Google found me through the Internet. I don't believe I was made to sign an NDA about that email, so I can say that much.
I guess they found some of my old rants on the kernel mailing list, when I was too young to know better. (And yes, it was with this email address.) Nothing came of it, but it was still a pleasant surprise.
I can't really find a good thread to attach this to, but it has to be said -- this isn't just about "turning the tables" and scoping out management. If you're going to screen out negative candidates, also screen for positive ones.
At least, some of the things I've said on Slashdot have been moderated insightful, interesting, and informative. Maybe they actually are. I'm really not sure what kind of dirt you could dig up on me (and David is a common enough name), but I intend to counteract that not by hiding, but by putting out more good stuff.
If the highest ranked page on a Google search of me is my Github account, I'm doing something right.
The incredible mish-mash of CSS/HTML/Javascript/ASP.Net/AJAX/JQuery can do it sure. But it is a royal pain in the ass.
Sounds like a lot of whining.
Using a decent design, and unobtrusive javascript via jQuery, it's really not that hard. And you're in a controlled environment -- why force IE and Silverlight, instead of, say, Firefox 3 and standards compliance -- probably automatically working on Opera, and likely everywhere except IE?
By switching to Silverlight we gain all of the UI features (and more) in a single language that is in the core-competency of the development team.
That has its advantages, sure. But while you're at it, might I suggest another switch -- use CouchDB, instead of SQL Server. Then you can write your map/reduce functions in.NET, using a JSON parsing library.
Of course, SQL is useful, as a domain-specific language, partly because of the queries you can do, but mostly because of the software you're running. CouchDB is new and unproven; there are many existing SQL servers, and while it's not necessarily easy to port, at least SQL Server isn't completely alien once you've used MySQL, or vice versa.
I would argue that many of the same properties apply to JavaScript and HTML. HTML is a domain-specific language for displaying content. CSS is one for styling content -- and it's clearly better at that than JavaScript is, since you have JS to do that anyway.
Using either Silverlight or Flash in inappropriate situations is dumb.
I would argue that any case where you could use an open standard to do the same job, with little or no ill effects, it's dumb to use Silverlight or Flash.
But the existence of Silverlight and the competition it creates for Flash will truly only improve the functionality of rich browser applications (what I've been trying to get coined as 'Chubby clients').
Certainly. But I hope it does so not directly, but by putting pressure on people interested in HTML5 and other interesting things to get it done.
For example, the existence of Flash helped bring about YouTube, which now forces us to think about the video tag. But there's no way I'd prefer maintaining Flash to allow YouTube to work into posterity, when for that purpose, the video tag and simple AJAX is easily superior and likely more compatible in the long run, as it doesn't depend on the way Adobe behaves.
Silverlight really doesn't change that much. The simplest example is patents. There really aren't any patents likely to hit a video tag implementation which provides Vorbis and Theora, and once you have that, it's reasonably easy to separate codecs and distribute them to people who either have permission, or live in a place where US patents don't apply, or simply want to take the risk.
Worst case, you limit yourself to sites supporting Vorbis and Theora. If the Video tag sees wide adoption, I would imagine those codecs might start to -- after all, patent royalties could be a large-ish expense for a site which has to do any sort of video encoding.
However, with Silverlight, it's quite possible Microsoft could drop the patent hammer and kill Mono and Moonlight at any moment. Failing that, since Silverlight is the canonical implementation, and there is no standard, and Microsoft is known not to follow standards (even their own standards), it then follows that Microsoft can always release a new version of Silverlight that is incompatible with Moonlight, and not provide the needed expertise to bring Moonlight (or Mono) up to the task.
Consider that Microsoft's current motivation for not doing these things, and for actually supporting Moonlight, is to position Silverlight as a Flash killer, while claiming to be cross-platform. It certainly looks attractive -- it's probably more cross-platform than Flash is, at the moment. But that's only because they need it to be -- there's no reason to assume it won't go the way of IE for the Mac (or IE for Unix) as soon as Silverlight dominates the marketplace.
I've tried at length, and Linux (or OSX, for that matter) don't offer anything comparing to the ease-of-use and efficiency of running a tablet PC in Vista with OneNote for academic settings.
Fair enough. I prefer typing, but that's not really an excuse if the handwriting recognition isn't working.
Yes, Linux is a great OS, but it simply doesn't have photoshop or anything that compares to it. GIMP is a clumsy hack and is frankly like Paint in comparison.
At the same time, last I checked, photoshop ran faster on Windows (64-bit support), and does work under wine. And if the Gimp is clumsy compared to Photoshop, it is still far ahead of paint -- and it's not the only option.
Gnome, KDE and Explorer have nothing on the frankly revolutionary changes Mac has seamlessly implemented in the last few years. There are a lot of poorly implemented whizbang features like Time Machine's GUI or Safari 4's Top Pages, but there are also features like Spotlight, Expose, the new stacks in the Dock, and Quick Look,
As has been pointed out elsewhere, these are all implemented on Linux.
And, for that matter, take a compositing GUI -- not only is it there, but it can be toggled on and off easily when performance is needed (there's actually a widget, in the Dashboard sense, that does this) -- and it can operate on far less hardware than Vista, and far less than is included with a modern OS X machine.
Call me back when Linux works with my hardware out of the box
Call me back when OS X works with my hardware out of the box.
What's that, you say? Buy a Mac? Fine, then you can buy hardware that works with Linux -- or even a machine that comes with it preloaded. Dell sells them now.
