I agreee 100% -- I recognize that in my particular circumstances it's easy, but I have done hard concurrent programming and it's not like my dod died or my hair caught on fire.
The issue, as I see it, is that *testing* a game is hard in ways that other application programming often isn't.
Right now I'm trying to solve an issue with the physics engine ( the Open Dynamic Engine ) in which it goes haywire in certain collision circumstances -- a positive feedback correction harmonic, for lack of a better way to describe it. The tough thing is, to recreate the bug you've got to drive the vehicle in just the right way each time.
Actually, for what it's worth I'm writing a game in my free time which already splits rendering and (physics/game logic) into two threads. The idea being that the physics runs while the rendering thread is blocking on an opengl vsync. While the behavior is synchronous, it runs beautifully on both single and dual processor machines.
In principle this should have detrimental effects on single processor machines but my relatively meager 1.3 ghz powerbook plays beautifully at 30 fps and 60 physics frames per second.
Anyway, this isn't *really* what's being discussed since the behavior is in fact synchronous, but I'm just saying it ain't hard. I'm surprised more games aren't multithreaded. It's not as if MT is hard. You just have to be careful.
I don't know how much GTK+ breaks with major revisions. I do recall, for what it's worth, being able to port non-trivial programs from Qt 1.x to 2.x and then to 3.x with ( in most cases ) no more than updates to the build system and a few one-liners.
Now, I'll be the first to say that GTK+ as an API is broken from the start -- I tried it and no sir I didn't like it. It is *not* a well designed API -- but that said, it does work and I imagine that if 2.x were completely incompatible with 1.x then nobody would have updated apps from 1.x, XMMS not-withstanding. People did port, and I recall many documents on gnome.org explaining how to transition code.
My understanding is that GTK 2.x, while not 100% source compatible with 1.x, was at the very least semantically compatible, which means it's not *that* hard to port.
Going from an API founded upon the notion of nested widgets with clear update rects and paint clipping to a scene-graph canvas approach would be a significant difference. It might even have a significant impact on event delegation. This would be non-trivial to squish into GTK or Qt, and if it weren't forced into their models, it would result in a new API that simply didn't *work* like the old one.
Regarding #3.. well, no shit. But the discussion was in part about Havoc and he works for redhat.
Regarding your final statement. Well, yes my argument is based on the notion of incompatibility and for good reason. I know I wouldn't want to rewrite nautilus or inkscape or whatever for an entirely new API, and neither would you or the people maintaining hundred kloc repositories. Your statement that the next gen toolkits thow away compatibility -- well, only at a superficial level ( and of course internally they change ) but the semantics of the interface will at the very least be similar to aid in porting.
That's part of the problem. While Raster's done a magnificent job -- and frankly, it's pretty mind-blowing -- he's completely not concerned about backwards compatibility regarding toolsets. Again, what he's done is amazing but it's basically a canvas, not a traditional toolkit.
The OSS GUI world is so deeply rooted in Qt/KDE and GTK+/Gnome that there's no chance at *all* that people will adopt his APIs for the next gen display system.
Red Hat's people are concerned with achieving this kind of stuff without too deeply breaking source compatibility. If they can pull that off, my hat's off to them.
That said, red hat's people can learn a *lot* by working with Raster. Clearly, his code is fast, and his technical design's good. But the model is likely inapplicable to traditional widget toolkits.
When you get used to biking to work, you'll wonder how you ever lived without doing so. It helps mentally separate your work from your *life*, it's good for your body, and it's a lot of fun.
I work in downtown Washington DC, and live in Arlington, the corner cut off of DC in the 19th century, so it's not as if my commute is very long -- only about 2.5 miles ( I walk when the weather's really nice and I'm not in a rush ) -- but I tell you it's a blast. I can avoid traffic completely, and the view in the mornings on Key Bridge overlooking the Potomac is breathtaking.
Perhaps it's too cold right now for you to start biking to work, but start soon!
P.S. If you're in or near a city, wear a helmet. I've been hit by cars three times in four years. None actually hurt me, but... well... I can't count on luck forever.
P.P.S. Also, I agree 100% that if it's hard to cut emissions now, why would it be easier ten years from now? Criminy.
Also, keep in mind that our natural eyes are actually pretty shitty cameras ( at least, from the standpoint of what we expect from a CCD ). It turns out our brains are interpolating an awful lot from the spotty signals coming in from our eyes.
