We use SubEthaEdit for taking notes during Steve Jobs keynotes. It's brilliant for that. Alas, EtherPad won't work there because Apple always blocks Internet access in the keynote room.
Actually, the server in question was an Intel-based Xserve running Apache 2, MySQL, and a bunch of custom stuff. We believe the problem may have been drive failure.:-(
Yes, I believe this is basically a bandwidth problem. Although a passenger in the car definitely distracts, it's nowhere near as bad as trying to have a cell phone conversation where the brain is filling in for the massive drop in bandwidth (compared to someone sitting next to you). Think about it - the bandwidth of a cell phone call is less than 56 Kbps (can't remember, but that's what an analog POTS call is, and cell phones are much less, depending on how the carrier has shared out the bandwidth), whereas the bandwidth of an in-person conversation is huge, requiring much less fill-in processing to understand.
Count me among the people who feel that this is a mistake - speaking as an author, and as a publisher who has worked with hundreds of authors over the years, single authorship and control simply won't work in this sort of a situation. Maintaining content is hugely difficult and time-consuming, and not something that most authors do well (if at all). The beauty of the Wikipedia approach is that anyone who wants can contribute as much or as little as they want, as frequently as they want. If one person loses interest, there's always room for another to take over. There's also an implication in Manber's post that knols will be of high quality because of this authorial ownership. That will be true of some, but the reality of the situation is that most people, even if they are expert in some topic, can't write out their way of the proverbial paper bag. Many won't even have the necessary skills to organize the source material - this stuff isn't nearly as easy as it sounds.
A few other questions: What happens when there are copyright infringement claims against knols that plagiarize content from elsewhere? Will knol authors start by just stealing Wikipedia articles, and will Google act to prevent that? Will Google's policies disallowing specific content for services like Google Groups apply to Google Knol? What happens when a knol author gets busy, becomes bored with a knol, or dies? Will Google be able to argue in international court that it has no oversight over illegal content created using its own service? There's nothing new here, but the bigger the company, the bigger the target.
One great thing that Engst mentioned (but it was sort of glossed over by the writer) was this: these e-things are promoted in TidBITS. This gives them -the- major advantage over most eBooks: PR.
You're absolutely right. We've spent 14 years publishing TidBITS every week for free to tens of thousands of readers, which makes all this possible. We're not complacent about it; we're certainly working to expand the audience beyond TidBITS in both traditional and non-traditional ways, but it's good to have a base from which to work. To be clear, we generally promote the ebooks in TidBITS with an excerpted article that we would have published happily even if there had been no ebook to publish. So hopefully even our marketing is useful.
Anyone else think these books are overpriced, considering the low-level topics? Ten dollars for an electronic pamphlet on how to share files in Panther? I just bought a three hundred page (real) book on DVD Studio Pro for twenty!
Only you can decide if any one book is overpriced, of course, but "Take Control of Sharing Files in Panther" is 104 pages, highly detailed, completely searchable, and comes with free updates for that $10. And it's not a low-level book by any means; it has instructions for reconfiguring httpd.conf to turn on WebDAV, twiddling smb.conf to share specific folders, and editing the AppleShare.plist file to turn off guest access. Perhaps you know all of those things, but plenty of people don't and have been grateful for the assistance provided in the book.
One of the points of the Take Control model is that we can write about topics that aren't big enough to fill an entire paper book. That doesn't make them less interesting, or readers less desirous of reading them, just shorter.
I think they'd make more money by providing the books for the cost of download ($2 or $3 a e-book) and then offering exclusive paperbacks/hardbacks to people who want them at $20 or $30 a pop, or they can offer books that'll last forever for mroe. After a year, they begin throwing books up free to download with advertisements in the front and back for paperbacks. The good books will gross a lot of sales for paperbacks, as the demand for them is still there. If I had the money I'd certainly buy or print a number of books I have, and a few (scabbed wings of abadon, www.rantradio.ca for that one) I *REALLY* want a paperback of, not only because the book is so damn kewl, but so I can loan it to friends.
You may actually be more prescient than you know.:-) However, keep in mind a few things. Paperbacks in the computer book world may last for years, but they're obsolete in months in many cases. Ebooks can last forever, in contrast, because they can be updated.
