Didn't bother to look up the legal term, but this kind of thing HAS been done -- where one plaintiff went and sued a bunch of people, independently, for essentially the same crime. They then fought back, collectively.
Yes, Microsoft would sue me, an individual. And then I'd go on Slashdot with my evidence, get people to donate to my Paypal, and pool resources with other individuals they'd sued.
you can choose whatever method you like when your whole IP stack is SSL encrypted, like mine.
That doesn't make it secure. SSH also has an authentication protocol, and it's per-user. If yours is per-IP-stack, you already lost -- both because we already have that (in the form of VPNs and ipsec) and because it's not secure (anyone who can connect to the server can authenticate).
So once again: What method of RPC would you use instead of SSH? Telnet? Last I checked, it won't do RSA authentication -- it's really hardly more than netcat.
Also, the difference between 1-2s poweron and 30s poweron is a total loss of the "mental moment".
Hmm. 20s, more like. And really?
Consider also that 10s of that is likely due to it being a bad time (a nightly cron job was running).
Which is why the Flash is worth the small improvement: while the disk is just spinning up, the Flash can be done reloading the image to RAM.
How much?
Because most Flash I know isn't that fast. Fast seeks, slow throughput.
The big advantage would be that you could save a smaller image -- really just swap out what you have to, and otherwise purge your cache. Then, things would be loaded on demand. The reason hibernate doesn't work like this already is -- well, actually, it used to, but the problem was, your disk would thrash for another minute or two as various programs would wake up and have to be pulled off the disk. Flash, with 0 seek time, would make that fast.
The OS should compute an image that is persistent between boots that can be saved and save it, even before the command to shutdown. Then, at shutdown, only the part of the image that needs updating from the new state should be updated and stored. Then, at poweron, just reload the whole image, and update the parts that need updating for current time, DHCP, etc.
Well, actually, most of what wastes disk during boot is pure waste, not redundancy. There doesn't need to be a distinction between loading an app and loading an image, for instance. Consider the difference between OpenOffice and a more modular approach -- a shell script, say. OpenOffice loads a huge chunk of the app off the disk, while a shell script loads a collection of small programs, each on demand.
As for redundancy, well, it actually doesn't take that much time to run DHCP, update the clock, check what monitor is plugged in, etc. Really, most of what's done during boot would be faster if it was decently parellized -- even moreso if it was on a medium with 0 seek time.
We clearly have 4 "restart" scenarios: "reboot" from scratch (regenerate executable image, regenerate user state); de/activate from execution (save/load executable image, persistent user state); "reset" to flush corrupt state (regenerate executable image, save/load user state); and "restart" (save/load executable image - or just retain it, persistent user state).
You forgot "reinstall".
Reboot, reinstall, and reset are all essentially hacks to deal with bad software. Even the concept of a running program vs a binary on disk is an abstraction we don't need.
What we do need is "de/activate", or suspend, but not in the form of a completely blind dump to RAM -- this should be more like "save/load user state". With a flash-based disk and well-written software, there'd be no performance difference between intelligently saving/loading state and blindly dumping RAM.
And by "saving/loading state", I don't actually mean that there should be some magic signal that every app should obey. I mean apps should be written to use common libraries that will do orthogonal persistence for them, that understand the difference between volatile and nonvolatile storage, and between critical data and various types of cache. There would be other performance gains of such an architecture -- when the system needs RAM, it can flush everyone's cache, of all kinds of things, not just the FS cache -- same goes for disk cache.
Imagine that classic Unix idea -- imagine everything is a file, including your running programs. But how do we think of files? You mount the filesystem, which takes almost no time, and then you can get inside and poke around. Or unmount it, and it's static on disk, and you can't get to it. But when you make changes to the filesystem, they stay. And the system makes it impossible for you to make changes that damage the FS itself, though you can certainly screw up files, and delete
Sorry about the OO comment, apparently I was confusing you with someone else there -- there was someone who doesn't like the "everything is an object" way.
I've seen some pretty strangely formatted code in my time - everyone has their own style. I didn't like Python's style, and since I didn't have an option, I didn't pursue it.
What this tells me is that you grew up with C, or something similar. It doesn't explain why you've chosen to regard code indentation as a sacred element of "style", and not, say, semicolons at the end of every statement, curly braces for blocks, parentheses for both overriding precedence and passing arguments, and so on.
I get that you don't like it, but I don't think syntax alone should be a major reason to ignore a language, although it may make the learning curve a bit harder.
Perl appeared to do everything that Python could do, plus some good friends programmed in Perl, so I could work on stuff with them, plus the Perl community is huge compared to the Python community. There were plenty of reasons. There still are.
Most of those, I consider valid. I use Perl. But really, the indentation is hardly worth mentioning.
Imagine if it was the other way around -- if Perl forced indentation. Wouldn't you still use it, for all the other reasons you stated?
Web 'interfaces' suck.
All interfaces suck. The important thing here is, why do web interfaces suck?
I hate to be trendy, but XMLHttpRequest (AJAX) would work very well here.
I've spent a hell of a lot of time making sure the libraries will work well over slow network connections.
I remember when the Web worked reasonably well over dialup. And that was before widespread use of Javascript, let alone AJAX.
Gtk2::Ex::DBI has built-in record paging ( which I hacked into it with a plain-text editor when I was on holiday in Cambodia ).
