You're referring to pipelining. However, it doesn't help as much as you think. First of all, there are limits to how many files a browser will request in parallel; sites with a large number of resources still require multiple round trips. And second, you don't know what files you need until after you finish downloading the file that requests it (as you point out). So, you get some savings from pipelining, but you're still wasting a rather large amount of time with extra round trips.
It more than matches the speed of Safari, it destroys it. Safari is a traditional browser; establish a connection to the web server (some round trips right there), request and download the requested HTML page (another round trip), download any first-tier needed assets (JS, CSS, images, etc) (likely not all done in parallel, more round trips), download any second-tier assets (example, images from CSS, anything dynamically written by the JS, etc), and so on. All in all, you're probably adding in dozens of round trips at the least. The latency on the 3G link alone (ignoring internet latency) is probably 100+ms for a round trip, so you're adding multiple seconds worth of latency just by being on 3G.
Opera, on the other hand, does absolutely everything server-side. Any requests are being made from a connection that isn't sitting on the other side of a 100+ms wireless link, and they probably do a lot of caching on top of that. The actual data is sent to the client browser in the minimum number of round trips; enough to establish the connection and make the request. All content comes back in one single compressed glob. A page that might have taken 10 seconds to load before can suddenly load in half a second.
There are downsides, of course, to having no client-side javascript. Most web apps require connections to the server to do what was before a local operation. You're effectively streaming any changes to the page from the server to the client (presumably keeping the connection open while looking at the page in case any changes need to be sent), and this is not ideal.
Unfortunately, it's mandatory; Apple won't allow javascript execution locally.
By "malware updates", I mean antiviral or antimalware updates. I mean that you don't need to resort to a live CD, that the updates for these programs can be done overnight easily enough to make it unnecessary.
Malware updates are usually very tiny, and should only take 10-20 minutes to download.
It's not hard to configure the malware app to connect the dialup in the middle of the night, download the update, and disconnect the dialup modem when it's done.
Examples of things that benefit from more than two cores:
- Modern web browsers such as Chrome --- Multi-process architecture means that Flash sucks up a CPU all to itself, while various other tabs/domains are in different processes. Javascript-heavy web-apps or the user of other flash like plugins can easily make 3+ cores worthwhile - Most modern games --- Games are very CPU intensive. Most modern engines do a very good job of taking advantage of multiple cores. Some games even require 2+ cores in order to get any decent performance; MassEffect 2 (Unreal Engine 3) is unplayable on single-core processors - Video encoding --- GPU-accelerated (CUDA, OpenCL, etc) encoding is not yet useful, and isn't likely to be so any time soon. Existing hardware accelerated encoders are extremely limited in flexibility, and are usually of very poor quality (in terms of output). - Multitasking --- You scoff at it, but if I've got some demanding application running and I try to do something else at the same time (such as pause a game, alt-tab out of it, try to watch a video), CPU load starts to add up.
Most peoples' needs can be met with the low end dual-core processors with hyperthreading such as the i3 or i5 series, but these days it's not just anybody who can take advantage of a quad core CPU. Pretty much all gamers, for example.
You're clearly looking in the wrong place. For 26-27" LCD HDTVs, they go for about $280 USD, cheaper than your Xbox 360. Much less if you find them on sale. Clearly, with LCD HDTVs at that price, a 27" CRT is not worth anywhere near $200 (not that any local stores carry CRTs anymore).
So again, the question is, why would somebody pay $299 for a console and then saddle it with a TV that is worth a fraction as much? It'd be like putting an F1 engine into a 1993 Honda Civic.
Re:I actually kind of miss the old combat system
on
Review: Mass Effect 2
·
· Score: 1
It might not make sense logically, but it does encourage a certain type of gameplay that I've found to be a lot more engaging.
Ordinarily I'd say that the effort to do the port would not be worth it considering the small market share. However, ME2 uses Unreal Engine 3, which already has OpenGL bindings (it has to, it runs on platforms that use OpenGL ES like the PS3).
Instead, I'll just point out that you can play the game under Mac OS X under WINE (http://appdb.winehq.org/objectManager.php?sClass=version&iId=19125&iTestingId=49026), but that it will look like crap, run slow, and crashes on occasion.
Not only that, it makes even less sense for the Steam version. Not only is it not in-game (and the game keeps saying "New content available" even though I've downloaded it all), but it doesn't integrate even slightly into Steam. Steam has a whole system for handling DLC, but ME2 has you going to some EA website, registering stuff, and running installer apps to shove stuff into the Steam game directory.
