I agree with most of your post, but there is another solution to this. The latest version of Office presumably supports OOXML in its entirety[1]. You can reverse engineer that and be done with it.
True. But that is not what the standard says to do. I could quote you, over and over again... But I'm sure you can find it yourself, especially since we mostly agree.
I realize that standards need a reference implementation, to cover areas where it is unintentionally ambiguous, as this is inevitable. I object to the standard deliberately relying on the reference implementation, especially when said reference implementation is an expensive and proprietary product.
Also, realize that while we could "reverse engineer that and be done with it", that would only work until Microsoft changed the standard again. Of course, standards must change, but ODF makes that an open process -- Microsoft can actually come and be a part of that committee, I think. I know that we can't exactly have representatives from OpenOffice, KOffice, AbiWord, and iWork all come and sit in on the meetings at Microsoft that will determine the next OOXML.
Emacs asian girlfriend will cook, clean, balance your checkbook, do your taxes, and never, ever complain... but she weighs 300 lbs.
Vi asian girlfriend just stands there looking pretty, but if you thought you were going to get anything done, you're sadly mistaken. It'll take you a week to figure out how to get that dress off...
Vim asian girlfriend will do anything you ask, as soon as you learn the language. Fortunately, most of us know words like "Bukakke" already, and it doesn't take much.
This schedule does also explain why new kernel and X.Org support isn't generally added the same month as its release. If a new kernel at the start of the month breaks fglrx support, that month's driver is already far into the validation and beta stages, which prevents engineers from appending support to the branched driver.
This is why you should follow the new kernels and X.Org while in the development stage. Follow the CVS (or GIT or SVN).
It's true, that kernel may take longer to get out the door than you'd like -- but your new drivers should be backwards-compatible by at least one or two minor kernel versions, right? That way, even if your release is a month or so ahead of the official kernel release, you probably already support it by the time it does roll out.
It is also important to keep in mind that while the AMD release notes in every driver may not be as long as an end-user would like, the developers are actually working on things all of the time.
The end-user doesn't care about the release notes, probably doesn't read them. I know I don't read the release notes on my nVidia drivers, I just know that they just work, and just about as well as they do on Windows.
It is not that we expect you to be doing some set amount of work for us every month. It is that we expect you to not be so insanely, stupendously far behind your competition (Intel and nVidia) that I simply have to tell everyone building/buying a computer for use with Linux to get nVidia if they want games, and Intel if they don't, but never, ever buy ATI.
The entire year we have seen only one formal Linux display driver release from NVIDIA for the GeForce series, along with two beta releases (100.14.03 and 100.14.06) and finally another legacy release (1.0-7185) for their older graphics cards.
And all of them worked.
On the AMD side, however, there have been five Linux drivers this year with another seven expected by year's end.
All of which are lacking certain features I've taken for granted elsewhere (AIGLX, for one), perform worse, and generally suck more.
It's been a LONG time since I tried ATI/AMD for video, and it will take a lot more than a pathetic "It's not our fault, really!" article to make me try them again.
As far as the NVIDIA Linux driver itself goes, it is separated into independent x86 and x86_64 packages and no distribution-specific scripts are included with the mainstream driver. NVIDIA Corporation does not rely upon distribution vendors and community maintainers for allowing the end-user to generate distribution-specific packages, which AMD relies upon heavily.
Irrelevant. My nvidia drivers shipped with Ubuntu, so I really couldn't give a damn who it was that took the x86_64 package and turned it into an Ubuntu package.
In fact, while I do encourage testing on various distros, really, just provide one easy, reasonably transparent installation package -- in the form of a tarball or something close to it -- and have the license allow distros to repackage them for you.
Also, you're forgetting Intel. Intel is going to be the really ugly competition here, because they are the only game in town with a fully supported, fully open driver. If Intel ever releases a video card capable of taking on ATI or nVidia in my price range, I'll take it. (In the past they have done integrated crap that doesn't really compare to the $100-200 range of nVidia/ATI cards, but they've told us that's changing.)
This is the perfect opportunity for AMD, by the way: Just open up your code, and I'm there. The performance difference in hardware isn't significant, but better drivers are, and open drivers would be better.
Try and explain to an average person why all the typing they just did cannot even be viewed in a Web browser, they will not get it.
Sure they will. In fact, they'll assume it.
Basically, they will either assume that they can't make it into a webpage (because it's a word document, and that's different somehow), or they'll assume that making a webpage is too hard for them.
However, if they have a nice WYSIWYG editor or CMS, they'll copy/paste from the word processor onto the webpage, and that will work. If they have FTP access, they can just upload the.doc file, and that might work for what they want -- remember that all their friends will probably be on IE+Office, which means it'll just open up in their browser.
It's not ideal, but trust me, real idiots will out-idiot your expectations. And the ones who are smart enough to realize they're being "conned" might actually find "HTML" in the file->save feature of their word processor.
Saving the user's typing as DOC or ODF is a con.
Wow, I've just been conned out of... oh... $0! Zero dollars and zero cents.
Or would you care to clarify that?
Your document format is ready it is HTML 4.01 Strict, CSS 2.1, and JS 1.5, there is nothing in the 1980's technology of MS Word that cannot be stored this way.
I actually tried this, recently. I wrote a script to generate reports for some poor bastard who's stuck on FileMaker. It generates HTML and uses CSS and JS, and does almost everything he wants, except we can't control the page margins as well as he could in FileMaker. Yes, I know about the CSS margin properties, and browser support for them sucks, and even if you crank them down to zero, the browser likes to add margins of its own.
Point is, there are subtle and fundamentally different problems to be solved by each. Maybe someday they'll converge, but right now, word processing programs are designed to make it easy to physically lay something out on a piece of paper. HTML is designed to lay something out on a piece of software, often in a fluid way, with dynamic components.
Coders should move on to the word processing interface WHICH FUCKING SUCKS.
Yes, it does. It's just the best we've got.
Make users the word processor that is wanted by both Firefox/Safari and Firefox/Safari users.
Should I even try to parse that?
I guess I should make gamers the game that is wanted by both Xbox/PC and Xbox/PC gamers?
Please the humans with a great writing interface and easy document-construction tools
Yeah, great. That's called a word processor.
and please the browsers by storing all of the work as really plain HTML+CSS+JS.
That's great for the browsers, but sucks for the main point of word processors, which is: Printing! Yes! On paper!
Designers can also generate templates for such a word processor out of their own HTML+CSS+JS authoring tools. Programmers can integrate their work with everyday work processing documents if they are stored as HTML+CSS+JS. You put the office typewriter into the Web's tool chain and it will be good for everyone.
Please explain to me how this is different or better than Google Office + ODF.
Oh, by the way: ODF and HTML+CSS are actually pretty close. Close enough that I've got less than 1k worth of Ruby code to convert between them, which includes all sorts of stuff that was specific to the last company I worked at.
OOXML is built in to Office 2007. The files are zips, if you unzip them, they're generally plaintext XML. Which is to say, JUST AS READABLE AS ODF.