If I understand "record paging" here, it looks like it means you can perform some query and get only partial results, and then, transparently, pull in more results as needed as the user "pages" (scrolls?) through the records. Not having used your software, let alone dug into the code, I'm only guessing here...
But if so, that would be easy with a Web interface. User requests next page, so an HTTP request gets sent to the server -- it comes back with a small chunk of XML (or JSON) representing more records to be added to the aperture.
Now, if you meant "paging to disk", as in a local cache, you've got me -- cookies certainly aren't enough storage here.
you send back big chunks of updates ( and / or inserts ) when you're done.
Again, no problem. Send em back as another chunk of custom XML (or JSAN), describing the operations to be performed. Because of the "A" in AJAX, you could even let the user keep working while this is happening.
I realise that you need to install some libraries on the client PCs, whereas with a web-based solution you wouldn't have to
It's not just "Install some libraries." It's "Install, maintain, and update some 20-30 different libraries."
On Linux, that's not a problem -- we have good package managers. But on Windows, unless you basically implement your own package manager -- and likely waste disk space, unless you cooperate with all the other apps out there that might use your library. On OS X, you have to deal with the fact that probably half the libraries are already installed, but out of date. You could use Fink, but it's likely going to be out of date, or break in some subtle way.
A web interface here means that all the client-side issues go away.
it's more than work it, just for the incredible control over the GUI and huge feature-set that it brings.
I think one of the points of confusion here seems to be that most people don't realize that while something is built into a motherboard it doesn't have some magical interface that makes the bits fly differently than if it was in a slot.
Actually, some of them do -- I seem to remember PHY (for network cards) is only available if the card is part of the motherboard. Also, if you buy a gigabit PCI card, you'll likely saturate your PCI bus.
What you wanted was: Just because it's built in doesn't necessarily mean it has a magical interface. It might, or it might not.
I believe both GNOME and KDE have something rudimentary set up here, so that if an app supports this, and you log out (telling it to "save your session", which may be the default), it will save your session. At which point, there is no point to putting it on flash -- disk is more than fast enough. It's also a bit more flexible this way -- log out and switch users becomes much faster.
Here's the problem:
1) The on is no longer instant. Sure, it's quick to shut down, but now the app must be "rebooted", which means reading the entire thing off disk. The reason you think Flash is fast for this is because it doesn't have seek problems -- but normal PC hibernate doesn't, either.
2) Every app has to support it, or there's no point. Unless you want to be checking a particular app for "hibernate support", what usually ends up happening here is either the app gets killed completely on hibernate, or it gets started again properly, but without its state stored.
Now, suspend to RAM is fast enough. Suspend to disk is actually almost as fast for me, were it not for the other hassles of boot. I'm going to test this right now:
Very rough estimate:
Boot -> OS is some 25 seconds. This includes at least a few pauses in both BIOS and OS designed to give me the opportunity to press some keystroke to enter some setup -- I believe that's BIOS, RAID, and the Grub bootloader. Lilo and Macs have the right idea -- why not just let me hold a particular key combo while the box is booting? Even Windows was better about this, sometimes -- when resuming from hibernate, I can press F8 if I want to cancel the resume and do something else.
Maybe it's not intuitive that you can hold Apple+C (or is it Apple+D?) to boot from CD, but it's also something that has to be done maybe one in a thousand boots, and I'd rather have the other 999 boots be as fast as possible.
Then, it's still some 5-10 seconds while Linux churns and loads drivers, etc, before it actually gets to resuming.
The actual act of resuming is about 10 seconds flat, after which my desktop is pretty close to ready to use. Maybe 20 seconds, tops, counting disk thrashing, but that may also be my updatedb cron job.
But even 20 seconds, I could live with. What I hate is that there's another 30 seconds or so that's completely wasted, so it ends up being a minute to resume from Hibernate, when it could've been maybe 20 seconds. (It takes even longer to actually boot.)
Magnetic hard disks are certainly not the limitation here.
What about those first however many seconds that the kernel takes to boot, then the initramfs loads even more...
I mean, yes, Upstart is cool, but right now, fairly irrelevant -- hibernate is stupidly fast when you have any kind of RAID. But the BIOS and the initial Linux boot (during which a fresh kernel loads all the hardware support it will need to... load the hibernate-image-kernel) do still take entirely too long.
Then again, it might help to simply have hibernate images supported directly by bootloaders.
Seriously, every language has similar constraints. I'm confused that nobody ever complains about having to have a semicolon at the end of every line, even when it's obvious to the compiler/interpreter that it is the end of a line -- yet when you're told to indent your code the way you should anyway, it's oppressive? What?
As for your app, assuming you are actually using object-oriented development under the hood -- the fact that you were scared by an "everything is an object" language scares me, frankly -- it looks like it could be VERY interesting. Any plans for a web interface?
I know this is often not a relevant issue, as most closed software, for awhile there, was sufficiently reliable that you wouldn't need to fix bugs. Most open software has some pretty big problems -- Firefox's memory leaks, for instance.
But closed software still has some of the biggest, stupidest bugs that would NOT stand in an open system.
I do HD-DVD development. I use Microsoft's HDiSim. Here are two showstopper bugs:
1) No vista support. Even Microsoft doesn't support Vista. I had to wait a week for this laptop to be downgraded to XP because of this.
