Someone has to say it: $150.00 won't buy much programming time. Even if it was possible to spread the cost over two years of cost savings ($150 x 24 months = $2,400 total), that's still not a lot of programmer time--maybe 30-40 hours at most. And that would have to include design, planning, programming, testing, etc., etc.. That's one short week of a programmer's time to try and implement the change, and think of the number of browsers that would need to be tested. I propose that maybe a full XHTML facelift for/. is not the best place to invest time, energy, and resources. Now editorial training, spell (and grammar) checking, better dupe-checking... that's where some serious improvements can be made!
Of course, a full XHTML 2 version of/., with easily replaceable and editable CSS would be truly sweet. Slash code is Perl, and there are great Perl XML and XHTML tools widely available. Maybe someone will volunteer to do the work. Hmmmm... to bad I've got to work this weekend.
The problem with allowing editing of posts before they're moderated or replied to is there's no way of knowing when someone's actually started to reply. That would require the server to lock the post as soon as someone hit the "Reply to this" link, until they either submit or cancel (timeout, whatever) and the reply/moderation is saved. Since it can take a few minutes to type a reply, and several more if the preview option is used, this would impose huge locking overhead. Then there's the problem of making an appropriate choice of a timeout value: If it's set too short, then slow (or long-winded) typists would be penalized by having their lock removed. Too long and it would tie up the comment in a locked state even when the poster decides to just browse somewhere else, disconnect, or otherwise just not finish their reply.
I guess it would be possible to set a flag when someone hits either the reply or moderate links and check the flag before allowing someone to edit their comment. This would eliminate the database lock overhead, but it would still add a good bit of code devoted to tracking the status of the flag: Is it set? Has the comment/moderation that set the flag in the first place actually been saved? How long has it been since the flag was set? What if someone clicks a comment link, sets the flag, and takes all afternoon to finish her comment and then submit it? If the flag has timed out in the meantime, should Slash discard the comment? Inform the user that the comment has changed? And this doesn't address the issue of validating which users can edit which comments (though that would probably be simple, using the existing user validations). With thousands of users reading, commenting, moderating, and replying, I think adding the option to edit comments as well would add huge overhead and slow everything else down.
I guess I would go along with the suggestion to add a little "P.S." box to a post, but even then it would be possible to change the post enough to invalidate the contents of any moderations or comments.
All that aside, I'm glad/. is up and running again, and it may be crazy wishful thinking, but it seems more responsive and seems to render better in Firefox as well. Woo hoo!
Wow, since last summer Microsoft has been sending me updates directly in email. That saves me a lot of time! And Yahoo, and ebay, and even my bank have jumped on the bandwagon too--emailing me with updates (Of course, they're all spyware, adware, viruses, and obvious attempts to hijack my computer and/or steal my money...).
I too only load IE to check updates and run House Call. The regular, major updates seem to come through the Windows Update daemon (urrrm... I mean "service") fine, but I still check for driver updates and the like with IE, and then switch back to Firefox for actual browsing. The hardest part is finding the IE icon since I took it off my desktop and start menu.
This law isn't about punishing American malefactors. I would bet the problem of pirating films with vidcams isn't even serious in the USA. This law was passed so that we can show our international trading partners (East Asia, among others) that we're serious about intellectual property and hence so should they be. The USA has to create and impose some way of protecting intellectual property, because nowadays, it's among our major exports.
The fact that it's easy enough for anyone with a CD-R to "produce" the same end-user product (whether it be a copy of Win XP or a copy of Dogma) means we have to resort to draconian measures to enforce ownership of the product. This means doing things like changing the word from "pirating" to "stealing". And, unfortunately, it also means people (who own IP rights) think we need to pass laws like this, that force courts to mete out punishments that are out of all proportion to the crime. If we don't have these laws, how can we expect Indonesia, Hong Kong, China, or any other country to enforce similar laws.
Watch for more of this. IP as a major export is a bit of a paradigm shift, and we're likely to see a lot of crazy and wrong-headed laws before figuring out how to actually make it work. And note that the owners of IP are also among "the wealthy". This type of law is passed to protect them.
There's so much wrong with your post, I won't even bother to address it point-by-point. I will say that just because some bad action has the potential to cause problems for the authorities / corporations / individuals, and this kind of disorder causes you or your clueless friends to be terrified does not make it terrorism (though it may define you as hopelessly neurotic). This distinction (is it a crime or is it terrorism?) has been co-opted by the current American justice department so they can use the "streamlined" definition of due-process supplied by the Patriot Act in the prosecution of "terrorists". You and your friends should calm down, take some deep breaths, and change the channel when the news starts talking about "terrorism." Consider installing and learning a secure operating system on your computer.
