There is not a single brake pedal! And worse, the W3C or MS or Mozilla or whoever could introduce a new gas pedal, and you the website operator have to filter out the new gas pedal when it's introduced.
Undid my mods, but I had to post this.
There used to be a break pedal. I think it was Firefox 1.5 where this code didn't evaluate any tags:
element.append(document.createTextNode(sText));
The solution, therefore, was to manually parse italic/bold/a tags, to append those elements - and then create a text node inside. A perfect working DHTML/DOM solution, properly sanitized!
However, with Firefox 3, text nodes now evaluate HTML tags. This handy function went out with eval usage for local callbacks.:/ Opera and Chrome also evaluate some(all?) tags for appended text.
You are aware that TF2 was written for PC, right? XBox was almost an afterthought.
Kind of. I'm sure they were considering a console port when they first designed it. No company just out of the blue says "lets port our game to a totally foreign architecture!"
I chose TF2 rather than L4D, because TF2 has less hackers. L4D's experience is pretty bad on every platform unless you have lots of friends and go pub-stomp.:P
I once claimed that xvid was better than h.264. Boy was I wrong! Slashdotters set me right almost immediately, and then I started researching it.
The h.264 4.0 profile isn't that good. I believe Youtube uses that, with optimizations like CABAC and B/Ref frames turned off, and motion estimation quite low. During my research, and after days of tweaking, I put together some ludicrously good x264 settings(very tweaked 5.1 profile) which yielded incredible results for FRAPS'd test vids. I was getting Youtube HD's quality(2mbit) at just 640-1280 kbit.
I started by uploading 4mbit version Youtube, for comparison.
One video was a low-motion RTS (Kohan II) with gaudy ground textures. It looked almost identical to Youtube at just 640kbit. 768-1024kbit was easily better quality. I noted that tweaking aq_strength in one direction led to massive blocking on the ground textures (distracting), and the other way preserved the details. (but then units and spells looked slightly worse) I've never seen a video before where aq_strength had such an impact on where bits are spent, but I suppose that's the nature of this one game.
Another video was of Left4Dead. High motion, somewhat gaudy textures, but lots of stuff moving around and quite a few bullet/fog effects. I needed roughly 1280kbit to match Youtube's 2mbit. At sub-1024kbit, blocking was noticeable on some textures. I also discovered that by increasing the gamma in L4D when recording from a replay, you can turn wall and ground textures into low-colour varients, while brightening the screen. This improved the quality of bullet/fog effects at ~1024kbit, but the walls really did look horrible.
All in all, I learned a lot - none of it conclusive, but all of it pointing to H.264 being far more awesome than I originally thought. Too bad Youtube's encoder favours speed over quality!
SMplayer is the best MPlayer frontend I've tried. I still prefer MPC-HC + KLite for the GPU shaders, but I can't deny that SMplayer and MPlayer are quality software! Based on CPU usage when playing stuff, I'd bet that the GPU acceleration/decoding is fully enabled and working.
It's up to the media player to ensure the streams are accelerated by picking a proper codec. It's also up to the media player to understand the container format. These things aren't very difficult, because of the codec frameworks that exist. On Windows, the most common one is DirectShow. (or whatever they've renamed it in Vista/Win7)
The media player has to pipe the stream data through to wherever it has to go - the Codec handles this, so once the media player picks a hardware accelerated codec, you're set!
VLC usually just sends it to its own CPU-based codecs, but other media players (like MPC, loaded up with directshow codecs for different formats) will send parts of it to the GPU to be decoded/accelerated. MPC-HC also has GPU shaders that can enhance the quality, regardless of the codec.
H.264 will be accelerated in.MKV,.MOV, and.MP4 unless your media player doesn't know what to do, which is unlikely because of the codec frameworks. The biggest issue is either going to be a missing codec(solved by using a pack like the klite mega codec pack) or your media player of choice(VLC) favouring compatibility over performance. VLC likes to choose CPU-decode codecs rather than GPU-decode ones. As far as I know, it also lacks GPU shaders.
Side-note: Recently I was uploading H.264/AAC to Youtube. There was a glitch on Youtube's end that it thought VBR-AAC was longer than it really was, so it rejected the video. After switching to.mp4(h.264/mp3), I had problems with audio desyncing. Then I switched to.mkv(h.264/mp3), and it worked fine. Seems like youtube has solid mkv support, just like most desktop software I've tested.
