I can't really see why a customer would want to use it though. Why not just use a real web host? Amazon S3 has is no minimum monthly fee, has redundancy built into it, guaranteed availability.
Compared with Dreamhost (say) which has a bundle for almost 10 USD/mo. That deal has 20GB + 1TB transfers. For the same amout on Amazon S3 you only 5 GB + 64GB transfers, and doesn't have FTP nor SSH access, nor your own domain, etc etc.
The point of Amazon's service is to complement a hosted web app. Say I want to start a Flickr competitor for movies (like video.google.com), I sign up for Dreamhost's $8/mo (parent is wrong at $10/mo) deal. My website, my (mysql) database, my php scripts and cron jobs are on Dreamhost, but oh crap, I've suddenly run out of disk space.
For $32 - 60 G, $64 - 90 G, then what?
Dreamhost is unusual in it's range of options, look at yahoo's hosting plan's: $9-5G, $20-10G, $40-20G. Yahoo's "Merchant" plans offer no more than 20G either, even their $300/mo plan.
There's also a real benefit in paying only for what you're using.
This begs the question as to why yahoo doesn't have a more flexible disk storage pricing plan for hosting, but I think that most web hosting companies aren't in the massive data storage business. Dreamhost and Serverbeach, etc spend their money on big pipes, not administering SANS.
I think that there is a real need for this type of service and the choice of interface (SOAP and REST) demonstrates the market. Not personal backup, not even backup at all, rather immediately accessible storage of user data to hosted web applications.
This isn't a fair criticism, because you should dump all before upgrading your database. PostgreSQL just enforces this behavior. MySQL lets you do it half-assed with the possibility of losing your databases.
1) Right now we are going through another bubble I think with venture capital. Too many stupid ideas are getting funded. ...
2) At least they fail cheaply.
Doesn't anyone notice the contradiction here?
What are the high-profile web2.0 sites/businesses that have succeeded? Flickr, del.icio.us, writely, basecamp/ta-da lists, Upcoming,...
What are the high-profile web2.0 sites that have failed? Where is web2.0's pets.com?
Sure ajax and web2.0 are dumb monikers, but the biggest and, arguably, smartest web companies are snapping up these new Ajax sites.
[Stallman and the FSF] are "zealots" for lack of a better term. For them, free software is less about open source and open development and more about a form of political agenda.
Free as in libre. Open source software is inherently political.
The 2nd result for this query is a negative article:
Asktheheadhunter.com: Bernard Haldane Professional headhunter explains how Bernard Haldane has been "busted" in several states by promising jobhunters access to the "secret job market" for a...
Based off what I've seen in the past, I predict del.icio.us is about to become useless crap, infested with ads, and a generally unpleasant experience.
Every time I've used a service which Yahoo has come on board to, it becomes unuseable crap. And shorly after they come on board I end up stopping using it because it is no longer as good as it once was.
Which is really too bad, because I was beginning to find del.icio.us exceedingly useful to me. Now I'll probably have to export all of my bookmarks and see what else I can find.
I'm not sure what yahoo can do to make del.icio.us less desirable.
I use firefox plugins to bookmark sites and the rss feed to review my bookmarks. I rarely, actually visit http://del.icio.us/ I couldn't care what they do to delicious' front page.
Moodle and Sakai simply don't do the same things on the same scale as WebCT and presumably Blackboard. It's like comparing Dia to Visio, of course we'd all rather use Dia, but we go with the more functional product.
WebCT "Campus Edition" vs WebCT "Vista"
Campus Edition was this hacked together organically grown POS. I worked a little with the web services functionality of Vista and I must say that it's well-done. All of Vista's functionality is accessible through an Apache Axis layer. Admittedly it's complicated, but that because it's designed for VL educational institutions.
WebCT Vista is a thoroughly engineered modern product. Those of you complaining about the UI aren't treating it fairly. One could literally write their own web UI by hooking into Vista without editing product code at all.