And your subject, "ambigous term" (read it again, slowly) reminds me of the Roman in Monty Python's Life of Brian, named Bigous Dickus. Go ahead and laugh. It's just a typo, not terrorism.
You're right, describe doesn't work. I'm too used to Oracle. (I could have sworn I typed it into psql before I posted it, but alas, I'm wrong). \d table_namedoes work as well. It's also possible to get the information by querying the system catalog tables, though in the last 10 minutes, I've proven to myself that your complaint is valid. This is all less than obvious to a newcomer, and finding it in the documentation is less than trivial.
If you've got access to a web server that has PHP, it's well worth it to install phpPgAdmin, which makes all this incredibly simple. PgAccess also works well, and there's a windows version as well as a linux version, and it doesn't require a webserver or PHP.
The thing that makes PostgreSQL great is every time you define a view, you never have to write it again in code! Just query the view. And every time you define a function that returns a value, you can use it in any subsequent query. And every trigger you define is there whenever you insert or update or delete a record, not just when you remember to do it correctly in your code. MySQL is great to use when you're slapping out the code to get the job done. PostgreSQL is great when you've forgotten exactly how everything works (for me, that's a couple of weeks at most) and you've got to make a change or add some functionality. Keep trying, it's not easy to learn, but it's worth the trouble!
And finally, your point about "update" is well taken. I use that feature in ADODB all the time, and it saves writing a lot of code.
Answer to third question: If you're in psql interpreter, and connected to the database, type \dt and voila, a list of the tables. There are various ways to get the same list by querying the system tables, but I never learned them (I admit to using phpPgAdmin which is superb for administering PostgreSQL databases, and which makes those kinds of things easy).
Answer to fourth question: in psql type describe your_tablename and it will show you the structure. Many of the commands in psql are the same as in SQL*Plus in Oracle, and it wouldn't be a bad idea to get a nutshell guide to that for this kind of problem.
I agree, views are great, but wait until the first time you create an inline function... Mmmmm... true database goodness.
And I don't know where you get the idea that replace is any better than update... set... where... functionality. I guess I just don't think in "replace" anymore (FoxPro has it, and I used to use it all the time).
Damn, that's a beautiful post! I'm in tears laughing. Bravo. Not only have you tossed an asbestos blanket over the smoldering flame war, but did it with some humor.
"Online", in this case, is intentionally vague, and means "deep pockets". I'm willing to bet (a lot) there's a lawyer involved in this, who will do his damndest to haul into court the NY Housing Authority, or the security company, or the NYPD or the manufacturer of the surveillance equipment or the property owner or the website owner or HUD or anyone else who might have both some implied control over the property or the tapes and an insurance policy that will cover the liability.
And in the name of protecting the privacy of the victim (who needs no protection, and probably didn't give a rat's ass about it at the time he shot himself) the lawyer will show the tape over and over (traumatizing the family, who I doubt needed to see the tape) to a judge and/or jury to convince them that somehow a grave wrong has been perpetrated by someone somewhere (who would show the tape to anyone else, or email it, or publish it on a website, or who didn't stop the act in the first place). If the lawyer does his job and finds someone with both responsibility and the means to pay, some court will direct that funds be moved to the lawyer.... errr, I mean to the aggrieved party.
This is why there are insurance companies, why there are lawyers, and why there are civil courts. This isn't a story about suicide, privacy, race, youth, or public housing. It's economics. For every kid who blows his brains out on camera there are a whole bunch who do it in private. Damn shame. When it makes the news it's because some lawyer is making a move to collect money--most likely on contingency. This is a disgusting story, and the outcome is guaranteed to be just as disgusting.
200K Down, 33K up (it varies all over the place, but basically gets faster the longer I stay on). I've gotten about 1.2 gig in about 2.5 hours. Not too shabby.
The same thing started happening to me in the last couple months. My provider checked, and they're not hijacking the use of the mail server, they're just forging headers so it appears the mail is coming from my domain (They all pretend to be from some username@mydomain.com). There isn't any obvious solution. I've gotten about a hundred "bounces" in three months. I don't think this is the criteria AOL uses to block an email server, otherwise I assume I would stop getting the bounce messages. I can and do send emails to people on AOL.