2) No USB keyboard support. I'm on a laptop with an internal PS/2 keyboard, so this is inconvenient, but not a huge deal. But many Dells don't even come with PS/2 ports anymore -- it's USB or nothing. The best idea I've heard as to how to deal with this is to run it in a virtual machine, where I can set my real (USB) keyboard to be whatever I want it to be (USB, PS/2, Serial, whatever) -- but then, why not use Linux as the host OS?
Neither of these would stand in an open product. Not because "the community" is "better", but because I would go in and fix it myself.
Someone tell me, has Microsoft fixed the audio/net bug yet? The bug where playing music in Vista drops your network performance to 10%? (Drops it by 90%, to 10%.) Because if Linux ever had that problem, I'd fix it, too.
Actually, the reason you don't see them requested often is that they are easier to make work under Wine. Often, they "just work", and/or have native Linux apps that do the same job, as well or better.
Games are there because games are hard -- they don't just have to run, they have to perform well. And not just inside their own little process, the whole stack has to perform well, including the part where they emulate DirectX on OpenGL.
That said, games actually are working quite well. I boot Windows for LAN parties, once a month. The rest of my games are either on Linux or consoles, and I could lose the consoles.
Sure, if Windows is hardly a monopoly. There are reasonable arguments that both have sufficient competition that they shouldn't be considered monopolies, and, I think, reasonable arguments that neither really has sufficient competition that there's a real choice of moving off one, onto another.
Does iTunes work with any other player? Does the iTunes Music Store work with any other player?
Even if it's not a complete monopoly, my point is that Apple does far more lock-in, DRM, avoiding standards, and so on, than Microsoft ever does. Or rather, they are more extreme about it -- not being a complete monopoly, they have far less opportunity to do these things.
And thus, my point is that were Apple the dominant OS vendor, or the dominant in any field, they'd be even more proprietary and anticompetitive than Microsoft is -- in fact, they already are. The only real advantage would be, the Apple stuff would work. It would actually be a good product, without there having to be sufficient competition and years of R&D and beta testing on the public before they have a kinda sorta OK product.
However, the question is whether Apple has done something like that. I don't think we have a clear answer.
I realize that iff it's merely collateral damage, there's no problem, though it would've been nice of them not to break unlocked phones if they could help it. But that might not be the case.
Laptops absolutely have their appropriate uses- but desktops work just fine for a huge percentage of people.
That works just as well the other way around. Desktops have their appropriate uses, but laptops work just fine for a huge percentage of people.
Their components are cheaper, more easily replaced, and usually superior in performance.
Well, the company paid for my laptop, so "cheaper" doesn't matter to me -- or to other people, in a similar situation.
My desktop recently died. It was likely either the CPU or the motherboard, but I figured I should probably replace both, since new CPUs are cheap -- like $60 for a new dual-core 64-bit. This was made even less of an issue by my friend, who was looking to upgrade his CPU, so he gave me his old one -- so it made even more sense just to rip the motherboard and replace.
But the new motherboard was DDR2 only, and I had DDR400 in the old machine.
Thus, I ended up replacing CPU, motherboard, and RAM. I think I might've kept the same PSU, and also video card, power supply, case, monitor, and hard drives.
However, if you add it all up... It's an old video card. I could probably buy a newer, faster one for less than $100.
Hard drives are about the most replaceable things on laptops. Power supplies are cheap. So about the only real loss to buying a new laptop, versus the upgrade I made to my desktop, was not being able to upgrade the computer without upgrading the monitor.
Realistically, most people don't do it that way. New desktops are generally cheaper whole, and by the time something dies, or you just feel like you want an upgrade, pretty much everything in that box is going to be obsolete, to the point you'll want to replace it.
So, I like having my options open, yes. But the modularity inside the case is generally irrelevant, to most people.
Nevermind that forcing you to sit in front of the computer, as opposed to being available to you in bed, on your couch, on your porch, etc- means you're more prone to wasting more time on the internet.
And less able to.
Why is that a good thing?
I can do all manner of repairs and data recovery very, very easily on a desktop. Laptops are a total mixed bag ranging from "the company will have a tech here tomorrow morning" to "ARRRRG its going to take an hour to get the damn thing apart."
This can also range from a serious problem to a minor nuisance to a huge problem, depending on whether IT is done right at your company.
Ours basically outsources IT. Basically, all the stuff that's actually irreplaceable on my laptop is stored somewhere else. My code is on a server somewhere, my email and documents are on another. I can still work on my own, since the code is checked out on my hard drive, but so long as I remember to check in regularly, all the data is safe.
I also have a disk image of my hard drive in its first usable state, since it was a pain to get drivers, etc working in XP. It took me less than half an hour to get Linux working, including waiting for the installer.
What this means is, if my laptop was stolen tomorrow, they'd get no confidential information (that's all encrypted), and I'd be up and running again in maybe an hour or two with a new laptop, provided it was available.
If it actually died, I could send it back to the manufacturer, let them take their time trying to recover (it's likely under warranty), while I sit comfortably working with a new laptop. Once the old one is properly repaired, it can become the spare.
And if it means a pain in the ass for you, so what? You're on the clock. Unless the vast majority of your time is spent doing nothing until there's a problem, this shouldn't even annoy you -- really, is there some other aspect of IT you'd rather be doing than disassembling a laptop?
I say Apple should go down this time because they behave like bastards.
Only this time?