I've tried it on five laptops and two desktops in the last couple years, most of those very recently, and it never Just Works
I have never found a single OS that Just Works on all laptops. I have found that Linux does more frequently than XP. I haven't tried Vista much.
But again, you generally don't reinstall the OS on your laptop by yourself, and then expect everything to work with no tinkering. You buy a laptop with the OS you want preinstalled, and if it doesn't work out of the box, you call the manufacturer.
To suggest otherwise isn't a comment on the relative merits of each OS, it's to set up a stacked playing field.
Call me back when Linux or Windows have system-wide drag-and-drop that lets me drag an image off a webpage or into an chat window
Just dragged an image from Konqueror to a folder. What else you got?
or from my desktop into the Mail icon to start a new mail with an attachment
Well, you've got me there. However, once I do have a message I'm composing, I can drag it to that message to add it as an attachment.
or from an email to a filesystem icon which pops open, lets me browse my hard drive by hovering and dropping where I want, and then goes away.
Close. That might be a worthwhile suggestion to make to the KDE team -- however, I can drag it to any open filesystem window, and get the same "move here / copy here" prompt I get for dragging files anywhere else.
And really, these seem like minor things compared to my own gripes. For example: How, in OS X or on Windows, can I bind a keystroke to things like "pack right"? I can hold a key, and then toss the active window against the edge of a screen with my arrow keys -- or against the next window.
How about focus-follows-mouse? Or sloppy focus? Why must I click on everything on OS X and Windows?
OS X finally got "spaces", and they're still well behind what most Linux WMs provide. The only real advantage might have been the spiffy animations, which Compiz already did when Leopard came out.
In my experience, I've found that the troll you're replyi
I found the poison zombies and headcrabs to be more something that kept me on edge, gave me a shot of adrenaline... Scary, yes, but I could always dodge the headcrabs, whether thrown or not. I hear them screech, I move, they fly past me -- I can do that all day.
Consider the first encounter -- that screech is like nothing you've heard so far, in a game full of weird sounds. You see them coming over the rooftops, then you see the pipes start moving... Finally they pop out and you have less than a second to fill their face with lead if you want to live (and be ready for the next one).
Unlike most other creatures in the game, you really don't have an opportunity to outsmart it, or run, or dodge. There is no amount of superior play that will save you, even if it wasn't for the fact that you're stuck on a rooftop with nowhere to go but down.
So, claustrophobic, for many reasons. Poison zombies, and headcrabs, I can always just run, until I get a better angle, or let them crawl slowly towards me so I can time my crowbar perfectly. Switching to a different genre, even the Flood in Halo, which are physically similar, are still possible to dodge. But a Fast Zombie, there's no escape from.
Oh, and they look (and sound) somewhat like Reavers, which are still scary.
I guess it's a personal thing. If you're a crack shot, but not great at dodging, you're probably terrified of the Guardians (Antlion bosses). And you're right, poison headcrabs are pretty scary -- first encounter with that is "I'm at 1 health -- what the fuck just hit me?!"
Oh well. Scariest creature in the game is Dog. Let's just leave it at that, and be glad he's on our side.
I would always save ammo "just in case" so I'd never end up using the fun guns all that much....
I had to make a point to -- pace myself, and then say "Ok, it's time for something to explode!" and toss a grenade. The situation didn't really call for it, but I'd know there would be another dozen grenades I wouldn't need somewhere...
I agree -- Doom 3 did a lot of things that people tend to forget (including me) when discussing how it wasn't scary. After all, while a little duct tape would've helped, the flashlight/weapon trick was very well done, the sound was brilliant...
But, it did have that problem -- it became a little too predictable, to where it was almost cliche'd where stuff would pop out of. If you creep along slowly, it's not as scary as it could be, because you're expecting stuff to pop out everywhere. If you charge in like the timedemo shows, it's not scary, because you may as well be playing a deathmatch -- too much adrenaline, too much going on, before long it's just a blur of monsters.
But at least a few of these, like the lighting... It's nice, these days, to have a horror game that doesn't need to make everything dark, but then, the moments traveling with the scientist in an area with frequent blackouts...
I don't know, the ready availability of things like saw blades made it possible to kill all of them.
But it's true -- it was in the setting, and the sounds, and the Fast Zombie screech. The environment was also, in many ways, set up to feel claustrophobic -- even moreso than crawling around vents in the original Half-Life. Some of this was the actual, physical environment -- having to sit in a corner and wait for the zombies to walk into your trap.
And some of it was the fact that even when you had a moderately open area, you would be surrounded by zombies, herding you towards the next obstacle.
So, ammo was part of it, but I don't think it was a large part. I had a friend complain that Half-Life 1 was mostly running around with no ammo -- HL2 tends to have enough, as long as you don't do anything stupid.
'P2P file sharing is designed to cause network congestion,' says the company.
Yes! Clearly, when designing a P2P protocol, my first concern was to make absolutely sure that your network would be congested, because I hate the Internet!
This isn't all about you, ISPs. It's about us, and what we want to use our bandwidth for. And yes, P2P filesharing does have design goals other than clogging your tubes.
But if you consider what the post is actually saying, this is not a cave on amazon's part. They're actually holding their ground.
I don't see how.
This basically says that if a publisher or author decides they don't want TTS on kindle, then they have to make that decision on a choice by choice basis.
