Now, if javelin throwers started targeting people on the field, how much do you think that statistic would change?
Hardly at all, because nobody would be stupid enough to put himself in the way of a javelin and they fly slow enough to give plenty of time to run away. A javelin used in sport is simply not build as a weapon, you can kill people with it, just like you can kill people with hammer, but they are not optimized for the task at all and you will have a though time carrying them concealed. As said before: Care to provide some statistics on sport javelin used in crime?
All of that is completely different for handguns, they are small enough to carry concealed, precise and fast enough to kill multiple targets in a short amount of time, quick to reload, small, easy to carry, etc. and deadly crime is committed with them on a regular basis. If you want to do sport with guns, you can use a much saver air pistol or laserpointer gun, there is no reason to use a weapon for sports.
It [Bittorrent] is one of the most, if not THE most, efficient file transfer protocols invented.
Only when it comes to large static files, for everything else it sucks.
I was talking about the design intent of this particular tool... what this particular software tool was originally used for.
Googling around a bit, it seems that MP2P is an evolution of Gnutella with some anonymity and speed ups thrown in, i.e. regular old school illegal file sharing.
Do you trust your OS updates to come out of p2p cloud?
Cryptographic signatures exist, so getting secure updates from an untrusted sources is quite possible.
Their customers are already using p2p so it was a natural fit.
I haven't checked WoW in a while, but I can't remember ever having to touch a Bittorrent client for it, it's all handled internally by the updater. So while that is a use of a P2P protocol, it's not exactly a justification for PirateBay and friends.
You will not see video on p2p due to the nature of how videos work. You need the front part before you can watch the later parts. So streaming is what they do.
And what stops you from developing a P2P protocol that can handle that?
Also for things like github it makes little sense for p2p there.
Github has an incredible small quota for their free service, so a free P2P alternative would make perfect sense to host larger files. Also storage is cheap, so a handful of users hanging their TB drives into the P2P cloud could go a long way to host a ton of projects.
You are confusing usage with 'right or wrong' moral choices. They are 'right or wrong' design decisions. Very different.
I am not passing a judgment, heck, I wouldn't mind if copying would be legal. What I do mind is all those lies and half truth that people throw around to defend P2P.
Also the entire internet was designed up front to be p2p.
The Internet has always operated on a client/server model not P2P, the only P2P about that was that every client could also be a server in the early days, which isn't true today due to NAT. However, currently day use of the term P2P generally implies some short of resource/bandwidth sharing between peers, none of the mainstream FTP/HTTP part of the Internet was ever build for that, one could make an argument for Usenet, but even there that wasn't really P2P, but more a collection of central servers that copied data to each other.
Ah sure, bit torrent is useless for installing updates?
Wake me when you have a tool that is actually in standard use by a major distribution, not just some random proof of concept that is used by nobody and apperently hasn't updated their webpage in the last four years.
GitHub, Google Code, Youtube or plain old commercial hosting all have significant operating expenses in terms of servers, bandwidth and people power,
The thing is even with all those disadvantages people still vastly prefer them over the P2P alternatives, even if that means paying their own money for hosting. If P2P is so awesome, why isn't Ubuntu and Debian using it for distribution software updates? Why are all the podcast I listen to, even those from the Free Software community and those under free CC licenses, hosted on regular for-pay HTTP servers, not on P2P? Why are almost all the videos I watch hosted on vimeo, blip, youtube and Co. not on bittorrent? Why doesn't my Firefox support bittorrent?
What would be the operating, server and bandwidth costs of hosting the full contents of The Pirate Bay on a centralized server?
Hosting large amounts of illegal content would cost money, P2P makes it easier to share large amount of illegal content. I don't doubt that. That's my point: P2P tools like Bittorrent are optimized for uses which rarely appear in legal use, but which are frequent when it comes to illegal use.
The ability to share frequently changing collections of small files would be far more valuable for legal users, then sharing large blobs of static data, yet in all those years there has never been a P2P tool optimized for the former.
Please explain to me how you make a long thin stick that you throw at high velocities not deadly?
