Well, some cities require that smoking sections be sealed and separately ventilated. There's a greasy-spoon diner where I live that has the bar sealed-off from the rest of the restaurant for the smokers. Its pretty much the main reason people go there.
That and the wifi alone are the reasons I want this thing. The fact is this: you have a console with a) each player has his own screen, no splitscreen bullshit b) large scale networking, including large flash-storage for network spawning c) a real pointing device.
Who cares about two screens, or the microphone, or any other frill features of the damn thing. To me, the implication is obvious: here's a handheld console with a featureset makign it ideal for just about every awesome PC title for the past decade.
I will be sorely disappointed if I don't see a decent library of solid RTS, team-oriented FPS (a-la Team Fortress or Tribes), 4X, and multiplayer Tactical/Stealth titles before summer.
Well, it sounds like its all touchscreen - I hope they have some way to have a physical interface, some buttons or a joystick or something - touchscreens are generally terrible for joystick interaction, unless you're using it as a single large analog control (like Metroid Hunters does). After all, without looking at the screen, how does the pilot know what control he's pressing?
Right - so why wouldn't it be a problem that it could allow hackers and spammers to ramp up their activities to a larger scale? Spamming and hacking actions are usually very high in the output and low in the feedback - hitting every port until one acks, net sending to a thousand users, etc. Why wouldn't multicast be useful for that? Especially since most users have a bigger downpipe to handle responses than up-pipe.
And as for P2P, imagine you've a shared session with a bunch of users, and you make a change to teh shared document - you can multicast the change out to all the users rather than sending it to each one or bouncing it off of a server.
Look at it this way - you can pay for their gym memberships now, or you can pay for their heart pills and triple-bypass surgery later. I think we both know which is cheaper.
The fundamental problem is this: RSS was simple. That's why it was adopted - its easy to serve, easy to implement.
Unfortunately, that's also why it sucks. It just doesn't scale.
My dream RSS replacement:
-serves out HTML, so you use it with a modified web browser instead of an RSS aggregator. -uses a p2p system like JXTA to send out update notifications to clients, rather than polling. -uses a p2p system or multicasting to disseminate content, minimizing server load on updates. -sends out only differential information from the last user request. With each update to the page, it logs the current page and the last diff - users requesting updates receive either an agggregate of the diffs since last request, or the full page if it would be smaller. Updates are sent gzipped.
With features like that, you could serve comment threads off of the damn thing, updated per comment, much less news headlines. Only catch is that you'd want the site to have one single page shared between all users (like Fark) so that it wouldn't have to log a diff history per-user.
Agreed. I've always thought that the web needed better support for real-time, and RSS was such an unpleasant band-aid solution for the problem. For one thing, sending XML in plaintext form is a fugly waste of bandwidth. And polling, imho, is always a sign that something is wrong.
There are many new open P2P systems that promise to give users a way around port forwarding (jxta for example) - so really, having the server dispatch notification of a new update to all the clients is doable. The same P2P framework could then be used to disseminate the information to clients who request details on the update, with the feed acting as a tracker.
IMHO, I think a better use than simply sending random little headlines would be to work with the existing framework of HTML - have it send out updates to a shared dynamic page. Rather than resending all the html files, send a diff. Basically, eliminate the "refresh" button mashing and streamline the process. Make the Fark comments page into a hybrid of web and irc. Of course, it would have poor ping compared to IRC, and you wouldn't want to track individualised pages for each user like Slashdot maeks (heh, I could see Frames coming back to compensate for this), but a systme like that would be so much cooler - making the Web actually *interactive* - than just RSS.
But nobody's gonna do it. RSS succeeded _because_ it had a small, refined scope that made it easy to implement.
The only catch with multicast is that it would spawn a whole new world of spamming. Consider multicasting portscans, Net Send spam, and a host of other malevolent activities.
Still, yeah, multicast is important. To me, I think one big nice use is in P2P networking - if you have a fixed peergroup, you can message all the peers with the ease of messaging only one.
Ick. The b5 movies were horrible. The only good one was "in the beginning". 3rd space was excruciatingly terrible, and "A call to arms" was rather poor, imho.
This reflects a fundamental difference between English and Spanish - in Spanish, spelling is inextricably tied to pronounciation, so when a new word is imported into the language its spelling is revised to match its pronounciation. In English, the word is simply written as is (if possible), and then its pronouncation is perverted gradually with time.
Actually, that thing is teh suck. It doesn't seem to actually swap nouns, adverbs and verbs with "smurf, smurfy, smurfing" or anthing, it just inserts Smurf phrases into the paragraph, and occaisionally the word "Smurf" into nonsensical places. That's pretty weak.
Not only that, but imho splash screens need to be proper windows - that is, minimizable, draggable, windows. Not just a pretty picture. That way, I can launch the program and minimize it and then get on with my day, and go back to it when its ready.