Apple locks you down way more than MS. The only reason they get away with it is that they actually make good products, so for the most part, people are happy with Apple as a monopoly (where it is), whereas everyone needs to get away from MS as fast as they can.
According to your analogy, it's perfectly acceptable for the firmware update to not brick my player because of any technical reason, but to actually contain code like:
if (coffee_maker_detected()) { brick(); } else { do_what_customer_fucking_bought_it_for(); }
Now, I am not positive that it was done this way, but I'm fairly sure Apple has done things like this in the past, where they deliberately broke the homebrew stuff, for no other reason than to break the homebrew stuff (or maybe to tighten their DRM, with the iPod, for instance...)
It's not tagged "colddayinhell" because nothing could be better than OS X.
It's tagged "colddayinhell" because Amiga is vaporware. Since Vista has actually been released, consider the new joke to be, "Amiga 5 WILL be released... and it will be bundled with Duke Nukem Forever."
They only have to disclose the claims to the party they choose to sue as part of the legal filings.
Of course, they only have to.
Which is a bit like the Corombite Maneuver. It doesn't matter that Microsoft doesn't have a single patent, it matters that people think they do.
When the day comes that some of these patents are real (assuming any are) are you going to dig into your checkbook and pay the hundreds of millions in damages?
First, they're not real. Don't you think Microsoft would've been happy to name them, if they are?
Second, surely you've heard of a "class action" lawsuit? Or do you really think Microsoft could demonstrate that I've cost them the hundreds of millions? Or that any other single person is as guilty? Surely, that blame rests with more than just one person, right?
Finally, given the quality of most software patents, I'd probably simply refuse to pay. Civil disobedience. I cannot patent the wheel, and Amazon cannot patent "one-click", and IBM cannot patent "outsourcing".
So, why don't you spine up. Form a company selling a fairly complete distro, say a respin of Redhat, and preemptively sue Microsoft for damaging your business with their false claims of patent infringets.
What, to prove a point?
I have better things to do with my time. Don't you?
Otherwise your bolded demands for anything are meaningless and irrelevant.
And so is your pitiful ad hominem.
Right back at ya -- if you're so convinced I'm a spineless coward, why don't you sue me? For anything -- anything at all? Surely, if I'm such a spineless coward, I'll just fork over the money screaming "don't hurt me"? Even better, what if I changed the rules so that you have to form a corporation, first?
Time period and technology posessed by humans in Halo are completely out of whack with what the Marathon story and backstory would imply. Technology behind AI's is entirely different. Technology behind cyborgs is entirely different...
Well, let's see...
Time period is different. But it's my understanding that in Halo, cyborgs have generally been outlawed, which is why they did the Spartan program -- they can't _quite_ make him cyborg, but they can genetically engineer him and give him all kinds of toys.
AI is also not incredibly different. You could easily claim that Cortana is deliberately limited, and certainly the AI (what was her name? Leela?) that you start out dealing with is much more like Cortana. At least until Cortana is left alone in that Covenant capital city...
And then Halo 2 ends, so I'd better go play Halo 3 and get the complete picture.
Shutting up means PHBs get to vaguely remember something about it, in an almost-subconscious way, that makes them feel that much more uncomfortable with the idea of Linux and open source.
We want them to admit that they were lying, publicly.
That, or if there really are patent claims, show us what they are.
I don't mean another game that's called Halo, and is set in exactly the same storyline.
More like one set in the same universe. It's my understanding that basically every shooter they've ever put out, including the Marathon games, is set in the same universe. Maybe even the shooter/rpg before that... what was it called?
But imagine being a Marathon fan leading up to the launch of the original Halo. Cortana actually sent out emails to various fan sites ahead of time, so people were analyzing all the ways in which her emails could be related to a sequel to Marathon... and then they got Halo, which was so completely different, yet in the same universe, with kind of the same theme.
I would love to see them do something like that again. Imagine something as much better than Halo as Halo was better than Marathon.
corrupted encrypted data will simply result in corrupted data when decrypted.
True, but I know that at least for compression, and probably encryption, corrupted data is more than just a single bit flipped, making a slightly different sound. It's a whole block unreadable.
it's more then likely it will perfectly cope with a certain amount off data loss, just like most internet streaming media do.
Ok, RealPlayer, and maybe some realtime chat systems, seem to deal with this by sounding staticy, not sure where that comes from.
But as I understand it, most streaming media doesn't work like that AT ALL. It simply uses standard TCP, if your connection can't keep up, you're dropped. But if a packet is dropped, there's still a chance it'll be resent, and you'll keep playing from your own local buffer until you catch up.
Try this one, rip a full length audio cd to wav files, then try to burn these files onto a cd-recordable. It won't fit, go and find out why and you may learn a lot about what digital really means.
As I understand it, burning it as wav _files_ won't work, but burning it with, say, cdrecord, should work just fine.
"go find out why" is no help at all. At least gimme a Wikipedia article or something.
And most important, all that is totally unimportant to me, because I can never really lose ME - and this is imortality, which is granted to everybody to realize when they are ready.
I don't see how this is connected to James Randi attempting to disprove people like Uri Gellar and (say) Monster Cable.
What you've just described is a philosophy, and obviously not something that can be tested. And that it cannot be tested puts it entirely outside the realm of science.
If you're willing to cast off dogma, then you can really be free and play with life.
How about your "I'm immortal" dogma?
Science, which is originating from the Vedic science
Oh, really?