They are already pretty non-deadly as they are, as they make it impossible to aim at large distances and fly slow enough to give the victim possibility to dodge if aware of the spear. Surviving a hit is also quite possible.
Analysis by National Center for Catastrophic Sport Injury Research of NCAA and high school injury data found that between 1990 and 2007, eight individuals were struck by a thrown javelin. There were no fatalities. The NCCSIR found that more serious injuries and the largest number of deaths in track and field are incurred during pole vaulting.
Of course that's just regular sport use, not actual abuse of the tool in armed rubbery and murder, but then I can't think of any cases when they where ever abused for any of the later. Care to provide some statistics?
P2P helps people break the law in the very same way as FTP ad HTTP do.
Both FTP and HTTP have ton's legal uses and only a tiny fraction of illegal uses, with most P2P stuff it's exactly the reverse, they are optimized for illegal sharing and quite useless for legal sharing.
Bittorrent for example is nice for downloading isos, but that's something one does every few month or even years. For the much more regular tasks, such as regular software updates, bittorrent is rather useless, as it has no real support for frequently changing collections of small files, thus all the OS updates have to come over FTP and HTTP.
Or how about we look a little further, what do people actually need for their legal use, people such as video producers, podcast producers and software creators. Is my Git repository hosted on P2P? No. Do you see podcasts on P2P? No. Do you see videos on P2P? Very rarely. Can by Firefox play video directly from P2P? No.
All those needs are currently not full filled by P2P, but by things like GitHub, Google Code, Youtube or plain old commercial hosting. So in essence, if the goal of P2P was to make it easier for users to share legal stuff, it does an incredible shitty job at it.
Handguns do not need to have a "good" use, if an individual has the right to own/use one that is sufficient.
In legal terms, sure, that's what we have rights for. However as a point for discussion your argument is totally useless, as it's just an appeal to authority.
Speaking about DNS blocking and DNS names. How large would a full dump of the whole DNS system actually be? From the numbers I could gather it be in the low GB range for all the top level domains and easily fit on a DVD, i.e. a rather trivial size in the days of movie streaming. How much bigger would it get by including all the subdomains (I assume you'd need a spider to actually gather those)? How big would daily updates be? In essence would it be possible to just completely bypass the classic DNS and move to one big hosts file on the local computer? Also is it possible to actually publicly download the TLD zone file? Verisign seems to offer it, but not publicly.
For encryption to be used and valuable, it has to happen transparently to the user, none of the encryption plugins I ever tried to use got anywhere near that, they added so many junk into the process that it just wasn't worth the effort. Heck, the fact that they where plugins in the first place, not build in features of the mail reader alone is enough to make encryption a failure. And no, this isn't an issue with individual lazyness, but the problem is that encryption is only really safe when everybody is using it. If you are the only guy in town using encryption, that just makes you suspicious, makes you look like you have something to hide. Furthermore, email encryption already sucks on the concept levels, subjects headers are often not encrypted at all, thus just inviting accidental information leakage. The biggest problem with email encryption however is that To and From headers are not encrypted, which for a lot of big brother like surveillance is really the only information they are looking for anyway, as who communicates with whoem already gives them most of the information they need. This renders email encryption near useless and the To/From header issue can't be fixed by moving to a different protocol anyway.
The sad part in all this is that PGP/GPG could have added value to email not by encryption, but via signing. And while I do see signatures a lot of times on mailing list or private mail, where it really serves little to no purpose, I never see it in the places where it would be needed: Mail from Paypal, Amazon, a bank, etc. It's completely idiotic that I have to fish Paypal mail out of my Spam filter, when a simple signature could easily distinguish junk from legitimate mail. Further more it would make phishing a good bit harder, if a signature would allow to distinguish scam from mail automatically.
All that said, I do use encryption when its available an easy, the Pidgen encryption plugin is sort of usable, but even there just barely, as messages get lost in there on a regular basis when you communicate with somebody who changes clients between communication or other mishaps happens. It's annoying, and just goes to show that usability, which is among the most important things with encryption, always gets the least amount of thought.