I hate having my domain name associated with V*I*A*G*R*A (the bounced messages are almost all online medication related, though with forged headers, I assume they're all just scams to get people to send their credit card numbers), but I doubt anyone receiving those messages pays much attention to the originating mail domain, even a forged one. I hate to say it, but I think email is broken beyond repair, and someone needs to create a replacement system.
The PostgreSQL vs MySQL debate always seems to hinge on features, but rarely on adequacy to the task at hand. A good programmer can knock together a data-centric application that has great data integrity and works as required using.TXT files! When it comes to "enterprise" class, though, the implication is that there is a division of labor between the front-end program(s) that is the "source" of the data and the database itself.
In the first case, where the data is completely managed by one front-end (viz many of the small to medium web applications out there), what's needed is a small, fast, simple database backend that will store and retrieve data quickly and reliably. The programmer (or small team) controls the tricky issues, like integrity, consistency, concurrency, etc., all in the program.
In the second case, where there's a different person/team responsible for the database per se, or more than one set of programs uses the data, then the advanced features like triggers, views, stored procedures, sub-selects, become more important. DBA's don't (many can't) read code. A team of programmers writing a new front-end (or a totally new application) that accesses the data can't be expected to replicate all of the integrity/validation/transaction processing logic from another application into the new system. They can use advanced database features (like triggers, foreign key integrity constraints, etc.) to apply rules and consistency checks outside of the front end programs that are actually collecting/manipulating the data. This ability to control the data independently of any front end program is a great feature. It is, however, completely overkill for many day-to-day tasks.
Take it from someone whose been writing data-centric front-ends for "databases" from dBase II to Oracle every day for the last 15+ years. It doesn't take a huge database server to make an effective application. And not every application requires the ability of a DBA to enforce rules from outside of the application itself. Maybe a simple rule of thumb is "the more people working on the data, the more complex the DBMS that's required." PostgreSQL is a great product, but if it's used without the stored procedures (well, functions), integrity constraints, and triggers, I don't see it as any improvement over MySQL (which I believe, on strictly limited and anecdotal evidence, to be slightly faster).
I think he meant dBASE2, the long forgotten ancestor of Foxpro. In fact he probably meant db3, which was a mysql-like database-like PC application with built-in programming language and reasonably fast access using flat files and indexes.
Otherwise, I agree with your post. I work in an Oracle shop, and though it's a great product to work with--gobs of support software and 3rd party support--it's pricing is ridiculous. And DB2 seems to be getting more interesting with cross-platform support and that huge IBM support behemoth behind it.
The solution is simple. The editors just have to set the article's evil bit "ON" whenever they post an article that can inspire bad behavior. Then people can set filters to keep that crap out of their browsers. Maybe/. could even change the SlashCode so that moderators could toggle on the evil bit when the editors couldn't see the full implications of their stories at the time of original posting. No more worries. It's all good...
If MS had tried this, they would likely have tested it with the LG drive/driver, found the problem, and then decided, based on the number of units in the field how to code around the problem. I think this happens more than anyone is aware. I remember reading about a Linux installation problem several years ago where someone upgrading from Win95 couldn't install Linux because it wouldn't detect his drives correctly. Eventually it was discovered that his IDE hard drives weren't configured correctly--either both were masters or both were slaves. Apparently MS had seen the problem so often that they wrote their code/drivers in such a way that it didn't matter how the drives were configured. The configuration problem would only show up when trying to install a different OS.
It is probably well within the skills of the MandrakeSoft programmers to code around this particular flaw. Unfortunately, MandrakeSoft can't afford to test on every possible configuration and hardware/firmware revision combination. Too bad... I blame LG in this case (for not adhering to standards, the only alternative to testing every combination), but there's nothing like thorough testing to catch that kind of problem. Too bad testing like that is not FREE! As in beer!
The fact that so many spammers don't have to invest in any actual product has a positive effect on their bottom line as well. That's why they're so evasive about their address/phone/contact information. Most, if not all, of the 100+ spam emails I receive every day are obviously completely fraudulent. I find it amazing that anyone actually responds to any of it.
The spammers are outlaws, but it would be good if the few actual identifiable vendors who profit from Spam could have their feet held to the fire. They can't ALL be overseas. For the rest, I say block them all if that's the only way.
As for the article, I'm afraid we've/.'ed the poor website and I won't get to read it for another 12 hours or so... *sigh*
From the article: Several MacFixIt readers have reported an inability to wake from sleep since installing the 10.2.8 update.
