An interesting flip side to this argument is that the fast loading of pages that users find so handy implies that their bandwidth usage is probably lower than you might expect. After all, the reason the pages load so fast is because they're quite sparse and don't contain that much information. Some quick math suggests that their bandwidth usage probably isn't as high as you'd think: 2e7 pages per month times 1e5 bits per page (rough guess) gives about 2e12 bits per month. Since a month is about 2e6 seconds, that's of order 1e6 bits per second, or roughly T1 speed. Putting banner ads on their pages could wind up greatly increasing their outbound bandwidth.
Actually, though, their big bandwidth usage is probably their web crawler. Indexing a billion pages, even if it's not monthly, probably consumes several orders of magnitude more bandwidth than serving the search pages does. That actually suggests that they should be able to grow their business pretty easily; if indexing the web is the major cost, every added user is a win. That suggests that their business model (work on growing customer base first, then worry about making money on each customer) makes a lot of sense.
Of course a properly designed piece of censorware should be able to deal with this problem. They just have to be a bit cleverer about dealing with URLs. How do I know this? Because Junkbuster is capable of screening out Akamized content by looking at the whole URL. When it sees "ads.doublclick.net" in the URL, even if it isn't the base domain, it knows to filter it. A reasonably designed censorware proxy could do the same thing with "www.pr0nstars.com" or, sadly, "www.peacefire.org". Actually, you could turn Junkbuster into a Censorware proxy very easily by feeding it a different list of sites to block...
I think that the author gets one big point but misses another. He is correct that today an operating system is generally considered to contain more than just a kernel and a shell. What he misses is that Unix has grown more expansive in exactly the way that he suggests. For instance, he comments that a real OS needs to have "hundreds of utilities". That's exactly what Unix provides; most people think of grep, sed, awk, find, ps, etc. as being essential parts of Unix even though they obviously aren't part of the shell or kernel. Similarly, X11R6 is a key part of Unix as it is now perceived.
This is, IMO, a big part of the reason that Mac OSX won't be Unix even though it's based on underlying Unix technology. It doesn't incorporate all of the other stuff that's really part of Unix as it is understood to be, so the use of a Unix kernel doesn't make it Unix.
BUT, I think anyone who would ever go so far as to say the books are for idiots are wrong.
And that's why I don't like the title. After all, the publisher is going right out an saying that the books are for idiots. I think that it's part of a sad, but surprisingly general, confusion of ignorance and stupidity. I personally find it insulting to be told that I'm a dummy for not already knowing what the book is supposed to teach me. I think I'd personally be more inclined to buy the series if it were re-named [Blank} for Novices, rather than [Blank] for Dummies. They both get across the same idea, but one is much friendlier.
Bah! I've seen README file larger than the memory in my first computer. I've seen comment blocks larger than the memory in my first computer. Many, many computers today have larger block sizes on their hard drives than my first computer's entire memory. What's really scary is that my first hard drive wasn't large enough to hold the entire current source code for the Linux kernel- gzipped.
People who
would normally be too embarassed to go in to a liquor store or a peep show have no problem getting porno on the net.
The internet makes people do what they would not normally do.
The net does not make people do what they otherwise would not do; it allows them to do so. It's not as though the net creates a desire in those people to view pornography, engage in vandalism, or do any of the other socially unacceptable acts you mention. The desire was pre-existing but kept under cover by the fear of public scrutiny.
Part of the reason that people are so attached to the anonymity of the net is because it gives them the opportunity to act out on their socially unacceptable desires without fear of public scorn. That's not to say that the net's influence this way is necessarily desirable for unleashing those feelings, but it's not causing them.
Sadly not true. I was able to get to the site listed, but I'm often not able to get to other sites mentioned in YROL because of filtering. It's just more evidence of how ineffective internet filtering is, that some sites that really seem as though they might be blocked by filtering software aren't but others (like ESR's home page on my system) are.
Sure, and after a while people tend to discover that some sources are especially reliable, and they pay extra attention to them. The fact that people will give that extra attention, and in many cases extra business, to the most reliable sources is a key part of the reason that the press is as trustworthy as it is. In a real sense it's like the way that peer review of Open Source code helps to ensure that nobody deliberately slips in security bugs. The risk of being caught is enough to keep people from even trying. The result is that we know that we know that we can generally trust the facts presented in the Wall Street Journal, that ZDnet is less reliable, and the Weekly World News is completely unreliable.