Would this be the same "Vedic science" preached by Maharishi Mahesh Yogi?
Science should not be about power-struggles, defending dogmas and engaging in logically fallible witch-hunts.
So what would you call what you just now did?
James Randi is uncool -- translation: I want you to listen to me, and not him. I want your power.
We are energy. We are immortal. Science originates from Vedic science. All because I said so. Follow my dogma.
Spritituality and things like yoga, tai-chi, quigong etc., you can practice as the atheist, religious, human-ethic or scientific type.
However, if you claim to know what is happening in Yoga, and define it in terms of Qi or "energy flows", you most certainly are believing in something paranormal.
This doesn't mean you can't do Yoga, just don't claim to know that it's more than it is. If you do, you're no better than James Randi, who seems to believe Yoga cannot be anything more than some interesting stretches. (He's open-minded in the sense that if you actually start levitating due to Yoga, he'll have to admit something else is going on.)
We know so much about our external world, but our internal world we almost never experience, just in sleep. This really seems like a stupid way to go about it.
Actually, I experience it all the time.
That is to say, I am a programmer. Everything I do, all day long, is translating things from my inner world (my mind) to a somewhat less inner world (a computer).
But I don't claim to know it as a separate world -- for all I know, my thoughts are merely electrochemical reactions, no more real than a program in a computer.
Then again, the "real" world could be no more real either -- see "The Matrix".
So why is it that science focuses on the real world, and things we can perceive with our real senses?
Because that's all we can agree on, the only place our observations match up 100%. Scientists will frequently argue over the reason a particular thing was observed, but they rarely argue over whether it was observed. When they want to challenge it, they simply repeat the experiment...
So of course, unrepeatable observations don't belong in science, which limits it. However, it is the best we've got.
Slashdot Mirror
Didn't bother to look up the legal term, but this kind of thing HAS been done -- where one plaintiff went and sued a bunch of people, independently, for essentially the same crime. They then fought back, collectively.
Yes, Microsoft would sue me, an individual. And then I'd go on Slashdot with my evidence, get people to donate to my Paypal, and pool resources with other individuals they'd sued.
That doesn't make it secure. SSH also has an authentication protocol, and it's per-user. If yours is per-IP-stack, you already lost -- both because we already have that (in the form of VPNs and ipsec) and because it's not secure (anyone who can connect to the server can authenticate).
So once again: What method of RPC would you use instead of SSH? Telnet? Last I checked, it won't do RSA authentication -- it's really hardly more than netcat.
Hmm. 20s, more like. And really?
Consider also that 10s of that is likely due to it being a bad time (a nightly cron job was running).
How much?
Because most Flash I know isn't that fast. Fast seeks, slow throughput.
The big advantage would be that you could save a smaller image -- really just swap out what you have to, and otherwise purge your cache. Then, things would be loaded on demand. The reason hibernate doesn't work like this already is -- well, actually, it used to, but the problem was, your disk would thrash for another minute or two as various programs would wake up and have to be pulled off the disk. Flash, with 0 seek time, would make that fast.
Well, actually, most of what wastes disk during boot is pure waste, not redundancy. There doesn't need to be a distinction between loading an app and loading an image, for instance. Consider the difference between OpenOffice and a more modular approach -- a shell script, say. OpenOffice loads a huge chunk of the app off the disk, while a shell script loads a collection of small programs, each on demand.
As for redundancy, well, it actually doesn't take that much time to run DHCP, update the clock, check what monitor is plugged in, etc. Really, most of what's done during boot would be faster if it was decently parellized -- even moreso if it was on a medium with 0 seek time.
You forgot "reinstall".
Reboot, reinstall, and reset are all essentially hacks to deal with bad software. Even the concept of a running program vs a binary on disk is an abstraction we don't need.
What we do need is "de/activate", or suspend, but not in the form of a completely blind dump to RAM -- this should be more like "save/load user state". With a flash-based disk and well-written software, there'd be no performance difference between intelligently saving/loading state and blindly dumping RAM.
And by "saving/loading state", I don't actually mean that there should be some magic signal that every app should obey. I mean apps should be written to use common libraries that will do orthogonal persistence for them, that understand the difference between volatile and nonvolatile storage, and between critical data and various types of cache. There would be other performance gains of such an architecture -- when the system needs RAM, it can flush everyone's cache, of all kinds of things, not just the FS cache -- same goes for disk cache.
Imagine that classic Unix idea -- imagine everything is a file, including your running programs. But how do we think of files? You mount the filesystem, which takes almost no time, and then you can get inside and poke around. Or unmount it, and it's static on disk, and you can't get to it. But when you make changes to the filesystem, they stay. And the system makes it impossible for you to make changes that damage the FS itself, though you can certainly screw up files, and delete
Sorry about the OO comment, apparently I was confusing you with someone else there -- there was someone who doesn't like the "everything is an object" way.
What this tells me is that you grew up with C, or something similar. It doesn't explain why you've chosen to regard code indentation as a sacred element of "style", and not, say, semicolons at the end of every statement, curly braces for blocks, parentheses for both overriding precedence and passing arguments, and so on.
I get that you don't like it, but I don't think syntax alone should be a major reason to ignore a language, although it may make the learning curve a bit harder.
Most of those, I consider valid. I use Perl. But really, the indentation is hardly worth mentioning.