I'd both agree and disagree. The DLC weapons/armour/etc were essentially crap. The DLC armour doesn't let you customize anything, and the DLC weapons are better than your default weapons, but very quickly get outpaced.
But the Normandy crash site mission was a nice little link to the first game (and about the only place that the death of the XO gets any coverage), and Zaeed, while useless in combat, has a lot of fun stories to tell if you go back to him after each mission. Also, you can't play with the trash compactor without him to unlock that room on the ship:P
Yes. One of the biggest issues with all the ME1 side quests were that they all shared a very limited subset of game maps. Clearing out the same damned underground bunker for the umpteenth time wasn't really made any more interesting by "Oh, this time they're crazed biotics that you need to quell" or "This time it's a crazy computer".
Of those kinds of side quests, perhaps 15 hours into the game, I have not had a single one that duplicated content. One side quest, in fact, had multiple locations, and completing one aspect of it unlocked the next part.
There are still the random tiny things, like "Oh, so and so lost his credit chit, look around to find it" or "Chef wants better food, buy him some", but those are the kind of quests that virtually solve themselves; you'll get the assignment and at some point you'll say "Oh, here's the food I was supposed to buy." It just sort of gives a little bit of depth to the world without really taking any effort.
So, the situation is quite improved over ME1.
Re:I actually kind of miss the old combat system
on
Review: Mass Effect 2
·
· Score: 2, Insightful
The in-game explanation was that the guns shaved off an extremely small sliver of metal from a block, and then accelerated them to extreme speeds in order to get the same energy as a traditional projectile. The ammo was unlimited because the shavings from the ammo block were incredibly small.
The in-game explanation as to why they added "ammo" was that they decided that dissipating the excess heat from weapons was limiting their firing speed too much (you could only fire as fast as you could cool). The new system dumps the heat directly into a sink (the cartridges), which you discard. In one sense, you're not emptying your ammo clip when you fire, you're filling it.
So, they do explain this all in the story, through the codex.
It's important to remember that the ammo that you pick up is universal; you pick up one clip and all your weapons get more ammo. I found that this changed the weapons dynamic in a way that I actually enjoyed. In ME1, I used nothing but the assault rifle and on rare occasions the sniper rifle. In ME2, because they made the weapons more unique, I really find myself changing up my weapons based on the type of enemy, the distance to the enemy, etc. in order to conserve ammunition. The fact that there is only one ammo type for all weapons (except heavy) simplifies this to make it work much better.
Re:I actually kind of miss the old combat system
on
Review: Mass Effect 2
·
· Score: 1
A better question is why you were out of cover to begin with. My shields go down often, and I usually have enough health to make it to cover. The rare few times that I don't, it's usually my own damned fault for letting myself get exposed.
What you're really complaining about is that ME2 forces you to use cover, and you don't like that. ME2 doesn't let you run and gun, it punishes you for that.
The author could not be more wrong. Looking up some blog posts from Asa Dotzler, 2-5% of Firefox users are on Linux, 7% are on OS X, and the rest are Windows. Right off the bat, that means that 95-98% of Firefox users are on platforms that either do or will eventually support h.264 out of the box.
The "eventually" part is important because at some point, the percentage of people using a version before Windows 7 will be small enough that either Firefox will drop support or they will be statistically insignificant. For example, Firefox 3 and above do not run on Win98.
Of the Linux users, some will already have some sort of h.264 support installed. I'll emphasize the already installed part because it circumvents the "jump through arcane hoops to download and install software (whose use is legally questionable)" bit, as in those users would have had it installed anyhow, for whatever reason.
So if we make up a number and assume that half of Linux users have done so, we could come to a figure of 1.75% of Firefox users eventually supporting h.264 without any effort on the users' part.
Now we turn it around; 98.25% of users can play h.264, it is the 1.75% of users who are selfish for refusing to let the VAST majority enjoy higher quality video. The selfish tiny free software minority are holding back technological progress for the rest of us!
And all users with Windows 7 have already paid for the h.264 license, and have h.264 decoding built-in. It won't be long before the majority of Firefox's userbase will have a valid h.264 license.
So, what's the problem with Firefox supporting h.264 if a local option is available? Supporting h.264 for (eventually) the majority of clients is better than not supporting it at all.