/etc/apt/sources.list has supported multiple repositories forever.
Multiple repositories don't address the underlying problem, as you are still installing everything in a single dependency and filesystem tree. If you have two repository with packages of the same name, you have a conflict that current package managers have no proper way of dealing with, aside from forcing you to use whatever has the highest version number (i.e. the Gnome2 vs Gnome3 problem).
The only viable alternative to the repository system is to either
There are far more ways to solve the issue. You could for example start by separating software installation from software availability, i.e. copy files to/packages/foobar-2.0/bin/foobar instead of right into/usr/bin/foobar. And then just place a symlink/start-script in/usr/bin/ to make the software available. This allows users to have different versions of the same software. Libraries can be dealt with by adjusting the LD_LIBRARY_PATH on a per program level.
Yeah, that's bad and Linux has it's analog "dependency hell", which is just as much of a problem.
A package manager that is more flexible then what we have right now could fix that and no, that doesn't mean libraries would fly everywhere, it simply would mean that it would have more advanced mechanisms to resolve naming conflicts then forcing you to always use the newest package from the repository and nothing else, which is what current package managers do.
This is what you get when you have infrastructure that is build around one centrally maintained dependency tree, you are slave to whatever decisions they make. It's not even a new problem, similar software removals against the users will have happened with Gnome2 vs Gnome3 and even back with Gnome1 vs Gnome2 and counterless times when a working version of Gimp was replaced with a broken one and only fixed month later. This one seems a bit more sinister as from the looks of it it seems they remove it in a regular software update, not a dist-upgrade, but it's essentially the same issue. And to all those "This isn't a problem"-sayers, the existence of complicated time consuming workarounds by manual compilation/installation, thus by-passing the binary package distribution, is part of the problem, not the solution.
It should really be time for Debian to move to a more flexible, more free form of package distribution that doesn't depend on a single dependency tree and fixed locations in the file system.
Desura is the prime example of the Windows Mindset creeping into Linux (Ubuntu's PPA are the other example).
Kind of, but Linux people have nobody else to blame for that then themselves. The lack of a cross-platform packaging format isn't exactly a new issue, it has been a major annoyance when shipping binary on Linux for over a decade and yet progress in that area has been slim to none, it's still a complete cluster fuck. It of course doesn't help that major distro specific packaging formats themselves are also lacking in features (deb can't install multiple versions of software, can't install software as user, etc.).
So as imperfect as Desura might be at the moment, at least it makes live a hell of a lot easier for single-user machines.
For those that have missed it: Seems like the Humble Bundle is getting a bit competitions, a few weeks ago the IndieRoyale Bundles got launched, they follow a similar model of multiple games for an almost-pay-what-you-want price (min around $3). it however doesn't have the charity and it only sometimes has Linux versions of the games. Also their game selections seems to be not so great most of the times, however they include a gem every now and then.
The scrollbar, which shows exactly where you are in the document and how much of the document is on screen, has just moved by half its length.
Doesn't provide enough information on a long document.
The cursor, which shows what line you're in, has just jumped from the bottom of the screen to halfway up it.
Yes, it's however still disorienting to follow a constantly jumping cursor. Furthermore, not all applications have a cursor, i.e. your webbrowser has none.
And since you are not a baby or an idiot, after the second time you trigger this, you grasp that it scrolls half a page at a time and know intuitively where things have moved to.
Actually no, I never did. I consider Emacs default scrolling completely fucked up and unusable, which is why I switched Emacs to line-by-line scrolling a decade ago. That said, smooth scrolling alone wouldn't have fixed Emacs either, but it would have made it a little less disorienting.
Emacs is what a non-condescending UI looks like - and it's clear that it's not the right answer for everyone.
Emacs GUI isn't exactly non-condescending, for most part it's just an ugly broken mess, slapped together by programmers with absolutely no regard to usability. Which is a little sad, as some parts of rest of the interface are really really good from a usability point of view (type-ahead search, all functions accessible via M-x, documentation for each functions, direct jump to source code of each functions, etc.).