Of probably greater impact than the use of this kind of technology in the news media is its use in criminal justice. People are very heavily swayed by the perceived reliability of videotaped evidence. The fact that tapes can now be falsified with considerable ease, and that in many cases tapes of relevance in criminal cases will be unique and not subject to this kind of peer review leaves a very big and dangerous place for falsification. Imagine the police taping an interview with a criminal suspect, for instance, and then changing their tape to show the suspect waiving his right to counsel and confessing to the crime. It sounds as though that's perfectly possible, and even comparatively easy, with this technology. That's far scarier than a news program changing the logo on the side of a building.
Of course this is actually the way that things have always been. If you can't directly witness something yourself you are inherently trusting someone else to pass it on to you. You have no real knowledge of, to pick an example, whether there really was a Russian sub lost in the Barents Sea or whether it was an elaborate hoax. You're trusting that the people who bring you news are being honest and not showing you a bunch of crap.
Certainly this has always been the case in print media, hence the saying that you can't believe everything that you read. This invention doesn't really change anything, except that it makes the need for trust in your news deliverer more explicit- and in some well publicized cases so far showing how untrustworthy some of those news deliverers actually are.
The real costs are in promotions, marketing, and productions.
And under the current system, the artists actually wind up paying for it! If you read about the current system, the record company will advance the musicians money against their royalties to produce, promote, and market the album. This means that for the most part the musicians don't actually receive any royalties unless the album sells at least a million copies. If it doesn't, the musicians wind up in debt to the record company. It's basically like share cropping, where the system is stacked so that it's almost impossible to break away from one company.
In any case, it's dubious that promotions and marketing are really an essential part of the process, or that they can't be solved by going outside the traditional music industry approach. If you don't think that Napster is changing the promotions and marketing aspect of things, you really need to pay more attention.
I'm consistantly impressed with the candor and openness that Linus exhibits when he talks about a project that is, in
a big way, his baby.
This is actually surprisingly common among the truly great, at least among scientists, where I spend most of my time. I think that there are two facets to the fact that so many really great people are modest.
One thing is that part of what makes them great at what they do is maintaining an honest assesment of their strengths and weaknesses. This is very helpful because it means that, as in the case of the Mindcraft test, they're able to go back and correct genuine errors rather than completely blowing them off as unfair.
The other thing is that they're able to be if not exactly content at least pleased enough with what they have done that they have the self confidence to admit what they haven't. This can actually enhance ones public image if you do it well, since people who are aware of ones actual accomplishments and don't need to be reminded of them are often impressed by ones apparent nonchalance at such an impressive achievement.
Metallica shouldn't be able to force me to only play their music on a standalone CD player, despite the fact that my computer's cdrom drive is just as capable of playing it.
But Metallica hasn't been saying that you can't play MP3s, or burn backup copies of their CD's for personal archiving, or play their CD's using your computer's CD player. They've said that they don't like people giving away copies of their albums online. They have the right under copyright (and under Napster's use policy, for that matter) to keep people from doing that.
You need to separate what Metallica has said ("Don't trade songs from our albums on Napster, but bootlegs from our concerts are just fine.") from what the RIAA has said ("The only legal way to get recorded music is to buy our CDs at our prices. No copying for personal use, making MP3s for personal use, storing on computer, etc."). Metallica has expressed an interest in expanding into online music distribution (on their terms) once their current contractual obligations are completed. The RIAA has dragged its feet at every turn and proposed draconian limitations when on-line music formats have been discussed. One of those is reasonable, and the other one isn't.
Actually, there is a PowerPoint for Mac, so at least one of those programs isn't exclusive to Windows. More to the point, though, every time a Free Software hacker decides to create a free version of a commercial program, he's implicitly stating that he likes the original. After all, he likes what it does well enough to copy it!
I do find it ironic, though, that it is *Gnome* taking this step. Could anybody have possibly imagined this when Gnome started? Weren't they the "hacker desktop"? Didn't they have all the "desktop for the people" principles? Hmm... times change, I guess.
Those who don't understand the Free Software ideal (as very distinct from Open Source) are doomed to say really stupid things about it. It seems really obvious to me that a person who would make the quote above doesn't understand the principles that he's complaining about. Complaints about QT were not about corporate involvement (as the above quote seems to suggest is the big sin among the FSF crowd) but about lack of programming freedom.