Idunno, I like the toolbarless look, but that's just because I think its nice looking and I'm one of those people who never ever touches the toolbar. I find that the Gnome people in general don't like toolbars and tend to prefer right-click-menus. This is good for simple apps (like the file browser) but a poor decision for more complicated apps.
Still, I agree that the "new window for each folder" thing is a bad idea. Why not follow FireFox's success and go with a rocker/radial approach? Middle-click = open in new window, rclick + scrollup = up one level, stuff like that? Just have the context-menu list the rocker gestures and hotkeys alongside the command names.
That's interesting. Since you seem to be keeping tabs on Palm's OS plans, any idea if OS6 will be backwards compatible with the OS5 ARM Palms? Or did I pick up a Zire with future-proofing in mind for nothing?
Total Annihilation was pretty good for that - thought the units were a little stupider than you'd like, they made up for it by being simple (no worrying about spells) and having long graphical itineraries (you could see their patrol route onscreen). By default, any unit on patrol would do intelligent things while out there, and any unit guarding another unit would automatically do intelligent things for its target (aid in construction, repair target, attack target's foes, etc).
Dunno, I use Office 2k and OOo, and there are a few features where OOo whups O2k. OO Draw, OO Formula are two big ones - the corresponding drawing and formula packages under Office 2k are piss-poor. Plus, for some reason I have horrible stability in Office 2k when opening remote documents (like shares over wifi) - Office 2k really doesn't crash gracefully, so I get corrupt files.
Besides, OOo will always have one big feature that defeats O2k - non-retarded autocorrect. How many times has office fucked up perfectly good tabbing by converting shit to bullets? Or various other creative formatting on the fly?
So I think "very small set of features" is an overstatement. Unless you mean in comparison to Office XP, but I don't know anybody who actually owns that (including businesses).
Still, not to say that OOO doens't have its own share of problems.
Slashdot Mirror
Well, some cities require that smoking sections be sealed and separately ventilated. There's a greasy-spoon diner where I live that has the bar sealed-off from the rest of the restaurant for the smokers. Its pretty much the main reason people go there.
Yes. In fact, we should take the crying babies and eat them. But only if they're Irish.
Oh, and Twilight Imperium sucks.
That and the wifi alone are the reasons I want this thing. The fact is this: you have a console with
a) each player has his own screen, no splitscreen bullshit
b) large scale networking, including large flash-storage for network spawning
c) a real pointing device.
Who cares about two screens, or the microphone, or any other frill features of the damn thing. To me, the implication is obvious: here's a handheld console with a featureset makign it ideal for just about every awesome PC title for the past decade.
I will be sorely disappointed if I don't see a decent library of solid RTS, team-oriented FPS (a-la Team Fortress or Tribes), 4X, and multiplayer Tactical/Stealth titles before summer.
Besides that, if we're splurging for monitors, why beat around the bush?
Go CAVE
Is it good?
.
.
.
or is it whack?
Hmm, keep in mind that they just decided to launch FireFox as "Netscape"... maybe they have something interesting planned.
Grass is Greener? Hmm? I thought the line when talking about Japanese gadgets was "If there's grass on the field, play ball!"
Cripes. For that price, just make the damn thing an external USB device and use a laptop or a tablet PC to control it.
Well, it sounds like its all touchscreen - I hope they have some way to have a physical interface, some buttons or a joystick or something - touchscreens are generally terrible for joystick interaction, unless you're using it as a single large analog control (like Metroid Hunters does). After all, without looking at the screen, how does the pilot know what control he's pressing?
Right - so why wouldn't it be a problem that it could allow hackers and spammers to ramp up their activities to a larger scale? Spamming and hacking actions are usually very high in the output and low in the feedback - hitting every port until one acks, net sending to a thousand users, etc. Why wouldn't multicast be useful for that? Especially since most users have a bigger downpipe to handle responses than up-pipe.
And as for P2P, imagine you've a shared session with a bunch of users, and you make a change to teh shared document - you can multicast the change out to all the users rather than sending it to each one or bouncing it off of a server.
Look at it this way - you can pay for their gym memberships now, or you can pay for their heart pills and triple-bypass surgery later. I think we both know which is cheaper.
The fundamental problem is this: RSS was simple. That's why it was adopted - its easy to serve, easy to implement.
Unfortunately, that's also why it sucks. It just doesn't scale.
My dream RSS replacement:
-serves out HTML, so you use it with a modified web browser instead of an RSS aggregator.
-uses a p2p system like JXTA to send out update notifications to clients, rather than polling.
-uses a p2p system or multicasting to disseminate content, minimizing server load on updates.
-sends out only differential information from the last user request. With each update to the page, it logs the current page and the last diff - users requesting updates receive either an agggregate of the diffs since last request, or the full page if it would be smaller. Updates are sent gzipped.
