
Are you honestly that clueless?
The FBI, DoD, etc. often require US citizens for the support as well. In fact, for some projects you have to have security clearance to offer tech support, since you might have to be disclosed some sensitive infrastructure information.
The GP poster had it dead on.
because Nvidia/ATI/whoever won't and the distro makers can't
Nvidia and ATI can't either. The drivers contain technology licensed from other companies (such as S3). And in many cases they'd be damn near useless without that technology.
DON'T BUY HARDWARE THAT ONLY HAS PROPRIETARY DRIVERS
In other words -- don't buy 3D graphics cards and expect them to work. Yes, I'm aware of the OSS ATI driver. I also know it sucks. Poor speed, poor compatibility, and poor stability. None of which have been improving.
Sorry, but your "take home lesson" is utterly and completely devoid of use in the real world. The reality is that proprietary hardware and software must be supported sometimes. Who that support should fall to is the next question, and right now nobody is willing to do so. Making it fall to the users just means that the users are likely to say "fuck this" and go to a solution where it is supported -- namely Windows.
Am I the only one who does not like Google collecting surfing habits or using email to decide what ads to send my way.
Yes. You're the only one who has ever had any misgivings over it. I'm so glad you finally said something, because certainly nobody else has ever thought it before.
Seriously - is your tin foil covering your entire body now? Or are you just that much of an egotist?
Will Google one day sell this information to employers?
Not without either violating their privacy policy or significantly changing it. But don't let that get in the way of your paranoia.
Will credit card companies and banks join a data mining company to share collected information?
No that would never happen.
Sorry, I was too lazy to link every single letter to a different data mining company, otherwise I could've included ones that operate predominantly outside of the US (although the big 3 all have non-US operations).
Your concerns on this issue are about 50 years out of date. And, somehow, I doubt that you know that much about the system as a whole either (and yes, I do).
Can people imagine if their bank, ISP, and employer joined forces to paint a complete profile of a person?
A rather large amount of that information, particularly the financial data, is already available. See above. If I pull a report on you from the credit bureaus then I can already tell a great deal about you -- where you live, how badly you are in debt (or if you're not), how much you're paid (roughly), possibly what kind of car you drive.
Can that data, when taken as a whole, be used to predict things like how much a person will cost in health insurance, and that data be used to not hire a person?
Not bloody likely. Even in Right to Work states you'd have a hard time pulling that one off. You might try, but if you were ever found out then you'd lose far, far more in legal bills than you'd ever gain in insurance savings. Not to mention that you'd get your ass sued off for invasion of privacy -- no matter what waivers you had employees (or potential employees) sign.
Say, smart, zoned HVAC, super-efficient insulation
How do you know he doesn't have smart, zoned HVAC? It's not horribly uncommon anymore, and it's even DIY-able.
As for the insulation -- if you're building a new house, you have options. Spray foam insulation or (even better) prefabbed walls with integrated insulation are reasonable and likely to pay for themselves. If you're not building a new home, however, then they simply won't give you a return on your money. Close the cracks, add attic insulation, and insulate any outside walls that aren't already insulated. Beyond that, you're just wasting money.
super-efficient insulation and windows (say, even the LCD dimmable variety)
Even the most effectively insulated window may as well be treated as a hole in the wall -- they're just not that effective. And by "effectively insulated" I'm talking about double or triple paned with aerogel sandwiching. That's high tech. That's also an R value of about 5. Whoop de doo. If you have a huge solarium with HVAC then it might pay to go this way, but for your average window pane? You'll never recoup the costs over a "standard" high efficiency window (vinyl or wood frame, double or triple paned with good seals and a tint). Even that is going to take a long time to pay off over an average window -- really, think of a window as a hole in the house and you've got it.
ust a general attention to using advanced materials, design and techniques in the construction of the building itself
There are a lot of these, but they're not going to be done by your average builder. Take a look at any of the advanced, energy efficient showcase homes and you'll see this kind of stuff. And realize that most of the stuff in the showcase home is already available. Honestly, if you use spray foam insulation or the pre-fab insulated walls then you're not going to find much more in the way of energy savings (of course, the flipside is that you're also screwed if you want to run new cable/pipes through them -- plan ahead!). They're already hideously efficient. We're talking about a 5000 sq ft house in the Northeast US with heating bills of $100 in the middle of winter (IIRC).