Odd, that was my reaction as well. And I don't even use a Mac! The number of affected users (low) and the complexity of the fix (plug into a 100base-T socket instead of a 10base-T) indicate not much to worry about. These could be huge problems for average home users (network? what network?), but how many of them are downloading OS updates anyway?
As a vereran of "dll hell" and several (admittedly early) problems with failed RPM's, I see this Apple problem as very minor. The "evil twist" in this story is that when the problem becomes apparent, it's not possible to connect to the internet to find a diagnosis / solution. I can almost hear all those users growling "ohhh shit!"
In a related note, I've noticed that OS vendors and ISV's are recently (in the last couple of years) referring a lot more to some mysterious "backup" which, while it reflects a very good practice, is not necessarily practiced universally. The vendors used to try and work around the problem ("here's how to back up what you'll need to recover in case this install fails"). That didn't serve them well, so referencing a "backup" without going into the minutiae of how to actually get one is really good for them to do. It makes the installation instructions much shorter, and face it--anyone who would uprade their OS without a backup doesn't really have anything important on their computer anyhow.
This is probably offtopic, but I suggest you tell the management-type "folks above you" that "have fundamental issues" with users saving their data onto servers instead of locally that the data belongs to the company! It's not a complicated idea. Simple as that. Anyone who saves anything but temp data to their local drive is risking the company's data, and its value.
Personally, I hate the whole exercise, but if the data has value, it belongs to the company, and if it's stored on a local hard drive it's vulnerable to local failure.
Also, my experience is that the best backup solutions, from the point of view of speed, cost effectiveness, and completeness, are invariably based on the server. It's not that hard to teach users to save their work to the network instead of to their local hard drive. Part of the reason corporations like thin clients is because of this.
The Blaster fix didn't require reboot, at least on my machine. Of course, there's still a hitch, because it's requested that I install it at least 4 times. But so far so good. No Blaster and no reboots.
Right on. My experience was the same. I was immunized from BLASTER on July 17th according to the log from MS Update. It's very hip and au courant to ignore MS Updates, because they're a pain, and their Service Packs don't have a great reputation. But updating early and often has kept me out of trouble.
When I started getting Sobig emails on Tuesday, I even took the time to call two of my friends (who subscribe to some of the same lists I do) to warn them not to trust emails with attachments. I had to explain the whole concept to them, but they got it. I got 40 the first day, 20 the second and only a handful since. And I had no desire to open any of them.
The biggest threat that Windows poses is that from users who are totally clueless... they turn on their machine thinking it's some kind of "email machine" and nothing else. Not a clue there are threats or risks out there. And no indication from Windows, or Outlook, or IE that anything they do could be unsafe. Windows update works, at least this time it did. They're not going to get more saavy, so there's no harm in telling people to use windows update.
Tell your friends:
1. Don't preview email
2. Delete email you don't know or trust
3. Don't open attachments if they're not absolutely known and expected
3. Update early and often
The article is right, Windows is dangerous. MS isn't going to tell the consumer, because that would threaten their (considerable) cash flow.
If it takes a week or two for the DBA to learn the ropes of the new system you've easily paid for MS SQL.
Last time I checked, an internet version of MSSQL cost more than a couple of weeks of the DBA's time. I could be wrong about this, because it's been awhile since I looked a prices, but I thought it was about $20K plus annual maintenance/support. Oracle costs even more.
It seems like the issue in scaling is to break up the software that's running (DB, Web Sever, Transaction Processing, User Application, Backup, whatever) into their own separate boxes as much as possible, and tune them separately for their tasks. Segregating the applications will allow identifying and rectifying problems and bottlenecks within each separate process. Throwing RAM and disks at the DB will probably have a great effect, though probably not so for IIS, so why keep them together on the same box. Gigabit Ethernet will likely enhance communication between IIS and MSSQL, but most likely be useless enhancing IIS throughput if you're hooked to the internet at 1Mbs. I know, more boxes mean more MS Server licenses, but at least then you can identify and fix bottlenecks where they happen, rather than just, e.g.add more servers to a cluster. And heck, that requires more MS licenses as well.
Nope. The problem is they (AOL and now AOL/TW) have been bankrolling the development of Mozilla for years and now they've got a world class product (Netscape 7.1, based on the Mozilla engine) that they can't really sell, because it's free. That and the fact that because AOL/TW is going through some "lean" times, everyone in the corp has to cost justify the big expenditures. The Mozilla development team has got to be a big expenditure (see the posts above), and the Netscape product doesn't have a great sales future. Sadly, it's the way corporations are funded. I doubt it's much better in the movie or magazine branches of the business.
