Keep in mind that the copper network is going to be absolutely useless in a few years when the exchanges get switched off, at which point anyone who does not have fibre/satellite/mobile is going to be in trouble.
You really should RTFA. Google is the hypocrite here.
Basically, Google was all, "Patents are evil! h.264 is patented and evil! Use WebM instead, we're giving it away free, out of the goodness of our patent-hating, royalty-eschewing hearts!" Remember that?
Now Google, by way of Motorola, is proving its own point by charging obscene percentage-based royalties on h.264. Meaning, instead of charging a set rate, it charges a percentage of the sale price of the device. So apparently, to Motogoogle, the exact same h.264 decoder function is worth 3x as much in a $300 phone than in a $100 phone.
Apple is has filed complaints against Motogoogle over the exact same thing, but of course Apple is evil, like Microsoft, and unlike Google, which Does No Evil.
Forgive me if I'm wrong, but isn't this the exact same strategy that was proposed in a previous/. article as a means of allowing FOSS use of FRAND-patented technologies? Say what you will about patents, FRAND-terms could clean this mess up a bit, especially if they are compatible with FOSS usage.
Once this comes to market, economies of scale will make the hardware relatively cheap. The rest is software. If they go with something like Android, it's open source. If they don't, it'll only be a matter of time before someone reverse engineers it to run an open-source stack.
Consider that we already have $100 USB EEG headsets, and there's nothing stopping you from using your own software with those either. I would rather have Google working on these than someone like Apple any day.
Archos sells Android tablets for $140*. You can get Chinese ones as cheap as $85, but Archos is the cheapest I know of that's actually GPL compliant. (If anyone knows of a cheaper, GPL-compliant one, I'm all ears.)
* Yes, the hardware is nowhere near that of the TouchPad. But we're looking at sustainable pricing here, not clearances.
I'm surprised it doesn't support this already. The public transport site for my area has supported that as long as I can remember - it makes it easy to find out how to get from A to B with minimal fuss.
It's the right idea, but Ubuntu on ARM is nowhere near ready. It's crazy buggy, and you're going to miss out on hardware accelerated graphics for the vast majority of applications, because most apps still expect OpenGL, and can't take advantage of OpenGL ES.
This isn't limited to Ubuntu. Hardware acceleration is probably the hardest thing to get working on ARM Linux as it stands, because the drivers are binary blobs and Android has a completely different architecture (e.g. no X server). Everything else (bluetooth, audio, etc.) has a similar architecture which means that even if the driver is a binary blob, you can still just load the kernel module and use it. Realistically, getting hardware acceleration working would require a massive amount of reverse engineering, and would take several years to complete for just one chipset, by which point its successors would have been released.
The other problem is that devices like the Atrix, while an interesting concept, aren't really ready to host desktop Linux yet. The performance just isn't there yet. I suspect that the next crop of smartphones, with dual core A15s or quad-core A9s, those will probably do a decent job at it.
Desktop Ubuntu might be a poor fit, but a customized Linux distro should run fine. Keep in mind that the N900 ran a derivative of Debian on 600 MHz with 256 MB RAM. A Tegra 2 will run Kubuntu comfortably (as long as you don't open too many tabs in Firefox).
My experience stems from running Kubuntu on my Asus Transformer, so take that as you will.
If I can make a Droid 4 run Ubuntu I'll buy one ASAP. Once you can run a regular GNU/Linux distro you can customize it to do anything the hardware is capable of. That's the only problem with my N900, the hardware's old and out of date.
Modifying an Android kernel to run Ubuntu isn't very hard, as long as you have the ability to flash the new kernel. There just isn't much interest in doing so right now given the lack of suitable mobile operating systems (compared to certain Android tablets, which can get by with desktop OSs), but I think it'll pick up once Kubuntu Mobile is released.
Worse, ReCAPTCHA's idea of the "right" answer is crowdsourced. It's possible for bots to pollute the ReCAPTCHA database, by providing the same wrong answer more than once. You only have to get one of the words right, so if you can read one, a junk response for the other works. This goes into the database as a vote for the "right answer", to be presented to someone else later. I sometimes type "whatever" when one of the images is unreadable.
Not just bots - humans can (unintentionally) do it as well. Sparkfun (an electronics hobbyist site) recently had a giveaway in order to stress test their servers. Several thousand people were solving CAPTCHAs as quickly as possible. There was a noticeable drop in the accuracy of the answers required, since a lot of people were taking shortcuts in entering them.