Isn't it much simpler, easier for the user, more intuitive, to simply "go one page down" when you press "page down"?
Without a transition animation you simply see two screens that look completely different and have no feedback for why they look differently. Say you accidentally pressed the up instead of the down button, with a transition animation you would notice it instantly, without one it could take quite a bit to notice that something went wrong.
Slow animations would help if there was an emacs bug where it scrolled down a completely random number of lines each time.
Your webbrowser will scroll a random number of lines when you reach the end of the webpage. It only scrolls one full screen as long as you are in the middle of the page, but once you reach the end of a webpage it will scroll however long it takes to reach the end, anything from a single line to a full screen. That's confusion that could be a avoided with smooth scrolling.
Of course that is all just theoretically speaking, in practical terms I agree that most transition animation in software today do far more harm then good. That doesn't make them fundamentally evil, but it certainly does mean that they have to be used with care.
The concept of transition animation is perfectly valid, as it tells you what changed on the screen and where your last positions was. To see how bad things can get when you don't have them, try a vanilla Emacs. Emacs will by default scroll through text not per line and not per page, but per half-page, this makes it extremely easy to lose track in a document, as you have no real indication for how far it has scrolled. Proper smooth scrolling can easily fix that, as it provides you with guidance where the text went, how far it has scrolled and where your last reading positions was.
Where it becomes a problem is when the animation is non-skipable and blocking user input, that is something that should never happen. If the user pressed page-flip, the computer should flip the page, if it's in a transition animation, it should simply scroll faster. And of course it should never take to long to scroll to begin with. But even if you just have an animation that is 100ms long, which is what humans normally accept as instantly, that's still gives you 6 frames that you can use for a transition animation that can provide valuable information for the user.
Slashdot Mirror
Now, if javelin throwers started targeting people on the field, how much do you think that statistic would change?
Hardly at all, because nobody would be stupid enough to put himself in the way of a javelin and they fly slow enough to give plenty of time to run away. A javelin used in sport is simply not build as a weapon, you can kill people with it, just like you can kill people with hammer, but they are not optimized for the task at all and you will have a though time carrying them concealed. As said before: Care to provide some statistics on sport javelin used in crime?
All of that is completely different for handguns, they are small enough to carry concealed, precise and fast enough to kill multiple targets in a short amount of time, quick to reload, small, easy to carry, etc. and deadly crime is committed with them on a regular basis. If you want to do sport with guns, you can use a much saver air pistol or laserpointer gun, there is no reason to use a weapon for sports.
It [Bittorrent] is one of the most, if not THE most, efficient file transfer protocols invented.
Only when it comes to large static files, for everything else it sucks.
I was talking about the design intent of this particular tool... what this particular software tool was originally used for.
Googling around a bit, it seems that MP2P is an evolution of Gnutella with some anonymity and speed ups thrown in, i.e. regular old school illegal file sharing.
Do you trust your OS updates to come out of p2p cloud?
Cryptographic signatures exist, so getting secure updates from an untrusted sources is quite possible.
Their customers are already using p2p so it was a natural fit.
I haven't checked WoW in a while, but I can't remember ever having to touch a Bittorrent client for it, it's all handled internally by the updater. So while that is a use of a P2P protocol, it's not exactly a justification for PirateBay and friends.
You will not see video on p2p due to the nature of how videos work. You need the front part before you can watch the later parts. So streaming is what they do.
And what stops you from developing a P2P protocol that can handle that?
Also for things like github it makes little sense for p2p there.
Github has an incredible small quota for their free service, so a free P2P alternative would make perfect sense to host larger files. Also storage is cheap, so a handful of users hanging their TB drives into the P2P cloud could go a long way to host a ton of projects.
You are confusing usage with 'right or wrong' moral choices. They are 'right or wrong' design decisions. Very different.
I am not passing a judgment, heck, I wouldn't mind if copying would be legal. What I do mind is all those lies and half truth that people throw around to defend P2P.
Also the entire internet was designed up front to be p2p.