And that's why libraries are distributed under LGPL, not GPL ?
Not all libraries are. Readline for instance, is GPL only. The Linux kernel, which could be considered a library of sorts, is GPL. There are many libraries distributed under the GPL. Are you seriously unaware of this?
And the LGPL is hardly a real fix. The wording in it is sufficiently unclear that many companies that write proprietary code will not consider linking to anything under it. The differentiation between interface and object code is particularly muddy when you consider functions implemented in the header files, functions that may be inlined by the compiler (comingling your code and the library code), and so forth.
This is why my company, which is a very large software company (top 5 software companies in the world, and in the Fortune 500) has such strict limitations on what OSS licenses are acceptable. Too many people don't understand the potential legal issues, and too many licenses are written too sloppily. You do not give up rights unless you explicitly disclaim them, and most licenses don't.
A few weeks ago some coworkers were pointing at some code they found online. The author appears to have posted it intending for others to use. But there's absolutely no license info anywhere -- no separate file, nothing in the code or header files. As such it's copyrighted material and cannot be used by anyone, in any project, even for personal, non-distrbution use (although the odds of you being sued on that are pretty much zero). But I had to explain this repeatedly... and I still don't think some of them understood the idea.
but in fact you need to provide source only if someone requests it
That's a bogus argument because you're only safe then if nobody ever requests it, and that's not a viable position to take as a company that sells proprietary code. You'd get your ass laughed out of the room if you were to propose that as a reason to use such code to a manager or lawyer.
How can referencing someone else's library, regardless of their license, encumber my source code?
If that library is under the GPL and you distribute your executable then you are violating the GPL unless you distribute your source as well.
If the code is under some other license then you can still violate the license by not complying to it -- e.g., if it's under the BSD/MIT license and you don't include the original license text or try to represent the code as your own, original work. That doesn't encumber your source, but it's still illegal.
Yes, if you paste OSS code into your software project, you will need to follow their license.
If you link to a library you have to follow the license as well. And, sadly, this is where many (I'd even say most) developers just don't get it. They don't understand the differences between the varying licenses, or the difference between OSS licensing and commercial licensing. Most developers are familiar with commercial licenses where you buy the code once and you can now link to the library, using the documented API, and incur no further obligations. There are some commercial licenses that require you to pay additional money, but hey, that's handled by management.
OSS is different. People see the code out on the 'net and think they can just use it -- that it's free (as in beer) with no restrictions whatsoever. Excepting code that is in the public domain, that's just not true.
Hell, just yesterday I had to correct a fellow developer on the usage of the TinyXML library. No, it's not shareware. No, it's not freeware. In fact, when I looked at the license (png/zlib for the version we have) I realized that we need to go contact the lawyers. Because it's not one of the two licenses approved for corporate use (BSD and MIT are the only ones approved w/o special permission). I don't expect we'll have an issue here, but we were acquired last year, the new rules about licensing went into place immediately, and up until right now nobody even bothered to check the licensing of the OSS libraries we were using. And I'm not the only OSS aware developer here (there is, to my knowledge, precisely one other in our group of about two dozen developers). Who knows what some of the other projects are using (about another 75-100 developers).
The fact of the matter is that there's a disconnect between management and developers when it comes to OSS, and it's pretty easy for an ignorant programmer to include OSS code into a project without management knowledge or consent. Yes, that's a failure of both parties, but the fact remains that it's simply easier with OSS code than it is with proprietary code.
And to be clear -- I prefer to use OSS whenever possible. But I know the issues behind various licenses and will make sure we comply to them. All the article says is to do the same diligence on OSS that you would with a commercial license.
I believe the traditional solution to the egg problem is some sort of energy-absorbing shell- much like what was used in TFA.