It usually isn't that bad, but then Akiyuki Shinbo happens.
For those who didn't get the reference, Shinbo is a notable Shaft (an anime studio) director. Most of his works involve rather abstract imagery and sometimes flashing walls of text on screen for under a second. ASS subs are so capable that they can be used to add a subtitle overlay that matches the original video sufficiently well to be indistinguishable in most scenarios.
Agreed. Just yesterday I was trying to debug some audio drivers under Kubuntu, and was using about 5 virtual desktops with 2 monitors. Gnome 3 just doesn't work for a dual monitor setup. In comparison, KDE was laughably easy to configure to suit my needs. The only way you can meet the needs of the largest number of people is if you make your product configurable, and make it easy to do so. This is the very reason why the Linux kernel is so widely used - it's easy to port to different platforms; to configure it to suit different people's needs.
IMO, the KDE Project has superior vision as well. While Gnome 3 and Unity were the first to market, KDE is taking their time with Plasma Active, and designing it from the start to work for both tablets and mobiles, with reference platforms for each. Given that Maemo/Meego has imploded and Android is in many ways very limiting compared to a desktop OS, I believe that Plasma Active will become increasingly relevant among geeks.*
* While Gnome 3 nay have been motivated by a desire to be more accessible to regular users, it goes without saying that the vast majority of people using Linux do so because they either are geeks, or it was recommended to them by one. By alienating a large number of us, they're eating the seed corn.
The problems with putting Plasma Active on different tablets are almost the same as the problems with putting Cyanogen on different phones. That hasn't stopped Cyanogen from becoming extremely popular.
The problem with putting Plasma Active on any number of tablets is the nature of ARM platforms themselves:
- No two boards are alike, all different in how GPIOs are hooked up and what they do, even if the SoC is the same.
This is true. However, this is handled in the kernel, and since Android uses a fork of the GPLed Linux kernel, recompiling the kernel with support for the requisite filesystems, etc. is trivial. (Some effort may be required to handle differences between Android and Linux architectures for things like sound drivers.)
- Only the hardware vendor gets the sources for some drivers, usually GPU. Thus putting something like Xorg on these devices is exceedingly hard. Spark works around it with a completely reverse-engineered Mali driver.
Xorg isn't hard - it doesn't even need a GPU. The hard part is hardware acceleration, without which you can say goodbye to video playback at any decent resolution. This is probably the biggest problem right now. The only solution I can think of is for someone like the nouveau project to create drivers for a common chipset like the Tegra series, which would take a lot of resources and a considerable amount of time.
For other devices, like bluetooth/wifi modules, the binary blobs from Android can be used as is.
- Many tablets put up a fight when it comes to installing 3rd party firmware, unlike most PMPs/Routers. Even the Transformer Prime is locked down and ASUS has yet to release the promised unlock.
I agree that this can be problematic, and that we are at the manufacturer's mercy in this regard (unless someone is able to take the tablet apart and read the keys directly off the chip's fuses). Asus announced that they would release the unlock this month, and as long as they do so they're far ahead of their competitors in this regard. (It's only a flashboot unlock and not one for nvflash, which means you can't repartition the drive, but it's still better than nothing.) It's likely that they will continue to release unlocks for all future devices, which would provide a partial solution to this problem.
- Tablets with Android tend to have older kernels, with little to nothing having gone upstream thus making adaptation harder.
So it's a nice thought, and there are attempts, but it's nowhere near as easy in the tablet and handset space as it is in the PMP/Router space.
There is some work at the moment on merging some of the changes in the Android kernel (particularly those involving power management and drivers) into the mainline, which is due to be completed by 3.4. In the meantime, Android didn't fork the kernel that long ago - what changes have their been made since then that offer a substantial impediment to running Plasma Active (or any other modern distro)?
Running Plasma Active on numerous tablets may not be as simple as running an installation program, but it's definitely not as hard as many would believe. The only reason we haven't seen this happening yet is because it's incomplete.
Rockby is reasonably good for analog parts, though their website's search is somewhat lacking. Jaycar has a minuscule selection in comparison. The reality is that for electronic components, the demand for the really interesting ones is just too low to justify a brick and mortar shop keeping them in stock. Resistors, capacitors, etc. you'll find, but good luck finding a microprocessor more powerful than an Arduino or a display better than a simple 16x2 character display.