With features like that, you could serve comment threads off of the damn thing, updated per comment, much less news headlines. Only catch is that you'd want the site to have one single page shared between all users (like Fark) so that it wouldn't have to log a diff history per-user.
Agreed. I've always thought that the web needed better support for real-time, and RSS was such an unpleasant band-aid solution for the problem. For one thing, sending XML in plaintext form is a fugly waste of bandwidth. And polling, imho, is always a sign that something is wrong.
There are many new open P2P systems that promise to give users a way around port forwarding (jxta for example) - so really, having the server dispatch notification of a new update to all the clients is doable. The same P2P framework could then be used to disseminate the information to clients who request details on the update, with the feed acting as a tracker.
IMHO, I think a better use than simply sending random little headlines would be to work with the existing framework of HTML - have it send out updates to a shared dynamic page. Rather than resending all the html files, send a diff. Basically, eliminate the "refresh" button mashing and streamline the process. Make the Fark comments page into a hybrid of web and irc. Of course, it would have poor ping compared to IRC, and you wouldn't want to track individualised pages for each user like Slashdot maeks (heh, I could see Frames coming back to compensate for this), but a systme like that would be so much cooler - making the Web actually *interactive* - than just RSS.
But nobody's gonna do it. RSS succeeded _because_ it had a small, refined scope that made it easy to implement.
The only catch with multicast is that it would spawn a whole new world of spamming. Consider multicasting portscans, Net Send spam, and a host of other malevolent activities.
Still, yeah, multicast is important. To me, I think one big nice use is in P2P networking - if you have a fixed peergroup, you can message all the peers with the ease of messaging only one.
Ick. The b5 movies were horrible. The only good one was "in the beginning". 3rd space was excruciatingly terrible, and "A call to arms" was rather poor, imho.
So, a B5 movie....
Who here saw 3rd Space?
This could suck. Even B5 was seriously hit-or-miss.
This reflects a fundamental difference between English and Spanish - in Spanish, spelling is inextricably tied to pronounciation, so when a new word is imported into the language its spelling is revised to match its pronounciation. In English, the word is simply written as is (if possible), and then its pronouncation is perverted gradually with time.
Actually, that thing is teh suck. It doesn't seem to actually swap nouns, adverbs and verbs with "smurf, smurfy, smurfing" or anthing, it just inserts Smurf phrases into the paragraph, and occaisionally the word "Smurf" into nonsensical places. That's pretty weak.
Not only that, but imho splash screens need to be proper windows - that is, minimizable, draggable, windows. Not just a pretty picture. That way, I can launch the program and minimize it and then get on with my day, and go back to it when its ready.
Funny, The GIMP was exactly the app I had in mind with my "complex apps" complaint. That, and vague memories of an old version of Dia.
Seriously, I think radial-context-menus, if done properly (read: not like the awful ones in FireFox), could be the future.
precipitate
Idunno, I like the toolbarless look, but that's just because I think its nice looking and I'm one of those people who never ever touches the toolbar. I find that the Gnome people in general don't like toolbars and tend to prefer right-click-menus. This is good for simple apps (like the file browser) but a poor decision for more complicated apps.
Still, I agree that the "new window for each folder" thing is a bad idea. Why not follow FireFox's success and go with a rocker/radial approach? Middle-click = open in new window, rclick + scrollup = up one level, stuff like that? Just have the context-menu list the rocker gestures and hotkeys alongside the command names.
That's interesting. Since you seem to be keeping tabs on Palm's OS plans, any idea if OS6 will be backwards compatible with the OS5 ARM Palms? Or did I pick up a Zire with future-proofing in mind for nothing?
Total Annihilation was pretty good for that - thought the units were a little stupider than you'd like, they made up for it by being simple (no worrying about spells) and having long graphical itineraries (you could see their patrol route onscreen). By default, any unit on patrol would do intelligent things while out there, and any unit guarding another unit would automatically do intelligent things for its target (aid in construction, repair target, attack target's foes, etc).
Dunno, I use Office 2k and OOo, and there are a few features where OOo whups O2k. OO Draw, OO Formula are two big ones - the corresponding drawing and formula packages under Office 2k are piss-poor. Plus, for some reason I have horrible stability in Office 2k when opening remote documents (like shares over wifi) - Office 2k really doesn't crash gracefully, so I get corrupt files.
Besides, OOo will always have one big feature that defeats O2k - non-retarded autocorrect. How many times has office fucked up perfectly good tabbing by converting shit to bullets? Or various other creative formatting on the fly?
So I think "very small set of features" is an overstatement. Unless you mean in comparison to Office XP, but I don't know anybody who actually owns that (including businesses).
Still, not to say that OOO doens't have its own share of problems.