The lowest any game got was 67ms of latency, and Unreal Tournament 3 measured 100ms-133ms.
There's a *LOT* of leeway in there to account for 20ms of network latency.
The problem with the PC Perspective article, other than that he was using a borrowed account, was that the OnLive beta has a limited number of PoPs active; the beta is limited to the geographical areas near the PoPs.
The author of the article alternatively either had OnLive refuse to connect since the latency was too high, or received warnings that the latency was too high; he ignored them. It's obvious that that would seriously impact performance.
I'll do some math... I'm sitting on a fibre connection, pinging a fibre connection 1752 miles away, and getting peak latency of 60ms. Let's assume the user is on DSL and add 10ms of latency to account for that, 70ms. OnLive has a maximum range of 1000 miles, so they'd have 34+10ms of latency, or 44ms of network latency. You're already over one frame worth of latency, but this is worst-case for a user at the edge of the range of a PoP.
Now, if the author of the article was at least twice as far away, you'd be expecting him to get at least 78ms of latency or more... We're starting to talk about 2-3 frames of lag at least. And that is where responsiveness starts going down the tubes.
He complained about going half a screen past enemies in UT3; if we assume that a twitch action to rotate by 90 degrees takes one fifth of a second (which is probably slower than reality), and the author's ill-adviced "preview" of OnLive has added perhaps 50ms of latency over the intended experience. That would mean that he would overshoot his mark by about 23 degrees, which sounds about right from his experience.
OnLive will always have problems with twitch games; I suspect games like UT3 will work for users who are quite close to the PoP and experience good conditions. But most other games that are not latency sensitive should work fine for anyone in range of a PoP. The author, however, was far enough away that even those games start to have issues.
If the author had done all his previewing at the home of this "friend of a friend", we might have seen something more representative of what the final service will behave like!
Well, the thing is, on those kinds of devices, often what you describe is the only thing people want to do with them anyhow. To account for a few of the odd cases such as wanting to play video from a USB stick, Chrome OS has support for that added.
The argument is, if that's all you're going to do with a netbook or smartbook, why waste resources on the rest of the OS (window manager, all those services, etc) when you don't need them? Why not just strip it down right to what you'll actually use.
It's not for everyone, but a lot of people don't need anything more.
In terms of "Google closing it down, and you're done.", would you want to keep using a certain version of Windows if Microsoft suddenly stopped releasing security patches for it? The mechanism of death would be different, the result would be the same.
I've seen more than a few netbooks (including some of Asus) that come close enough to their battery claims to consider them not exaggerated. When they said that the Eee PC 1000HE would get a 7 hour battery life, it actually did.
The frequency is irrelevant; the power just isn't there. Even if you captured ALL energy from the wifi device and converted it into electricity with 100% efficiency, it would still take more than a month to charge a laptop battery. A foot from the device, it would take decades.
The truth is that every mobile Intel CPU since at least the Pentium M has featured SpeedStep, with OS support dating back to WinXP (although widgets could enable support on Win2K). Back then, SpeedStep would dynamically clock the CPU between 1GHz and 1.6GHz based on CPU load. The voltage would also change accordingly. These days, all their products support it, even desktop and server processors. What Asus might be doing is underclocking further to try to eke out additional savings. That's not really terribly noteworthy.
The thing that's "new" here is the dynamic real-time switching between an IGP and discrete GPU... but that doesn't produce the power savings. After all, the vast majority of notebooks ship with *only* the IGP, so they're already getting the "maximum" power savings for graphics.
The 12 hour battery life probably comes from a variety of sources:
1) Probably using an LED backlight, which consume less power than traditional cold cathode backlights
2) Slight savings from underclocking the CPU more than SpeedStep normally does
3) Various other settings might be tweaked (HDD power settings) to be more optimal than default
4) Big battery
5) Consciously choosing lower power components. Lower power (slower) HDDs, avoiding discrete controllers if the chipset can do it, etc.
This notebook seems focused on getting decent performance combined with good battery life. However, for those of us who just want the good battery life and aren't as concerned with performance, it's not very interesting.
Slashdot Mirror
There's still only one round trip (plus possibly extra if they're using TCP). 100 + 100 < 100 * 10
You're referring to pipelining. However, it doesn't help as much as you think. First of all, there are limits to how many files a browser will request in parallel; sites with a large number of resources still require multiple round trips. And second, you don't know what files you need until after you finish downloading the file that requests it (as you point out). So, you get some savings from pipelining, but you're still wasting a rather large amount of time with extra round trips.