Part of the FSF ethic is that anyone is free to hack on the programs, and that "everyone" includes big corporations. To tell Sun and HP that they mayn't become involved is actually more contrary to the spirit of the GPL than accepting big corporate money. The problems come when people try to place restraints on the code. Given that everything is going to be kept open by the GPL, I don't see this as being a big problem.
Sure. All you have to do is to be your own creditor. Imagine the following steps:
You set up two companies, popularandusefulinfo.com and The Dot Com Loan Company, Inc.
Dot Com Loan loans popularandusefulinfo.com a big stack of money at an unreasonable interest rate.
popularandusefulinfo.com sets up their web site with an iron-clad privacy guarantee and starts gathering customer info.
Any income from popularanduseful winds up going to pay the outrageous interest on their loan, shoveling any potential profit back to Dot Com Loan.
popularanduseful goes bankrupt because they have no prayer of making money and they eventually pay all their money back to Dot Com Loan in the form of interest. Their assets (read customer info) are sold to cover their debt, namely the principal on the interest.
Bingo, you've now managed to get big piles of customer info under false pretenses and profit from it.
The only way I can see this making any sense is if MS has resigned themselved to being split into MS/OS and
MS/Applications as per the initial DOJ v. MS ruling.
Not necessarily resigned to being split up, but considering it as a strong possibility. Microsoft would have to be far dumber than anyone has ever accused them of being to believe that there's no chance of them being split up; they'd have to be awfully stupid to believe that the chance is even particularly small.
Given that the break up is a real possibility, it makes sense to look into porting MS apps to a wider range of OS's than they currently serve. It may even be a reasonable plan if they don't wind up being split up. If other operating systems (particularly Linux and the BSD's) continue to invade the desktop, sources of income other than Windows will become increasingly important. In that case the Apps division had damn well better be able to generate some sales to UNIXoid OSes, or the whole company is going to go into the tank.
(I've even thought of basing one on Latin; its working name is SPQL, but I
don't know if I'll ever get around to creating it)
Shouldn't that really be SPQR? I know that it would be a bit more difficult to come up with a good excuse for doing so, but it would obviously be a lot funnier.
Me, I live over 800 miles from the nearest place
a foreign language might actually be useful. So obviously the need just isn't as high.
You must use different miles from everyone else. There are plenty of places in the Bay Area where knowing at least one other language could be quite helpful. If you want to go a bit further afield, it's less than 400 miles to LA, where "Se Hablo Espanol" signs are so common that you nobody notices them anymore, and it could be quite useful to speak either Mandarin or Cantonese.
Sure, but you can't easily tell the nationality of the singer, just of the composer. It's somewhat eerie hearing Japanese singing German music and sounding indistinguishable from native German speakers.
Arguably the only valid testing is to get it out there and wait for problems to appear. They will anyway, regardless of how long Debian
has taken.
Or, to put it another way, you can't really guarantee a 1 year uptime until you've actually run a few servers for a year without them going down. You never know if there's going to be a subtle, slow acting bug of some kind that happens to kill systems after 6 months until you test those systems for that kind of time scale.
But when one distributes the modified forms, the source must still be distributed under the terms in section 3. That means that it still must exist as cannonical TT form + patches, rather than as a complete integrated source package. The need to maintain all changes as separate patches against a cannonical source presents a formidable challenge to any attempt to fork the code. It is still theoretically possible to do so, but it would be difficult enough to be practically impossible.
Not to spark a big religious war, but I think that this points out a potential advantage of the GPL over BSD as a license. The vendors in the first case were able to make their own proprietary, incompatible extensions because the BSD license allows such behavior. The GPL places much sharper limits on the extent to which you can do so; you can develop proprietary programs that run under a GPL environment, but you can't make the environment itself proprietary.
I also think that this will have advantages for encouraging cooperation. With a license like BSD that allows proprietary development, there's a strong disincentive for commercial vendors to continue to release their hard work as Open Source, since it allows freeloaders to make their proprietary systems stronger. With the GPL, though, your competitors can't "steal" your hard work to the same extent; if they want to expand on it, you still have the opportunity to take advantage of their work.
Chapter and verse of the QPL that forbids forking, please?