It's even pluggable. It's relatively easy to write multiple authentication/session modules. Does Sakai even have LDAP integration possibilities?
The last thing that I want to do is to disparage the f/oss efforts, but from reading the current posts one would question why anyone uses these real, enterprise-capable course management systems instead of these less functional, less capable and less proven f/oss packages.
1) why would anyone want to rely on a net connection to be able to write a letter,
2) or trust a remote company to hold their data,
3) The world of users was ebullient when it shook off the shackels of having to connect to a mainframe to do work; why would they want to give that freedom up?
1) Computers are unreliable and if the internet is down, the computer might as well be down to, 'cause letter = email.
2) The average computer user's letter is a lot more secure and persistant on a google server than on a home pc hard drive. Over the last ten years, my parents lost all of the data on their pc an average of once every 1.5 years. I don't think that this is that uncommon of a statistic. What if google still had your college essays?
Definitely the trust is an issue, but it has to be (and will be) weighed against the benefits.
3) How much does a mediocre computer cost that can run outlook, Word, Powerpoint, Internet Explorer, an antivirus program and the round of anti-malware programs? $500+? (Including monitor and accessories.)
When people talk about the digital divide, this is what they're talking about. $500 bucks is a weeks paycheck for a majority of the households in the southern US at least, not to mention developing nations.
Give me a 17 inch flat panel with a network card, a couple of USB ports, and a DVD reader and I'll sign a two-year contract for $60/month. Hell yeah!
My wife would love it. When something went wrong, she could just call the local _ company and they'd ship us a new one.
4: Google acquires Project Gutenberg and expands on their free, public domain, efforts. PROFIT - at least if you're associated with PG.
5: Public Domain is strengthened for all of us because works in PD are now more accessable to everyone. PROFIT - more traffic to Google to get these works, and society overall is richer!
I'm as big a fan of google as the average stereotypical slashdotter.
But, we really need to rein in our imaginations.
When rumors of google print first popped up on the net, it got compared to the Library at Alexandria . I think that we need multiple, accessible repositories of the world's knowledge. But google ain't ever gonna be it. Why? Because google may not be evil, but they aren't good either. They're trying to store every printed book, but what are they going to do with such extremely valuable property? They're gonna let you see a few paragraphs at a time next to non-standard html tables containing advertisements. WTF? This is what mankind is going to do with the single largest collection of knowledge ever assembled?
Honestly, google scares me. Microsoft is too short-sighted to grasp at the kind of power that google is reaching for. The world's information, in one place, owned by one company.
The world needs a Library of Alexandria. Become a participating user at wikipedia.
4: Google acquires Project Gutenberg and expands on their free, public domain, efforts. PROFIT - at least if you're associated with PG.
5: Public Domain is strengthened for all of us because works in PD are now more accessable to everyone. PROFIT - more traffic to Google to get these works, and society overall is richer!
I'm as big a fan of google as the average stereotypical slashdotter. Google isn't evil (yet).
But, we really need to rein in our imaginations.
When rumors of google print first popped up on the net, it got compared to the Library at Alexandria which reputedly held all the world's knowledge. I think that we need multiple, accessible repositories of the world's knowledge. But google ain't never gonna be it. Why? Because google may not be evil, but they aren't good either. They're trying to store every printed book, but what are they going to do with such extremely valuable property? They're gonna let you see a few paragraphs at a time next to non-standard html tables containing advertisements. WTF? This is what mankind is going to do with the single largest collection of knowledge ever assembled?
Honestly, google scares me. Microsoft is too short-sighted to grasp at the kind of power that google is reaching for. The world's information, in one place, owned by one company.
The world needs a Library of Alexandria. Become a participating user at wikipedia.
This seems like a common problem, how does one protect again appending sensitive information from a protected document into an ordinary text or non-sensitive file? Is there a technology out there that can mark the data so it can not be copied into another file even though it is accessible to some. Apparently the 'now retired faculty member' had access to the file. Probably used cut and paste to imbed it into a file he/she could access from home/laptop etc. We had lots of problems like this at government locations I worked at
"Marking data so it can not be copied into another file" is the wrong way to think about this issue. That's a microsoft approach.