I usually just order stuff from Sparkfun, or AustralianRobotics if I need it sooner. I'd probably use Digikey as well if they didn't charge $35 for shipping.
That said, I don't see how this keyboard can possibly work if it only has 32 combinations, unless there's some extra modifier keys that you use with the other hand or something.
Apparently he's planning to extend it to include a 6th key that can be activated by the left-hand thumb. That gives 64 chars, and could be extended to 96 if you added a 7th that was adjacent and mutually exclusive to the 6th. Realistically, I don't think it's intended as a complete replacement for the conventional keyboard layout; if it's used for everything except the less commonly used symbols that would be sufficient.
That's not why magnet links are useful. You can be certain that the provision of magnet links will be held to be the facilitation of copyright infringement at some point or another, since from a legal perspective they're no different to torrent files - both effectively contain instructions on how to get the content, it's just that magnet links added a 'follow the directions you find here' step. It doesn't matter how many links there are, only whether the dominant usage is illegal.
Magnet links are useful because they offer increased decentralization. We used to need trackers. Then there was DHT. We used to need to the entire torrent file. Now we just need the hash. We currently need indexers. Once the technology in Tribler takes off, we won't even need that.
A completely decentralized P2P network has no central points of failure (e.g. trackers) that can be sued and taken down. All it has is users, and while the RIAA has demonstrated a willingness to go sue individual users, they have not had much success in countries which didn't appreciate the clear abuses of the judicial system, nor in cases where the individuals took them to court (you're dreaming if you think they'll recover enough via damages to cover legal costs).
To that end, I'd love to see a 100% Linux phone, nothing relying on Google with it's bits locked down or tracking
They're also working on 'Kubuntu Mobile' (alpha release's available from the site), which presumably uses Plasma Active as well. The reference platform for that is also the Nokia N900, though as modifying an Android kernel to run Ubuntu doesn't take a great deal of effort, it won't be long at all before someone has it running on the Galaxy Nexus (or its successor).
Australia and North America, on the other hand, are much more spread out, and the number of GSM cells that would be needed to provide acceptable coverage to rural areas would be too expensive for the likely revenue from them.
As an Australian, I'm fairly certain we standardized on GSM (2G) and UMTS (3G). A quick Google indicates that whatever vestigial CDMA reception we had was shutdown in 2008.
I thought prohibiting students from sharing past copies of tests was a standard and acceptable method. Is it because they are using copyright to attack the practice?
This is news to me. Maybe it's different here in Australia, but it's common practice for universities* to make the last decade's worth of exams available to students. Admittedly they are under copyright and not publicly accessible, but that true of most of the course content. It's rather beneficial as well, since hey constitute the main form of revision for most students once their done reading through their notes. Most lecturers even spend the last lecture or two going over the answers to the previous year's paper.
*The Department of Education did the same thing with the standardized year 12 exams, which were made publicly available.
I'm an engineering student and I have noticed that most of the time, the general theorem that applies is that the interestingness of lecture is inversely proportional to the technologic level used.
In other words, someone in the theatre who'll use blackboard/scribbled projection tend to be almost universally amazing, those that use common "fill in gaps" projections tend to be OK , and lecturers using powerpoint tend to be the "gouge out eyes" sort of boring.
I'm an (Australian) engineering student, and almost all of my lecturers will use Powerpoint slides and end up writing all over them, to the extent that they often added mostly blank slides or wide margins purely for that purpose. The Powerpoints are published on the university website beforehand, and recorded copies of the lectures are usually made available afterwards.
The only subjects I've ever had trouble with have been ones where the lecturers avoided Powerpoint and just scribbled on the projector. Their structure wasn't as clear, and it was harder to decipher their handwriting.
When I studied law, it was pretty similar, except instead of scribbling they just talked more.
Slashdot Mirror
Keep in mind that the copper network is going to be absolutely useless in a few years when the exchanges get switched off, at which point anyone who does not have fibre/satellite/mobile is going to be in trouble.
You really should RTFA. Google is the hypocrite here.
Basically, Google was all, "Patents are evil! h.264 is patented and evil! Use WebM instead, we're giving it away free, out of the goodness of our patent-hating, royalty-eschewing hearts!" Remember that?