Very well:
3. You may make modifications to the Software and distribute your modifications, in a form that is separate from the Software, such as patches. The
following restrictions apply to modifications:
This means that there is a single cannonical version of the software- the one released by TrollTech- and anything else is only allowed to exist as patches. That is not an absolute prevention of forking, but it puts up a sufficiently big obstacle to a fork that it is for all intents and purposes not possible.
It's pretty clearly become a hobby. Some people get started in overclocking because they see it as a way of getting a given performance for cheaper. After a while, the idea of getting cheap performance sort of falls by the wayside, and the idea is to see how much performance you can get out of a given system, whether it's cheaper to get it that way or not. It's fun playing with cool equipment, trying to get the absolute maximum out of a computer, and having bragging rights about how fast you can get your processor to go. It makes about as much sense as programming yet another Napster client clone; people do it as much to have fun doing it as because they actually expect the product to be particularly useful compared to the stuff that's already out there.
Slashdot Mirror
An interesting flip side to this argument is that the fast loading of pages that users find so handy implies that their bandwidth usage is probably lower than you might expect. After all, the reason the pages load so fast is because they're quite sparse and don't contain that much information. Some quick math suggests that their bandwidth usage probably isn't as high as you'd think: 2e7 pages per month times 1e5 bits per page (rough guess) gives about 2e12 bits per month. Since a month is about 2e6 seconds, that's of order 1e6 bits per second, or roughly T1 speed. Putting banner ads on their pages could wind up greatly increasing their outbound bandwidth.
Actually, though, their big bandwidth usage is probably their web crawler. Indexing a billion pages, even if it's not monthly, probably consumes several orders of magnitude more bandwidth than serving the search pages does. That actually suggests that they should be able to grow their business pretty easily; if indexing the web is the major cost, every added user is a win. That suggests that their business model (work on growing customer base first, then worry about making money on each customer) makes a lot of sense.
Of course a properly designed piece of censorware should be able to deal with this problem. They just have to be a bit cleverer about dealing with URLs. How do I know this? Because Junkbuster is capable of screening out Akamized content by looking at the whole URL. When it sees "ads.doublclick.net" in the URL, even if it isn't the base domain, it knows to filter it. A reasonably designed censorware proxy could do the same thing with "www.pr0nstars.com" or, sadly, "www.peacefire.org". Actually, you could turn Junkbuster into a Censorware proxy very easily by feeding it a different list of sites to block...
I think that the author gets one big point but misses another. He is correct that today an operating system is generally considered to contain more than just a kernel and a shell. What he misses is that Unix has grown more expansive in exactly the way that he suggests. For instance, he comments that a real OS needs to have "hundreds of utilities". That's exactly what Unix provides; most people think of grep, sed, awk, find, ps, etc. as being essential parts of Unix even though they obviously aren't part of the shell or kernel. Similarly, X11R6 is a key part of Unix as it is now perceived.
This is, IMO, a big part of the reason that Mac OSX won't be Unix even though it's based on underlying Unix technology. It doesn't incorporate all of the other stuff that's really part of Unix as it is understood to be, so the use of a Unix kernel doesn't make it Unix.
And that's why I don't like the title. After all, the publisher is going right out an saying that the books are for idiots. I think that it's part of a sad, but surprisingly general, confusion of ignorance and stupidity. I personally find it insulting to be told that I'm a dummy for not already knowing what the book is supposed to teach me. I think I'd personally be more inclined to buy the series if it were re-named [Blank} for Novices, rather than [Blank] for Dummies. They both get across the same idea, but one is much friendlier.
Bah! I've seen README file larger than the memory in my first computer. I've seen comment blocks larger than the memory in my first computer. Many, many computers today have larger block sizes on their hard drives than my first computer's entire memory. What's really scary is that my first hard drive wasn't large enough to hold the entire current source code for the Linux kernel- gzipped.
The net does not make people do what they otherwise would not do; it allows them to do so. It's not as though the net creates a desire in those people to view pornography, engage in vandalism, or do any of the other socially unacceptable acts you mention. The desire was pre-existing but kept under cover by the fear of public scrutiny.
Part of the reason that people are so attached to the anonymity of the net is because it gives them the opportunity to act out on their socially unacceptable desires without fear of public scorn. That's not to say that the net's influence this way is necessarily desirable for unleashing those feelings, but it's not causing them.