The 2 million bucks they paid toward creating the Foundation is really chicken-feed in the overall AOL/TW budget, and it's a way for them to cast off the product line without killing it outright. It's more generous than they had to be, but it's clear the Netscape browser isn't strategically important to the corporation in any way. It's probably inflammatory to say it this way, but one of the problems with "free software" is its price. I'm hoping it doesn't mean the product's demise. I'm starting to love Firebird, and hoping that pop-up free, tabbed browsing will become the internet platform of the future. My thanks and wish for good luck to all the Mozilla developers.
Slashdot Mirror
Someone has to say it: $150.00 won't buy much programming time. Even if it was possible to spread the cost over two years of cost savings ($150 x 24 months = $2,400 total), that's still not a lot of programmer time--maybe 30-40 hours at most. And that would have to include design, planning, programming, testing, etc., etc.. That's one short week of a programmer's time to try and implement the change, and think of the number of browsers that would need to be tested. I propose that maybe a full XHTML facelift for /. is not the best place to invest time, energy, and resources. Now editorial training, spell (and grammar) checking, better dupe-checking... that's where some serious improvements can be made!
Of course, a full XHTML 2 version of /., with easily replaceable and editable CSS would be truly sweet. Slash code is Perl, and there are great Perl XML and XHTML tools widely available. Maybe someone will volunteer to do the work. Hmmmm... to bad I've got to work this weekend.
All I can say is thank God it wasn't the other way around!
The problem with allowing editing of posts before they're moderated or replied to is there's no way of knowing when someone's actually started to reply. That would require the server to lock the post as soon as someone hit the "Reply to this" link, until they either submit or cancel (timeout, whatever) and the reply/moderation is saved. Since it can take a few minutes to type a reply, and several more if the preview option is used, this would impose huge locking overhead. Then there's the problem of making an appropriate choice of a timeout value: If it's set too short, then slow (or long-winded) typists would be penalized by having their lock removed. Too long and it would tie up the comment in a locked state even when the poster decides to just browse somewhere else, disconnect, or otherwise just not finish their reply.
I guess it would be possible to set a flag when someone hits either the reply or moderate links and check the flag before allowing someone to edit their comment. This would eliminate the database lock overhead, but it would still add a good bit of code devoted to tracking the status of the flag: Is it set? Has the comment/moderation that set the flag in the first place actually been saved? How long has it been since the flag was set? What if someone clicks a comment link, sets the flag, and takes all afternoon to finish her comment and then submit it? If the flag has timed out in the meantime, should Slash discard the comment? Inform the user that the comment has changed? And this doesn't address the issue of validating which users can edit which comments (though that would probably be simple, using the existing user validations). With thousands of users reading, commenting, moderating, and replying, I think adding the option to edit comments as well would add huge overhead and slow everything else down.
I guess I would go along with the suggestion to add a little "P.S." box to a post, but even then it would be possible to change the post enough to invalidate the contents of any moderations or comments.
All that aside, I'm glad /. is up and running again, and it may be crazy wishful thinking, but it seems more responsive and seems to render better in Firefox as well. Woo hoo!
Wow, since last summer Microsoft has been sending me updates directly in email. That saves me a lot of time! And Yahoo, and ebay, and even my bank have jumped on the bandwagon too--emailing me with updates (Of course, they're all spyware, adware, viruses, and obvious attempts to hijack my computer and/or steal my money...). I too only load IE to check updates and run House Call. The regular, major updates seem to come through the Windows Update daemon (urrrm... I mean "service") fine, but I still check for driver updates and the like with IE, and then switch back to Firefox for actual browsing. The hardest part is finding the IE icon since I took it off my desktop and start menu.
This law isn't about punishing American malefactors. I would bet the problem of pirating films with vidcams isn't even serious in the USA. This law was passed so that we can show our international trading partners (East Asia, among others) that we're serious about intellectual property and hence so should they be. The USA has to create and impose some way of protecting intellectual property, because nowadays, it's among our major exports.
The fact that it's easy enough for anyone with a CD-R to "produce" the same end-user product (whether it be a copy of Win XP or a copy of Dogma) means we have to resort to draconian measures to enforce ownership of the product. This means doing things like changing the word from "pirating" to "stealing". And, unfortunately, it also means people (who own IP rights) think we need to pass laws like this, that force courts to mete out punishments that are out of all proportion to the crime. If we don't have these laws, how can we expect Indonesia, Hong Kong, China, or any other country to enforce similar laws.