Imagine if it was the other way around -- if Perl forced indentation. Wouldn't you still use it, for all the other reasons you stated?
All interfaces suck. The important thing here is, why do web interfaces suck?
I hate to be trendy, but XMLHttpRequest (AJAX) would work very well here.
I remember when the Web worked reasonably well over dialup. And that was before widespread use of Javascript, let alone AJAX.
If I understand "record paging" here, it looks like it means you can perform some query and get only partial results, and then, transparently, pull in more results as needed as the user "pages" (scrolls?) through the records. Not having used your software, let alone dug into the code, I'm only guessing here...
But if so, that would be easy with a Web interface. User requests next page, so an HTTP request gets sent to the server -- it comes back with a small chunk of XML (or JSON) representing more records to be added to the aperture.
Now, if you meant "paging to disk", as in a local cache, you've got me -- cookies certainly aren't enough storage here.
Again, no problem. Send em back as another chunk of custom XML (or JSAN), describing the operations to be performed. Because of the "A" in AJAX, you could even let the user keep working while this is happening.
It's not just "Install some libraries." It's "Install, maintain, and update some 20-30 different libraries."
On Linux, that's not a problem -- we have good package managers. But on Windows, unless you basically implement your own package manager -- and likely waste disk space, unless you cooperate with all the other apps out there that might use your library. On OS X, you have to deal with the fact that probably half the libraries are already installed, but out of date. You could use Fink, but it's likely going to be out of date, or break in some subtle way.
A web interface here means that all the client-side issues go away.
I haven't seen one feat
Actually, some of them do -- I seem to remember PHY (for network cards) is only available if the card is part of the motherboard. Also, if you buy a gigabit PCI card, you'll likely saturate your PCI bus.
What you wanted was: Just because it's built in doesn't necessarily mean it has a magical interface. It might, or it might not.
I believe both GNOME and KDE have something rudimentary set up here, so that if an app supports this, and you log out (telling it to "save your session", which may be the default), it will save your session. At which point, there is no point to putting it on flash -- disk is more than fast enough. It's also a bit more flexible this way -- log out and switch users becomes much faster.
Here's the problem:
1) The on is no longer instant. Sure, it's quick to shut down, but now the app must be "rebooted", which means reading the entire thing off disk. The reason you think Flash is fast for this is because it doesn't have seek problems -- but normal PC hibernate doesn't, either.
2) Every app has to support it, or there's no point. Unless you want to be checking a particular app for "hibernate support", what usually ends up happening here is either the app gets killed completely on hibernate, or it gets started again properly, but without its state stored.
Now, suspend to RAM is fast enough. Suspend to disk is actually almost as fast for me, were it not for the other hassles of boot. I'm going to test this right now:
Very rough estimate:
Boot -> OS is some 25 seconds. This includes at least a few pauses in both BIOS and OS designed to give me the opportunity to press some keystroke to enter some setup -- I believe that's BIOS, RAID, and the Grub bootloader. Lilo and Macs have the right idea -- why not just let me hold a particular key combo while the box is booting? Even Windows was better about this, sometimes -- when resuming from hibernate, I can press F8 if I want to cancel the resume and do something else.
Maybe it's not intuitive that you can hold Apple+C (or is it Apple+D?) to boot from CD, but it's also something that has to be done maybe one in a thousand boots, and I'd rather have the other 999 boots be as fast as possible.
Then, it's still some 5-10 seconds while Linux churns and loads drivers, etc, before it actually gets to resuming.
The actual act of resuming is about 10 seconds flat, after which my desktop is pretty close to ready to use. Maybe 20 seconds, tops, counting disk thrashing, but that may also be my updatedb cron job.
But even 20 seconds, I could live with. What I hate is that there's another 30 seconds or so that's completely wasted, so it ends up being a minute to resume from Hibernate, when it could've been maybe 20 seconds. (It takes even longer to actually boot.)
Magnetic hard disks are certainly not the limitation here.
What about those first however many seconds that the kernel takes to boot, then the initramfs loads even more...
I mean, yes, Upstart is cool, but right now, fairly irrelevant -- hibernate is stupidly fast when you have any kind of RAID. But the BIOS and the initial Linux boot (during which a fresh kernel loads all the hardware support it will need to... load the hibernate-image-kernel) do still take entirely too long.
Then again, it might help to simply have hibernate images supported directly by bootloaders.
Just a little bit...
Seriously, every language has similar constraints. I'm confused that nobody ever complains about having to have a semicolon at the end of every line, even when it's obvious to the compiler/interpreter that it is the end of a line -- yet when you're told to indent your code the way you should anyway, it's oppressive? What?
As for your app, assuming you are actually using object-oriented development under the hood -- the fact that you were scared by an "everything is an object" language scares me, frankly -- it looks like it could be VERY interesting. Any plans for a web interface?
I know this is often not a relevant issue, as most closed software, for awhile there, was sufficiently reliable that you wouldn't need to fix bugs. Most open software has some pretty big problems -- Firefox's memory leaks, for instance.
But closed software still has some of the biggest, stupidest bugs that would NOT stand in an open system.
I do HD-DVD development. I use Microsoft's HDiSim. Here are two showstopper bugs:
1) No vista support. Even Microsoft doesn't support Vista. I had to wait a week for this laptop to be downgraded to XP because of this.