Now Google, by way of Motorola, is proving its own point by charging obscene percentage-based royalties on h.264. Meaning, instead of charging a set rate, it charges a percentage of the sale price of the device. So apparently, to Motogoogle, the exact same h.264 decoder function is worth 3x as much in a $300 phone than in a $100 phone.
Apple is has filed complaints against Motogoogle over the exact same thing, but of course Apple is evil, like Microsoft, and unlike Google, which Does No Evil.
Forgive me if I'm wrong, but isn't this the exact same strategy that was proposed in a previous /. article as a means of allowing FOSS use of FRAND-patented technologies? Say what you will about patents, FRAND-terms could clean this mess up a bit, especially if they are compatible with FOSS usage.
Once this comes to market, economies of scale will make the hardware relatively cheap. The rest is software. If they go with something like Android, it's open source. If they don't, it'll only be a matter of time before someone reverse engineers it to run an open-source stack.
Consider that we already have $100 USB EEG headsets, and there's nothing stopping you from using your own software with those either. I would rather have Google working on these than someone like Apple any day.
Archos sells Android tablets for $140*. You can get Chinese ones as cheap as $85, but Archos is the cheapest I know of that's actually GPL compliant.
(If anyone knows of a cheaper, GPL-compliant one, I'm all ears.)
* Yes, the hardware is nowhere near that of the TouchPad. But we're looking at sustainable pricing here, not clearances.
Maemo will be reincarnated as Kubuntu Mobile - it's being used as the foundation, as I understand it.
I'm surprised it doesn't support this already. The public transport site for my area has supported that as long as I can remember - it makes it easy to find out how to get from A to B with minimal fuss.
It's the right idea, but Ubuntu on ARM is nowhere near ready. It's crazy buggy, and you're going to miss out on hardware accelerated graphics for the vast majority of applications, because most apps still expect OpenGL, and can't take advantage of OpenGL ES.
This isn't limited to Ubuntu. Hardware acceleration is probably the hardest thing to get working on ARM Linux as it stands, because the drivers are binary blobs and Android has a completely different architecture (e.g. no X server). Everything else (bluetooth, audio, etc.) has a similar architecture which means that even if the driver is a binary blob, you can still just load the kernel module and use it. Realistically, getting hardware acceleration working would require a massive amount of reverse engineering, and would take several years to complete for just one chipset, by which point its successors would have been released.
The other problem is that devices like the Atrix, while an interesting concept, aren't really ready to host desktop Linux yet. The performance just isn't there yet. I suspect that the next crop of smartphones, with dual core A15s or quad-core A9s, those will probably do a decent job at it.
Desktop Ubuntu might be a poor fit, but a customized Linux distro should run fine. Keep in mind that the N900 ran a derivative of Debian on 600 MHz with 256 MB RAM. A Tegra 2 will run Kubuntu comfortably (as long as you don't open too many tabs in Firefox).
My experience stems from running Kubuntu on my Asus Transformer, so take that as you will.
If I can make a Droid 4 run Ubuntu I'll buy one ASAP. Once you can run a regular GNU/Linux distro you can customize it to do anything the hardware is capable of. That's the only problem with my N900, the hardware's old and out of date.
Modifying an Android kernel to run Ubuntu isn't very hard, as long as you have the ability to flash the new kernel. There just isn't much interest in doing so right now given the lack of suitable mobile operating systems (compared to certain Android tablets, which can get by with desktop OSs), but I think it'll pick up once Kubuntu Mobile is released.
Worse, ReCAPTCHA's idea of the "right" answer is crowdsourced. It's possible for bots to pollute the ReCAPTCHA database, by providing the same wrong answer more than once. You only have to get one of the words right, so if you can read one, a junk response for the other works. This goes into the database as a vote for the "right answer", to be presented to someone else later. I sometimes type "whatever" when one of the images is unreadable.
Not just bots - humans can (unintentionally) do it as well. Sparkfun (an electronics hobbyist site) recently had a giveaway in order to stress test their servers. Several thousand people were solving CAPTCHAs as quickly as possible. There was a noticeable drop in the accuracy of the answers required, since a lot of people were taking shortcuts in entering them.
It usually isn't that bad, but then Akiyuki Shinbo happens.
For those who didn't get the reference, Shinbo is a notable Shaft (an anime studio) director. Most of his works involve rather abstract imagery and sometimes flashing walls of text on screen for under a second. ASS subs are so capable that they can be used to add a subtitle overlay that matches the original video sufficiently well to be indistinguishable in most scenarios.