Watch for more of this. IP as a major export is a bit of a paradigm shift, and we're likely to see a lot of crazy and wrong-headed laws before figuring out how to actually make it work. And note that the owners of IP are also among "the wealthy". This type of law is passed to protect them.
There's so much wrong with your post, I won't even bother to address it point-by-point. I will say that just because some bad action has the potential to cause problems for the authorities / corporations / individuals, and this kind of disorder causes you or your clueless friends to be terrified does not make it terrorism (though it may define you as hopelessly neurotic). This distinction (is it a crime or is it terrorism?) has been co-opted by the current American justice department so they can use the "streamlined" definition of due-process supplied by the Patriot Act in the prosecution of "terrorists". You and your friends should calm down, take some deep breaths, and change the channel when the news starts talking about "terrorism." Consider installing and learning a secure operating system on your computer.
And your subject, "ambigous term" (read it again, slowly) reminds me of the Roman in Monty Python's Life of Brian, named Bigous Dickus. Go ahead and laugh. It's just a typo, not terrorism.
You're right, describe doesn't work. I'm too used to Oracle. (I could have sworn I typed it into psql before I posted it, but alas, I'm wrong). \d table_name does work as well. It's also possible to get the information by querying the system catalog tables, though in the last 10 minutes, I've proven to myself that your complaint is valid. This is all less than obvious to a newcomer, and finding it in the documentation is less than trivial.
If you've got access to a web server that has PHP, it's well worth it to install phpPgAdmin, which makes all this incredibly simple. PgAccess also works well, and there's a windows version as well as a linux version, and it doesn't require a webserver or PHP.
The thing that makes PostgreSQL great is every time you define a view, you never have to write it again in code! Just query the view. And every time you define a function that returns a value, you can use it in any subsequent query. And every trigger you define is there whenever you insert or update or delete a record, not just when you remember to do it correctly in your code. MySQL is great to use when you're slapping out the code to get the job done. PostgreSQL is great when you've forgotten exactly how everything works (for me, that's a couple of weeks at most) and you've got to make a change or add some functionality. Keep trying, it's not easy to learn, but it's worth the trouble!
And finally, your point about "update" is well taken. I use that feature in ADODB all the time, and it saves writing a lot of code.
Answer to third question: If you're in psql interpreter, and connected to the database, type \dt and voila, a list of the tables. There are various ways to get the same list by querying the system tables, but I never learned them (I admit to using phpPgAdmin which is superb for administering PostgreSQL databases, and which makes those kinds of things easy).
Answer to fourth question: in psql type describe your_tablename and it will show you the structure. Many of the commands in psql are the same as in SQL*Plus in Oracle, and it wouldn't be a bad idea to get a nutshell guide to that for this kind of problem.
I agree, views are great, but wait until the first time you create an inline function... Mmmmm... true database goodness.
And I don't know where you get the idea that replace is any better than update... set... where... functionality. I guess I just don't think in "replace" anymore (FoxPro has it, and I used to use it all the time).
Damn, that's a beautiful post! I'm in tears laughing. Bravo. Not only have you tossed an asbestos blanket over the smoldering flame war, but did it with some humor.
Thanks."Online", in this case, is intentionally vague, and means "deep pockets". I'm willing to bet (a lot) there's a lawyer involved in this, who will do his damndest to haul into court the NY Housing Authority, or the security company, or the NYPD or the manufacturer of the surveillance equipment or the property owner or the website owner or HUD or anyone else who might have both some implied control over the property or the tapes and an insurance policy that will cover the liability.
And in the name of protecting the privacy of the victim (who needs no protection, and probably didn't give a rat's ass about it at the time he shot himself) the lawyer will show the tape over and over (traumatizing the family, who I doubt needed to see the tape) to a judge and/or jury to convince them that somehow a grave wrong has been perpetrated by someone somewhere (who would show the tape to anyone else, or email it, or publish it on a website, or who didn't stop the act in the first place). If the lawyer does his job and finds someone with both responsibility and the means to pay, some court will direct that funds be moved to the lawyer.... errr, I mean to the aggrieved party.