2) No USB keyboard support. I'm on a laptop with an internal PS/2 keyboard, so this is inconvenient, but not a huge deal. But many Dells don't even come with PS/2 ports anymore -- it's USB or nothing. The best idea I've heard as to how to deal with this is to run it in a virtual machine, where I can set my real (USB) keyboard to be whatever I want it to be (USB, PS/2, Serial, whatever) -- but then, why not use Linux as the host OS?
Neither of these would stand in an open product. Not because "the community" is "better", but because I would go in and fix it myself.
Someone tell me, has Microsoft fixed the audio/net bug yet? The bug where playing music in Vista drops your network performance to 10%? (Drops it by 90%, to 10%.) Because if Linux ever had that problem, I'd fix it, too.
Actually, the reason you don't see them requested often is that they are easier to make work under Wine. Often, they "just work", and/or have native Linux apps that do the same job, as well or better.
Games are there because games are hard -- they don't just have to run, they have to perform well. And not just inside their own little process, the whole stack has to perform well, including the part where they emulate DirectX on OpenGL.
That said, games actually are working quite well. I boot Windows for LAN parties, once a month. The rest of my games are either on Linux or consoles, and I could lose the consoles.
RDP? VNC? RSH???
Sure, if Windows is hardly a monopoly. There are reasonable arguments that both have sufficient competition that they shouldn't be considered monopolies, and, I think, reasonable arguments that neither really has sufficient competition that there's a real choice of moving off one, onto another.
Does iTunes work with any other player? Does the iTunes Music Store work with any other player?
Even if it's not a complete monopoly, my point is that Apple does far more lock-in, DRM, avoiding standards, and so on, than Microsoft ever does. Or rather, they are more extreme about it -- not being a complete monopoly, they have far less opportunity to do these things.
And thus, my point is that were Apple the dominant OS vendor, or the dominant in any field, they'd be even more proprietary and anticompetitive than Microsoft is -- in fact, they already are. The only real advantage would be, the Apple stuff would work. It would actually be a good product, without there having to be sufficient competition and years of R&D and beta testing on the public before they have a kinda sorta OK product.
I realize he said no such thing.
However, the question is whether Apple has done something like that. I don't think we have a clear answer.
I realize that iff it's merely collateral damage, there's no problem, though it would've been nice of them not to break unlocked phones if they could help it. But that might not be the case.
That works just as well the other way around. Desktops have their appropriate uses, but laptops work just fine for a huge percentage of people.
Well, the company paid for my laptop, so "cheaper" doesn't matter to me -- or to other people, in a similar situation.
My desktop recently died. It was likely either the CPU or the motherboard, but I figured I should probably replace both, since new CPUs are cheap -- like $60 for a new dual-core 64-bit. This was made even less of an issue by my friend, who was looking to upgrade his CPU, so he gave me his old one -- so it made even more sense just to rip the motherboard and replace.
But the new motherboard was DDR2 only, and I had DDR400 in the old machine.
Thus, I ended up replacing CPU, motherboard, and RAM. I think I might've kept the same PSU, and also video card, power supply, case, monitor, and hard drives.
However, if you add it all up... It's an old video card. I could probably buy a newer, faster one for less than $100.
Hard drives are about the most replaceable things on laptops. Power supplies are cheap. So about the only real loss to buying a new laptop, versus the upgrade I made to my desktop, was not being able to upgrade the computer without upgrading the monitor.
Realistically, most people don't do it that way. New desktops are generally cheaper whole, and by the time something dies, or you just feel like you want an upgrade, pretty much everything in that box is going to be obsolete, to the point you'll want to replace it.
So, I like having my options open, yes. But the modularity inside the case is generally irrelevant, to most people.
And less able to.
Why is that a good thing?
This can also range from a serious problem to a minor nuisance to a huge problem, depending on whether IT is done right at your company.
Ours basically outsources IT. Basically, all the stuff that's actually irreplaceable on my laptop is stored somewhere else. My code is on a server somewhere, my email and documents are on another. I can still work on my own, since the code is checked out on my hard drive, but so long as I remember to check in regularly, all the data is safe.
I also have a disk image of my hard drive in its first usable state, since it was a pain to get drivers, etc working in XP. It took me less than half an hour to get Linux working, including waiting for the installer.
What this means is, if my laptop was stolen tomorrow, they'd get no confidential information (that's all encrypted), and I'd be up and running again in maybe an hour or two with a new laptop, provided it was available.
If it actually died, I could send it back to the manufacturer, let them take their time trying to recover (it's likely under warranty), while I sit comfortably working with a new laptop. Once the old one is properly repaired, it can become the spare.
And if it means a pain in the ass for you, so what? You're on the clock. Unless the vast majority of your time is spent doing nothing until there's a problem, this shouldn't even annoy you -- really, is there some other aspect of IT you'd rather be doing than disassembling a laptop?
I'd rather back up! Duh!
Or just leave everything on network storage, and use the laptop as a dumb terminal!
It's not rocket surgery, people!
(Well, ok, it might be for Joe User...)
Only this time?
Apple locks you down way more than MS. The only reason they get away with it is that they actually make good products, so for the most part, people are happy with Apple as a monopoly (where it is), whereas everyone needs to get away from MS as fast as they can.