Yes, I understand. However, this is also something that Amazon should have some pretty firm legal basis for saying "Fuck you, we're doing TTS anyway."
What happened is, instead of standing that ground, they caved, figuring it was cheaper than a lawsuit, even if they win. The only thing they've "stood their ground" on is not abandoning TTS on all titles, which I don't think anyone was demanding in the first place.
I've gotta give props to that move.
For standing firm in the face of really no opposition! Clearly they deserve mad props!
If it wasn't for viruses, there wouldn't be antivirus vendors. If it wasn't for online scam artists, there wouldn't be any cyber-crimes divisions in law enforcement.
Broken-window economics don't work, of course. If people weren't employed in these professions, they'd be doing something else.
If the government really wanted to shut down SPAM, it can easily do it by making up bullshit laws and detaining people indefinitely.
How would you like to spend a year creating a document, and then your boss decides to take the document without paying you?
I wouldn't. Just how do you think DRM would protect you from this?
In fact, why would you need it? Just sue your boss for the past year of wages. Not that it was very smart of you to work for free for a year...
Oh, but I forgot -- this is a bad analogy for something completely different.
In essence that's what happens to authors every time someone takes a book. It's stolen labor.
Ah, yes, because every time I take a book, I'm really forcing you to work for a year without getting paid. It's totally the same thing.
In fact, maybe this is even true if I don't actually take it -- if I just copy it, that's the same thing, right? Because taking a photograph of you is just like kidnapping you, right?
What I object to is when a product stops working. Like when Walmart turned-off their DRM music servers. In that case consumers should have a right to demand a refund since the product is no longer functioning as advertised. Billion-dollar walmart can certainly afford it.
Ah, so ripping people off is ok, so long as they have billions of dollars? Good to know.
In the case of Wall-Mart, it does make sense that they should either do that, or open up the music. But how, then, should a publisher implement DRM? If they assume that when they shut down the DRM store, they must provide a full refund for everything sold, ever, then no one would ever start a DRM store in the first place. If they assume that they'll have to open up the DRM, then people like you would never do business with them, because clearly, the second the DRM servers shut down, everyone will start swapping your files like it's 1999.
Look, I'd have more sympathy if DRM ever actually worked -- though even then, it's still a massive inconvenience for legitimate consumers. As it is, DRM has driven me to piracy more than any other issue, including money. If I can't buy and download your book in a format I can easily read on Linux -- like, say, PDF -- I will pirate and download that book in such a format, or I will not read it at all. Either way, you're out of a sale.
It's not that I'm cheap. I have bought music through channels where I can get it for free, especially lossless. I've got my mother using Amazon MP3, for example. But we're not going to pay for a crippled product, just so you can pretend it won't get pirated. In fact, I, for one, am going to do everything in my power to ensure that people like you make less money, the more DRM you use -- and not just because you're spending money on snake-oil to some little company that pretends to know crypto.
It's not a complete DRM break, but it is significant -- especially considering that it means the only way to get the version you want is to download it.
This means that if Amazon ever successfully forces it to get to that stage, as some forms of DRM are at the moment (Blu-Ray), you're also forcing anyone who wants a bit more freedom -- or a feature you've disabled -- to pirate the media in order to get it, whether or not they were ever a legitimate customer.
The reason why is simple: those TCL/TK-and-then-call-a-commandline-tool apps never, never, never deal with error conditions correctly. In other words,
In other words, they're no worse than a GUI app with a library which doesn't check for error conditions either. In any case, it's not the methodology you hate, it's the implementation.
when the shit hits the fan, as it invariably does, you must still go out and delve into the commandline.
All things considered, that is better than having to delve into an API.
So just Don't Do It, Ok?
If that's referring to my subject, I wasn't indicating this as the only way to tie things together -- at least, not intentionally. An obvious example is Erlang -- it tends to avoid allowing anything to link directly into a running Erlang VM, and rather exposes a C library that allows other languages to send messages to a running Erlang program. Thus, you could build a bunch of small tools which communicate with each other over local Erlang sockets...
But, consider that Erlang is actively implemented in telephone switches which are not allowed to have downtime, even for upgrades. It is doubtful the Erlang community isn't going to handle errors correctly.
Ultimately, I suppose they could screw it up, just as I could ignore the exit code for a commandline app. In each case, it's programmer error -- I don't really see anything to make the environment inherently fragile.
Consider that any protocol you use will have to have an implementation in every language you're using, otherwise there's no point. That means that to tweak the protocol, you need to tweak the same code in each language, unless you use a shared library -- say, a C library that each language now has bindings to. But if you do that, you still have some of the same problem (maintaining the bindings), and a lot less of the point -- why have a protocol at all, instead of just a common API?
For that reason, I'd avoid inventing your own protocol from scratch, unless you have the hubris to think there's something wrong with all the existing ones, and that you'll do better. (I know I do, but for now, I can't really find too much wrong with HTTP and things like JSON or YAML for most purposes.) Stick to things that have common libraries, making them easy to interface with from any language. And try to stick to things that are simple enough that you could write your own library... not that you want to.
Your "Unix Way" is a wheel that's being reinvented as SOA, etc.
Here's the thing: It is possible for one language to be good enough for nearly everything, especially if you pick one with good support for internal DSLs (I like Ruby). Also, while message-passing is a good idea, it's usually slow, and you probably don't want to be designing your own text-based format every time.