Also, if you backload all your recompense for a book (giving an ebook away to goose print sales), you as the author have to be able to survive for a long time with no income from that particular work. That's because it would take six months from the time the print book came out to see any of those royalties with the royalty payment schedules of many print publishers (they pay quarterly, for the previous quarter). I'm not saying it's a bad idea, just that it's one that requires an author who's willing to invest a lot of time for little known reward (until such a model could be proven) or reward that's far in the future. It's a cash flow problem, in essence.
All that said, watch this space - we have lots of things we want to try.
You can have Bayesian filtering in Mail, with SpamSieve from Michael Tsai.
You might also be interested in reading Joe Kissell's just-released ebook Take Control of Spam with Apple Mail, which explains the common accuracy problems with Mail's Junk filter and how to optimize it for better results. Joe also recommends SpamSieve as an alternative to Mail's Junk filter in those instances where Mail proves inadequate.
Yes, slashdot is a great and mighty force before which old and meek Macintosh Web servers must bow.:-)
(Though it is back up now, and serving 75 simultaneous connection as fast as it can.)
Development resources being what they are, it's taking us quite some time to move to Mac OS X running on a Xserve, which would likely address this problem (and the fact that I can't get email while the Web server is serving so many simultaneous connections).
I suppose asking why Slashdot doesn't automatically cache these linked pages and redirect queries automatically to a cached version if the remote site crumbles under the load wouldn't be helpful?:-)
We have several industries that are unfair: Those industries built on selling information. Authors, entertainors, software producers, musicians, etc. have been raking in the dough by dealing in information. They have a create once, sell many times scam going.
Careful there. I'm as apalled at the venality of the Content Cartel as the next person, but I think there's a big difference between the people who actually create content (remember, the ones who were supposed to be protected by copyright) and the industries that have grown up to exploit the creations of these people. The vast majority of authors, entertainers, programmers, and musicians are lucky if they can even begin to earn a living by writing, performing, programming, or playing.
If anything, I think one of the things we individuals can do is to support the content creators who are not part of the Content Cartel. If a few more of the little people can start to make a living from their content, we'll all benefit. Personally, I've made a policy of buying CDs from musicians I hear live whose music I like - no major decision, just if I like the music, I buy the CD without any worrying if it's worth it. That applies to groups I see in concerts, restaurants, and so on, as well as street performers.
We do cache all frequently accessed content out of our database, but the database server is behind a 128K ISDN line (long and ugly story related to DSL firms self-destructing), so the turnaround time was just too long even on the article file.
Now we have the article cached on our main server, so all the database server has to do is redirect hits to the main server. That's working fine - even the 128K ISDN line can do that. Our main server is handling 75 simultaneous connections at the moment - I had it up to 100, but brought it back down after a crash. That will remain the bottleneck - digital.forest has an OC-12.
Keep in mind, folks, that our hardware, software, and bandwidth solutions have arisen in a situation where we're trying to do things in a way that's as cheap, appropriate to our primary audience (savvy Mac users), and simple as possible. As such, all this was put together over the last four or five years and is changed only when necessary, not just because there's newer hardware or software available.
So the database server is a Performa 6400 running WebSTAR 3.1 and serving data out of a FileMaker database (don't get me started) via Lasso; our main server is a Power Mac 7600 running WebSTAR 3.0 and serving static files.
And yes, we'd like to move everything to a coherent Mac OS X solution running on an Xserve, but when you've built a huge amount of infrastructure using strings, twigs, and baling wire, it's not an easy thing to do while still trying to put out a weekly publication.:-)
We're working on it - we can normally handle up to 45 simultaneous connections on our database server, but it's behind a slow line and, to paraphrase Monty Python, "No one ever expects the Spanish Slashdot!"
We're moving that particular article to our main server, which can handle more simultaneous connections and has way more bandwidth thanks to digital.forest's huge pipes. Should be up soon.
Slashdot Mirror
Google Docs is great, but it doesn't update in real time. There's always a lag that gets in the way for quick collaboration.
We use SubEthaEdit for taking notes during Steve Jobs keynotes. It's brilliant for that. Alas, EtherPad won't work there because Apple always blocks Internet access in the keynote room.