It more than matches the speed of Safari, it destroys it. Safari is a traditional browser; establish a connection to the web server (some round trips right there), request and download the requested HTML page (another round trip), download any first-tier needed assets (JS, CSS, images, etc) (likely not all done in parallel, more round trips), download any second-tier assets (example, images from CSS, anything dynamically written by the JS, etc), and so on. All in all, you're probably adding in dozens of round trips at the least. The latency on the 3G link alone (ignoring internet latency) is probably 100+ms for a round trip, so you're adding multiple seconds worth of latency just by being on 3G.
Opera, on the other hand, does absolutely everything server-side. Any requests are being made from a connection that isn't sitting on the other side of a 100+ms wireless link, and they probably do a lot of caching on top of that. The actual data is sent to the client browser in the minimum number of round trips; enough to establish the connection and make the request. All content comes back in one single compressed glob. A page that might have taken 10 seconds to load before can suddenly load in half a second.
There are downsides, of course, to having no client-side javascript. Most web apps require connections to the server to do what was before a local operation. You're effectively streaming any changes to the page from the server to the client (presumably keeping the connection open while looking at the page in case any changes need to be sent), and this is not ideal.
Unfortunately, it's mandatory; Apple won't allow javascript execution locally.
Why would the womp rats be stationary? You're just beating a dead womp rat.
By "malware updates", I mean antiviral or antimalware updates. I mean that you don't need to resort to a live CD, that the updates for these programs can be done overnight easily enough to make it unnecessary.
Malware updates are usually very tiny, and should only take 10-20 minutes to download.
It's not hard to configure the malware app to connect the dialup in the middle of the night, download the update, and disconnect the dialup modem when it's done.
- Preemptable (Linux isn't)
Sure it is:
http://en.wikipedia.org/wiki/RTLinux
Or just a patchset:
http://rt.wiki.kernel.org/index.php/CONFIG_PREEMPT_RT_Patch
http://www.linuxatemyram.com/
Examples of things that benefit from more than two cores:
- Modern web browsers such as Chrome
--- Multi-process architecture means that Flash sucks up a CPU all to itself, while various other tabs/domains are in different processes. Javascript-heavy web-apps or the user of other flash like plugins can easily make 3+ cores worthwhile
- Most modern games
--- Games are very CPU intensive. Most modern engines do a very good job of taking advantage of multiple cores. Some games even require 2+ cores in order to get any decent performance; MassEffect 2 (Unreal Engine 3) is unplayable on single-core processors
- Video encoding
--- GPU-accelerated (CUDA, OpenCL, etc) encoding is not yet useful, and isn't likely to be so any time soon. Existing hardware accelerated encoders are extremely limited in flexibility, and are usually of very poor quality (in terms of output).
- Multitasking
--- You scoff at it, but if I've got some demanding application running and I try to do something else at the same time (such as pause a game, alt-tab out of it, try to watch a video), CPU load starts to add up.
Most peoples' needs can be met with the low end dual-core processors with hyperthreading such as the i3 or i5 series, but these days it's not just anybody who can take advantage of a quad core CPU. Pretty much all gamers, for example.
You're clearly looking in the wrong place. For 26-27" LCD HDTVs, they go for about $280 USD, cheaper than your Xbox 360. Much less if you find them on sale. Clearly, with LCD HDTVs at that price, a 27" CRT is not worth anywhere near $200 (not that any local stores carry CRTs anymore).
So again, the question is, why would somebody pay $299 for a console and then saddle it with a TV that is worth a fraction as much? It'd be like putting an F1 engine into a 1993 Honda Civic.
It might not make sense logically, but it does encourage a certain type of gameplay that I've found to be a lot more engaging.
Ordinarily I'd say that the effort to do the port would not be worth it considering the small market share. However, ME2 uses Unreal Engine 3, which already has OpenGL bindings (it has to, it runs on platforms that use OpenGL ES like the PS3).
Instead, I'll just point out that you can play the game under Mac OS X under WINE (http://appdb.winehq.org/objectManager.php?sClass=version&iId=19125&iTestingId=49026), but that it will look like crap, run slow, and crashes on occasion.