The Internet has always operated on a client/server model not P2P, the only P2P about that was that every client could also be a server in the early days, which isn't true today due to NAT. However, currently day use of the term P2P generally implies some short of resource/bandwidth sharing between peers, none of the mainstream FTP/HTTP part of the Internet was ever build for that, one could make an argument for Usenet, but even there that wasn't really P2P, but more a collection of central servers that copied data to each other.
Ah sure, bit torrent is useless for installing updates?
Wake me when you have a tool that is actually in standard use by a major distribution, not just some random proof of concept that is used by nobody and apperently hasn't updated their webpage in the last four years.
GitHub, Google Code, Youtube or plain old commercial hosting all have significant operating expenses in terms of servers, bandwidth and people power,
The thing is even with all those disadvantages people still vastly prefer them over the P2P alternatives, even if that means paying their own money for hosting. If P2P is so awesome, why isn't Ubuntu and Debian using it for distribution software updates? Why are all the podcast I listen to, even those from the Free Software community and those under free CC licenses, hosted on regular for-pay HTTP servers, not on P2P? Why are almost all the videos I watch hosted on vimeo, blip, youtube and Co. not on bittorrent? Why doesn't my Firefox support bittorrent?
What would be the operating, server and bandwidth costs of hosting the full contents of The Pirate Bay on a centralized server?
Hosting large amounts of illegal content would cost money, P2P makes it easier to share large amount of illegal content. I don't doubt that. That's my point: P2P tools like Bittorrent are optimized for uses which rarely appear in legal use, but which are frequent when it comes to illegal use.
The ability to share frequently changing collections of small files would be far more valuable for legal users, then sharing large blobs of static data, yet in all those years there has never been a P2P tool optimized for the former.
People who think that knives are less dangerous than guns have been watching too many movies.
Or they might have actually looked at the numbers: Murders in 2008 in the USA: 9484 by gun, 1897 by knife.
Please explain to me how you make a long thin stick that you throw at high velocities not deadly?
They are already pretty non-deadly as they are, as they make it impossible to aim at large distances and fly slow enough to give the victim possibility to dodge if aware of the spear. Surviving a hit is also quite possible.
Risks & Hazards Involved in Javelin Throwing:
Of course that's just regular sport use, not actual abuse of the tool in armed rubbery and murder, but then I can't think of any cases when they where ever abused for any of the later. Care to provide some statistics?
P2P helps people break the law in the very same way as FTP ad HTTP do.
Both FTP and HTTP have ton's legal uses and only a tiny fraction of illegal uses, with most P2P stuff it's exactly the reverse, they are optimized for illegal sharing and quite useless for legal sharing.
Bittorrent for example is nice for downloading isos, but that's something one does every few month or even years. For the much more regular tasks, such as regular software updates, bittorrent is rather useless, as it has no real support for frequently changing collections of small files, thus all the OS updates have to come over FTP and HTTP.
Or how about we look a little further, what do people actually need for their legal use, people such as video producers, podcast producers and software creators. Is my Git repository hosted on P2P? No. Do you see podcasts on P2P? No. Do you see videos on P2P? Very rarely. Can by Firefox play video directly from P2P? No.
All those needs are currently not full filled by P2P, but by things like GitHub, Google Code, Youtube or plain old commercial hosting. So in essence, if the goal of P2P was to make it easier for users to share legal stuff, it does an incredible shitty job at it.
Handguns do not need to have a "good" use, if an individual has the right to own/use one that is sufficient.
In legal terms, sure, that's what we have rights for. However as a point for discussion your argument is totally useless, as it's just an appeal to authority.
Sport?
Sport doesn't need a deadly weapon, it would be trivial to build a replacement that you can still do sports with that wouldn't be deadly.
Spoons and cars actually provide a useful service for society which makes them worth the cost, guns don't.
Speaking about DNS blocking and DNS names. How large would a full dump of the whole DNS system actually be? From the numbers I could gather it be in the low GB range for all the top level domains and easily fit on a DVD, i.e. a rather trivial size in the days of movie streaming. How much bigger would it get by including all the subdomains (I assume you'd need a spider to actually gather those)? How big would daily updates be? In essence would it be possible to just completely bypass the classic DNS and move to one big hosts file on the local computer? Also is it possible to actually publicly download the TLD zone file? Verisign seems to offer it, but not publicly.