The problem was that the format was a text file and that the only way the faculty member had access to the data was by being emailed to him/her from the dean.
A data modeling/analysis program or custom web interface that allows authorized people (where the correct people are authorized) is the solution to this problem. This professor didn't want a text file of every student's information. S/he probably wanted something like a comparison of GPA's between different campus organized by majors, or something like that.
The problem wasn't that there wasn't enough meta-data and a required file reader restricting a user to what they can do with their data, but rather that a Dean wanted a person to have access to certain data and the only way to give it was to send a text file over.
The need for true-to-life authorization & useful data analysis tools are what caused this problem.
As a web programmer for a large US university, I can easily see how this can happen. IT staff aren't the people with sensitive data. In a university this large, there are alot of people with a business need for enrollment information and student identifiers. It's unfortunate that ssn's are the prevalent identifiers, but many institutions are moving away from ssn's because of this type of cautionary story and press coverage.
I think it's interesting that Miami doesn't know if this data was accessed by unauthorized people.
All they'd have to do was grep through their web server access logs and look at ips. There's a small, but important, possibility that no one ever accessed this file other than the professor. Or at least, they'd know ips for those who had accessed the file.
The people who make the budgets and track the money often seem openly hostile to hearing from the classroom teachers -- they want to set budgets without asking us what the priorities are from our vantage.
If you look at university or college level administration, they were all grad students and teachers at some point. The way to become a dean or the president of a college is to get a PhD in something, teach it for a few years and then try to move up. If you compare this to a highschool/middle/elementary school, the way to become a principal or high-level administrator is to get a degree in Education Administration, no years-long period of teaching required.
It's true that Solaris 10 has a few handy features Linux may be currently missing.
But Linux has *far* more developers, and will continue to until Solaris 10 becomes portable to over 20 architectures and begins to include tons of hardware support.
And while they're working on closing that gap, Linux will tie up any loose ends. When Sun surfaces from their driver-writing festival to get their bearings, they're going to be eating Linux's dust.
The few handy features that solaris 10 has that Linux doesn't, are system-wide features that the open-source/distributed development (bazaar) model have difficulty with. Like dTrace. The linuxy model would end up with three competing methods that admins and developers would have to sort through and none of which would be as complete.
This is the same reason, they won't give up stewardship of java.
It seems to me that Sun has never gotten very serious about putting out Free Software for x86, not only because they can't make any money off it, but also because it cuts into the profits from their products that they can make money off of.
Additionally, Sun knows that Sparc is dying, that's why they're handing over a lot of control to Fujitsu. Sun'll be concentrating on x86 in the next five years, and I believe will start to make inroads in the commercial server market.
I've even seen Slurp get stuck on some of my pages (www.math.purdue.edu) and continusouly request the same page. There was a month or two when Slurp was generating a serious amount of our traffic.
I'm the pleased owner of a dual G5 powerpc and a g4 powerbook, but if apple does advance intel's drm, I'll bail ship and swtich fulltime to... Solaris 10 and ubuntu?
Support sites, don't use adblock, yada yada google yada yada...
One cannot easily block google ads because they're simple html tables and text. One the first five things I do with a computer I'm using is (after installing Thunderbird, Firefox and web developer extension ) install Adblock and find an ad from ad.doubleclick.net to block all images from their servers.
My point is that I don't go to the trouble to block text ads, but I get an obstinate pleasure out of blocking banners and flash ads.
Now perhaps google really is evil, but merely aware of this attitude, or perhaps google really does try to not be evil.
However, in 5-10 years if the PATRIOT act is still around, I believe things will change greatly. Once the US stops chasing people around the globe these very convenient changes in rights and law will be used against everyone equally.
Of course you're right that the USA PATRIOT Act will be used against everyone. Already the term "terrorism" has less specific meaning. Am I terrorizing the 7-11 clerk, when I rob the store? Yes? Bam, Patriot Act.