The massive licensing and merchandising program on tap will coincide with the new series launch for Family Guy on May 1st, and is expected to be one of the company's largest product blitzes to date.
Product blitzes! Mechandising programs! That sounds like an AWESOME game.
I agree. TV programming is crap enough already. But as the actual process becomes more hostile, perhaps people will be encouraged to watch less. I mean, I don't have a DVR ( or cable for that matter ) but if there were something I'd like to see but I have to work late or I'd prefer to be a bar with friends, and my DVR told me I couldn't record it to watch it, well frankly I'd be pretty ticked.
I'm not a high-falutin kill-your-TV type, but frankly, we all could do with a little less TV and a little more reading and getting out of the apartment/house. This kind of stuff is probably one more step towards the eventual irrelevance of television.
Never made sense to me either. The girls I knew in high-school had *terrible* driving records. Multiple accidents each, speeding tickets six ways to sunday, etc.
Now, I know my high-school isn't a good sample, but nonetheless. I mean, I go into shock when my girlfreind drives -- tailgating in *my* car on 95, in the rain with ice on the ground. And when I say -- "hey, tailgate in your car, not mine" she calls me a grandma.
( don't get me wrong, we have a great relationship )
I think the problem here is that 30 years ago, girls *were* better drivers -- for whatever reason, probably sociological. Today, however, girls are kicking the pants off men in everything from math, & science, to employability, university degrees, and so on ( My GF's a neuroscientist, btw ) -- so they might as well be winning the fight for irresponsibly fast driving.
Give the insurance companies a few years and they'll realize that today girls are every bit as aggressive and stupid as guys. Way to lower the standard, ladies!
This brings up old cartoon imagery but I was wondering... since the propulsion comes from the emission of gas from the surface of the sail, couldn't you bring your energy source *with* you?
I mean, you could have a nuclear reactor and a microwave emitter along with the payload. I know this would add to the mass, but it seems you could then have the energy to decelerate when you reach your destination.
We probably were all wishing for the same, but that said, "opening" the format is probably harder than it sounds.
My understanding of the Office document formats -- which comes entirely from reading rants by OO.org and other projects to write office suites so take it with a big grain of salt -- is that the format itself is made up of serializations of stuff like activex control states. In other words, non-trivial.
I don't know if you or anybody here ever wrote a BeOS "replicant", but it was sort of like ActiveX in that they were serializable classes which could be instantiated by any program, by dlopening the replicant's source executable and running the exported code with the serialized state as initialization parameters. It was really cool -- an app could send a replicant to another app and whammo, you had stuff like a web-browser embedded on the desktop running in the desktop process, or a tray-item using your app's code, but running in the deskbar's process.
Anyway, given that Office uses this kind of approach, it would be near 100% *impossible* to get the state out without the source activex component. Unless the state itself is described in a 100% abstract manner. Which I doubt. The data is almost certainly just a serialization of the internal state of the activex control which created/modified/rendered it.
Now, I know that this kind of stuff only applies to Office when Word or Excel is embedding charts or whatnot from other parts of the office suite, but the fact is this is a useful ( and good ) way to get interoperability, even if it means that it's completely non-portable. Given MS's history, I doubt they've taken a simple approach.
I'm sure there could be better ways, and I imagine OO.org is taking a maximum-interoperability approach...
Anyway, I'm just saying. I don't think MS *could* open the format -- at least not as regards document embedding.
SideTrack 1.0 was rock-stable for me on my 3rd gen PB 12".
That said, I've had *three* kernel panics since installing SideTrack 1.1 -- that's three more than I've had since purchasing this powerbook in august. SideTrack is *excellent*, and frankly indispensable for me, but I think 1.1 is rough about the edges. I'm going to downgrade to 1.0, and wait for a bugfix release.
I agree -- but here's an observation. Often I think of windows users as being people subject to a kind of Stockholm syndrome: They've been abused for so long that they can't imagine life without that abuse. And when challenged, they defend their abuser because they're afraid of losing the abuse.
Now, OS X users don't have the same sorts of problems to deal with as Windows users, except for the damned one-button mouse. But Apple users have used one-button mice for so long they've become accustomed to it as well.
For reference, I'm a die-hard apple user -- but I had to adapt 3 years ago coming from linux on a thinkpad, where the trackpoint had a blessed three buttons. it was heaven, being able to use X11 properly with three buttons.
And now, I can't imagine life with two buttons, let alone three.