Security pretty much ends after a logon password. A password that changes every two weeks, must begin with a capital letter, may be 6-8 characters long, and must have 2 numbers within it.
After all, we wouldn't want to narrow the parameters down to make it easier to hack, would we?
The last virus I got piggy-backed a firefox XPI. But that was Firefox 1.5
Viruses are sneaky. What you're thinking of is called crapware. If you want a fine example of bundled crapware, check out the CCleaner installer, or perhaps the MediaCoder one.
You can uninstall Google Toolbar fully from the control panel. I don't mind it, because it seems to remove that infobar-refreshpage-runaround to download files.
The update notifier is so annoying that I turned it off completely, and manually run it every couple days.
I've had it pop-over a video playing in Totem no less than 6 times. That is SO irritating. Sometimes it also grabs input without appearing over totem, so the pause button and volume controls stop working.:/
At some point I'll look up the update command and add it to a script launched by crontab -e
In UI responsiveness, perhaps - assuming adequate specs to run both operating systems. I have a feeling Win7 will be horribly slow if you actually got it to run on a P3 with 256MB of RAM.
XP = not fast enough for netbooks and definitely not the OLPC
OLPC 1.5 has a 1.0ghz C7-M, I thought? That's quite a bit beefier than the laptops XP originally shipped on. And if I read the specs correctly, 1GB of RAM too?
XP is faster than Ubuntu, and people try to run that on their netbooks. Seems to me both are decently suited to the task. You could also strip XP down with a tool like nLite. I did that for a relative that had an old Thinkpad 390e, which turned it into a speed demon. Booted in about 30-35 seconds, started Firefox 2 in about 12, and OpenOffice 2.x in ~15. Warm starts were only a second or two. Hibernation only took ~20 seconds from power-on to desktop.
To compare, I have Ubuntu installed on a 1.2ghz Via Eden computer, and it's nowhere near that fast. Booting takes a long time, and OpenOffice takes forever to start up. Firefox is faster, at just 8 seconds, but Firefox is also newer, so that's not a fully valid comparison. Memory usage at boot is 3x higher, at just over 130MB.
However, I'm not a linux guru, and I haven't finished tweaking Ubuntu yet. To give it a fair shake, I need to be able to claim equal expertise at Ubuntu modding as I have with XP modding.
Also have to consider the real cost of windows is, MS ULTRA CHEAP OMG WINDOWZ LIZENZES are actually meant to be paid for real later...
Windows isn't suitable for the OLPC. Microsoft has very corporate motivations rather than charitable ones. The main reason it isn't suitable isn't cost, but rather how quickly they drop support.
Microsoft is unwilling to customize XP for such a device, because they'd have to support it - and yet they want XP on it, because it's great marketing.
As far as I'm concerned the OLPC project should stay the hell away from Microsoft - but XP still makes a fine choice for netbooks, especially for someone willing to dig in and mod it a bit.:)
I'm thinking back to when I played Kotor. The virtual memory usage would rise into the gigabytes, and then it'd crash.
I determined it was a memory leak caused from loading saved games. I theorized that it wasn't freeing resources properly on load. How did I do it? By watching VRAM usage in the task manager.
I posted it on the forums, and never got a response - but at least I notified them of a bug, and pinpointed it pretty accurately. You could say that having access to the source would be a boon here, but the truth is a novice non-coder isn't going to be able to fix it even with the source, and FOSS and proprietary software are just as bad at not fixing user reported bugs.
That leads me to conclude that if I had had access to the source code, it wouldn't have made any difference - except perhaps I would have wasted time looking into it. Source code is great for learning programming, but there's a lot of basics you have to understand first, before you can make sense of stuff like that.
I respect the goals of the OLPC project, but I'm not convinced fully 100% open source is necessary for children to learn computers.
Not in the realm of developing usable applications and OSes, they weren't.
I was downloading Linux distros back in '06. But short of installing them to see if they worked on my hardware, I never really dove in and used one seriously.
Then Ubuntu matured, and I hopped over. It's not being used as my primary desktop yet, but it is working fine as a Samba NAS. Quite fun to hack together.:)
To be worth switching, you have to show how things are going to be BETTER. You have to show that you can do 100% of your job, and that it'll be better. Otherwise, it really isn't worth it.