Sadly not true. I was able to get to the site listed, but I'm often not able to get to other sites mentioned in YROL because of filtering. It's just more evidence of how ineffective internet filtering is, that some sites that really seem as though they might be blocked by filtering software aren't but others (like ESR's home page on my system) are.
Sure, and after a while people tend to discover that some sources are especially reliable, and they pay extra attention to them. The fact that people will give that extra attention, and in many cases extra business, to the most reliable sources is a key part of the reason that the press is as trustworthy as it is. In a real sense it's like the way that peer review of Open Source code helps to ensure that nobody deliberately slips in security bugs. The risk of being caught is enough to keep people from even trying. The result is that we know that we know that we can generally trust the facts presented in the Wall Street Journal, that ZDnet is less reliable, and the Weekly World News is completely unreliable.
Of probably greater impact than the use of this kind of technology in the news media is its use in criminal justice. People are very heavily swayed by the perceived reliability of videotaped evidence. The fact that tapes can now be falsified with considerable ease, and that in many cases tapes of relevance in criminal cases will be unique and not subject to this kind of peer review leaves a very big and dangerous place for falsification. Imagine the police taping an interview with a criminal suspect, for instance, and then changing their tape to show the suspect waiving his right to counsel and confessing to the crime. It sounds as though that's perfectly possible, and even comparatively easy, with this technology. That's far scarier than a news program changing the logo on the side of a building.
Of course this is actually the way that things have always been. If you can't directly witness something yourself you are inherently trusting someone else to pass it on to you. You have no real knowledge of, to pick an example, whether there really was a Russian sub lost in the Barents Sea or whether it was an elaborate hoax. You're trusting that the people who bring you news are being honest and not showing you a bunch of crap.
Certainly this has always been the case in print media, hence the saying that you can't believe everything that you read. This invention doesn't really change anything, except that it makes the need for trust in your news deliverer more explicit- and in some well publicized cases so far showing how untrustworthy some of those news deliverers actually are.
And under the current system, the artists actually wind up paying for it! If you read about the current system, the record company will advance the musicians money against their royalties to produce, promote, and market the album. This means that for the most part the musicians don't actually receive any royalties unless the album sells at least a million copies. If it doesn't, the musicians wind up in debt to the record company. It's basically like share cropping, where the system is stacked so that it's almost impossible to break away from one company.
In any case, it's dubious that promotions and marketing are really an essential part of the process, or that they can't be solved by going outside the traditional music industry approach. If you don't think that Napster is changing the promotions and marketing aspect of things, you really need to pay more attention.
This is actually surprisingly common among the truly great, at least among scientists, where I spend most of my time. I think that there are two facets to the fact that so many really great people are modest.
One thing is that part of what makes them great at what they do is maintaining an honest assesment of their strengths and weaknesses. This is very helpful because it means that, as in the case of the Mindcraft test, they're able to go back and correct genuine errors rather than completely blowing them off as unfair.
The other thing is that they're able to be if not exactly content at least pleased enough with what they have done that they have the self confidence to admit what they haven't. This can actually enhance ones public image if you do it well, since people who are aware of ones actual accomplishments and don't need to be reminded of them are often impressed by ones apparent nonchalance at such an impressive achievement.
But Metallica hasn't been saying that you can't play MP3s, or burn backup copies of their CD's for personal archiving, or play their CD's using your computer's CD player. They've said that they don't like people giving away copies of their albums online. They have the right under copyright (and under Napster's use policy, for that matter) to keep people from doing that.
You need to separate what Metallica has said ("Don't trade songs from our albums on Napster, but bootlegs from our concerts are just fine.") from what the RIAA has said ("The only legal way to get recorded music is to buy our CDs at our prices. No copying for personal use, making MP3s for personal use, storing on computer, etc."). Metallica has expressed an interest in expanding into online music distribution (on their terms) once their current contractual obligations are completed. The RIAA has dragged its feet at every turn and proposed draconian limitations when on-line music formats have been discussed. One of those is reasonable, and the other one isn't.
Actually, there is a PowerPoint for Mac, so at least one of those programs isn't exclusive to Windows. More to the point, though, every time a Free Software hacker decides to create a free version of a commercial program, he's implicitly stating that he likes the original. After all, he likes what it does well enough to copy it!