I agree, but there is something odd about using OWA on a windows machine: When you first visit ( say, on a clean machine ) it seems to ask you for your Office 2k cd-rom.
I have to assume this is because it's expecting ActiveX controls that office provides. Still, how creepy is it that going to a *website* prompts your browser for your office install cd.
Thank heaven OWA allows other browsers and still provides an acceptable, if degraded, interface.
Exactly. A lot of us geeks would rather express our geek sides by dicking around writing code and making our computers to *intersting* things -- not by shelling out bucks for the latest and greatest XYZ card to give me 64000 channel surround sound or whatever.
Personally, I feel great pleasure in optimizing my code. Just this morning I've got my robotics simulator ( in this case doing quadruped simulation, with many motors, sensors and whatnot, but it can simulate just about anything you can describe to it ) running at 100 Hz physics and 30 fps using less than 10% CPU -- on my meager 12" powerbook. Now, *that* is being a geek. Soon I'll be able to simulate swarms of robotic spiders, each with its own brain and with all with realistic physics. And all for fun.
I think a lot of people today mistake consumerism for geekery. A lot of people I hear being referred to as "experts" are really just people who know how to go to CompUSA or whatever and buy a card, stick it in, and run the windows installer for its driver. [sarcasm]Way to go. That's some HARD stuff. You must be really, really l33t[/sarcasm]
What made you think it was turning around? I'm serious. If they had had this kind of stuff 40 years ago they'd have done this then, too. The popular music industry has always been just that: An industry, read: factories, profit, and anonymous (interchangeable) labor.
Slashdot Mirror
Uhh... what's so controversial about Google Toolbar 3? I'm serious, what have I missed here?
Looks like yet another toolbar I don't want and can't use ( I use Safari ).
I agreee 100% -- I recognize that in my particular circumstances it's easy, but I have done hard concurrent programming and it's not like my dod died or my hair caught on fire.
The issue, as I see it, is that *testing* a game is hard in ways that other application programming often isn't.
Right now I'm trying to solve an issue with the physics engine ( the Open Dynamic Engine ) in which it goes haywire in certain collision circumstances -- a positive feedback correction harmonic, for lack of a better way to describe it. The tough thing is, to recreate the bug you've got to drive the vehicle in just the right way each time.
You should read the overview, then. This is some *serious* and hairy shit.
It's like Apple's cocoa bindings, but... well... more so. I guess you'd say it's like automatic data/event bindings with semantic layout for HI.
I repeat: Serious shit.
Actually, for what it's worth I'm writing a game in my free time which already splits rendering and (physics/game logic) into two threads. The idea being that the physics runs while the rendering thread is blocking on an opengl vsync. While the behavior is synchronous, it runs beautifully on both single and dual processor machines.
In principle this should have detrimental effects on single processor machines but my relatively meager 1.3 ghz powerbook plays beautifully at 30 fps and 60 physics frames per second.
Anyway, this isn't *really* what's being discussed since the behavior is in fact synchronous, but I'm just saying it ain't hard. I'm surprised more games aren't multithreaded. It's not as if MT is hard. You just have to be careful.
Reharding #1 and #2
I don't know how much GTK+ breaks with major revisions. I do recall, for what it's worth, being able to port non-trivial programs from Qt 1.x to 2.x and then to 3.x with ( in most cases ) no more than updates to the build system and a few one-liners.
Now, I'll be the first to say that GTK+ as an API is broken from the start -- I tried it and no sir I didn't like it. It is *not* a well designed API -- but that said, it does work and I imagine that if 2.x were completely incompatible with 1.x then nobody would have updated apps from 1.x, XMMS not-withstanding. People did port, and I recall many documents on gnome.org explaining how to transition code.
My understanding is that GTK 2.x, while not 100% source compatible with 1.x, was at the very least semantically compatible, which means it's not *that* hard to port.
Going from an API founded upon the notion of nested widgets with clear update rects and paint clipping to a scene-graph canvas approach would be a significant difference. It might even have a significant impact on event delegation. This would be non-trivial to squish into GTK or Qt, and if it weren't forced into their models, it would result in a new API that simply didn't *work* like the old one.
Regarding #3.. well, no shit. But the discussion was in part about Havoc and he works for redhat.