Where you're wrong is assuming that the US will stop chasing people around the globe. The "War on Terrorism" (like the "War on Drugs", neither of which are wars) will never end. All we can hope for is a government that ackowledges this and doesn't try to change the nature of American citizenship because of it.
Real robust enterprise web site design uses table-driven database-back-end solid server-side code with (x)html templating
I'm trying to explore some templating options for a smaller web site within a university's site. The setup is solaris, apache 1.3, and postgresql. Currently, the site is running php4 using smarty for templating.
My initial reaction to smarty is: wth? we're simply adding a layer of complexity (and another layer of computation) by putting a different programming language in the html page.
It seems to me that templating systems often end up arbitrarily complicated; hence the desire for a more standard templating structure.
I would like to ask what other professional web masters/developers/designers use to accomplish templating.
Slashdot Mirror
Your point being???
Compared with Dreamhost (say) which has a bundle for almost 10 USD/mo. That deal has 20GB + 1TB transfers. For the same amout on Amazon S3 you only 5 GB + 64GB transfers, and doesn't have FTP nor SSH access, nor your own domain, etc etc.
The point of Amazon's service is to complement a hosted web app. Say I want to start a Flickr competitor for movies (like video.google.com), I sign up for Dreamhost's $8/mo (parent is wrong at $10/mo) deal. My website, my (mysql) database, my php scripts and cron jobs are on Dreamhost, but oh crap, I've suddenly run out of disk space.
For $32 - 60 G, $64 - 90 G, then what?
Dreamhost is unusual in it's range of options, look at yahoo's hosting plan's: $9-5G, $20-10G, $40-20G. Yahoo's "Merchant" plans offer no more than 20G either, even their $300/mo plan.
There's also a real benefit in paying only for what you're using.
This begs the question as to why yahoo doesn't have a more flexible disk storage pricing plan for hosting, but I think that most web hosting companies aren't in the massive data storage business. Dreamhost and Serverbeach, etc spend their money on big pipes, not administering SANS.
I think that there is a real need for this type of service and the choice of interface (SOAP and REST) demonstrates the market. Not personal backup, not even backup at all, rather immediately accessible storage of user data to hosted web applications.
This isn't a fair criticism, because you should dump all before upgrading your database. PostgreSQL just enforces this behavior. MySQL lets you do it half-assed with the possibility of losing your databases.
2) At least they fail cheaply.
Doesn't anyone notice the contradiction here?
What are the high-profile web2.0 sites/businesses that have succeeded? Flickr, del.icio.us, writely, basecamp/ta-da lists, Upcoming, ...
What are the high-profile web2.0 sites that have failed? Where is web2.0's pets.com?
Sure ajax and web2.0 are dumb monikers, but the biggest and, arguably, smartest web companies are snapping up these new Ajax sites.
Free as in libre. Open source software is inherently political.
The bazaar development model itself is political.
The 2nd result for this query is a negative article:
Asktheheadhunter.com: Bernard Haldane
Professional headhunter explains how Bernard Haldane has been "busted" in several states by promising jobhunters access to the "secret job market" for a
Now go ahead and post this message down, fanboy
I'm not sure what yahoo can do to make del.icio.us less desirable.
I use firefox plugins to bookmark sites and the rss feed to review my bookmarks. I rarely, actually visit http://del.icio.us/ I couldn't care what they do to delicious' front page.
It seems that digg userbase is largely n00bs.
If you're just reading the articles, nothing beats diggdot.us
It's Slashdot's top stories, digg's top stories and del.icio.us/popular links.
There are some stories that slashdot posts a few hours before digg.com, but digg had this story on their front page, days ago
Moodle and Sakai simply don't do the same things on the same scale as WebCT and presumably Blackboard. It's like comparing Dia to Visio, of course we'd all rather use Dia, but we go with the more functional product.