Those who don't understand the Free Software ideal (as very distinct from Open Source) are doomed to say really stupid things about it. It seems really obvious to me that a person who would make the quote above doesn't understand the principles that he's complaining about. Complaints about QT were not about corporate involvement (as the above quote seems to suggest is the big sin among the FSF crowd) but about lack of programming freedom.
Part of the FSF ethic is that anyone is free to hack on the programs, and that "everyone" includes big corporations. To tell Sun and HP that they mayn't become involved is actually more contrary to the spirit of the GPL than accepting big corporate money. The problems come when people try to place restraints on the code. Given that everything is going to be kept open by the GPL, I don't see this as being a big problem.
Sure. All you have to do is to be your own creditor. Imagine the following steps:
Bingo, you've now managed to get big piles of customer info under false pretenses and profit from it.
Not necessarily resigned to being split up, but considering it as a strong possibility. Microsoft would have to be far dumber than anyone has ever accused them of being to believe that there's no chance of them being split up; they'd have to be awfully stupid to believe that the chance is even particularly small.
Given that the break up is a real possibility, it makes sense to look into porting MS apps to a wider range of OS's than they currently serve. It may even be a reasonable plan if they don't wind up being split up. If other operating systems (particularly Linux and the BSD's) continue to invade the desktop, sources of income other than Windows will become increasingly important. In that case the Apps division had damn well better be able to generate some sales to UNIXoid OSes, or the whole company is going to go into the tank.
Of course you could simplify this even more by using:
[mark@pate mark]$ su -c "cp index.html /home/httpd/html"
Password:
cp: overwrite '/home/httpd/html/index.html'? y
[mark@pate mark]$
Shouldn't that really be SPQR? I know that it would be a bit more difficult to come up with a good excuse for doing so, but it would obviously be a lot funnier.
You must use different miles from everyone else. There are plenty of places in the Bay Area where knowing at least one other language could be quite helpful. If you want to go a bit further afield, it's less than 400 miles to LA, where "Se Hablo Espanol" signs are so common that you nobody notices them anymore, and it could be quite useful to speak either Mandarin or Cantonese.
Sure, but you can't easily tell the nationality of the singer, just of the composer. It's somewhat eerie hearing Japanese singing German music and sounding indistinguishable from native German speakers.
Or, to put it another way, you can't really guarantee a 1 year uptime until you've actually run a few servers for a year without them going down. You never know if there's going to be a subtle, slow acting bug of some kind that happens to kill systems after 6 months until you test those systems for that kind of time scale.
But when one distributes the modified forms, the source must still be distributed under the terms in section 3. That means that it still must exist as cannonical TT form + patches, rather than as a complete integrated source package. The need to maintain all changes as separate patches against a cannonical source presents a formidable challenge to any attempt to fork the code. It is still theoretically possible to do so, but it would be difficult enough to be practically impossible.
Not to spark a big religious war, but I think that this points out a potential advantage of the GPL over BSD as a license. The vendors in the first case were able to make their own proprietary, incompatible extensions because the BSD license allows such behavior. The GPL places much sharper limits on the extent to which you can do so; you can develop proprietary programs that run under a GPL environment, but you can't make the environment itself proprietary.
I also think that this will have advantages for encouraging cooperation. With a license like BSD that allows proprietary development, there's a strong disincentive for commercial vendors to continue to release their hard work as Open Source, since it allows freeloaders to make their proprietary systems stronger. With the GPL, though, your competitors can't "steal" your hard work to the same extent; if they want to expand on it, you still have the opportunity to take advantage of their work.
Very well:
This means that there is a single cannonical version of the software- the one released by TrollTech- and anything else is only allowed to exist as patches. That is not an absolute prevention of forking, but it puts up a sufficiently big obstacle to a fork that it is for all intents and purposes not possible.
It's pretty clearly become a hobby. Some people get started in overclocking because they see it as a way of getting a given performance for cheaper. After a while, the idea of getting cheap performance sort of falls by the wayside, and the idea is to see how much performance you can get out of a given system, whether it's cheaper to get it that way or not. It's fun playing with cool equipment, trying to get the absolute maximum out of a computer, and having bragging rights about how fast you can get your processor to go. It makes about as much sense as programming yet another Napster client clone; people do it as much to have fun doing it as because they actually expect the product to be particularly useful compared to the stuff that's already out there.