WMV support is built into Windows Media Player. The files are binary, if you look at them in a hex editor, they're generally plain numbers. Which is to say, JUST AS READABLE AS MPEG.
Except not, unless you are a fucking moron. I'm sorry, but XML is not magic open interoperability pixie dust.
MS looked at ODF, but felt that since it didn't support some functions of legacy Office applications, they wanted a broader definition set, which led to the ginormous OOXML standard.
Nope. Nice try, though.
What actually happened was, MS looked at ODF, but felt that since it threatened their monopoly of Office applications, they wanted their own "standard" that they could control.
Or maybe they did it by accident. (Yeah, right.)
ODF was designed to be all things to all office suites. OOXML was designed to basically be an XML dump of MS Office documents, and from what I have heard (and seen), it's little more than a straight 1:1 conversion of the binary Office format into XML.
I suggest you go actually try to read the OOXML "open standard", and understand why it is neither. It has little to do with the 6000 pages, it's about how little is actually in that 6000 page document.
Now, you can complain (not without significant justification) that OOXML is a hugely bloated standard
No. We complain that it is not a standard, and not suitable for implementation in anything but MS Office.
due to it's trying to be all things to all iterations of Office
The problem is not that it supports all these various iterations of Office, and even older things (WordPerfect, etc). The problem is that they support these by creating some sort of tag or attribute or something which flags a section as being formatted for Word95 or somesuch, and then don't define how to do that. They basically say it's "beyond the scope of this document", and that you should emulate the behavior of the software in question.
And this is not the right way to design a standard format anyway. Suppose different versions of Word came with different default heading styles. You could just put <word95heading> tags around something -- or you could use a format that supports defining custom styles.
but your comment that only MS can use MS formats is a red-herring.
That's true, we can reverse-engineer MS formats, and have done so. Most open office suites (OpenOffice, KOffice, AbiWord, etc) support the binary Office formats quite well. But it's still reverse-engineered, and still not complete.
It would be entirely possible to make a document standard that is just as flexible, concise, and transparent as ODF, but support all of the crap that OOXML does. The difference is, it would be much more difficult for MS to support such a standard, and much easier for everyone else. As it is, OOXML is much easier for Microsoft to implement than for anyone else.
Consider that, in order to fully support OOXML, you have to actually go and buy all of those different versions of Office, plus random crap like WordPerfect, and reverse-engineer their behavior. So OOXML is not any better than the binary formats, because in reality, you may actually have to reverse-engineer MORE products in order to make it work.
Do you people really want Government to have a hand in this?
In its own internal affairs? Hell yes!
You do realize that once they start "regulating" they start taking control?
Great. What's the alternative? Should the city council hold a vote every time they want to author a new document, to determine what format the town wants it in that time?
Even if you wanted that, who decides how to take the vote? Maybe they'll write ballots in MS Word, maybe in ODF.
Or, do you think Government should step in and make your decisions for you?
First, you're either trolling or stupid. This is about the government deciding what format should be used for its own documents. If you really, really want to spend $200 for an OS and $300 for an office suite, you're free to do so -- and I'll stick with my free OS and free word processor.
Second, this argument is generally made by astroturfers or outright shills for various companies who want things which are about to be restricted. For example: No one in their right mind who knows about Net Neutrality honestly believes it to be a bad thing. ISP-controlled packet shapers bad, egalitarian Internet good -- it's almost a reflex. So no one who argues against Net Neutrality even brings up the packet shapers anymore -- they bring up the old "Do you really want the government to interfere?" argument.
Well, I'm sorry, but the whole fucking point of having a government is to interfere. That's why laws exist. That's why murder is illegal, for example. But apparently this anarchistic drivel fools enough people that you actually managed to redefine the phrase "Net Neutrality" to mean not letting the government interfere -- which is the exact OPPOSITE of what it originally meant.
Yes, Point b) might not be exactly legal, but you see where I'm coming from.
Point b is called "civil disobedience", and I think we should remember this and point it out.
I rip DVDs and watch them on Linux. Often, someone in the house will rent a DVD for everyone to watch, but I'm busy, so I rip it and watch it later, once the disc is back in the store. I acknowledge that all of this is illegal, and if caught, I may well go quietly. I am deliberately disobeying this law, however, to express that I do not agree with it -- and to do the things I should be able to do anyway.
Just like Rosa Parks on the bus. You can argue magnitude if you like -- that I could just choose not to watch DVDs, or I could choose to use Windows and approved, DRM-enabled solutions. Right -- and Rosa Parks could've chosen to not ride that bus, or to give up her seat.
The average user cattle doesn't care about the data format war, only the technical folks.
It is important, though.
Let me put it this way: In 100 years, Microsoft will most likely not exist, nor will any of our current word processors. (That's not meant to be a political statement; Linux won't exist, or won't be recognizable.) Which format do you think is more likely to be readable then?
I think it's ODF, but in any case, it's an important question. If you're doing scientific work, presumably you do hope to publish some hugely important paper that people will still want to read in 50 or 100 years -- something like General Relativity. And of course, even if it's not successful, presumably at some point, you or someone else in the field will want to refer to your earlier work.
It's a bit like DRM. The average user doesn't care about DRM, but they hate it when their brand new HDTV goes black for some reason, or when they can't skip the previews on a DVD, or when they have to buy the same damned thing in 20 different formats for no reason.
It's not that users don't care about technical debates. What's really going on is, users do care, and do get pissed off when things don't work. But they don't want to know why something doesn't work, they just want it fixed. That's why we see people who don't maintain their computers at all, get loaded up with spyware, and then just buy a new one or pay GeekSquad to run Spybot.
It's a bit like never changing the oil in your car, and constantly raving that you shouldn't have to change the oil, and buying a new car when your old engine dies because you refused to learn anything about it.
So it's not that users don't care, it's that users desperately want to avoid ever having to think about something that isn't their area of expertise.
Personally, I think users should care about the format war if they care about the consequences of the format war, and most do.
I expected you to lead up to "high-performance applications need a barrier - ensure operations this process performed before the barrier hit disk before operations it performs after the barrier()". Or for even higher performance, a more precise barrier that takes inodes or file descriptors: "the operations performed before on inode A hit disk before the operations performed after the barrier on inode B if a call to".
That might work, and I think some filesystems might have implemented it. It's also closer to working with existing POSIX stuff than what I might've come up with. But it also requires the app to do quite a lot that I don't think it has to, and requires some duplication of code.
For the simple, small-file example, it works reasonably well. Write to the new file, barrier, rename to the old, barrier. Essentially have barrier calls in most places you'd have sync calls -- and yes, maybe fbarrier or fdatabarrier, like fsync/fdatasync. Note that sync-ing already does basically the same thing, just slower. And you might also want to have a call which explicitly says you're going to be using barriers, telling the FS it can reorder as much as it wants.