WebCT "Campus Edition" vs WebCT "Vista"
Campus Edition was this hacked together organically grown POS. I worked a little with the web services functionality of Vista and I must say that it's well-done. All of Vista's functionality is accessible through an Apache Axis layer. Admittedly it's complicated, but that because it's designed for VL educational institutions.
WebCT Vista is a thoroughly engineered modern product. Those of you complaining about the UI aren't treating it fairly. One could literally write their own web UI by hooking into Vista without editing product code at all.
It's even pluggable. It's relatively easy to write multiple authentication/session modules. Does Sakai even have LDAP integration possibilities?
The last thing that I want to do is to disparage the f/oss efforts, but from reading the current posts one would question why anyone uses these real, enterprise-capable course management systems instead of these less functional, less capable and less proven f/oss packages.
2) or trust a remote company to hold their data,
3) The world of users was ebullient when it shook off the shackels of having to connect to a mainframe to do work; why would they want to give that freedom up?
1) Computers are unreliable and if the internet is down, the computer might as well be down to, 'cause letter = email.
2) The average computer user's letter is a lot more secure and persistant on a google server than on a home pc hard drive. Over the last ten years, my parents lost all of the data on their pc an average of once every 1.5 years. I don't think that this is that uncommon of a statistic. What if google still had your college essays?
Definitely the trust is an issue, but it has to be (and will be) weighed against the benefits.
3) How much does a mediocre computer cost that can run outlook, Word, Powerpoint, Internet Explorer, an antivirus program and the round of anti-malware programs? $500+? (Including monitor and accessories.)
When people talk about the digital divide, this is what they're talking about. $500 bucks is a weeks paycheck for a majority of the households in the southern US at least, not to mention developing nations.
Give me a 17 inch flat panel with a network card, a couple of USB ports, and a DVD reader and I'll sign a two-year contract for $60/month. Hell yeah!
My wife would love it. When something went wrong, she could just call the local _ company and they'd ship us a new one.
5: Public Domain is strengthened for all of us because works in PD are now more accessable to everyone. PROFIT - more traffic to Google to get these works, and society overall is richer!
I'm as big a fan of google as the average stereotypical slashdotter.
But, we really need to rein in our imaginations.
When rumors of google print first popped up on the net, it got compared to the Library at Alexandria . I think that we need multiple, accessible repositories of the world's knowledge. But google ain't ever gonna be it. Why? Because google may not be evil, but they aren't good either. They're trying to store every printed book, but what are they going to do with such extremely valuable property? They're gonna let you see a few paragraphs at a time next to non-standard html tables containing advertisements. WTF? This is what mankind is going to do with the single largest collection of knowledge ever assembled?
Honestly, google scares me. Microsoft is too short-sighted to grasp at the kind of power that google is reaching for. The world's information, in one place, owned by one company.
The world needs a Library of Alexandria. Become a participating user at wikipedia.
5: Public Domain is strengthened for all of us because works in PD are now more accessable to everyone. PROFIT - more traffic to Google to get these works, and society overall is richer!
I'm as big a fan of google as the average stereotypical slashdotter. Google isn't evil (yet).
But, we really need to rein in our imaginations.
When rumors of google print first popped up on the net, it got compared to the Library at Alexandria which reputedly held all the world's knowledge. I think that we need multiple, accessible repositories of the world's knowledge. But google ain't never gonna be it. Why? Because google may not be evil, but they aren't good either. They're trying to store every printed book, but what are they going to do with such extremely valuable property? They're gonna let you see a few paragraphs at a time next to non-standard html tables containing advertisements. WTF? This is what mankind is going to do with the single largest collection of knowledge ever assembled?
Honestly, google scares me. Microsoft is too short-sighted to grasp at the kind of power that google is reaching for. The world's information, in one place, owned by one company.
The world needs a Library of Alexandria. Become a participating user at wikipedia.
"Marking data so it can not be copied into another file" is the wrong way to think about this issue. That's a microsoft approach.