Actually, the server in question was an Intel-based Xserve running Apache 2, MySQL, and a bunch of custom stuff. We believe the problem may have been drive failure. :-(
cheers... -Adam
Yes, I believe this is basically a bandwidth problem. Although a passenger in the car definitely distracts, it's nowhere near as bad as trying to have a cell phone conversation where the brain is filling in for the massive drop in bandwidth (compared to someone sitting next to you). Think about it - the bandwidth of a cell phone call is less than 56 Kbps (can't remember, but that's what an analog POTS call is, and cell phones are much less, depending on how the carrier has shared out the bandwidth), whereas the bandwidth of an in-person conversation is huge, requiring much less fill-in processing to understand.
Count me among the people who feel that this is a mistake - speaking as an author, and as a publisher who has worked with hundreds of authors over the years, single authorship and control simply won't work in this sort of a situation. Maintaining content is hugely difficult and time-consuming, and not something that most authors do well (if at all). The beauty of the Wikipedia approach is that anyone who wants can contribute as much or as little as they want, as frequently as they want. If one person loses interest, there's always room for another to take over. There's also an implication in Manber's post that knols will be of high quality because of this authorial ownership. That will be true of some, but the reality of the situation is that most people, even if they are expert in some topic, can't write out their way of the proverbial paper bag. Many won't even have the necessary skills to organize the source material - this stuff isn't nearly as easy as it sounds.
A few other questions: What happens when there are copyright infringement claims against knols that plagiarize content from elsewhere? Will knol authors start by just stealing Wikipedia articles, and will Google act to prevent that? Will Google's policies disallowing specific content for services like Google Groups apply to Google Knol? What happens when a knol author gets busy, becomes bored with a knol, or dies? Will Google be able to argue in international court that it has no oversight over illegal content created using its own service? There's nothing new here, but the bigger the company, the bigger the target.
More in my TidBITS article at: http://db.tidbits.com/article/9360
cheers... -Adam
One great thing that Engst mentioned (but it was sort of glossed over by the writer) was this: these e-things are promoted in TidBITS. This gives them -the- major advantage over most eBooks: PR.
You're absolutely right. We've spent 14 years publishing TidBITS every week for free to tens of thousands of readers, which makes all this possible. We're not complacent about it; we're certainly working to expand the audience beyond TidBITS in both traditional and non-traditional ways, but it's good to have a base from which to work. To be clear, we generally promote the ebooks in TidBITS with an excerpted article that we would have published happily even if there had been no ebook to publish. So hopefully even our marketing is useful.
cheers... -Adam (Take Control publisher)
Anyone else think these books are overpriced, considering the low-level topics? Ten dollars for an electronic pamphlet on how to share files in Panther? I just bought a three hundred page (real) book on DVD Studio Pro for twenty!
Only you can decide if any one book is overpriced, of course, but "Take Control of Sharing Files in Panther" is 104 pages, highly detailed, completely searchable, and comes with free updates for that $10. And it's not a low-level book by any means; it has instructions for reconfiguring httpd.conf to turn on WebDAV, twiddling smb.conf to share specific folders, and editing the AppleShare .plist file to turn off guest access. Perhaps you know all of those things, but plenty of people don't and have been grateful for the assistance provided in the book.
One of the points of the Take Control model is that we can write about topics that aren't big enough to fill an entire paper book. That doesn't make them less interesting, or readers less desirous of reading them, just shorter.
cheers... -Adam (Take Control publisher)
I think they'd make more money by providing the books for the cost of download ($2 or $3 a e-book) and then offering exclusive paperbacks/hardbacks to people who want them at $20 or $30 a pop, or they can offer books that'll last forever for mroe. After a year, they begin throwing books up free to download with advertisements in the front and back for paperbacks. The good books will gross a lot of sales for paperbacks, as the demand for them is still there. If I had the money I'd certainly buy or print a number of books I have, and a few (scabbed wings of abadon, www.rantradio.ca for that one) I *REALLY* want a paperback of, not only because the book is so damn kewl, but so I can loan it to friends.