You're absolutely correct. That's one of the reasons XP is still my desktop.
However, I just finished setting up an Ubuntu NAS - it was quite fun hacking it together! I wouldn't use Ubuntu as a primary desktop, but I am enjoying learning some Linux.:)
NetworkManager just died on me yesterday. I had to manually configure/etc/network/interfaces and/etc/resolv.conf
Apparently NetworkManager connects after boot, which was messing up Samba. (no connection detected, so no daemons started) I deleted the connection from it because it wouldn't go static, but the add button wouldn't add any back, so my ethernet connection was gone. I had to manually configure stuff to get it working. Good thing I had a secondary XP box on hand.:P
In the short-term, binary drivers are often a better choice. Some driver is better than no driver, right? But in the long run, free/open source drivers are usually better. Just look at how crashy and difficult to work with the Nvidia drivers have been. Or how severely limited the ATI drivers were.
Intel and VIA/S3 are fine examples of this. I wouldn't want to use their binary drivers - the FOSS ones are so much more advanced.
nVidia gets away with it because they do a good job - not because they're large. (although being large helps, it didn't work for Intel)
Slashdot Mirror
There is not a single brake pedal! And worse, the W3C or MS or Mozilla or whoever could introduce a new gas pedal, and you the website operator have to filter out the new gas pedal when it's introduced.
Undid my mods, but I had to post this.
There used to be a break pedal. I think it was Firefox 1.5 where this code didn't evaluate any tags:
element.append(document.createTextNode(sText));
The solution, therefore, was to manually parse italic/bold/a tags, to append those elements - and then create a text node inside. A perfect working DHTML/DOM solution, properly sanitized!
However, with Firefox 3, text nodes now evaluate HTML tags. This handy function went out with eval usage for local callbacks. :/ Opera and Chrome also evaluate some(all?) tags for appended text.
Actually, it hits VIA pretty hard - they make a lot of chips, including USB 2.0 controllers.
Intel is moving in on basically everything they do.
You are aware that TF2 was written for PC, right? XBox was almost an afterthought.
Kind of. I'm sure they were considering a console port when they first designed it. No company just out of the blue says "lets port our game to a totally foreign architecture!"
I chose TF2 rather than L4D, because TF2 has less hackers. L4D's experience is pretty bad on every platform unless you have lots of friends and go pub-stomp. :P
And yes, L4D still plays better on PC.
That would be sensationalism. So far, it is measured in cm; by the turn of the century (90 years from now) it is projected to be a few meters,
I would think both of those outcomes would be awful for a few dozen cities on our planet that are only a foot above sea level.
Ahh!...bitter, are we? ;)
Well, truth be told, I agree - although many console ports(like TF2) still play better on PC.
Absolutely agree! And so do these guys!
I once claimed that xvid was better than h.264. Boy was I wrong! Slashdotters set me right almost immediately, and then I started researching it.
The h.264 4.0 profile isn't that good. I believe Youtube uses that, with optimizations like CABAC and B/Ref frames turned off, and motion estimation quite low. During my research, and after days of tweaking, I put together some ludicrously good x264 settings(very tweaked 5.1 profile) which yielded incredible results for FRAPS'd test vids. I was getting Youtube HD's quality(2mbit) at just 640-1280 kbit.
I started by uploading 4mbit version Youtube, for comparison.
One video was a low-motion RTS (Kohan II) with gaudy ground textures. It looked almost identical to Youtube at just 640kbit. 768-1024kbit was easily better quality. I noted that tweaking aq_strength in one direction led to massive blocking on the ground textures (distracting), and the other way preserved the details. (but then units and spells looked slightly worse) I've never seen a video before where aq_strength had such an impact on where bits are spent, but I suppose that's the nature of this one game.
Another video was of Left4Dead. High motion, somewhat gaudy textures, but lots of stuff moving around and quite a few bullet/fog effects. I needed roughly 1280kbit to match Youtube's 2mbit. At sub-1024kbit, blocking was noticeable on some textures. I also discovered that by increasing the gamma in L4D when recording from a replay, you can turn wall and ground textures into low-colour varients, while brightening the screen. This improved the quality of bullet/fog effects at ~1024kbit, but the walls really did look horrible.