Read the rules he posted -- the competition isn't so much to get an egg to fall w/o breaking (although that is required), but to take as much time as possible in doing so. The longest amount of time in the air wins.
Unsure how a parachute is "out of the box" thinking though.
The traditional egg drop contest went the way of the dodo when someone realized that a piece of bread and some creamy peanut butter could beat just about any other design out there. The bread provides a small, lightweight cushion and the peanut butter has rather stunning impact distribution and absorption capabilities.
historicaly, the term university, has been reserved for large schools that are divided into multiple colleges
Actually, it's usually used to distinguish the difference between a school that has a post-graduate program from one that doesn't.
This isn't always true anymore, of course, since some "Colleges" offer both undergraduate and graduate programs.
when it comes to hiring, there are people out there who know app Y but not app X.
Exactly. And, more to the point, does it make sense to force them to learn a new app? Taking someone you pay $30-100k a year and forcing them to relearn the tools of their trade, when those tools cost $1000/year in their time as they relearn the toolset. And that's assuming that the new tool can replace the old one with no loss of functionality.
I'm all for switching to a non-Windows desktop when it makes sense... but the fact of the matter is that will not occur until the industry standard apps are available (either ported or via work-a-like OSS apps). There are some rare instances where the *nix tools are actually superior, but those instances have become more and more uncommon over the past decade. Particularly when you consider that you can actually get a lot of the *nix tools on the Windows platform as well.
is there anybody who has spent some time on X11 with a decent window manager who thinks that the Windows window manager is more useable?
More usable? No. Roughly equally usable? Yup. It takes some 3rd party software to get there, but if you're going to object to that then realize that pretty much any XWM you choose falls into "3rd party software" as well as most of the widgets and other software that you probably use on *nix.
And yes, I am a Unix C++ developer. But except for one position (my first, quite some time ago) all of my development work has been done on Windows workstations.
almost fake virtual desktops
Why are you "almost" faking virtual desktops? Why don't you get a program that does them for real? I have no experience w/ OS X, but for Windows I can recommend VirtuaWin and Deskwin. And they're both OSS too! I find VirtuaWin better -- more options, less incidences of lost apps, and less issues with apps that aren't responding well to desktop switching.
Want a decent shell? Cygwin gives it to you, in any flavor you could want. There's also MKS (we use it here), but I really recommend Cygwin if you can work around its wonkiness with the Windows drive structure (if we could freaking disable that we could ditch MKS... sigh). Want perl/python/etc? Nearly any you want are also available on Windows, both via Cygwin and in natively compiled versions.
Want other window manager nice-ities? Nearly anything you want for X is also available in Windows, for free, via 3rd party programs. All it takes is some Google searches.
By necessity, the center of mass (radially from the surface of the Earth) must be at or near geosynchronous orbit, so it naturally remains centered over its ground anchor
For the simple case, yes. But (IIRC) Robert Forward proposed a modified concept that utilized solar sails to stabalize the orbit and allow for them to be in other orbits. Or it may have just allowed for non-equatorial placement, or both -- I don't recall exactly and I'm certainly not a rocket scientist/orbital mechanics expert.
I've not installed any extension (unless you count Flash as an extension).
Flash is a plugin, and plugins do not get unloaded when the page using them is no longer displayed. Furthermore it's widely believed that the Flash plugin is a huge memory leak in and of itself.
Uninstall the Flash plugin and I bet your memory usage will remain stable.
Note that Adobe Reader and Java are also plugins, and they can (and do) use large amounts of memory.
I only have adblock and bugmenot, so it's not extensions causing the problem.
No, it is extensions causing the problem. Specifically adblock. Adblock appears to be a horribly written piece of shit -- it leaks memory all over the damn place. Use Adblock Plus instead -- I think it still leaks memory a bit, but I can surf for several days without reloading Firefox and be at <200M usage, while I'd hit 400M with Adblock in a day.
And I've actually added a couple extensions since switching to Adblock Plus, so if anything my memory should be up, not down. Adblock Plus still works with FilterSet.G, and it also adds whitelists. There are a few esoteric features it doesn't have compared to Adblock, but I certainly haven't missed them.