There are a couple of problems with that approach, I think. The most obvious one is that essentially, your application is trying to implement a transaction, and your filesystem will, if it's smart, group these into a transaction. In other words, if I read from file a, write a new version to file tmp, barrier, and then rename that tmp to a, then barrier again, all before the FS writes it out to disk, a smart enough FS might figure out that I want to atomically update a, and skip the tempfile altogether. Skipping the tmpfile means skipping dealing with allocating a new inode, setting a dozen timestamps, then unlinking it, when it might be faster to simply use the FS's own journal. (Or it might use the tmpfile anyway, if it thinks that's faster.)
At the same time, your application, if it's more than just vim, might have more than one file it wants updated atomically. For instance, it might want to go through a whole directory tree, doing some operation to, say, half the files in the tree. This starts to make it more complicated. The simplest way I can think of with barriers would be something like: Copy everything to a temporary tree, do all of your update operations there. Barrier. Rename the original tree to something like orig.bak, and move the new tree to the original tree. Barrier.
The above may not always work, of course. It ignores open filehandles or working directories within the tree by other processes. It also means that, even if you use hardlink tricks, you're probably copying far more than you need to.
And consider a database, or any large file that you might update random chunks in the middle of. For that to work, you're either going to need to copy the database around a LOT -- not feasible if it's a few gigs in size -- or you're going to have to maintain some sort of journal yourself.
So, again -- the FS already has a journal and a concept of an atomic operation. Your app is trying to implement an atomic operation. It seems to me that you should communicate via atomic operations. It also doesn't even seem that hard -- consider that ZFS supports writable copy-on-write snapshots of the entire filesystem. If I were implementing a filesystem, that's about how I'd do it -- provide a new API that resembles POSIX (and maybe is backwards-compatible), but allows each operation to be tagged as belonging to a certain transaction. It would look like this (excuse the pseudo-C code, I don't do a lot of C):
foo = begin_transaction();
write(some_file, "some data", 9, foo);
write(some_other_file,.... foo);
rename ('a', 'b', foo);
unlink ('c', foo);
commit_transaction(foo); // or you could do
sync_transaction(foo);
Roughly, of course. The idea is, anything read or written in the context of that transactio
I would say, be aware of where you're re-using passwords. The reason not to re-use a password is to prevent a compromise of one account on one site leading to a compromise of another account on another site -- and that compromise may come from inside.
For example, I really don't give a damn if MySpace can get into my free New York Times account, but that's basically what using the same password on both implies -- if someone 0wns MySpace, or MySpace itself becomes corrupt, they can get my password and use it elsewhere.
But, for example, my passwords to my bank account, or to various other things that I actually care about, are entirely unique. And most of my access to things I care about (remote servers that I admin) is done with SSH keys instead of passwords.
The problem with power is, if something accidentally gets stuck on that button, hold it down for 5 seconds and the machine is OFF. I know you can intercept a single press/release, but if you press and hold, on most boxes, it will eventually shut off no matter what you're doing.
On the last laptop I had Linux on, I played with the lid-close event. That was a bit more useful...
Only problem: AV is based on the assumption that we know what a virus looks like. There are enough false positives that the heuristics can't be working well, and the very existence of a signature means someone must've been infected already.
AV is a bit like the rifle, because it's the last line of defense, and a pretty damned weak one. I'd say anytime your AV hits, if you didn't see it coming with that particular file, you're doing something wrong.
As Los Angeles County sheriff's deputies and Secret Service investigators try to track down the crooks, Carson has fielded calls from officials worried about the security of municipal coffers. "They want to know how they can prevent this," Avilla said.
I know it's not going to fix anything, but there are a few simple, simple steps:
Linux. If you can't make that work, get a Mac, but really, do give Linux some serious consideration. Especially if you can standardize on things in the normal repositories, you basically kill any equivalent of the most common and easiest Windows attack vectors.
Never let it out of your sight. If it's a desktop, it stays in a room that only you and trusted people have access to, like your office. When you're not there, lock the door. If it's a laptop, either keep it locked in a similar room, or carry it with you. If you MUST let it out of your sight, get one of those stupid-looking laptop locks and lock it to something solid. When you get back, check for tampering.
Don't let anyone have unlimited access to it. If someone MUST use your computer, every time they touch it, it should be under some limited account, not yours. When they're done, nuke the account. And again, be in the room, paying enough attention that you'll notice if they try to open the case or unplug anything.
Lock it down. Linux/Mac is part of the above, but even if you MUST use Windows, turn on the firewall, download some good, free antivirus and antispyware (and pay for some if you can't get it free, due to many of the "free" ones being free only for home use), and turn off AutoRun, even if you never plan to play music CDs. You could go farther, too -- on Mac/Windows, BitLocker/FileVault. On Linux, you could encrypt the entire disk except your boot partition, and you could put that on a removable flash thumbdrive. You could also use SELinux, which, on a distro that supports it, is complete overkill even for this -- every process has a set of rules defining what it can and cannot do.
Use a secure browser, which basically means anything except IE. If you're on Vista, maybe IE 7, but I still prefer open source. And even then, disable crap you don't need, run Flash on a per-page click-to-play basis, and pay very close attention to the URLs you visit when accessing your bank.
Use at least two-factor authentication. A thumbprint reader, a smartcard reader, or even a simple thumb-drive with a keyfile on it.
Don't be stupid with passwords. Don't give them out for chocolate (has happened before). It is not enough to name it after your dog and add a year, your Fido1993 will be cracked in two minutes with a dictionary cracker, if you even bothered to capitalize the F. Make it hard enough that you have to write it down, and then make sure where you write it is sufficiently protected -- for example, on something in your pocket, or have the browser remember on that encrypted hard drive. (The encrypted drive, of course, will always have the same password, and that should be a hard one that you bite the bullet and memorize anyway. Or a very-obfuscated one that you can remember, for example, 2b||!2b could read "To be or not to be" (to a programmer), but beware that being predictable (such as pulling it out of my Slashdot comment) can make hard obfuscation easy.)
This is common sense stuff. Some of it is a bit tinfoil-hat (SELinux, secure hardware), but really, most of the above can be done very cheaply, and in the long run, won't take any significant amount of time or brainpower to maintain.
And though I've never been a cracker, it still pisses me off when, instead of responding by paying attention to common-sense security (as I've just described), they'll attempt to buy a magic bullet -- they'll buy ONE product, probably something standard like Windows Defender, and then get lazy again. Or sometimes they'll try litigation, or both:
The treasurer said she is now determined to try to write legi
I really, really wish the people writing these stories would bother to at least try to get the jargon right. After all, there's no mention of the word "keylogger", or the fact that it's a really fucking obvious and common attack. I bet they either thought or are trying to pretend that this kind of thing has never happened before...
There's no mention of the method used to install the keylogger onto the treasurer's computer.
Yes there is.
Armed with a spyware program, the thieves tracked Avilla's moves on her laptop and obtained bank passwords.