Regarding your final statement. Well, yes my argument is based on the notion of incompatibility and for good reason. I know I wouldn't want to rewrite nautilus or inkscape or whatever for an entirely new API, and neither would you or the people maintaining hundred kloc repositories. Your statement that the next gen toolkits thow away compatibility -- well, only at a superficial level ( and of course internally they change ) but the semantics of the interface will at the very least be similar to aid in porting.
That's part of the problem. While Raster's done a magnificent job -- and frankly, it's pretty mind-blowing -- he's completely not concerned about backwards compatibility regarding toolsets. Again, what he's done is amazing but it's basically a canvas, not a traditional toolkit.
The OSS GUI world is so deeply rooted in Qt/KDE and GTK+/Gnome that there's no chance at *all* that people will adopt his APIs for the next gen display system.
Red Hat's people are concerned with achieving this kind of stuff without too deeply breaking source compatibility. If they can pull that off, my hat's off to them.
That said, red hat's people can learn a *lot* by working with Raster. Clearly, his code is fast, and his technical design's good. But the model is likely inapplicable to traditional widget toolkits.
When you get used to biking to work, you'll wonder how you ever lived without doing so. It helps mentally separate your work from your *life*, it's good for your body, and it's a lot of fun.
I work in downtown Washington DC, and live in Arlington, the corner cut off of DC in the 19th century, so it's not as if my commute is very long -- only about 2.5 miles ( I walk when the weather's really nice and I'm not in a rush ) -- but I tell you it's a blast. I can avoid traffic completely, and the view in the mornings on Key Bridge overlooking the Potomac is breathtaking.
Perhaps it's too cold right now for you to start biking to work, but start soon!
P.S. If you're in or near a city, wear a helmet. I've been hit by cars three times in four years. None actually hurt me, but... well... I can't count on luck forever.
P.P.S. Also, I agree 100% that if it's hard to cut emissions now, why would it be easier ten years from now? Criminy.
What? I thought all this time they were *Portable* Network Graphics. Well, the article says "Proprietary" so they must be right.
Also, keep in mind that our natural eyes are actually pretty shitty cameras ( at least, from the standpoint of what we expect from a CCD ). It turns out our brains are interpolating an awful lot from the spotty signals coming in from our eyes.
The massive licensing and merchandising program on tap will coincide with the new series launch for Family Guy on May 1st, and is expected to be one of the company's largest product blitzes to date.
Product blitzes! Mechandising programs! That sounds like an AWESOME game.
I agree. TV programming is crap enough already. But as the actual process becomes more hostile, perhaps people will be encouraged to watch less. I mean, I don't have a DVR ( or cable for that matter ) but if there were something I'd like to see but I have to work late or I'd prefer to be a bar with friends, and my DVR told me I couldn't record it to watch it, well frankly I'd be pretty ticked.
I'm not a high-falutin kill-your-TV type, but frankly, we all could do with a little less TV and a little more reading and getting out of the apartment/house. This kind of stuff is probably one more step towards the eventual irrelevance of television.
Never made sense to me either. The girls I knew in high-school had *terrible* driving records. Multiple accidents each, speeding tickets six ways to sunday, etc.
Now, I know my high-school isn't a good sample, but nonetheless. I mean, I go into shock when my girlfreind drives -- tailgating in *my* car on 95, in the rain with ice on the ground. And when I say -- "hey, tailgate in your car, not mine" she calls me a grandma.
( don't get me wrong, we have a great relationship )
I think the problem here is that 30 years ago, girls *were* better drivers -- for whatever reason, probably sociological. Today, however, girls are kicking the pants off men in everything from math, & science, to employability, university degrees, and so on ( My GF's a neuroscientist, btw ) -- so they might as well be winning the fight for irresponsibly fast driving.
Give the insurance companies a few years and they'll realize that today girls are every bit as aggressive and stupid as guys. Way to lower the standard, ladies!
I wish I had mod-points for you. Your argument is draconian, but frankly I agree 100%.
This brings up old cartoon imagery but I was wondering... since the propulsion comes from the emission of gas from the surface of the sail, couldn't you bring your energy source *with* you?
I mean, you could have a nuclear reactor and a microwave emitter along with the payload. I know this would add to the mass, but it seems you could then have the energy to decelerate when you reach your destination.
Anyway, I was just wondering...
We probably were all wishing for the same, but that said, "opening" the format is probably harder than it sounds.
My understanding of the Office document formats -- which comes entirely from reading rants by OO.org and other projects to write office suites so take it with a big grain of salt -- is that the format itself is made up of serializations of stuff like activex control states. In other words, non-trivial.