The problem was that the format was a text file and that the only way the faculty member had access to the data was by being emailed to him/her from the dean.
A data modeling/analysis program or custom web interface that allows authorized people (where the correct people are authorized) is the solution to this problem. This professor didn't want a text file of every student's information. S/he probably wanted something like a comparison of GPA's between different campus organized by majors, or something like that.
The problem wasn't that there wasn't enough meta-data and a required file reader restricting a user to what they can do with their data, but rather that a Dean wanted a person to have access to certain data and the only way to give it was to send a text file over.
The need for true-to-life authorization & useful data analysis tools are what caused this problem.
I think it's interesting that Miami doesn't know if this data was accessed by unauthorized people.
All they'd have to do was grep through their web server access logs and look at ips. There's a small, but important, possibility that no one ever accessed this file other than the professor. Or at least, they'd know ips for those who had accessed the file.
Presumably, they're not keeping logs this long.
If you look at university or college level administration, they were all grad students and teachers at some point. The way to become a dean or the president of a college is to get a PhD in something, teach it for a few years and then try to move up. If you compare this to a highschool/middle/elementary school, the way to become a principal or high-level administrator is to get a degree in Education Administration, no years-long period of teaching required.
At least, that seems to be the way its done.
But Linux has *far* more developers, and will continue to until Solaris 10 becomes portable to over 20 architectures and begins to include tons of hardware support.
And while they're working on closing that gap, Linux will tie up any loose ends. When Sun surfaces from their driver-writing festival to get their bearings, they're going to be eating Linux's dust.
The few handy features that solaris 10 has that Linux doesn't, are system-wide features that the open-source/distributed development (bazaar) model have difficulty with. Like dTrace. The linuxy model would end up with three competing methods that admins and developers would have to sort through and none of which would be as complete.
This is the same reason, they won't give up stewardship of java.
OpenOffice.org's GLOW directly competes with their Java Calendar Server software.
Sun developers work on Glow: http://groupware.openoffice.org/glow/team.html
Additionally, Sun knows that Sparc is dying, that's why they're handing over a lot of control to Fujitsu. Sun'll be concentrating on x86 in the next five years, and I believe will start to make inroads in the commercial server market.
I've even seen Slurp get stuck on some of my pages (www.math.purdue.edu) and continusouly request the same page. There was a month or two when Slurp was generating a serious amount of our traffic.
I'm the pleased owner of a dual G5 powerpc and a g4 powerbook, but if apple does advance intel's drm, I'll bail ship and swtich fulltime to ... Solaris 10 and ubuntu?
One cannot easily block google ads because they're simple html tables and text. One the first five things I do with a computer I'm using is (after installing Thunderbird, Firefox and web developer extension ) install Adblock and find an ad from ad.doubleclick.net to block all images from their servers.
My point is that I don't go to the trouble to block text ads, but I get an obstinate pleasure out of blocking banners and flash ads.
Now perhaps google really is evil, but merely aware of this attitude, or perhaps google really does try to not be evil.
LOL I voted for Kerry just to get bush out of office.
Of course you're right that the USA PATRIOT Act will be used against everyone. Already the term "terrorism" has less specific meaning. Am I terrorizing the 7-11 clerk, when I rob the store? Yes? Bam, Patriot Act.
Where you're wrong is assuming that the US will stop chasing people around the globe. The "War on Terrorism" (like the "War on Drugs", neither of which are wars) will never end. All we can hope for is a government that ackowledges this and doesn't try to change the nature of American citizenship because of it.
I'm trying to explore some templating options for a smaller web site within a university's site. The setup is solaris, apache 1.3, and postgresql. Currently, the site is running php4 using smarty for templating.
My initial reaction to smarty is: wth? we're simply adding a layer of complexity (and another layer of computation) by putting a different programming language in the html page.
It seems to me that templating systems often end up arbitrarily complicated; hence the desire for a more standard templating structure.
I would like to ask what other professional web masters/developers/designers use to accomplish templating.
How exactly would (x)html templating work?