That is, unless they don't know what the word "spyware" means. Being reporters, they might just assume that spyware means what it sounds like -- any software used to spy on you, including something picking up keystrokes from a physical keylogger.
But then, it also seems like it would be difficult to make a physical keylogger that communicates reliably with the outside world:
Each time Treasurer Karen Avilla logged into her laptop computer in the morning, someone was looking--virtually--over her shoulder, watching every keystroke.
That sort of implies it's being done in realtime. Of course, they could always mean it was a physical keylogger, which the "hacker" then collected and dumped...
Then again, it's a laptop. If you have physical access to a laptop for long enough and with enough tools to install a physical keylogger, it's probably easier to carry the thing off and hope there's something valuable on the hard drive.
First, to those who made comments about 128k encoding, you may be thinking of mp3. (Or maybe not, who knows...) From what I've heard, AAC, Vorbis, and AC3 all sound better than mp3 at similar bitrates.
Second, I remember there was a comment on Slashdot awhile back, before they actually came out with these, and I want to confirm... Apparently, CDs are recorded at a certain physical bitrate/frequency, and there are digital masters which are at a higher rate... it's late, so I'm not entirely coherent, but think of it as somewhat equivalent to resolution of a DVD (quality of video is proportional to resolution (HD vs normal) and bitrate). The point was that 256k may actually sound better than a CD, since it comes from a better source than a CD.
If so, this whole test is BS, since they did not do a comparison of CD vs 128k (either iTunes-DRM'd or custom-ripped) vs 256k (un-DRM'd, from the iTunes store). Specifically, I'd want to hear 256k vs CD. But at the same time, I don't know if any iPod, or specifically the one they are using, would be able to handle the higher resolution. If not, you'd have to specifically check your soundcard, too.
And finally, again vaguely remembering this from a Slashdot comment (so correct me if I'm wrong), but there was some comment about "The 30c may seem small, but imagine buying a whole album at these prices..." And I seem to remember that a full album is always $9.99. Still high compared to, for example, the minimum you'd pay for a FLAC-encoded album at MagnaTune, but if you're buying a whole album (and if that's accurate), you may as well just opt for un-DRM'd -- especially if it sounds better than a CD (which would probably cost more anyway.)
But then, of course, I'd like to hear a much bigger study, with more rigorous controls in place. 10 people is just not enough, no matter how you set it up.
And personally, if I had any money to spend on music, I'd be buying un-DRM'd stuff. But probably not from iTunes -- not till there's a web interface (or at least an API) that doesn't require me to download their client software. After all, if I'm buying a single file, the point of the client software is to implement the DRM, and if I buy the un-DRM'd version... Not that it shouldn't also work in the iTunes client, but it'd be nice for it to work natively in AmaroK, or just in a browser.
But to say that using a computer from 1986 is better than using what he have today merely because you wouldn't need AV or Firewalls is not accurate.
Agreed. Did I ever say that? Can you quote me where?
I'm just pointing out that I don't think firewalls/AV are things you should count when talking about how much more you can do with a modern computer.
And you're telling me that there's no such thing as an entry level Linux user who needs firewall/AV?
Well, no. There are a few entry-level Linux users who would be greatly appreciative if I just put an antivirus logo in the system tray, because unless I do that, they'll somehow delude themselves into thinking Linux is less secure.
Maybe because it's pretty hard to be an entry-level Linux user, period?
Another debate for another time, but my mother seems pretty comfortable whenever I put her on Linux. In fact, the most common problems I ever see anyone having with Linux these days are missing software. Most of the other complaints (having to edit config files, everything being complicated/weird/different for no reason, installation being hard/risky) are pretty much gone these days. (Even if installation IS hard, you can always buy Ubuntu on a Dell.)
What if something goes wrong with your Linux install? I'm sure an entry level user could fix it!
I'm sure that has something to do with our discussion about firewalls and AV, I just don't see it.
If something goes wrong with ANY install, entry-level users cannot fix it. It has nothing whatsoever to do with Linux.
How about if a huge corporation runs completely on Linux. No need for firewalls? No need for AV?
Yes, and yes. Although I would say that in such a corporation, you probably want the big corporate firewall sitting on the gateway, but that's as much to keep users in as to keep "hackers" out. For example, you don't want anyone at your company to start spewing random spam out to the Internet (intentionally or not), so blocking port 25 outbound seems fair.
But I'd argue that even the big-corporate-gateway-firewall is more often than not entirely about making the Windows boxes behind the firewall secure, and/or a need for NAT. The only other thing I can possibly see it preventing is, say, a ping flood, but all that does is make it lag your Internet connection instead of your internal network.
But AV? You're either clueless or joking.
Don't you think if *nix was as popular as Win32/64 there would be a lot more incentive to try to create malware and try to hack into the servers?
For the four billionth time, *nix is pretty fucking popular on servers. According to netcraft, Apache is still roughly twice as popular as IIS (guessing from the graft). I can't find the ratio of Windows to Linux servers, but I seem to remember it was something like less than 50% Windows, and roughly 30% Linux.
Don't you think, if it was so easy to hack, that someone would've already done so? Written a virus, built a botnet, but instead of using tons of home users overloaded with spyware, and with barely any bandwidth or performance... Don't you think they'd go after a dual-core or quad-core server with at least 10-100mbits of connectivity?
In other words, don't you think something like Slammer would've happened to Linux already, if it was just a popularity contest?
Linux is not more secure merely because it's obscure -- that's Microsoft's game, remember? Linux is more secure because of better design and a better development process, at least with regards to security.
Don't you think, just maybe, that there would be vulnerabilities found and exploited?
They're found, frequently. They're just found and patched ver
I'm not even going to read the rest of your post, you are so far wrong there that it's not funny.
I could say the same, but I'm a bit more persistent than that.
And before you say 'yeah, but you're not a 'normal' user', remember you said "...WAY more RAM than anyone needs. " I'm someone... and I need more, as do my colleagues here.
Well, I suppose it all depends how you work. I absolutely NEED at least a gig or two of RAM for games...
I do a bit of "light coding" myself, but I tend to use vim, or, worst case, Kate, for the actual text editing. I run Konqueror as a web browser. And I haven't touched Java since college.
I would say, if you can find a way to lose Eclipse, you'll have half your RAM problems solved right there.
Also, I'd suggest you read the rest of my post, given that wasn't even a main point. The main point is that background processes are generally irrelevant to most people, because they'll spend most of their time sitting in swap.
Note however that I don't depend directly on openSSL. I only use a class called "KSSL". Whatever it uses internally shouldn't be my problem.
That is true.
In Gentoo, I believe you can depend on specific USE flags being set. Also, autoconf and friends should catch things like this...
The right way to do this, I think, would be to fork off KSSL into its own library, so it can be compiled individually, and depended on individually, and not installed pointlessly with the rest of KDElibs just so I can run, say, AmaroK. The fact that it doesn't do this is half the reason Gentoo exists: huge packages like KDElibs which can only have certain support/dependencies added or removed at compile-time. That's the whole point of USE flags.