Now, you're still going to have DSLs and whole other languages forced on you, occasionally. For example, JavaScript is still the best language for AJAX clients, simply because no one has written a better language that compiles to JavaScript. (That's relative, of course -- if you like Java, then Google Web Toolkit will be perfect.) In fact, with web development, you'll want JavaScript, HTML, CSS, and probably another language (Ruby, Perl, Python, PHP, etc), and SQL, all in the same application.
But, each additional language is that much more for a newcomer to learn, and it's that much more glue. If you communicate with text, how much time are you spending writing text parsers instead of your app?
Of course, ideally, you provide a language-agnostic API, because you may need this application to interact with others. You might even find yourself writing multiple applications...
But the other big win of a huge application is the UI. The Unix commandline has made mashups of many small programs as easy as a pipe character. There's really no equivalent for the GUI -- users will relate better to one big monolith, even if it's just a frontend for a bunch of small tools.
So, I would split application by the UI concept, and share the small, common utilities via shared libraries. That's not far off from the Unix Way, either -- it's not hard to write a small commandline app with a shared library, if you find you need it. It can be annoyingly difficult to go the other way -- for example, Git bindings aren't as easy as they should be.
At the same time, if you find you're prone to that, contributing to an open source project may help avoid it.
I know that for me, it isn't about the money -- at least, not yet. Instead, it's about feeling useful, about actually doing something with your life other than vegetate in front of the TV -- or Slashdot. Speaking of which, it's about time for me to get useful.
Except that contributions to open source projects are both good resume-fodder and a good way to get noticed by potential employers.
I know a job search is hard work, but I'm not sure it's going to take as much time as a fulltime job... meaning that much more time to play with Xorg, or whatever else.
This is an XPS M1530. All Half-Life 2 offspring seem to run decently at 1920x1200, with quite a lot of the eye-candy enabled -- but only on Windows.
This really isn't surprising, by the way. Not only are Half-Life 1 games almost by necessity lower detail -- in poly count, textures, shaders, pretty much any way you measure it -- they also use OpenGL by default, whereas the Source engine (Half-Life 2) supports Direct3D only. I know Wine's Direct3D layer is good, but it's still not as good as native, any more than Microsoft's OpenGL-on-Direct3D implementation for Vista's compositing is as good as native.
It would mean having to dual boot, and that is, to me, not acceptable. I'm weird that way.
Weird indeed. Would you care to elaborate?
Unless it includes "not having a Windows license", I really don't see why. And certainly, the amount of effort it takes to get some games working on Linux is not worth it, if you make a decent wage, versus the cost of an XP license to run dual-boot (or in a VM, for older games).
Perhaps NVidia could get off their asses and develop a decent driver for Linux,
That is part of the problem. The other part of it is lack of Direct3D support on Linux, and pervasive use of Direct3D in games. While I don't doubt that nVidia could port it, I doubt very much that Microsoft would let them.
But I have to choose between Windows and playing all games.
I don't.
All I really have to do is be careful which MMOs and casual games I pick up. If I'm playing a game in a window, I'd like my IM working, and I wouldn't mind email, too -- and those are two things I don't do on Windows.
But, if I'm playing something like Natural Selection or a Half Life episode, I'm not likely to want any distractions outside of the game itself.
And by the way: Yes, they work on Linux, at a lower framerate and considerably more hassle than occasionally booting Windows. It's not just about framerate, either -- on Windows, I can play them fullscreen, 1920x1200, with most of the effects enabled. And that's on a laptop.
Oh, and there's the 64-bit problem. Sure, I'm on XP now, but if I ever do upgrade to a decent 64-bit Windows, quite a few games support it now. Wine doesn't, and it doesn't look like it's coming anytime soon.
Now, stuff that really works well on Wine, I'll play. For example, Warcraft 3 installed easily, runs fullscreen with all effects enabled, with no noticeable disadvantage to playing on Windows.
Partly because some of the nice things about Chrome (Webkit) aren't likely to ever be integrated into Firefox. At the very least, they would likely break plenty of existing extensions -- not even all Greasemonkey scripts would necessarily work out of the box if you wrote a Greasemonkey-like plugin for Chrome.
In Halo matches, you'll generally see people walk up to your corpse (after killing you) and tap crouch repeatedly for a few seconds until you fade to black and respawn.
That is, at least originally, intended to simulate teabagging.
The point being: This shit happens when you take it online. The way I see it, when thousands of 13-year-old boys can run around calling each other "faggot" in high-pitched voices, being able to say "I am a lesbian", even if it's not in a respectful tone, still isn't going to be as bad as what's already on there.
By not taking every opportunity to lie, cheat, and steal from my friends and family, I am at a disadvantage relative to those who do. By paying my taxes instead of hiding my income, I am at a disadvantage to those with, say, large Swiss bank accounts.
That is not an excuse, or a reason not to blame people. There is a reason we consider some things immoral, and that reason has nothing to do with what anyone else is doing.
If it's really "not that bad", that's another issue. But you don't get to say "At least we're not as bad as these guys." If "these guys" are Monsanto, that's not setting the bar very high.
Slashdot Mirror
WHY?!
Google found me through the Internet. I don't believe I was made to sign an NDA about that email, so I can say that much.
I guess they found some of my old rants on the kernel mailing list, when I was too young to know better. (And yes, it was with this email address.) Nothing came of it, but it was still a pleasant surprise.
I can't really find a good thread to attach this to, but it has to be said -- this isn't just about "turning the tables" and scoping out management. If you're going to screen out negative candidates, also screen for positive ones.