Considering that you can buy a new HDTV/monitor for about 2/3 the cost of your 360, I'm not sure your complaint is really all that valid.
Yes, if you use a $300 console with a $20 TV, you will not get an optimal experience.
Not only that, it makes even less sense for the Steam version. Not only is it not in-game (and the game keeps saying "New content available" even though I've downloaded it all), but it doesn't integrate even slightly into Steam. Steam has a whole system for handling DLC, but ME2 has you going to some EA website, registering stuff, and running installer apps to shove stuff into the Steam game directory.
I'd both agree and disagree. The DLC weapons/armour/etc were essentially crap. The DLC armour doesn't let you customize anything, and the DLC weapons are better than your default weapons, but very quickly get outpaced.
But the Normandy crash site mission was a nice little link to the first game (and about the only place that the death of the XO gets any coverage), and Zaeed, while useless in combat, has a lot of fun stories to tell if you go back to him after each mission. Also, you can't play with the trash compactor without him to unlock that room on the ship :P
Yes. One of the biggest issues with all the ME1 side quests were that they all shared a very limited subset of game maps. Clearing out the same damned underground bunker for the umpteenth time wasn't really made any more interesting by "Oh, this time they're crazed biotics that you need to quell" or "This time it's a crazy computer".
Of those kinds of side quests, perhaps 15 hours into the game, I have not had a single one that duplicated content. One side quest, in fact, had multiple locations, and completing one aspect of it unlocked the next part.
There are still the random tiny things, like "Oh, so and so lost his credit chit, look around to find it" or "Chef wants better food, buy him some", but those are the kind of quests that virtually solve themselves; you'll get the assignment and at some point you'll say "Oh, here's the food I was supposed to buy." It just sort of gives a little bit of depth to the world without really taking any effort.
So, the situation is quite improved over ME1.
The in-game explanation was that the guns shaved off an extremely small sliver of metal from a block, and then accelerated them to extreme speeds in order to get the same energy as a traditional projectile. The ammo was unlimited because the shavings from the ammo block were incredibly small.
The in-game explanation as to why they added "ammo" was that they decided that dissipating the excess heat from weapons was limiting their firing speed too much (you could only fire as fast as you could cool). The new system dumps the heat directly into a sink (the cartridges), which you discard. In one sense, you're not emptying your ammo clip when you fire, you're filling it.
So, they do explain this all in the story, through the codex.
It's important to remember that the ammo that you pick up is universal; you pick up one clip and all your weapons get more ammo. I found that this changed the weapons dynamic in a way that I actually enjoyed. In ME1, I used nothing but the assault rifle and on rare occasions the sniper rifle. In ME2, because they made the weapons more unique, I really find myself changing up my weapons based on the type of enemy, the distance to the enemy, etc. in order to conserve ammunition. The fact that there is only one ammo type for all weapons (except heavy) simplifies this to make it work much better.
A better question is why you were out of cover to begin with. My shields go down often, and I usually have enough health to make it to cover. The rare few times that I don't, it's usually my own damned fault for letting myself get exposed.
What you're really complaining about is that ME2 forces you to use cover, and you don't like that. ME2 doesn't let you run and gun, it punishes you for that.
The author could not be more wrong. Looking up some blog posts from Asa Dotzler, 2-5% of Firefox users are on Linux, 7% are on OS X, and the rest are Windows. Right off the bat, that means that 95-98% of Firefox users are on platforms that either do or will eventually support h.264 out of the box.
The "eventually" part is important because at some point, the percentage of people using a version before Windows 7 will be small enough that either Firefox will drop support or they will be statistically insignificant. For example, Firefox 3 and above do not run on Win98.
Of the Linux users, some will already have some sort of h.264 support installed. I'll emphasize the already installed part because it circumvents the "jump through arcane hoops to download and install software (whose use is legally questionable)" bit, as in those users would have had it installed anyhow, for whatever reason.
So if we make up a number and assume that half of Linux users have done so, we could come to a figure of 1.75% of Firefox users eventually supporting h.264 without any effort on the users' part.
Now we turn it around; 98.25% of users can play h.264, it is the 1.75% of users who are selfish for refusing to let the VAST majority enjoy higher quality video. The selfish tiny free software minority are holding back technological progress for the rest of us!
And all users with Windows 7 have already paid for the h.264 license, and have h.264 decoding built-in. It won't be long before the majority of Firefox's userbase will have a valid h.264 license.