This is why there are insurance companies, why there are lawyers, and why there are civil courts. This isn't a story about suicide, privacy, race, youth, or public housing. It's economics. For every kid who blows his brains out on camera there are a whole bunch who do it in private. Damn shame. When it makes the news it's because some lawyer is making a move to collect money--most likely on contingency. This is a disgusting story, and the outcome is guaranteed to be just as disgusting.
200K Down, 33K up (it varies all over the place, but basically gets faster the longer I stay on). I've gotten about 1.2 gig in about 2.5 hours. Not too shabby.
The same thing started happening to me in the last couple months. My provider checked, and they're not hijacking the use of the mail server, they're just forging headers so it appears the mail is coming from my domain (They all pretend to be from some username@mydomain.com). There isn't any obvious solution. I've gotten about a hundred "bounces" in three months. I don't think this is the criteria AOL uses to block an email server, otherwise I assume I would stop getting the bounce messages. I can and do send emails to people on AOL.
I hate having my domain name associated with V*I*A*G*R*A (the bounced messages are almost all online medication related, though with forged headers, I assume they're all just scams to get people to send their credit card numbers), but I doubt anyone receiving those messages pays much attention to the originating mail domain, even a forged one. I hate to say it, but I think email is broken beyond repair, and someone needs to create a replacement system.
The PostgreSQL vs MySQL debate always seems to hinge on features, but rarely on adequacy to the task at hand. A good programmer can knock together a data-centric application that has great data integrity and works as required using .TXT files! When it comes to "enterprise" class, though, the implication is that there is a division of labor between the front-end program(s) that is the "source" of the data and the database itself.
In the first case, where the data is completely managed by one front-end (viz many of the small to medium web applications out there), what's needed is a small, fast, simple database backend that will store and retrieve data quickly and reliably. The programmer (or small team) controls the tricky issues, like integrity, consistency, concurrency, etc., all in the program.
In the second case, where there's a different person/team responsible for the database per se, or more than one set of programs uses the data, then the advanced features like triggers, views, stored procedures, sub-selects, become more important. DBA's don't (many can't) read code. A team of programmers writing a new front-end (or a totally new application) that accesses the data can't be expected to replicate all of the integrity/validation/transaction processing logic from another application into the new system. They can use advanced database features (like triggers, foreign key integrity constraints, etc.) to apply rules and consistency checks outside of the front end programs that are actually collecting/manipulating the data. This ability to control the data independently of any front end program is a great feature. It is, however, completely overkill for many day-to-day tasks.
Take it from someone whose been writing data-centric front-ends for "databases" from dBase II to Oracle every day for the last 15+ years. It doesn't take a huge database server to make an effective application. And not every application requires the ability of a DBA to enforce rules from outside of the application itself. Maybe a simple rule of thumb is "the more people working on the data, the more complex the DBMS that's required." PostgreSQL is a great product, but if it's used without the stored procedures (well, functions), integrity constraints, and triggers, I don't see it as any improvement over MySQL (which I believe, on strictly limited and anecdotal evidence, to be slightly faster).
I think he meant dBASE2, the long forgotten ancestor of Foxpro. In fact he probably meant db3, which was a mysql-like database-like PC application with built-in programming language and reasonably fast access using flat files and indexes.
Otherwise, I agree with your post. I work in an Oracle shop, and though it's a great product to work with--gobs of support software and 3rd party support--it's pricing is ridiculous. And DB2 seems to be getting more interesting with cross-platform support and that huge IBM support behemoth behind it.
That was beautiful. Thanks.
The solution is simple. The editors just have to set the article's evil bit "ON" whenever they post an article that can inspire bad behavior. Then people can set filters to keep that crap out of their browsers. Maybe /. could even change the SlashCode so that moderators could toggle on the evil bit when the editors couldn't see the full implications of their stories at the time of original posting. No more worries. It's all good...
If MS had tried this, they would likely have tested it with the LG drive/driver, found the problem, and then decided, based on the number of units in the field how to code around the problem. I think this happens more than anyone is aware. I remember reading about a Linux installation problem several years ago where someone upgrading from Win95 couldn't install Linux because it wouldn't detect his drives correctly. Eventually it was discovered that his IDE hard drives weren't configured correctly--either both were masters or both were slaves. Apparently MS had seen the problem so often that they wrote their code/drivers in such a way that it didn't matter how the drives were configured. The configuration problem would only show up when trying to install a different OS.
It is probably well within the skills of the MandrakeSoft programmers to code around this particular flaw. Unfortunately, MandrakeSoft can't afford to test on every possible configuration and hardware/firmware revision combination. Too bad... I blame LG in this case (for not adhering to standards, the only alternative to testing every combination), but there's nothing like thorough testing to catch that kind of problem. Too bad testing like that is not FREE! As in beer!