I don't know if you or anybody here ever wrote a BeOS "replicant", but it was sort of like ActiveX in that they were serializable classes which could be instantiated by any program, by dlopening the replicant's source executable and running the exported code with the serialized state as initialization parameters. It was really cool -- an app could send a replicant to another app and whammo, you had stuff like a web-browser embedded on the desktop running in the desktop process, or a tray-item using your app's code, but running in the deskbar's process.
Anyway, given that Office uses this kind of approach, it would be near 100% *impossible* to get the state out without the source activex component. Unless the state itself is described in a 100% abstract manner. Which I doubt. The data is almost certainly just a serialization of the internal state of the activex control which created/modified/rendered it.
Now, I know that this kind of stuff only applies to Office when Word or Excel is embedding charts or whatnot from other parts of the office suite, but the fact is this is a useful ( and good ) way to get interoperability, even if it means that it's completely non-portable. Given MS's history, I doubt they've taken a simple approach.
I'm sure there could be better ways, and I imagine OO.org is taking a maximum-interoperability approach...
Anyway, I'm just saying. I don't think MS *could* open the format -- at least not as regards document embedding.
Rant over.
SideTrack 1.0 was rock-stable for me on my 3rd gen PB 12".
That said, I've had *three* kernel panics since installing SideTrack 1.1 -- that's three more than I've had since purchasing this powerbook in august. SideTrack is *excellent*, and frankly indispensable for me, but I think 1.1 is rough about the edges. I'm going to downgrade to 1.0, and wait for a bugfix release.
I agree -- but here's an observation. Often I think of windows users as being people subject to a kind of Stockholm syndrome: They've been abused for so long that they can't imagine life without that abuse. And when challenged, they defend their abuser because they're afraid of losing the abuse.
Now, OS X users don't have the same sorts of problems to deal with as Windows users, except for the damned one-button mouse. But Apple users have used one-button mice for so long they've become accustomed to it as well.
For reference, I'm a die-hard apple user -- but I had to adapt 3 years ago coming from linux on a thinkpad, where the trackpoint had a blessed three buttons. it was heaven, being able to use X11 properly with three buttons.
And now, I can't imagine life with two buttons, let alone three.
That said, I wish people would just let it drop.
Come on, that doesn't sound *nearly* expensive enough.
Normals infer trigonometry, to me.
Now, what I really hear, when somebody calls another a "normal" is a statement of self-superiority, based on perceived exclusivity.
Lots of people call others "normals". Goths, punks, etc. It's just childish name-calling.
Grow up, people.
May I suggest you NOT call them "normals". They're people, for f*ck's sake.
I agree, but there is something odd about using OWA on a windows machine: When you first visit ( say, on a clean machine ) it seems to ask you for your Office 2k cd-rom.
I have to assume this is because it's expecting ActiveX controls that office provides. Still, how creepy is it that going to a *website* prompts your browser for your office install cd.
Thank heaven OWA allows other browsers and still provides an acceptable, if degraded, interface.
Yeah, well, my old man used to call me broken-rubber. Seriously.
Exactly. A lot of us geeks would rather express our geek sides by dicking around writing code and making our computers to *intersting* things -- not by shelling out bucks for the latest and greatest XYZ card to give me 64000 channel surround sound or whatever.
Personally, I feel great pleasure in optimizing my code. Just this morning I've got my robotics simulator ( in this case doing quadruped simulation, with many motors, sensors and whatnot, but it can simulate just about anything you can describe to it ) running at 100 Hz physics and 30 fps using less than 10% CPU -- on my meager 12" powerbook. Now, *that* is being a geek. Soon I'll be able to simulate swarms of robotic spiders, each with its own brain and with all with realistic physics. And all for fun.
I think a lot of people today mistake consumerism for geekery. A lot of people I hear being referred to as "experts" are really just people who know how to go to CompUSA or whatever and buy a card, stick it in, and run the windows installer for its driver. [sarcasm]Way to go. That's some HARD stuff. You must be really, really l33t[/sarcasm]
Rant over.
Whoops. You're right. I was thinking of bad-think or whatever they called it. Been a long time since I read 1984.
The music industry might turn itself around;
What made you think it was turning around? I'm serious. If they had had this kind of stuff 40 years ago they'd have done this then, too. The popular music industry has always been just that: An industry, read: factories, profit, and anonymous (interchangeable) labor.
We're just doing it more efficiently, now.