For encryption to be used and valuable, it has to happen transparently to the user, none of the encryption plugins I ever tried to use got anywhere near that, they added so many junk into the process that it just wasn't worth the effort. Heck, the fact that they where plugins in the first place, not build in features of the mail reader alone is enough to make encryption a failure. And no, this isn't an issue with individual lazyness, but the problem is that encryption is only really safe when everybody is using it. If you are the only guy in town using encryption, that just makes you suspicious, makes you look like you have something to hide. Furthermore, email encryption already sucks on the concept levels, subjects headers are often not encrypted at all, thus just inviting accidental information leakage. The biggest problem with email encryption however is that To and From headers are not encrypted, which for a lot of big brother like surveillance is really the only information they are looking for anyway, as who communicates with whoem already gives them most of the information they need. This renders email encryption near useless and the To/From header issue can't be fixed by moving to a different protocol anyway.
The sad part in all this is that PGP/GPG could have added value to email not by encryption, but via signing. And while I do see signatures a lot of times on mailing list or private mail, where it really serves little to no purpose, I never see it in the places where it would be needed: Mail from Paypal, Amazon, a bank, etc. It's completely idiotic that I have to fish Paypal mail out of my Spam filter, when a simple signature could easily distinguish junk from legitimate mail. Further more it would make phishing a good bit harder, if a signature would allow to distinguish scam from mail automatically.
All that said, I do use encryption when its available an easy, the Pidgen encryption plugin is sort of usable, but even there just barely, as messages get lost in there on a regular basis when you communicate with somebody who changes clients between communication or other mishaps happens. It's annoying, and just goes to show that usability, which is among the most important things with encryption, always gets the least amount of thought.
/etc/apt/sources.list has supported multiple repositories forever.
Multiple repositories don't address the underlying problem, as you are still installing everything in a single dependency and filesystem tree. If you have two repository with packages of the same name, you have a conflict that current package managers have no proper way of dealing with, aside from forcing you to use whatever has the highest version number (i.e. the Gnome2 vs Gnome3 problem).
The only viable alternative to the repository system is to either
There are far more ways to solve the issue. You could for example start by separating software installation from software availability, i.e. copy files to /packages/foobar-2.0/bin/foobar instead of right into /usr/bin/foobar. And then just place a symlink/start-script in /usr/bin/ to make the software available. This allows users to have different versions of the same software. Libraries can be dealt with by adjusting the LD_LIBRARY_PATH on a per program level.
DLL HELL
Yeah, that's bad and Linux has it's analog "dependency hell", which is just as much of a problem.
A package manager that is more flexible then what we have right now could fix that and no, that doesn't mean libraries would fly everywhere, it simply would mean that it would have more advanced mechanisms to resolve naming conflicts then forcing you to always use the newest package from the repository and nothing else, which is what current package managers do.
This is what you get when you have infrastructure that is build around one centrally maintained dependency tree, you are slave to whatever decisions they make. It's not even a new problem, similar software removals against the users will have happened with Gnome2 vs Gnome3 and even back with Gnome1 vs Gnome2 and counterless times when a working version of Gimp was replaced with a broken one and only fixed month later. This one seems a bit more sinister as from the looks of it it seems they remove it in a regular software update, not a dist-upgrade, but it's essentially the same issue. And to all those "This isn't a problem"-sayers, the existence of complicated time consuming workarounds by manual compilation/installation, thus by-passing the binary package distribution, is part of the problem, not the solution.
It should really be time for Debian to move to a more flexible, more free form of package distribution that doesn't depend on a single dependency tree and fixed locations in the file system.
GPLv3 still allows you to run code on your servers without giving back, it's only the AGPLv3 that requires making the code available.
Looks like there are two more Indie Games Pack and Little Big Bunch.
That's exactly what they did, but they omitted allowing a multi-user installation.
Which is quite understandable, given that it's not exactly trivial to make a tool like Desura work across multiple user accounts.