Extensions are the really powerful bit of Firefox, and something matched by no other browser in ease of development and capability. Unfortunately they seem to also be prone to memory leaks. Firefox, in and of itself, doesn't leak memory (much), but a lot of extensions and plugins (Flash) do. If you want to test this, go for it -- start Firefox in Safe Mode and watch its memory usage over time. Note that plugins (fucking Flash) are still enabled, so if that's what's leaking memory (yes) then you'll still see memory usage increase over time. Surf to sites that don't require plugins and you shouldn't see much of a memory leak though. Remove the Flash plugin (or maybe use NoScript or Flashblock -- although I've had many issues with the latter, including conflicts with GreaseMonkey scripts) and you'll eliminate what's the second biggest memory leak.
But, really, ditch Adblock and replace it with Adblock Plus. You'll solve most of your problems right there. It's the biggest memory leak I've seen in a long, long time.
q[One more positive side is that open sourcing GSX may trigger few separate public projects based on it (depends on what license GSX sources will be provided under).]q
It's provided under the "here are the binaries; you may not reverse engineer them" license.
Read the topic again -- free as in beer, not free as in speech. Just because I give you the beer for free doesn't mean I have to provide you the recipe.
does anyone actually use these features and log-in to Google.com
/. instead of posting as an AC... your tin foil hat is slipping.
If you use gmail or Google Talk then you're logged in. Some people also use Personal Google which requires you to login to be of any use.
I'm unsure if Google News customization uses cookies or just login. I know it can use login -- if you're logged in then your customizations will be shared to any browser that you login under. I think it used to remember customization by cookies though (and may still if you're not logged in).
I'm not willing to make it any easier for big brother than it already is.
And yet you logged into
I haven't bothered looking at the new toolbar since it's IE only, but one think I've always looked for in "shareable bookmarks" is the ability to partially share -- to either mark some bookmarks as "do not share" or (better) different groupings that can be shared independantly. I know that there's some bookmarks on my home machines that I simply don't want coming anywhere near my work system. Ahem.
To my mind, there's a significant difference between buying a pair of Chinese-made shoelaces and offering a search engine that blocks links about Tibet and Taiwan.
You're splitting hairs.
In either case you're supporting a company that's willing to do business with a totalitarian regime that surpresses free speech. And I rather doubt that their native employees are given wide lattitude either.
Oh, and your comment about "shoelaces" is a rather odd attempt to devalue how much stuff is made in China now. Odds are a rather large percentage of the stuff you buy (excluding foods) was made in China. Electronics, cars, appliances, clothing, tools, you name it -- many are now made in whole or part in China. Or an equally repressive (but not yet as technologically advanced) country.
I'm not particularly happy with Google's decision, but given the circumstances I think it was the best decision they could make. If you really want to be pissed off at someone, be pissed off at the Chinese government, not at Google. Sure, it would've been nice if they could make a stand on principles, but I seriously question what good would've come of such a thing (unlike their stand on principles regarding search query results vs the US Government -- at least there there's some hope of setting precedant).
Then the question comes from the Microsoft employee that basically says these processes are a joke, and that "in the trenches" they are just going through the motions. Why didn't he answer the question?
What proof do you have that the AC was actually a MS employee? Their statement. That's nice... and I'm Steve Jobs. If I post AC (which I won't) then you can neither prove nor disprove that statement, much less a more ambiguous statement like "I work for company X" -- heck, I doubt you could prove or disprove that even if I don't post AC. And no, buzzwords aren't proof -- there's too much leaking of buzzwords onto the net, not to mention to friends.
If they are an MS employee then he provided some very reasonable methods for correcting the issue in the group the guy works in.
Honestly, recent MS software releases are better about security than the older ones were. Do they have a long way to go? Sure. But most of the security issues (like, oh say, the WMF bug) are in legacy code that's merely being brought forward. That's going to happen for awhile -- there are things you can do to alleviate it (such as enabling NX bits on CPUs that support them, which is an option in Server 2003), but without completely throwing out the code and rewriting it all you simply can't prevent it.