I like the Debian way of dealing with this, though, when it works: Split one source package into multiple binary packages, whenever possible, so that I can install the X libs without any kind of X server, for example. (I think this is possible on Gentoo now, but it took awhile, and basically didn't happen until X.org.)
But that still makes it a pain in the ass for developers.
So in summary, it's your fault for not depending on SSL support (however Gentoo does that these days), KDE's fault for even having one package called "KDElibs" that includes so much crap, and maybe the users' fault for actually disabling SSL in their USE flags. I wouldn't say it's actually Gentoo's fault, except maybe for not providing easier ways to resolve these problems, especially if they are the ones repackaging your software.
Also, assuming that it's an FPS, is there any new territory to cover?
Depends what you consider "new".
There's a lot of people who seem to think that in order to be "new" or "original", a game must have entirely new and original "gameplay", and then go debate about what that would be. Most of these people wouldn't argue about a few games -- DDR and Frequency/Amplitude were very much new, both in control style and actual gameplay.
First, new ground isn't necessary. I just got a new DDR game (having lost my old ones), and I haven't even gotten into the actually new features, like EyeToy and internet gameplay, but it already kicks ass. The gameplay is exactly the same as it always was, but there are a lot of good new songs, and the animated, dancing 3D chick is hotter and less pixelated than before, and also apparently no longer cell shaded.
I wouldn't have bought it if I didn't lose the old one, of course, because I don't think these things are worth shelling out $20-50 for some new songs. But that's all in the eye of the beholder, really. Half-Life 2: Episode 1 didn't really introduce any shockingly new gameplay since Half-Life 2, not even much in the way of new monsters. Alyx is hot as ever, but despite the hype, she's not all that much smarter than before. But I still think it's worth it, because the story was interesting enough, and there's nothing wrong with the gameplay.
For that matter, consider Counter-Strike. Really, nothing has changed since the original CS was released (free for download if you had Half-Life), yet people still play it every day. Or World of Warcraft -- are they really covering any new ground there?
I would say that FPS is just the interface, the medium, and the question you are asking is more like "Assuming it's just a video game, is there any new territory to cover?" There's always room for the truly odd and different, even if it's still technically an FPS. Consider Deus Ex, or Natural Selection. Even Cube.
Everything depends on resolution. The only "resolution independence" that we really have and use today is in apps which can be in resizable windows, which I'm pretty sure that Mac Plus had.
the processor-intensive audio and visual codecs we use everyday in music players and web browsers
I'd just like to take the opportunity to point out, again, how much Flash sucks. Take any video on YouTube, download it (may need a Firefox extension), and view it with VLC. On my box, it was 50% CPU at least in the browser, even postage-stamp-sized, and less than 1% in VLC, even fullscreen.
We should go back to 8-bit 640x480 displays because the Windows 3.1 pulldown menu seemed faster.
I think the point isn't that we should go back. The point was that, if you're not excited about higher resolution and more audio/video, there's not much reason to be excited about computing anymore. I'm not saying I agree with that, but let's see -- you mentioned firewalls. Great, but you can't go online with that '86 machine, so who cares? The only one you mention that makes any sense is spellchecking, which can be done tolerably well on any five-year-old machine, even realtime.
Maybe it is, but I honestly can't tell. Just took roughly one second to launch an instance of vim, with color-coding and everything, on a small python script. Quit and re-opened, and it's faster than I can see.
Slashdot Mirror
No, that would be Kate. KDevelop would be the wrapper (and all her toys).
True. But that is not what the standard says to do. I could quote you, over and over again... But I'm sure you can find it yourself, especially since we mostly agree.
I realize that standards need a reference implementation, to cover areas where it is unintentionally ambiguous, as this is inevitable. I object to the standard deliberately relying on the reference implementation, especially when said reference implementation is an expensive and proprietary product.
Also, realize that while we could "reverse engineer that and be done with it", that would only work until Microsoft changed the standard again. Of course, standards must change, but ODF makes that an open process -- Microsoft can actually come and be a part of that committee, I think. I know that we can't exactly have representatives from OpenOffice, KOffice, AbiWord, and iWork all come and sit in on the meetings at Microsoft that will determine the next OOXML.
I can't tell if you meant the machines of 95% of some number that is "most people", or, somehow, only 95% of each machine that belongs to most people.
Emacs asian girlfriend will cook, clean, balance your checkbook, do your taxes, and never, ever complain... but she weighs 300 lbs.
Vi asian girlfriend just stands there looking pretty, but if you thought you were going to get anything done, you're sadly mistaken. It'll take you a week to figure out how to get that dress off...
Vim asian girlfriend will do anything you ask, as soon as you learn the language. Fortunately, most of us know words like "Bukakke" already, and it doesn't take much.
From TFA:
This is why you should follow the new kernels and X.Org while in the development stage. Follow the CVS (or GIT or SVN).
It's true, that kernel may take longer to get out the door than you'd like -- but your new drivers should be backwards-compatible by at least one or two minor kernel versions, right? That way, even if your release is a month or so ahead of the official kernel release, you probably already support it by the time it does roll out.
The end-user doesn't care about the release notes, probably doesn't read them. I know I don't read the release notes on my nVidia drivers, I just know that they just work, and just about as well as they do on Windows.
It is not that we expect you to be doing some set amount of work for us every month. It is that we expect you to not be so insanely, stupendously far behind your competition (Intel and nVidia) that I simply have to tell everyone building/buying a computer for use with Linux to get nVidia if they want games, and Intel if they don't, but never, ever buy ATI.
And all of them worked.
All of which are lacking certain features I've taken for granted elsewhere (AIGLX, for one), perform worse, and generally suck more.
It's been a LONG time since I tried ATI/AMD for video, and it will take a lot more than a pathetic "It's not our fault, really!" article to make me try them again.
Irrelevant. My nvidia drivers shipped with Ubuntu, so I really couldn't give a damn who it was that took the x86_64 package and turned it into an Ubuntu package.
In fact, while I do encourage testing on various distros, really, just provide one easy, reasonably transparent installation package -- in the form of a tarball or something close to it -- and have the license allow distros to repackage them for you.
Also, you're forgetting Intel. Intel is going to be the really ugly competition here, because they are the only game in town with a fully supported, fully open driver. If Intel ever releases a video card capable of taking on ATI or nVidia in my price range, I'll take it. (In the past they have done integrated crap that doesn't really compare to the $100-200 range of nVidia/ATI cards, but they've told us that's changing.)
This is the perfect opportunity for AMD, by the way: Just open up your code, and I'm there. The performance difference in hardware isn't significant, but better drivers are, and open drivers would be better.
Sure they will. In fact, they'll assume it.
Basically, they will either assume that they can't make it into a webpage (because it's a word document, and that's different somehow), or they'll assume that making a webpage is too hard for them.
However, if they have a nice WYSIWYG editor or CMS, they'll copy/paste from the word processor onto the webpage, and that will work. If they have FTP access, they can just upload the .doc file, and that might work for what they want -- remember that all their friends will probably be on IE+Office, which means it'll just open up in their browser.