At least, some of the things I've said on Slashdot have been moderated insightful, interesting, and informative. Maybe they actually are. I'm really not sure what kind of dirt you could dig up on me (and David is a common enough name), but I intend to counteract that not by hiding, but by putting out more good stuff.
If the highest ranked page on a Google search of me is my Github account, I'm doing something right.
The incredible mish-mash of CSS/HTML/Javascript/ASP.Net/AJAX/JQuery can do it sure. But it is a royal pain in the ass.
Sounds like a lot of whining.
Using a decent design, and unobtrusive javascript via jQuery, it's really not that hard. And you're in a controlled environment -- why force IE and Silverlight, instead of, say, Firefox 3 and standards compliance -- probably automatically working on Opera, and likely everywhere except IE?
By switching to Silverlight we gain all of the UI features (and more) in a single language that is in the core-competency of the development team.
That has its advantages, sure. But while you're at it, might I suggest another switch -- use CouchDB, instead of SQL Server. Then you can write your map/reduce functions in .NET, using a JSON parsing library.
Of course, SQL is useful, as a domain-specific language, partly because of the queries you can do, but mostly because of the software you're running. CouchDB is new and unproven; there are many existing SQL servers, and while it's not necessarily easy to port, at least SQL Server isn't completely alien once you've used MySQL, or vice versa.
I would argue that many of the same properties apply to JavaScript and HTML. HTML is a domain-specific language for displaying content. CSS is one for styling content -- and it's clearly better at that than JavaScript is, since you have JS to do that anyway.
Using either Silverlight or Flash in inappropriate situations is dumb.
I would argue that any case where you could use an open standard to do the same job, with little or no ill effects, it's dumb to use Silverlight or Flash.
But the existence of Silverlight and the competition it creates for Flash will truly only improve the functionality of rich browser applications (what I've been trying to get coined as 'Chubby clients').
Certainly. But I hope it does so not directly, but by putting pressure on people interested in HTML5 and other interesting things to get it done.
For example, the existence of Flash helped bring about YouTube, which now forces us to think about the video tag. But there's no way I'd prefer maintaining Flash to allow YouTube to work into posterity, when for that purpose, the video tag and simple AJAX is easily superior and likely more compatible in the long run, as it doesn't depend on the way Adobe behaves.
Silverlight really doesn't change that much. The simplest example is patents. There really aren't any patents likely to hit a video tag implementation which provides Vorbis and Theora, and once you have that, it's reasonably easy to separate codecs and distribute them to people who either have permission, or live in a place where US patents don't apply, or simply want to take the risk.
Worst case, you limit yourself to sites supporting Vorbis and Theora. If the Video tag sees wide adoption, I would imagine those codecs might start to -- after all, patent royalties could be a large-ish expense for a site which has to do any sort of video encoding.
However, with Silverlight, it's quite possible Microsoft could drop the patent hammer and kill Mono and Moonlight at any moment. Failing that, since Silverlight is the canonical implementation, and there is no standard, and Microsoft is known not to follow standards (even their own standards), it then follows that Microsoft can always release a new version of Silverlight that is incompatible with Moonlight, and not provide the needed expertise to bring Moonlight (or Mono) up to the task.
Consider that Microsoft's current motivation for not doing these things, and for actually supporting Moonlight, is to position Silverlight as a Flash killer, while claiming to be cross-platform. It certainly looks attractive -- it's probably more cross-platform than Flash is, at the moment. But that's only because they need it to be -- there's no reason to assume it won't go the way of IE for the Mac (or IE for Unix) as soon as Silverlight dominates the marketplace.
I've tried at length, and Linux (or OSX, for that matter) don't offer anything comparing to the ease-of-use and efficiency of running a tablet PC in Vista with OneNote for academic settings.
Fair enough. I prefer typing, but that's not really an excuse if the handwriting recognition isn't working.
Yes, Linux is a great OS, but it simply doesn't have photoshop or anything that compares to it. GIMP is a clumsy hack and is frankly like Paint in comparison.
At the same time, last I checked, photoshop ran faster on Windows (64-bit support), and does work under wine. And if the Gimp is clumsy compared to Photoshop, it is still far ahead of paint -- and it's not the only option.
Gnome, KDE and Explorer have nothing on the frankly revolutionary changes Mac has seamlessly implemented in the last few years. There are a lot of poorly implemented whizbang features like Time Machine's GUI or Safari 4's Top Pages, but there are also features like Spotlight, Expose, the new stacks in the Dock, and Quick Look,
As has been pointed out elsewhere, these are all implemented on Linux.
And, for that matter, take a compositing GUI -- not only is it there, but it can be toggled on and off easily when performance is needed (there's actually a widget, in the Dashboard sense, that does this) -- and it can operate on far less hardware than Vista, and far less than is included with a modern OS X machine.
Call me back when Linux works with my hardware out of the box
Call me back when OS X works with my hardware out of the box.
What's that, you say? Buy a Mac? Fine, then you can buy hardware that works with Linux -- or even a machine that comes with it preloaded. Dell sells them now.
I've tried it on five laptops and two desktops in the last couple years, most of those very recently, and it never Just Works
I have never found a single OS that Just Works on all laptops. I have found that Linux does more frequently than XP. I haven't tried Vista much.
But again, you generally don't reinstall the OS on your laptop by yourself, and then expect everything to work with no tinkering. You buy a laptop with the OS you want preinstalled, and if it doesn't work out of the box, you call the manufacturer.