Hopefully Microsoft will eventually devote developers exclusively to security, and in nontrivial numbers.
No. No, no, no, no. That's completely ass-backwards and doesn't work. By doing that you presume that certain parts of the code are inherently immune to security threats, so you don't need a security-devoted developer working on them. That's bogus and you know it. He even touched on that idea in the interview. Instead you must train everyone (developers and testers) to think about security from the start, because otherwise things will slip through the gaps.
You can still have a "tiger team", but that's best done as a test group that bangs on products in the QA cycle. And even then you still need your regular testers to be doing basic security testing, because that should be part of the cycle now.
He's a VP at Microsoft, and treats his family like the BOFH!
No, the BOFH would tell his "friends" to keep using the unstable, insecure, and crappy products that were Windows95, Windows98, and WindowsME rather than moving to XP.
Or are you honestly going to try and say that XP isn't both more stable and more secure than those three products? It's not perfect, but it's a helluva lot better.
Ah yes, the secret undocumented Microsoft super-API
It's not undocumented. There were several books (not by MS of course) that documented the undocumented API calls in Windows 3.0, 3.1, Windows95, and Windows98. The better of them even listed what advantages they had over the documentd calls and which Microsoft programs used them. If something like Word or Excel used them then you could damn well count on them being available in later versions as well, even though they were not officially supported.
Were they super? Generally no... but they did have advantages in ease of use (mostly) or speed (less often). There were a few cases where a single API call replaced a hundred lines or more of convoluted logic to acheive the same thing.
Note that I don't have an issue with undocumented APIs -- they're common in software. And there are reasons you don't expose them, particularly if the API in question is actually a clever hack that may break later on because it takes advantage of unexpected behavior, or a "feature" that you should really get rid of at some point. The issue is that Microsoft's own application division appears to have had access to these calls (and their documentation) when nobody else did. At that point in time Microsoft had a de facto monopoly on the OS (which, realistically, was done in a fair marketplace), but not on application software (they were, at best, in distant second). By leveraging the OS monopoly they were able to create an application software monopoly -- and that's illegal.
Go ahead. Google for "Undocumented Windows API". There's plenty of hard evidence for this "meme".
You are not forced to use their products due to that lack of any others because of the actions of Microsoft
Yeah, and back in the early 1900s you didn't have to buy Standard Oil either. Sure, the only non-Standard Oil store was 50 miles away, but you had that choice!
The history of Windows (and MS products on it) is remarkably similar. Sure, you can use DR DOS. Oops, for some reason Windows 3.11 won't work with it! We're sure it's their problem though. Use MS-DOS instead. And you can use something other than Excel or Word too... except that those applications don't have the undocumented-yet-supported APIs that let Word/Excel run faster and more stable. Oh, and by the way -- Dell, Gateway? You can sell any operating systems you want. But you'll pay for Windows no matter what. Oh, sure, you don't have to do that -- but then we'll just charge you twice as much per license. That's fair, right?
Methinks you have a short memory for the "actions of Microsoft".
based upon actual Stanford materials, but with our EZ-Pass exams!
Hey, you can get the real exams too! You just need to get access to a fraternity/sorority's word file...
Question: If, as you say, Microsoft is a monopoly, are you than admitting that Linux and BSD are not viable operating systems? Of course that can't be true can it? We know in fact that Linux and BSD are viable operating systems.
Thanks for the sarcasm! Now go read up on the legal definition of monopoly. Pay particular attention to the concept of "de facto monopoly". In order to be a monopoly you do not need to be the only provider of a service -- merely an overwhelming majority. The other key part is that you must attain that status not through level competition, but by doing illegal things to your competitors (sabotaging products, not properly sharing information even handedly, giving preferential treatment to resellers based on whether or not they sell a competitor's product, etc).
And yes, MS was found guilty of being a de facto monopoly. That has no bearing on whether or not Linux, BSD, and Apple are "viable operating systems" or not.
If you're going to start talking about the facts, then it's useful to actually know them first.