It's not ideal, but trust me, real idiots will out-idiot your expectations. And the ones who are smart enough to realize they're being "conned" might actually find "HTML" in the file->save feature of their word processor.
Wow, I've just been conned out of... oh... $0! Zero dollars and zero cents.
Or would you care to clarify that?
I actually tried this, recently. I wrote a script to generate reports for some poor bastard who's stuck on FileMaker. It generates HTML and uses CSS and JS, and does almost everything he wants, except we can't control the page margins as well as he could in FileMaker. Yes, I know about the CSS margin properties, and browser support for them sucks, and even if you crank them down to zero, the browser likes to add margins of its own.
Point is, there are subtle and fundamentally different problems to be solved by each. Maybe someday they'll converge, but right now, word processing programs are designed to make it easy to physically lay something out on a piece of paper. HTML is designed to lay something out on a piece of software, often in a fluid way, with dynamic components.
Yes, it does. It's just the best we've got.
Should I even try to parse that?
I guess I should make gamers the game that is wanted by both Xbox/PC and Xbox/PC gamers?
Yeah, great. That's called a word processor.
That's great for the browsers, but sucks for the main point of word processors, which is: Printing! Yes! On paper!
Please explain to me how this is different or better than Google Office + ODF.
Oh, by the way: ODF and HTML+CSS are actually pretty close. Close enough that I've got less than 1k worth of Ruby code to convert between them, which includes all sorts of stuff that was specific to the last company I worked at.
WMV support is built into Windows Media Player. The files are binary, if you look at them in a hex editor, they're generally plain numbers. Which is to say, JUST AS READABLE AS MPEG.
Except not, unless you are a fucking moron. I'm sorry, but XML is not magic open interoperability pixie dust.
Nope. Nice try, though.
What actually happened was, MS looked at ODF, but felt that since it threatened their monopoly of Office applications, they wanted their own "standard" that they could control.
Or maybe they did it by accident. (Yeah, right.)
ODF was designed to be all things to all office suites. OOXML was designed to basically be an XML dump of MS Office documents, and from what I have heard (and seen), it's little more than a straight 1:1 conversion of the binary Office format into XML.
I suggest you go actually try to read the OOXML "open standard", and understand why it is neither. It has little to do with the 6000 pages, it's about how little is actually in that 6000 page document.
No. We complain that it is not a standard, and not suitable for implementation in anything but MS Office.
The problem is not that it supports all these various iterations of Office, and even older things (WordPerfect, etc). The problem is that they support these by creating some sort of tag or attribute or something which flags a section as being formatted for Word95 or somesuch, and then don't define how to do that. They basically say it's "beyond the scope of this document", and that you should emulate the behavior of the software in question.
And this is not the right way to design a standard format anyway. Suppose different versions of Word came with different default heading styles. You could just put <word95heading> tags around something -- or you could use a format that supports defining custom styles.
That's true, we can reverse-engineer MS formats, and have done so. Most open office suites (OpenOffice, KOffice, AbiWord, etc) support the binary Office formats quite well. But it's still reverse-engineered, and still not complete.
It would be entirely possible to make a document standard that is just as flexible, concise, and transparent as ODF, but support all of the crap that OOXML does. The difference is, it would be much more difficult for MS to support such a standard, and much easier for everyone else. As it is, OOXML is much easier for Microsoft to implement than for anyone else.
Consider that, in order to fully support OOXML, you have to actually go and buy all of those different versions of Office, plus random crap like WordPerfect, and reverse-engineer their behavior. So OOXML is not any better than the binary formats, because in reality, you may actually have to reverse-engineer MORE products in order to make it work.
In its own internal affairs? Hell yes!
Great. What's the alternative? Should the city council hold a vote every time they want to author a new document, to determine what format the town wants it in that time?
Even if you wanted that, who decides how to take the vote? Maybe they'll write ballots in MS Word, maybe in ODF.
First, you're either trolling or stupid. This is about the government deciding what format should be used for its own documents. If you really, really want to spend $200 for an OS and $300 for an office suite, you're free to do so -- and I'll stick with my free OS and free word processor.
Second, this argument is generally made by astroturfers or outright shills for various companies who want things which are about to be restricted. For example: No one in their right mind who knows about Net Neutrality honestly believes it to be a bad thing. ISP-controlled packet shapers bad, egalitarian Internet good -- it's almost a reflex. So no one who argues against Net Neutrality even brings up the packet shapers anymore -- they bring up the old "Do you really want the government to interfere?" argument.
Well, I'm sorry, but the whole fucking point of having a government is to interfere. That's why laws exist. That's why murder is illegal, for example. But apparently this anarchistic drivel fools enough people that you actually managed to redefine the phrase "Net Neutrality" to mean not letting the government interfere -- which is the exact OPPOSITE of what it originally meant.
Point b is called "civil disobedience", and I think we should remember this and point it out.
I rip DVDs and watch them on Linux. Often, someone in the house will rent a DVD for everyone to watch, but I'm busy, so I rip it and watch it later, once the disc is back in the store. I acknowledge that all of this is illegal, and if caught, I may well go quietly. I am deliberately disobeying this law, however, to express that I do not agree with it -- and to do the things I should be able to do anyway.
Just like Rosa Parks on the bus. You can argue magnitude if you like -- that I could just choose not to watch DVDs, or I could choose to use Windows and approved, DRM-enabled solutions. Right -- and Rosa Parks could've chosen to not ride that bus, or to give up her seat.
It is important, though.
Let me put it this way: In 100 years, Microsoft will most likely not exist, nor will any of our current word processors. (That's not meant to be a political statement; Linux won't exist, or won't be recognizable.) Which format do you think is more likely to be readable then?
I think it's ODF, but in any case, it's an important question. If you're doing scientific work, presumably you do hope to publish some hugely important paper that people will still want to read in 50 or 100 years -- something like General Relativity. And of course, even if it's not successful, presumably at some point, you or someone else in the field will want to refer to your earlier work.
It's a bit like DRM. The average user doesn't care about DRM, but they hate it when their brand new HDTV goes black for some reason, or when they can't skip the previews on a DVD, or when they have to buy the same damned thing in 20 different formats for no reason.
It's not that users don't care about technical debates. What's really going on is, users do care, and do get pissed off when things don't work. But they don't want to know why something doesn't work, they just want it fixed. That's why we see people who don't maintain their computers at all, get loaded up with spyware, and then just buy a new one or pay GeekSquad to run Spybot.
It's a bit like never changing the oil in your car, and constantly raving that you shouldn't have to change the oil, and buying a new car when your old engine dies because you refused to learn anything about it.
So it's not that users don't care, it's that users desperately want to avoid ever having to think about something that isn't their area of expertise.
Personally, I think users should care about the format war if they care about the consequences of the format war, and most do.