To suggest otherwise isn't a comment on the relative merits of each OS, it's to set up a stacked playing field.
Call me back when Linux or Windows have system-wide drag-and-drop that lets me drag an image off a webpage or into an chat window
Just dragged an image from Konqueror to a folder. What else you got?
or from my desktop into the Mail icon to start a new mail with an attachment
Well, you've got me there. However, once I do have a message I'm composing, I can drag it to that message to add it as an attachment.
or from an email to a filesystem icon which pops open, lets me browse my hard drive by hovering and dropping where I want, and then goes away.
Close. That might be a worthwhile suggestion to make to the KDE team -- however, I can drag it to any open filesystem window, and get the same "move here / copy here" prompt I get for dragging files anywhere else.
And really, these seem like minor things compared to my own gripes. For example: How, in OS X or on Windows, can I bind a keystroke to things like "pack right"? I can hold a key, and then toss the active window against the edge of a screen with my arrow keys -- or against the next window.
How about focus-follows-mouse? Or sloppy focus? Why must I click on everything on OS X and Windows?
OS X finally got "spaces", and they're still well behind what most Linux WMs provide. The only real advantage might have been the spiffy animations, which Compiz already did when Leopard came out.
In my experience, I've found that the troll you're replyi
I like it when C obliterates my foot every now and then.
Why?
I found the poison zombies and headcrabs to be more something that kept me on edge, gave me a shot of adrenaline... Scary, yes, but I could always dodge the headcrabs, whether thrown or not. I hear them screech, I move, they fly past me -- I can do that all day.
Consider the first encounter -- that screech is like nothing you've heard so far, in a game full of weird sounds. You see them coming over the rooftops, then you see the pipes start moving... Finally they pop out and you have less than a second to fill their face with lead if you want to live (and be ready for the next one).
Unlike most other creatures in the game, you really don't have an opportunity to outsmart it, or run, or dodge. There is no amount of superior play that will save you, even if it wasn't for the fact that you're stuck on a rooftop with nowhere to go but down.
So, claustrophobic, for many reasons. Poison zombies, and headcrabs, I can always just run, until I get a better angle, or let them crawl slowly towards me so I can time my crowbar perfectly. Switching to a different genre, even the Flood in Halo, which are physically similar, are still possible to dodge. But a Fast Zombie, there's no escape from.
Oh, and they look (and sound) somewhat like Reavers, which are still scary.
I guess it's a personal thing. If you're a crack shot, but not great at dodging, you're probably terrified of the Guardians (Antlion bosses). And you're right, poison headcrabs are pretty scary -- first encounter with that is "I'm at 1 health -- what the fuck just hit me?!"
Oh well. Scariest creature in the game is Dog. Let's just leave it at that, and be glad he's on our side.
I would always save ammo "just in case" so I'd never end up using the fun guns all that much....
I had to make a point to -- pace myself, and then say "Ok, it's time for something to explode!" and toss a grenade. The situation didn't really call for it, but I'd know there would be another dozen grenades I wouldn't need somewhere...
I agree -- Doom 3 did a lot of things that people tend to forget (including me) when discussing how it wasn't scary. After all, while a little duct tape would've helped, the flashlight/weapon trick was very well done, the sound was brilliant...
But, it did have that problem -- it became a little too predictable, to where it was almost cliche'd where stuff would pop out of. If you creep along slowly, it's not as scary as it could be, because you're expecting stuff to pop out everywhere. If you charge in like the timedemo shows, it's not scary, because you may as well be playing a deathmatch -- too much adrenaline, too much going on, before long it's just a blur of monsters.
But at least a few of these, like the lighting... It's nice, these days, to have a horror game that doesn't need to make everything dark, but then, the moments traveling with the scientist in an area with frequent blackouts...
I don't know, the ready availability of things like saw blades made it possible to kill all of them.
But it's true -- it was in the setting, and the sounds, and the Fast Zombie screech. The environment was also, in many ways, set up to feel claustrophobic -- even moreso than crawling around vents in the original Half-Life. Some of this was the actual, physical environment -- having to sit in a corner and wait for the zombies to walk into your trap.
And some of it was the fact that even when you had a moderately open area, you would be surrounded by zombies, herding you towards the next obstacle.
So, ammo was part of it, but I don't think it was a large part. I had a friend complain that Half-Life 1 was mostly running around with no ammo -- HL2 tends to have enough, as long as you don't do anything stupid.
'P2P file sharing is designed to cause network congestion,' says the company.
Yes! Clearly, when designing a P2P protocol, my first concern was to make absolutely sure that your network would be congested, because I hate the Internet!
This isn't all about you, ISPs. It's about us, and what we want to use our bandwidth for. And yes, P2P filesharing does have design goals other than clogging your tubes.
Hmm... Except for the part where it costs $1k for the "standard" version, or almost $500 for the "Web Server" version.
Vista Ultimate is $320, and that's retail. More like $120 more on a Dell.
So... Is it actually fast enough to justify spending hundreds or thousands of dollars on software, instead of hardware?
Or I'll just stick to Ubuntu, and spend the thousands on hardware.
But if you consider what the post is actually saying, this is not a cave on amazon's part. They're actually holding their ground.
I don't see how.
This basically says that if a publisher or author decides they don't want TTS on kindle, then they have to make that decision on a choice by choice basis.