Desura is the prime example of the Windows Mindset creeping into Linux (Ubuntu's PPA are the other example).
Kind of, but Linux people have nobody else to blame for that then themselves. The lack of a cross-platform packaging format isn't exactly a new issue, it has been a major annoyance when shipping binary on Linux for over a decade and yet progress in that area has been slim to none, it's still a complete cluster fuck. It of course doesn't help that major distro specific packaging formats themselves are also lacking in features (deb can't install multiple versions of software, can't install software as user, etc.).
So as imperfect as Desura might be at the moment, at least it makes live a hell of a lot easier for single-user machines.
For those that have missed it: Seems like the Humble Bundle is getting a bit competitions, a few weeks ago the IndieRoyale Bundles got launched, they follow a similar model of multiple games for an almost-pay-what-you-want price (min around $3). it however doesn't have the charity and it only sometimes has Linux versions of the games. Also their game selections seems to be not so great most of the times, however they include a gem every now and then.
The Humble Bundle doesn't need the Steam keys, they are just an optional addition, you can simply download the .bin/.exe/.tar.gz directly if you want.
The scrollbar, which shows exactly where you are in the document and how much of the document is on screen, has just moved by half its length.
Doesn't provide enough information on a long document.
The cursor, which shows what line you're in, has just jumped from the bottom of the screen to halfway up it.
Yes, it's however still disorienting to follow a constantly jumping cursor. Furthermore, not all applications have a cursor, i.e. your webbrowser has none.
And since you are not a baby or an idiot, after the second time you trigger this, you grasp that it scrolls half a page at a time and know intuitively where things have moved to.
Actually no, I never did. I consider Emacs default scrolling completely fucked up and unusable, which is why I switched Emacs to line-by-line scrolling a decade ago. That said, smooth scrolling alone wouldn't have fixed Emacs either, but it would have made it a little less disorienting.
Emacs is what a non-condescending UI looks like - and it's clear that it's not the right answer for everyone.
Emacs GUI isn't exactly non-condescending, for most part it's just an ugly broken mess, slapped together by programmers with absolutely no regard to usability. Which is a little sad, as some parts of rest of the interface are really really good from a usability point of view (type-ahead search, all functions accessible via M-x, documentation for each functions, direct jump to source code of each functions, etc.).
Isn't it much simpler, easier for the user, more intuitive, to simply "go one page down" when you press "page down"?
Without a transition animation you simply see two screens that look completely different and have no feedback for why they look differently. Say you accidentally pressed the up instead of the down button, with a transition animation you would notice it instantly, without one it could take quite a bit to notice that something went wrong.
Slow animations would help if there was an emacs bug where it scrolled down a completely random number of lines each time.
Your webbrowser will scroll a random number of lines when you reach the end of the webpage. It only scrolls one full screen as long as you are in the middle of the page, but once you reach the end of a webpage it will scroll however long it takes to reach the end, anything from a single line to a full screen. That's confusion that could be a avoided with smooth scrolling.
Of course that is all just theoretically speaking, in practical terms I agree that most transition animation in software today do far more harm then good. That doesn't make them fundamentally evil, but it certainly does mean that they have to be used with care.
The basic concept is a bug not a feature.
The concept of transition animation is perfectly valid, as it tells you what changed on the screen and where your last positions was. To see how bad things can get when you don't have them, try a vanilla Emacs. Emacs will by default scroll through text not per line and not per page, but per half-page, this makes it extremely easy to lose track in a document, as you have no real indication for how far it has scrolled. Proper smooth scrolling can easily fix that, as it provides you with guidance where the text went, how far it has scrolled and where your last reading positions was.
Where it becomes a problem is when the animation is non-skipable and blocking user input, that is something that should never happen. If the user pressed page-flip, the computer should flip the page, if it's in a transition animation, it should simply scroll faster. And of course it should never take to long to scroll to begin with. But even if you just have an animation that is 100ms long, which is what humans normally accept as instantly, that's still gives you 6 frames that you can use for a transition animation that can provide valuable information for the user.