That might work, and I think some filesystems might have implemented it. It's also closer to working with existing POSIX stuff than what I might've come up with. But it also requires the app to do quite a lot that I don't think it has to, and requires some duplication of code.
For the simple, small-file example, it works reasonably well. Write to the new file, barrier, rename to the old, barrier. Essentially have barrier calls in most places you'd have sync calls -- and yes, maybe fbarrier or fdatabarrier, like fsync/fdatasync. Note that sync-ing already does basically the same thing, just slower. And you might also want to have a call which explicitly says you're going to be using barriers, telling the FS it can reorder as much as it wants.
There are a couple of problems with that approach, I think. The most obvious one is that essentially, your application is trying to implement a transaction, and your filesystem will, if it's smart, group these into a transaction. In other words, if I read from file a, write a new version to file tmp, barrier, and then rename that tmp to a, then barrier again, all before the FS writes it out to disk, a smart enough FS might figure out that I want to atomically update a, and skip the tempfile altogether. Skipping the tmpfile means skipping dealing with allocating a new inode, setting a dozen timestamps, then unlinking it, when it might be faster to simply use the FS's own journal. (Or it might use the tmpfile anyway, if it thinks that's faster.)
At the same time, your application, if it's more than just vim, might have more than one file it wants updated atomically. For instance, it might want to go through a whole directory tree, doing some operation to, say, half the files in the tree. This starts to make it more complicated. The simplest way I can think of with barriers would be something like: Copy everything to a temporary tree, do all of your update operations there. Barrier. Rename the original tree to something like orig.bak, and move the new tree to the original tree. Barrier.
The above may not always work, of course. It ignores open filehandles or working directories within the tree by other processes. It also means that, even if you use hardlink tricks, you're probably copying far more than you need to.
And consider a database, or any large file that you might update random chunks in the middle of. For that to work, you're either going to need to copy the database around a LOT -- not feasible if it's a few gigs in size -- or you're going to have to maintain some sort of journal yourself.
So, again -- the FS already has a journal and a concept of an atomic operation. Your app is trying to implement an atomic operation. It seems to me that you should communicate via atomic operations. It also doesn't even seem that hard -- consider that ZFS supports writable copy-on-write snapshots of the entire filesystem. If I were implementing a filesystem, that's about how I'd do it -- provide a new API that resembles POSIX (and maybe is backwards-compatible), but allows each operation to be tagged as belonging to a certain transaction. It would look like this (excuse the pseudo-C code, I don't do a lot of C):
foo = begin_transaction(); .... foo);
// or you could do
write(some_file, "some data", 9, foo);
write(some_other_file,
rename ('a', 'b', foo);
unlink ('c', foo);
commit_transaction(foo);
sync_transaction(foo);
Roughly, of course. The idea is, anything read or written in the context of that transactio
I would say, be aware of where you're re-using passwords. The reason not to re-use a password is to prevent a compromise of one account on one site leading to a compromise of another account on another site -- and that compromise may come from inside.
For example, I really don't give a damn if MySpace can get into my free New York Times account, but that's basically what using the same password on both implies -- if someone 0wns MySpace, or MySpace itself becomes corrupt, they can get my password and use it elsewhere.
But, for example, my passwords to my bank account, or to various other things that I actually care about, are entirely unique. And most of my access to things I care about (remote servers that I admin) is done with SSH keys instead of passwords.
The problem with power is, if something accidentally gets stuck on that button, hold it down for 5 seconds and the machine is OFF. I know you can intercept a single press/release, but if you press and hold, on most boxes, it will eventually shut off no matter what you're doing.
On the last laptop I had Linux on, I played with the lid-close event. That was a bit more useful...
Only problem: AV is based on the assumption that we know what a virus looks like. There are enough false positives that the heuristics can't be working well, and the very existence of a signature means someone must've been infected already.
AV is a bit like the rifle, because it's the last line of defense, and a pretty damned weak one. I'd say anytime your AV hits, if you didn't see it coming with that particular file, you're doing something wrong.
I know it's not going to fix anything, but there are a few simple, simple steps:
This is common sense stuff. Some of it is a bit tinfoil-hat (SELinux, secure hardware), but really, most of the above can be done very cheaply, and in the long run, won't take any significant amount of time or brainpower to maintain.
And though I've never been a cracker, it still pisses me off when, instead of responding by paying attention to common-sense security (as I've just described), they'll attempt to buy a magic bullet -- they'll buy ONE product, probably something standard like Windows Defender, and then get lazy again. Or sometimes they'll try litigation, or both:
To my knowledge, "piracy" has two definitions:
I really, really wish the people writing these stories would bother to at least try to get the jargon right. After all, there's no mention of the word "keylogger", or the fact that it's a really fucking obvious and common attack. I bet they either thought or are trying to pretend that this kind of thing has never happened before...
Yes there is.
That is, unless they don't know what the word "spyware" means. Being reporters, they might just assume that spyware means what it sounds like -- any software used to spy on you, including something picking up keystrokes from a physical keylogger.
But then, it also seems like it would be difficult to make a physical keylogger that communicates reliably with the outside world:
That sort of implies it's being done in realtime. Of course, they could always mean it was a physical keylogger, which the "hacker" then collected and dumped...
Then again, it's a laptop. If you have physical access to a laptop for long enough and with enough tools to install a physical keylogger, it's probably easier to carry the thing off and hope there's something valuable on the hard drive.
First, to those who made comments about 128k encoding, you may be thinking of mp3. (Or maybe not, who knows...) From what I've heard, AAC, Vorbis, and AC3 all sound better than mp3 at similar bitrates.
Second, I remember there was a comment on Slashdot awhile back, before they actually came out with these, and I want to confirm... Apparently, CDs are recorded at a certain physical bitrate/frequency, and there are digital masters which are at a higher rate... it's late, so I'm not entirely coherent, but think of it as somewhat equivalent to resolution of a DVD (quality of video is proportional to resolution (HD vs normal) and bitrate). The point was that 256k may actually sound better than a CD, since it comes from a better source than a CD.
If so, this whole test is BS, since they did not do a comparison of CD vs 128k (either iTunes-DRM'd or custom-ripped) vs 256k (un-DRM'd, from the iTunes store). Specifically, I'd want to hear 256k vs CD. But at the same time, I don't know if any iPod, or specifically the one they are using, would be able to handle the higher resolution. If not, you'd have to specifically check your soundcard, too.
And finally, again vaguely remembering this from a Slashdot comment (so correct me if I'm wrong), but there was some comment about "The 30c may seem small, but imagine buying a whole album at these prices..." And I seem to remember that a full album is always $9.99. Still high compared to, for example, the minimum you'd pay for a FLAC-encoded album at MagnaTune, but if you're buying a whole album (and if that's accurate), you may as well just opt for un-DRM'd -- especially if it sounds better than a CD (which would probably cost more anyway.)
But then, of course, I'd like to hear a much bigger study, with more rigorous controls in place. 10 people is just not enough, no matter how you set it up.