According to your analogy, it's perfectly acceptable for the firmware update to not brick my player because of any technical reason, but to actually contain code like:
if (coffee_maker_detected()) {
brick();
} else {
do_what_customer_fucking_bought_it_for();
}
Now, I am not positive that it was done this way, but I'm fairly sure Apple has done things like this in the past, where they deliberately broke the homebrew stuff, for no other reason than to break the homebrew stuff (or maybe to tighten their DRM, with the iPod, for instance...)
It's not tagged "colddayinhell" because nothing could be better than OS X.
It's tagged "colddayinhell" because Amiga is vaporware. Since Vista has actually been released, consider the new joke to be, "Amiga 5 WILL be released... and it will be bundled with Duke Nukem Forever."
Of course, they only have to.
Which is a bit like the Corombite Maneuver. It doesn't matter that Microsoft doesn't have a single patent, it matters that people think they do.
First, they're not real. Don't you think Microsoft would've been happy to name them, if they are?
Second, surely you've heard of a "class action" lawsuit? Or do you really think Microsoft could demonstrate that I've cost them the hundreds of millions? Or that any other single person is as guilty? Surely, that blame rests with more than just one person, right?
Finally, given the quality of most software patents, I'd probably simply refuse to pay. Civil disobedience. I cannot patent the wheel, and Amazon cannot patent "one-click", and IBM cannot patent "outsourcing".
What, to prove a point?
I have better things to do with my time. Don't you?
And so is your pitiful ad hominem.
Right back at ya -- if you're so convinced I'm a spineless coward, why don't you sue me? For anything -- anything at all? Surely, if I'm such a spineless coward, I'll just fork over the money screaming "don't hurt me"? Even better, what if I changed the rules so that you have to form a corporation, first?
Well, let's see...
Time period is different. But it's my understanding that in Halo, cyborgs have generally been outlawed, which is why they did the Spartan program -- they can't _quite_ make him cyborg, but they can genetically engineer him and give him all kinds of toys.
AI is also not incredibly different. You could easily claim that Cortana is deliberately limited, and certainly the AI (what was her name? Leela?) that you start out dealing with is much more like Cortana. At least until Cortana is left alone in that Covenant capital city...
And then Halo 2 ends, so I'd better go play Halo 3 and get the complete picture.
Shutting up means PHBs get to vaguely remember something about it, in an almost-subconscious way, that makes them feel that much more uncomfortable with the idea of Linux and open source.
We want them to admit that they were lying, publicly.
That, or if there really are patent claims, show us what they are.
I don't mean another game that's called Halo, and is set in exactly the same storyline.
More like one set in the same universe. It's my understanding that basically every shooter they've ever put out, including the Marathon games, is set in the same universe. Maybe even the shooter/rpg before that... what was it called?
But imagine being a Marathon fan leading up to the launch of the original Halo. Cortana actually sent out emails to various fan sites ahead of time, so people were analyzing all the ways in which her emails could be related to a sequel to Marathon... and then they got Halo, which was so completely different, yet in the same universe, with kind of the same theme.
I would love to see them do something like that again. Imagine something as much better than Halo as Halo was better than Marathon.
True, but I know that at least for compression, and probably encryption, corrupted data is more than just a single bit flipped, making a slightly different sound. It's a whole block unreadable.
(Forgot this in my other post, sorry.)
Ok, RealPlayer, and maybe some realtime chat systems, seem to deal with this by sounding staticy, not sure where that comes from.
But as I understand it, most streaming media doesn't work like that AT ALL. It simply uses standard TCP, if your connection can't keep up, you're dropped. But if a packet is dropped, there's still a chance it'll be resent, and you'll keep playing from your own local buffer until you catch up.
As I understand it, burning it as wav _files_ won't work, but burning it with, say, cdrecord, should work just fine.
"go find out why" is no help at all. At least gimme a Wikipedia article or something.
I don't see how this is connected to James Randi attempting to disprove people like Uri Gellar and (say) Monster Cable.
What you've just described is a philosophy, and obviously not something that can be tested. And that it cannot be tested puts it entirely outside the realm of science.
How about your "I'm immortal" dogma?
Oh, really?
Would this be the same "Vedic science" preached by Maharishi Mahesh Yogi?
So what would you call what you just now did?
James Randi is uncool -- translation: I want you to listen to me, and not him. I want your power.
We are energy. We are immortal. Science originates from Vedic science. All because I said so. Follow my dogma.
However, if you claim to know what is happening in Yoga, and define it in terms of Qi or "energy flows", you most certainly are believing in something paranormal.
This doesn't mean you can't do Yoga, just don't claim to know that it's more than it is. If you do, you're no better than James Randi, who seems to believe Yoga cannot be anything more than some interesting stretches. (He's open-minded in the sense that if you actually start levitating due to Yoga, he'll have to admit something else is going on.)
Actually, I experience it all the time.
That is to say, I am a programmer. Everything I do, all day long, is translating things from my inner world (my mind) to a somewhat less inner world (a computer).
But I don't claim to know it as a separate world -- for all I know, my thoughts are merely electrochemical reactions, no more real than a program in a computer.
Then again, the "real" world could be no more real either -- see "The Matrix".
So why is it that science focuses on the real world, and things we can perceive with our real senses?
Because that's all we can agree on, the only place our observations match up 100%. Scientists will frequently argue over the reason a particular thing was observed, but they rarely argue over whether it was observed. When they want to challenge it, they simply repeat the experiment...
So of course, unrepeatable observations don't belong in science, which limits it. However, it is the best we've got.