So if Google wanted to they could have a tivoized phone with not a whiff of source for anything but the kernel out by the end of the day.
Most Android phones are tivoized, since the kernel is GPLv2.
Who needs a boundary? Just use two immiscible liquids.
Agreed. Just yesterday I was trying to debug some audio drivers under Kubuntu, and was using about 5 virtual desktops with 2 monitors. Gnome 3 just doesn't work for a dual monitor setup. In comparison, KDE was laughably easy to configure to suit my needs. The only way you can meet the needs of the largest number of people is if you make your product configurable, and make it easy to do so. This is the very reason why the Linux kernel is so widely used - it's easy to port to different platforms; to configure it to suit different people's needs.
IMO, the KDE Project has superior vision as well. While Gnome 3 and Unity were the first to market, KDE is taking their time with Plasma Active, and designing it from the start to work for both tablets and mobiles, with reference platforms for each. Given that Maemo/Meego has imploded and Android is in many ways very limiting compared to a desktop OS, I believe that Plasma Active will become increasingly relevant among geeks.*
* While Gnome 3 nay have been motivated by a desire to be more accessible to regular users, it goes without saying that the vast majority of people using Linux do so because they either are geeks, or it was recommended to them by one. By alienating a large number of us, they're eating the seed corn.
Gnome 2 and KDE do that as well. Better, in fact - the Windows command prompt can't handle filenames with non-ASCII characters like kanji.
The problems with putting Plasma Active on different tablets are almost the same as the problems with putting Cyanogen on different phones. That hasn't stopped Cyanogen from becoming extremely popular.
The problem with putting Plasma Active on any number of tablets is the nature of ARM platforms themselves:
- No two boards are alike, all different in how GPIOs are hooked up and what they do, even if the SoC is the same.
This is true. However, this is handled in the kernel, and since Android uses a fork of the GPLed Linux kernel, recompiling the kernel with support for the requisite filesystems, etc. is trivial. (Some effort may be required to handle differences between Android and Linux architectures for things like sound drivers.)
- Only the hardware vendor gets the sources for some drivers, usually GPU. Thus putting something like Xorg on these devices is exceedingly hard. Spark works around it with a completely reverse-engineered Mali driver.
Xorg isn't hard - it doesn't even need a GPU. The hard part is hardware acceleration, without which you can say goodbye to video playback at any decent resolution. This is probably the biggest problem right now. The only solution I can think of is for someone like the nouveau project to create drivers for a common chipset like the Tegra series, which would take a lot of resources and a considerable amount of time.
For other devices, like bluetooth/wifi modules, the binary blobs from Android can be used as is.
- Many tablets put up a fight when it comes to installing 3rd party firmware, unlike most PMPs/Routers. Even the Transformer Prime is locked down and ASUS has yet to release the promised unlock.
I agree that this can be problematic, and that we are at the manufacturer's mercy in this regard (unless someone is able to take the tablet apart and read the keys directly off the chip's fuses). Asus announced that they would release the unlock this month, and as long as they do so they're far ahead of their competitors in this regard. (It's only a flashboot unlock and not one for nvflash, which means you can't repartition the drive, but it's still better than nothing.) It's likely that they will continue to release unlocks for all future devices, which would provide a partial solution to this problem.
- Tablets with Android tend to have older kernels, with little to nothing having gone upstream thus making adaptation harder.
So it's a nice thought, and there are attempts, but it's nowhere near as easy in the tablet and handset space as it is in the PMP/Router space.
There is some work at the moment on merging some of the changes in the Android kernel (particularly those involving power management and drivers) into the mainline, which is due to be completed by 3.4. In the meantime, Android didn't fork the kernel that long ago - what changes have their been made since then that offer a substantial impediment to running Plasma Active (or any other modern distro)?
Running Plasma Active on numerous tablets may not be as simple as running an installation program, but it's definitely not as hard as many would believe. The only reason we haven't seen this happening yet is because it's incomplete.
http://arstechnica.com/gadgets/news/2012/01/new-kde-tablet-to-liberate-linux-enthusiasts-from-walled-garden.ars
Suddenly one of these is looking tempting for my tablet needs.
I did have an ASUS transformer for a few months but I sold it to a friend as I was unhappy with the way Android does things.
As the owner of an Asus Transformer currently running Kubuntu, I find this somewhat ironic.