Yes, I understand. However, this is also something that Amazon should have some pretty firm legal basis for saying "Fuck you, we're doing TTS anyway."
What happened is, instead of standing that ground, they caved, figuring it was cheaper than a lawsuit, even if they win. The only thing they've "stood their ground" on is not abandoning TTS on all titles, which I don't think anyone was demanding in the first place.
I've gotta give props to that move.
For standing firm in the face of really no opposition! Clearly they deserve mad props!
If it wasn't for viruses, there wouldn't be antivirus vendors. If it wasn't for online scam artists, there wouldn't be any cyber-crimes divisions in law enforcement.
Broken-window economics don't work, of course. If people weren't employed in these professions, they'd be doing something else.
If the government really wanted to shut down SPAM, it can easily do it by making up bullshit laws and detaining people indefinitely.
Your post advocates a
( ) technical (X) legislative ( ) market-based ( ) vigilante
approach to fighting spam. Your idea will not work....
Need I go on?
The most obvious reason is, it's too profitable for them, too hard to track them down, and there are too many countries for them to hide.
Sigh... Not this shit again...
How would you like to spend a year creating a document, and then your boss decides to take the document without paying you?
I wouldn't. Just how do you think DRM would protect you from this?
In fact, why would you need it? Just sue your boss for the past year of wages. Not that it was very smart of you to work for free for a year...
Oh, but I forgot -- this is a bad analogy for something completely different.
In essence that's what happens to authors every time someone takes a book. It's stolen labor.
Ah, yes, because every time I take a book, I'm really forcing you to work for a year without getting paid. It's totally the same thing.
In fact, maybe this is even true if I don't actually take it -- if I just copy it, that's the same thing, right? Because taking a photograph of you is just like kidnapping you, right?
What I object to is when a product stops working. Like when Walmart turned-off their DRM music servers. In that case consumers should have a right to demand a refund since the product is no longer functioning as advertised. Billion-dollar walmart can certainly afford it.
Ah, so ripping people off is ok, so long as they have billions of dollars? Good to know.
In the case of Wall-Mart, it does make sense that they should either do that, or open up the music. But how, then, should a publisher implement DRM? If they assume that when they shut down the DRM store, they must provide a full refund for everything sold, ever, then no one would ever start a DRM store in the first place. If they assume that they'll have to open up the DRM, then people like you would never do business with them, because clearly, the second the DRM servers shut down, everyone will start swapping your files like it's 1999.
Look, I'd have more sympathy if DRM ever actually worked -- though even then, it's still a massive inconvenience for legitimate consumers. As it is, DRM has driven me to piracy more than any other issue, including money. If I can't buy and download your book in a format I can easily read on Linux -- like, say, PDF -- I will pirate and download that book in such a format, or I will not read it at all. Either way, you're out of a sale.
It's not that I'm cheap. I have bought music through channels where I can get it for free, especially lossless. I've got my mother using Amazon MP3, for example. But we're not going to pay for a crippled product, just so you can pretend it won't get pirated. In fact, I, for one, am going to do everything in my power to ensure that people like you make less money, the more DRM you use -- and not just because you're spending money on snake-oil to some little company that pretends to know crypto.
It's not a complete DRM break, but it is significant -- especially considering that it means the only way to get the version you want is to download it.
This means that if Amazon ever successfully forces it to get to that stage, as some forms of DRM are at the moment (Blu-Ray), you're also forcing anyone who wants a bit more freedom -- or a feature you've disabled -- to pirate the media in order to get it, whether or not they were ever a legitimate customer.
And once they go pirate, they might not go back.
The reason why is simple: those TCL/TK-and-then-call-a-commandline-tool apps never, never, never deal with error conditions correctly. In other words,
In other words, they're no worse than a GUI app with a library which doesn't check for error conditions either. In any case, it's not the methodology you hate, it's the implementation.
when the shit hits the fan, as it invariably does, you must still go out and delve into the commandline.
All things considered, that is better than having to delve into an API.
So just Don't Do It, Ok?
If that's referring to my subject, I wasn't indicating this as the only way to tie things together -- at least, not intentionally. An obvious example is Erlang -- it tends to avoid allowing anything to link directly into a running Erlang VM, and rather exposes a C library that allows other languages to send messages to a running Erlang program. Thus, you could build a bunch of small tools which communicate with each other over local Erlang sockets...
But, consider that Erlang is actively implemented in telephone switches which are not allowed to have downtime, even for upgrades. It is doubtful the Erlang community isn't going to handle errors correctly.
Ultimately, I suppose they could screw it up, just as I could ignore the exit code for a commandline app. In each case, it's programmer error -- I don't really see anything to make the environment inherently fragile.
Your subject actually seems very insightful...
Consider that any protocol you use will have to have an implementation in every language you're using, otherwise there's no point. That means that to tweak the protocol, you need to tweak the same code in each language, unless you use a shared library -- say, a C library that each language now has bindings to. But if you do that, you still have some of the same problem (maintaining the bindings), and a lot less of the point -- why have a protocol at all, instead of just a common API?
For that reason, I'd avoid inventing your own protocol from scratch, unless you have the hubris to think there's something wrong with all the existing ones, and that you'll do better. (I know I do, but for now, I can't really find too much wrong with HTTP and things like JSON or YAML for most purposes.) Stick to things that have common libraries, making them easy to interface with from any language. And try to stick to things that are simple enough that you could write your own library... not that you want to.