And personally, if I had any money to spend on music, I'd be buying un-DRM'd stuff. But probably not from iTunes -- not till there's a web interface (or at least an API) that doesn't require me to download their client software. After all, if I'm buying a single file, the point of the client software is to implement the DRM, and if I buy the un-DRM'd version... Not that it shouldn't also work in the iTunes client, but it'd be nice for it to work natively in AmaroK, or just in a browser.
Agreed. Did I ever say that? Can you quote me where?
I'm just pointing out that I don't think firewalls/AV are things you should count when talking about how much more you can do with a modern computer.
Well, no. There are a few entry-level Linux users who would be greatly appreciative if I just put an antivirus logo in the system tray, because unless I do that, they'll somehow delude themselves into thinking Linux is less secure.
Another debate for another time, but my mother seems pretty comfortable whenever I put her on Linux. In fact, the most common problems I ever see anyone having with Linux these days are missing software. Most of the other complaints (having to edit config files, everything being complicated/weird/different for no reason, installation being hard/risky) are pretty much gone these days. (Even if installation IS hard, you can always buy Ubuntu on a Dell.)
I'm sure that has something to do with our discussion about firewalls and AV, I just don't see it.
If something goes wrong with ANY install, entry-level users cannot fix it. It has nothing whatsoever to do with Linux.
Yes, and yes. Although I would say that in such a corporation, you probably want the big corporate firewall sitting on the gateway, but that's as much to keep users in as to keep "hackers" out. For example, you don't want anyone at your company to start spewing random spam out to the Internet (intentionally or not), so blocking port 25 outbound seems fair.
But I'd argue that even the big-corporate-gateway-firewall is more often than not entirely about making the Windows boxes behind the firewall secure, and/or a need for NAT. The only other thing I can possibly see it preventing is, say, a ping flood, but all that does is make it lag your Internet connection instead of your internal network.
But AV? You're either clueless or joking.
For the four billionth time, *nix is pretty fucking popular on servers. According to netcraft, Apache is still roughly twice as popular as IIS (guessing from the graft). I can't find the ratio of Windows to Linux servers, but I seem to remember it was something like less than 50% Windows, and roughly 30% Linux.
Don't you think, if it was so easy to hack, that someone would've already done so? Written a virus, built a botnet, but instead of using tons of home users overloaded with spyware, and with barely any bandwidth or performance... Don't you think they'd go after a dual-core or quad-core server with at least 10-100mbits of connectivity?
In other words, don't you think something like Slammer would've happened to Linux already, if it was just a popularity contest?
Linux is not more secure merely because it's obscure -- that's Microsoft's game, remember? Linux is more secure because of better design and a better development process, at least with regards to security.
They're found, frequently. They're just found and patched ver
I could say the same, but I'm a bit more persistent than that.
Well, I suppose it all depends how you work. I absolutely NEED at least a gig or two of RAM for games...
I do a bit of "light coding" myself, but I tend to use vim, or, worst case, Kate, for the actual text editing. I run Konqueror as a web browser. And I haven't touched Java since college.
I would say, if you can find a way to lose Eclipse, you'll have half your RAM problems solved right there.
Also, I'd suggest you read the rest of my post, given that wasn't even a main point. The main point is that background processes are generally irrelevant to most people, because they'll spend most of their time sitting in swap.
That is true.
In Gentoo, I believe you can depend on specific USE flags being set. Also, autoconf and friends should catch things like this...
The right way to do this, I think, would be to fork off KSSL into its own library, so it can be compiled individually, and depended on individually, and not installed pointlessly with the rest of KDElibs just so I can run, say, AmaroK. The fact that it doesn't do this is half the reason Gentoo exists: huge packages like KDElibs which can only have certain support/dependencies added or removed at compile-time. That's the whole point of USE flags.
I like the Debian way of dealing with this, though, when it works: Split one source package into multiple binary packages, whenever possible, so that I can install the X libs without any kind of X server, for example. (I think this is possible on Gentoo now, but it took awhile, and basically didn't happen until X.org.)
But that still makes it a pain in the ass for developers.
So in summary, it's your fault for not depending on SSL support (however Gentoo does that these days), KDE's fault for even having one package called "KDElibs" that includes so much crap, and maybe the users' fault for actually disabling SSL in their USE flags. I wouldn't say it's actually Gentoo's fault, except maybe for not providing easier ways to resolve these problems, especially if they are the ones repackaging your software.
Depends what you consider "new".
There's a lot of people who seem to think that in order to be "new" or "original", a game must have entirely new and original "gameplay", and then go debate about what that would be. Most of these people wouldn't argue about a few games -- DDR and Frequency/Amplitude were very much new, both in control style and actual gameplay.
First, new ground isn't necessary. I just got a new DDR game (having lost my old ones), and I haven't even gotten into the actually new features, like EyeToy and internet gameplay, but it already kicks ass. The gameplay is exactly the same as it always was, but there are a lot of good new songs, and the animated, dancing 3D chick is hotter and less pixelated than before, and also apparently no longer cell shaded.
I wouldn't have bought it if I didn't lose the old one, of course, because I don't think these things are worth shelling out $20-50 for some new songs. But that's all in the eye of the beholder, really. Half-Life 2: Episode 1 didn't really introduce any shockingly new gameplay since Half-Life 2, not even much in the way of new monsters. Alyx is hot as ever, but despite the hype, she's not all that much smarter than before. But I still think it's worth it, because the story was interesting enough, and there's nothing wrong with the gameplay.
For that matter, consider Counter-Strike. Really, nothing has changed since the original CS was released (free for download if you had Half-Life), yet people still play it every day. Or World of Warcraft -- are they really covering any new ground there?
I would say that FPS is just the interface, the medium, and the question you are asking is more like "Assuming it's just a video game, is there any new territory to cover?" There's always room for the truly odd and different, even if it's still technically an FPS. Consider Deus Ex, or Natural Selection. Even Cube.
Computers from 1986 don't need that. In addition, secure OSes don't need that -- I'm on Linux, I have neither, and I've never been 0wned.
Where?
Everything depends on resolution. The only "resolution independence" that we really have and use today is in apps which can be in resizable windows, which I'm pretty sure that Mac Plus had.
I'd just like to take the opportunity to point out, again, how much Flash sucks. Take any video on YouTube, download it (may need a Firefox extension), and view it with VLC. On my box, it was 50% CPU at least in the browser, even postage-stamp-sized, and less than 1% in VLC, even fullscreen.I think the point isn't that we should go back. The point was that, if you're not excited about higher resolution and more audio/video, there's not much reason to be excited about computing anymore. I'm not saying I agree with that, but let's see -- you mentioned firewalls. Great, but you can't go online with that '86 machine, so who cares? The only one you mention that makes any sense is spellchecking, which can be done tolerably well on any five-year-old machine, even realtime.
Maybe it is, but I honestly can't tell. Just took roughly one second to launch an instance of vim, with color-coding and everything, on a small python script. Quit and re-opened, and it's faster than I can see.