Rockby is reasonably good for analog parts, though their website's search is somewhat lacking. Jaycar has a minuscule selection in comparison.
The reality is that for electronic components, the demand for the really interesting ones is just too low to justify a brick and mortar shop keeping them in stock. Resistors, capacitors, etc. you'll find, but good luck finding a microprocessor more powerful than an Arduino or a display better than a simple 16x2 character display.
I usually just order stuff from Sparkfun, or AustralianRobotics if I need it sooner. I'd probably use Digikey as well if they didn't charge $35 for shipping.
That said, I don't see how this keyboard can possibly work if it only has 32 combinations, unless there's some extra modifier keys that you use with the other hand or something.
Apparently he's planning to extend it to include a 6th key that can be activated by the left-hand thumb. That gives 64 chars, and could be extended to 96 if you added a 7th that was adjacent and mutually exclusive to the 6th.
Realistically, I don't think it's intended as a complete replacement for the conventional keyboard layout; if it's used for everything except the less commonly used symbols that would be sufficient.
Nope, that's Google Ride. Why would you drive a self-driving car?
That's not why magnet links are useful. You can be certain that the provision of magnet links will be held to be the facilitation of copyright infringement at some point or another, since from a legal perspective they're no different to torrent files - both effectively contain instructions on how to get the content, it's just that magnet links added a 'follow the directions you find here' step. It doesn't matter how many links there are, only whether the dominant usage is illegal.
Magnet links are useful because they offer increased decentralization. We used to need trackers. Then there was DHT. We used to need to the entire torrent file. Now we just need the hash. We currently need indexers. Once the technology in Tribler takes off, we won't even need that.
A completely decentralized P2P network has no central points of failure (e.g. trackers) that can be sued and taken down. All it has is users, and while the RIAA has demonstrated a willingness to go sue individual users, they have not had much success in countries which didn't appreciate the clear abuses of the judicial system, nor in cases where the individuals took them to court (you're dreaming if you think they'll recover enough via damages to cover legal costs).
To that end, I'd love to see a 100% Linux phone, nothing relying on Google with it's bits locked down or tracking
They're also working on 'Kubuntu Mobile' (alpha release's available from the site), which presumably uses Plasma Active as well. The reference platform for that is also the Nokia N900, though as modifying an Android kernel to run Ubuntu doesn't take a great deal of effort, it won't be long at all before someone has it running on the Galaxy Nexus (or its successor).
Australia and North America, on the other hand, are much more spread out, and the number of GSM cells that would be needed to provide acceptable coverage to rural areas would be too expensive for the likely revenue from them.
As an Australian, I'm fairly certain we standardized on GSM (2G) and UMTS (3G). A quick Google indicates that whatever vestigial CDMA reception we had was shutdown in 2008.
Use the mozplugger package - then it automatically embeds an instance of okular (or whatever PDF reader you use).
I thought prohibiting students from sharing past copies of tests was a standard and acceptable method. Is it because they are using copyright to attack the practice?
This is news to me. Maybe it's different here in Australia, but it's common practice for universities* to make the last decade's worth of exams available to students. Admittedly they are under copyright and not publicly accessible, but that true of most of the course content. It's rather beneficial as well, since hey constitute the main form of revision for most students once their done reading through their notes. Most lecturers even spend the last lecture or two going over the answers to the previous year's paper.
*The Department of Education did the same thing with the standardized year 12 exams, which were made publicly available.
Entirely this.
I'm an engineering student and I have noticed that most of the time, the general theorem that applies is that the interestingness of lecture is inversely proportional to the technologic level used.
In other words, someone in the theatre who'll use blackboard/scribbled projection tend to be almost universally amazing, those that use common "fill in gaps" projections tend to be OK , and lecturers using powerpoint tend to be the "gouge out eyes" sort of boring.
I'm an (Australian) engineering student, and almost all of my lecturers will use Powerpoint slides and end up writing all over them, to the extent that they often added mostly blank slides or wide margins purely for that purpose. The Powerpoints are published on the university website beforehand, and recorded copies of the lectures are usually made available afterwards.
The only subjects I've ever had trouble with have been ones where the lecturers avoided Powerpoint and just scribbled on the projector. Their structure wasn't as clear, and it was harder to decipher their handwriting.
When I studied law, it was pretty similar, except instead of scribbling they just talked more.