Your "Unix Way" is a wheel that's being reinvented as SOA, etc.
Here's the thing: It is possible for one language to be good enough for nearly everything, especially if you pick one with good support for internal DSLs (I like Ruby). Also, while message-passing is a good idea, it's usually slow, and you probably don't want to be designing your own text-based format every time.
Now, you're still going to have DSLs and whole other languages forced on you, occasionally. For example, JavaScript is still the best language for AJAX clients, simply because no one has written a better language that compiles to JavaScript. (That's relative, of course -- if you like Java, then Google Web Toolkit will be perfect.) In fact, with web development, you'll want JavaScript, HTML, CSS, and probably another language (Ruby, Perl, Python, PHP, etc), and SQL, all in the same application.
But, each additional language is that much more for a newcomer to learn, and it's that much more glue. If you communicate with text, how much time are you spending writing text parsers instead of your app?
Of course, ideally, you provide a language-agnostic API, because you may need this application to interact with others. You might even find yourself writing multiple applications...
But the other big win of a huge application is the UI. The Unix commandline has made mashups of many small programs as easy as a pipe character. There's really no equivalent for the GUI -- users will relate better to one big monolith, even if it's just a frontend for a bunch of small tools.
So, I would split application by the UI concept, and share the small, common utilities via shared libraries. That's not far off from the Unix Way, either -- it's not hard to write a small commandline app with a shared library, if you find you need it. It can be annoyingly difficult to go the other way -- for example, Git bindings aren't as easy as they should be.
At the same time, if you find you're prone to that, contributing to an open source project may help avoid it.
I know that for me, it isn't about the money -- at least, not yet. Instead, it's about feeling useful, about actually doing something with your life other than vegetate in front of the TV -- or Slashdot. Speaking of which, it's about time for me to get useful.
Except that contributions to open source projects are both good resume-fodder and a good way to get noticed by potential employers.
I know a job search is hard work, but I'm not sure it's going to take as much time as a fulltime job... meaning that much more time to play with Xorg, or whatever else.
This is an XPS M1530. All Half-Life 2 offspring seem to run decently at 1920x1200, with quite a lot of the eye-candy enabled -- but only on Windows.
This really isn't surprising, by the way. Not only are Half-Life 1 games almost by necessity lower detail -- in poly count, textures, shaders, pretty much any way you measure it -- they also use OpenGL by default, whereas the Source engine (Half-Life 2) supports Direct3D only. I know Wine's Direct3D layer is good, but it's still not as good as native, any more than Microsoft's OpenGL-on-Direct3D implementation for Vista's compositing is as good as native.
It would mean having to dual boot, and that is, to me, not acceptable. I'm weird that way.
Weird indeed. Would you care to elaborate?
Unless it includes "not having a Windows license", I really don't see why. And certainly, the amount of effort it takes to get some games working on Linux is not worth it, if you make a decent wage, versus the cost of an XP license to run dual-boot (or in a VM, for older games).
Perhaps NVidia could get off their asses and develop a decent driver for Linux,
That is part of the problem. The other part of it is lack of Direct3D support on Linux, and pervasive use of Direct3D in games. While I don't doubt that nVidia could port it, I doubt very much that Microsoft would let them.
But I have to choose between Windows and playing all games.
I don't.
All I really have to do is be careful which MMOs and casual games I pick up. If I'm playing a game in a window, I'd like my IM working, and I wouldn't mind email, too -- and those are two things I don't do on Windows.
But, if I'm playing something like Natural Selection or a Half Life episode, I'm not likely to want any distractions outside of the game itself.
And by the way: Yes, they work on Linux, at a lower framerate and considerably more hassle than occasionally booting Windows. It's not just about framerate, either -- on Windows, I can play them fullscreen, 1920x1200, with most of the effects enabled. And that's on a laptop.
Oh, and there's the 64-bit problem. Sure, I'm on XP now, but if I ever do upgrade to a decent 64-bit Windows, quite a few games support it now. Wine doesn't, and it doesn't look like it's coming anytime soon.
Now, stuff that really works well on Wine, I'll play. For example, Warcraft 3 installed easily, runs fullscreen with all effects enabled, with no noticeable disadvantage to playing on Windows.
Partly because some of the nice things about Chrome (Webkit) aren't likely to ever be integrated into Firefox. At the very least, they would likely break plenty of existing extensions -- not even all Greasemonkey scripts would necessarily work out of the box if you wrote a Greasemonkey-like plugin for Chrome.
In case you're not being sarcastic:
In Halo matches, you'll generally see people walk up to your corpse (after killing you) and tap crouch repeatedly for a few seconds until you fade to black and respawn.
That is, at least originally, intended to simulate teabagging.
The point being: This shit happens when you take it online. The way I see it, when thousands of 13-year-old boys can run around calling each other "faggot" in high-pitched voices, being able to say "I am a lesbian", even if it's not in a respectful tone, still isn't going to be as bad as what's already on there.
By not taking every opportunity to lie, cheat, and steal from my friends and family, I am at a disadvantage relative to those who do. By paying my taxes instead of hiding my income, I am at a disadvantage to those with, say, large Swiss bank accounts.
That is not an excuse, or a reason not to blame people. There is a reason we consider some things immoral, and that reason has nothing to do with what anyone else is doing.
If it's really "not that bad", that's another issue. But you don't get to say "At least we're not as bad as these guys." If "these guys" are Monsanto, that's not setting the bar very high.