All in all, I learned a lot - none of it conclusive, but all of it pointing to H.264 being far more awesome than I originally thought. Too bad Youtube's encoder favours speed over quality!
SMplayer is the best MPlayer frontend I've tried. I still prefer MPC-HC + KLite for the GPU shaders, but I can't deny that SMplayer and MPlayer are quality software! Based on CPU usage when playing stuff, I'd bet that the GPU acceleration/decoding is fully enabled and working.
You don't understand how acceleration works.
It's up to the media player to ensure the streams are accelerated by picking a proper codec. It's also up to the media player to understand the container format. These things aren't very difficult, because of the codec frameworks that exist. On Windows, the most common one is DirectShow. (or whatever they've renamed it in Vista/Win7)
The media player has to pipe the stream data through to wherever it has to go - the Codec handles this, so once the media player picks a hardware accelerated codec, you're set!
VLC usually just sends it to its own CPU-based codecs, but other media players (like MPC, loaded up with directshow codecs for different formats) will send parts of it to the GPU to be decoded/accelerated. MPC-HC also has GPU shaders that can enhance the quality, regardless of the codec.
H.264 will be accelerated in .MKV, .MOV, and .MP4 unless your media player doesn't know what to do, which is unlikely because of the codec frameworks. The biggest issue is either going to be a missing codec(solved by using a pack like the klite mega codec pack) or your media player of choice(VLC) favouring compatibility over performance. VLC likes to choose CPU-decode codecs rather than GPU-decode ones. As far as I know, it also lacks GPU shaders.
Side-note: Recently I was uploading H.264/AAC to Youtube. There was a glitch on Youtube's end that it thought VBR-AAC was longer than it really was, so it rejected the video. After switching to .mp4(h.264/mp3), I had problems with audio desyncing. Then I switched to .mkv(h.264/mp3), and it worked fine. Seems like youtube has solid mkv support, just like most desktop software I've tested.
Most corporations are that stupid, too.
Security pretty much ends after a logon password. A password that changes every two weeks, must begin with a capital letter, may be 6-8 characters long, and must have 2 numbers within it.
After all, we wouldn't want to narrow the parameters down to make it easier to hack, would we?
It's installed the same way as viruses;
The last virus I got piggy-backed a firefox XPI. But that was Firefox 1.5
Viruses are sneaky. What you're thinking of is called crapware. If you want a fine example of bundled crapware, check out the CCleaner installer, or perhaps the MediaCoder one.
You can uninstall Google Toolbar fully from the control panel. I don't mind it, because it seems to remove that infobar-refreshpage-runaround to download files.
Which is exactly the point. Look at the value you get for your new $100 HDD! You'll never fill that! ;)
I have an old Via C7 (Eden) 1.2ghz machine that goes from Power On to XP Desktop in 14 seconds.
I think it was a Phoenix BIOS, too...
It's absolutely nothing compared to my gaming computer. That thing takes 25-30 seconds to POST!
The update notifier is so annoying that I turned it off completely, and manually run it every couple days.
I've had it pop-over a video playing in Totem no less than 6 times. That is SO irritating. Sometimes it also grabs input without appearing over totem, so the pause button and volume controls stop working. :/
At some point I'll look up the update command and add it to a script launched by crontab -e
Way to miss the point.
Please clarify the point, then.
Anyway, windows 7 = "almost" as fast as XP
In UI responsiveness, perhaps - assuming adequate specs to run both operating systems. I have a feeling Win7 will be horribly slow if you actually got it to run on a P3 with 256MB of RAM.
XP = not fast enough for netbooks and definitely not the OLPC
OLPC 1.5 has a 1.0ghz C7-M, I thought? That's quite a bit beefier than the laptops XP originally shipped on. And if I read the specs correctly, 1GB of RAM too?
XP is faster than Ubuntu, and people try to run that on their netbooks. Seems to me both are decently suited to the task. You could also strip XP down with a tool like nLite. I did that for a relative that had an old Thinkpad 390e, which turned it into a speed demon. Booted in about 30-35 seconds, started Firefox 2 in about 12, and OpenOffice 2.x in ~15. Warm starts were only a second or two. Hibernation only took ~20 seconds from power-on to desktop.
To compare, I have Ubuntu installed on a 1.2ghz Via Eden computer, and it's nowhere near that fast. Booting takes a long time, and OpenOffice takes forever to start up. Firefox is faster, at just 8 seconds, but Firefox is also newer, so that's not a fully valid comparison. Memory usage at boot is 3x higher, at just over 130MB.