So, what's the problem with Firefox supporting h.264 if a local option is available? Supporting h.264 for (eventually) the majority of clients is better than not supporting it at all.
Apparently not as much as you think. Check out these measured latencies from XBox 360 games:
http://www.eurogamer.net/articles/digitalfoundry-lag-factor-article?page=3
The lowest any game got was 67ms of latency, and Unreal Tournament 3 measured 100ms-133ms.
There's a *LOT* of leeway in there to account for 20ms of network latency.
The problem with the PC Perspective article, other than that he was using a borrowed account, was that the OnLive beta has a limited number of PoPs active; the beta is limited to the geographical areas near the PoPs.
The author of the article alternatively either had OnLive refuse to connect since the latency was too high, or received warnings that the latency was too high; he ignored them. It's obvious that that would seriously impact performance.
I'll do some math... I'm sitting on a fibre connection, pinging a fibre connection 1752 miles away, and getting peak latency of 60ms. Let's assume the user is on DSL and add 10ms of latency to account for that, 70ms. OnLive has a maximum range of 1000 miles, so they'd have 34+10ms of latency, or 44ms of network latency. You're already over one frame worth of latency, but this is worst-case for a user at the edge of the range of a PoP.
Now, if the author of the article was at least twice as far away, you'd be expecting him to get at least 78ms of latency or more... We're starting to talk about 2-3 frames of lag at least. And that is where responsiveness starts going down the tubes.
He complained about going half a screen past enemies in UT3; if we assume that a twitch action to rotate by 90 degrees takes one fifth of a second (which is probably slower than reality), and the author's ill-adviced "preview" of OnLive has added perhaps 50ms of latency over the intended experience. That would mean that he would overshoot his mark by about 23 degrees, which sounds about right from his experience.
OnLive will always have problems with twitch games; I suspect games like UT3 will work for users who are quite close to the PoP and experience good conditions. But most other games that are not latency sensitive should work fine for anyone in range of a PoP. The author, however, was far enough away that even those games start to have issues.
If the author had done all his previewing at the home of this "friend of a friend", we might have seen something more representative of what the final service will behave like!
Well, the thing is, on those kinds of devices, often what you describe is the only thing people want to do with them anyhow. To account for a few of the odd cases such as wanting to play video from a USB stick, Chrome OS has support for that added.
The argument is, if that's all you're going to do with a netbook or smartbook, why waste resources on the rest of the OS (window manager, all those services, etc) when you don't need them? Why not just strip it down right to what you'll actually use.
It's not for everyone, but a lot of people don't need anything more.
In terms of "Google closing it down, and you're done.", would you want to keep using a certain version of Windows if Microsoft suddenly stopped releasing security patches for it? The mechanism of death would be different, the result would be the same.
I've seen more than a few netbooks (including some of Asus) that come close enough to their battery claims to consider them not exaggerated. When they said that the Eee PC 1000HE would get a 7 hour battery life, it actually did.
The frequency is irrelevant; the power just isn't there. Even if you captured ALL energy from the wifi device and converted it into electricity with 100% efficiency, it would still take more than a month to charge a laptop battery. A foot from the device, it would take decades.
The truth is that every mobile Intel CPU since at least the Pentium M has featured SpeedStep, with OS support dating back to WinXP (although widgets could enable support on Win2K). Back then, SpeedStep would dynamically clock the CPU between 1GHz and 1.6GHz based on CPU load. The voltage would also change accordingly. These days, all their products support it, even desktop and server processors. What Asus might be doing is underclocking further to try to eke out additional savings. That's not really terribly noteworthy.
The thing that's "new" here is the dynamic real-time switching between an IGP and discrete GPU... but that doesn't produce the power savings. After all, the vast majority of notebooks ship with *only* the IGP, so they're already getting the "maximum" power savings for graphics.
The 12 hour battery life probably comes from a variety of sources:
1) Probably using an LED backlight, which consume less power than traditional cold cathode backlights
2) Slight savings from underclocking the CPU more than SpeedStep normally does
3) Various other settings might be tweaked (HDD power settings) to be more optimal than default
4) Big battery
5) Consciously choosing lower power components. Lower power (slower) HDDs, avoiding discrete controllers if the chipset can do it, etc.
This notebook seems focused on getting decent performance combined with good battery life. However, for those of us who just want the good battery life and aren't as concerned with performance, it's not very interesting.