You may actually be more prescient than you know. :-) However, keep in mind a few things. Paperbacks in the computer book world may last for years, but they're obsolete in months in many cases. Ebooks can last forever, in contrast, because they can be updated.
Also, if you backload all your recompense for a book (giving an ebook away to goose print sales), you as the author have to be able to survive for a long time with no income from that particular work. That's because it would take six months from the time the print book came out to see any of those royalties with the royalty payment schedules of many print publishers (they pay quarterly, for the previous quarter). I'm not saying it's a bad idea, just that it's one that requires an author who's willing to invest a lot of time for little known reward (until such a model could be proven) or reward that's far in the future. It's a cash flow problem, in essence.
All that said, watch this space - we have lots of things we want to try.
cheers... -Adam (Take Control publisher)You can have Bayesian filtering in Mail, with SpamSieve from Michael Tsai.
You might also be interested in reading Joe Kissell's just-released ebook Take Control of Spam with Apple Mail, which explains the common accuracy problems with Mail's Junk filter and how to optimize it for better results. Joe also recommends SpamSieve as an alternative to Mail's Junk filter in those instances where Mail proves inadequate.
cheers... -Adam
Yes, slashdot is a great and mighty force before which old and meek Macintosh Web servers must bow. :-)
:-)
(Though it is back up now, and serving 75 simultaneous connection as fast as it can.)
Development resources being what they are, it's taking us quite some time to move to Mac OS X running on a Xserve, which would likely address this problem (and the fact that I can't get email while the Web server is serving so many simultaneous connections).
I suppose asking why Slashdot doesn't automatically cache these linked pages and redirect queries automatically to a cached version if the remote site crumbles under the load wouldn't be helpful?
cheers... -Adam
We have several industries that are unfair: Those industries built on selling information. Authors, entertainors, software producers, musicians, etc. have been raking in the dough by dealing in information. They have a create once, sell many times scam going.
Careful there. I'm as apalled at the venality of the Content Cartel as the next person, but I think there's a big difference between the people who actually create content (remember, the ones who were supposed to be protected by copyright) and the industries that have grown up to exploit the creations of these people. The vast majority of authors, entertainers, programmers, and musicians are lucky if they can even begin to earn a living by writing, performing, programming, or playing.
If anything, I think one of the things we individuals can do is to support the content creators who are not part of the Content Cartel. If a few more of the little people can start to make a living from their content, we'll all benefit. Personally, I've made a policy of buying CDs from musicians I hear live whose music I like - no major decision, just if I like the music, I buy the CD without any worrying if it's worth it. That applies to groups I see in concerts, restaurants, and so on, as well as street performers.
cheers... -Adam
We do cache all frequently accessed content out of our database, but the database server is behind a 128K ISDN line (long and ugly story related to DSL firms self-destructing), so the turnaround time was just too long even on the article file.
:-)
Now we have the article cached on our main server, so all the database server has to do is redirect hits to the main server. That's working fine - even the 128K ISDN line can do that. Our main server is handling 75 simultaneous connections at the moment - I had it up to 100, but brought it back down after a crash. That will remain the bottleneck - digital.forest has an OC-12.
Keep in mind, folks, that our hardware, software, and bandwidth solutions have arisen in a situation where we're trying to do things in a way that's as cheap, appropriate to our primary audience (savvy Mac users), and simple as possible. As such, all this was put together over the last four or five years and is changed only when necessary, not just because there's newer hardware or software available.
So the database server is a Performa 6400 running WebSTAR 3.1 and serving data out of a FileMaker database (don't get me started) via Lasso; our main server is a Power Mac 7600 running WebSTAR 3.0 and serving static files.
And yes, we'd like to move everything to a coherent Mac OS X solution running on an Xserve, but when you've built a huge amount of infrastructure using strings, twigs, and baling wire, it's not an easy thing to do while still trying to put out a weekly publication.
cheers... -Adam
We're working on it - we can normally handle up to 45 simultaneous connections on our database server, but it's behind a slow line and, to paraphrase Monty Python, "No one ever expects the Spanish Slashdot!"
We're moving that particular article to our main server, which can handle more simultaneous connections and has way more bandwidth thanks to digital.forest's huge pipes. Should be up soon.
cheers... -Adam