However, I'm not a linux guru, and I haven't finished tweaking Ubuntu yet. To give it a fair shake, I need to be able to claim equal expertise at Ubuntu modding as I have with XP modding.
Also have to consider the real cost of windows is, MS ULTRA CHEAP OMG WINDOWZ LIZENZES are actually meant to be paid for real later...
Windows isn't suitable for the OLPC. Microsoft has very corporate motivations rather than charitable ones. The main reason it isn't suitable isn't cost, but rather how quickly they drop support.
Microsoft is unwilling to customize XP for such a device, because they'd have to support it - and yet they want XP on it, because it's great marketing.
As far as I'm concerned the OLPC project should stay the hell away from Microsoft - but XP still makes a fine choice for netbooks, especially for someone willing to dig in and mod it a bit. :)
I'm thinking back to when I played Kotor. The virtual memory usage would rise into the gigabytes, and then it'd crash.
I determined it was a memory leak caused from loading saved games. I theorized that it wasn't freeing resources properly on load. How did I do it? By watching VRAM usage in the task manager.
I posted it on the forums, and never got a response - but at least I notified them of a bug, and pinpointed it pretty accurately. You could say that having access to the source would be a boon here, but the truth is a novice non-coder isn't going to be able to fix it even with the source, and FOSS and proprietary software are just as bad at not fixing user reported bugs.
That leads me to conclude that if I had had access to the source code, it wouldn't have made any difference - except perhaps I would have wasted time looking into it. Source code is great for learning programming, but there's a lot of basics you have to understand first, before you can make sense of stuff like that.
I respect the goals of the OLPC project, but I'm not convinced fully 100% open source is necessary for children to learn computers.
Not in the realm of developing usable applications and OSes, they weren't.
I was downloading Linux distros back in '06. But short of installing them to see if they worked on my hardware, I never really dove in and used one seriously.
Then Ubuntu matured, and I hopped over. It's not being used as my primary desktop yet, but it is working fine as a Samba NAS. Quite fun to hack together. :)
To be worth switching, you have to show how things are going to be BETTER. You have to show that you can do 100% of your job, and that it'll be better. Otherwise, it really isn't worth it.
You're absolutely correct. That's one of the reasons XP is still my desktop.
However, I just finished setting up an Ubuntu NAS - it was quite fun hacking it together! I wouldn't use Ubuntu as a primary desktop, but I am enjoying learning some Linux. :)
On what OS? If that's the behaviour on Linux, than no wonder I failed when attempting bash!
On Windows, s*it will match shit - only matching sit, ssit, sssit, ssssit, etc. would be weird.
I want Gnome's terminal to accept Ctrl+C and Ctrl+V, rather than Ctrl+Shift+C/V!
NetworkManager just died on me yesterday. I had to manually configure /etc/network/interfaces and /etc/resolv.conf
Apparently NetworkManager connects after boot, which was messing up Samba. (no connection detected, so no daemons started) I deleted the connection from it because it wouldn't go static, but the add button wouldn't add any back, so my ethernet connection was gone. I had to manually configure stuff to get it working. Good thing I had a secondary XP box on hand. :P
In the short-term, binary drivers are often a better choice. Some driver is better than no driver, right? But in the long run, free/open source drivers are usually better. Just look at how crashy and difficult to work with the Nvidia drivers have been. Or how severely limited the ATI drivers were.
Intel and VIA/S3 are fine examples of this. I wouldn't want to use their binary drivers - the FOSS ones are so much more advanced.
nVidia gets away with it because they do a good job - not because they're large. (although being large helps, it didn't work for Intel)
Sure you do. You just can't learn how real programmers made it.
But even before I learned how to code stuff, I could spot game bugs and guess what the reason was for their existence.
Same for UI glitches. And for older games, AI glitches.
Something like this really can't compete with a YikeBike.
More speed, looks more comfortable, more control.. probably cheaper too.
There's an EEE that costs $180 CAD. It comes on sale every few weeks at NCIX.com. I think it's one of the older Celeron/8GB-Flash versions.
Still, for a cheap email computer, not bad. I agree that OLPC has a ways to go, but I would take a C7 over a Celeron.