The fact that so many spammers don't have to invest in any actual product has a positive effect on their bottom line as well. That's why they're so evasive about their address/phone/contact information. Most, if not all, of the 100+ spam emails I receive every day are obviously completely fraudulent. I find it amazing that anyone actually responds to any of it.
/.'ed the poor website and I won't get to read it for another 12 hours or so... *sigh*
The spammers are outlaws, but it would be good if the few actual identifiable vendors who profit from Spam could have their feet held to the fire. They can't ALL be overseas. For the rest, I say block them all if that's the only way.
As for the article, I'm afraid we've
From the article: Several MacFixIt readers have reported an inability to wake from sleep since installing the 10.2.8 update.
Odd, that was my reaction as well. And I don't even use a Mac! The number of affected users (low) and the complexity of the fix (plug into a 100base-T socket instead of a 10base-T) indicate not much to worry about. These could be huge problems for average home users (network? what network?), but how many of them are downloading OS updates anyway?
As a vereran of "dll hell" and several (admittedly early) problems with failed RPM's, I see this Apple problem as very minor. The "evil twist" in this story is that when the problem becomes apparent, it's not possible to connect to the internet to find a diagnosis / solution. I can almost hear all those users growling "ohhh shit!"
In a related note, I've noticed that OS vendors and ISV's are recently (in the last couple of years) referring a lot more to some mysterious "backup" which, while it reflects a very good practice, is not necessarily practiced universally. The vendors used to try and work around the problem ("here's how to back up what you'll need to recover in case this install fails"). That didn't serve them well, so referencing a "backup" without going into the minutiae of how to actually get one is really good for them to do. It makes the installation instructions much shorter, and face it--anyone who would uprade their OS without a backup doesn't really have anything important on their computer anyhow.
This is probably offtopic, but I suggest you tell the management-type "folks above you" that "have fundamental issues" with users saving their data onto servers instead of locally that the data belongs to the company! It's not a complicated idea. Simple as that. Anyone who saves anything but temp data to their local drive is risking the company's data, and its value.
Personally, I hate the whole exercise, but if the data has value, it belongs to the company, and if it's stored on a local hard drive it's vulnerable to local failure.
Also, my experience is that the best backup solutions, from the point of view of speed, cost effectiveness, and completeness, are invariably based on the server. It's not that hard to teach users to save their work to the network instead of to their local hard drive. Part of the reason corporations like thin clients is because of this.
The Blaster fix didn't require reboot, at least on my machine. Of course, there's still a hitch, because it's requested that I install it at least 4 times. But so far so good. No Blaster and no reboots.
Right on. My experience was the same. I was immunized from BLASTER on July 17th according to the log from MS Update. It's very hip and au courant to ignore MS Updates, because they're a pain, and their Service Packs don't have a great reputation. But updating early and often has kept me out of trouble.
When I started getting Sobig emails on Tuesday, I even took the time to call two of my friends (who subscribe to some of the same lists I do) to warn them not to trust emails with attachments. I had to explain the whole concept to them, but they got it. I got 40 the first day, 20 the second and only a handful since. And I had no desire to open any of them.
The biggest threat that Windows poses is that from users who are totally clueless... they turn on their machine thinking it's some kind of "email machine" and nothing else. Not a clue there are threats or risks out there. And no indication from Windows, or Outlook, or IE that anything they do could be unsafe. Windows update works, at least this time it did. They're not going to get more saavy, so there's no harm in telling people to use windows update.
Tell your friends:
1. Don't preview email
2. Delete email you don't know or trust
3. Don't open attachments if they're not absolutely known and expected
3. Update early and often
The article is right, Windows is dangerous. MS isn't going to tell the consumer, because that would threaten their (considerable) cash flow.
I'll shut up now.
If it takes a week or two for the DBA to learn the ropes of the new system you've easily paid for MS SQL.
Last time I checked, an internet version of MSSQL cost more than a couple of weeks of the DBA's time. I could be wrong about this, because it's been awhile since I looked a prices, but I thought it was about $20K plus annual maintenance/support. Oracle costs even more.
It seems like the issue in scaling is to break up the software that's running (DB, Web Sever, Transaction Processing, User Application, Backup, whatever) into their own separate boxes as much as possible, and tune them separately for their tasks. Segregating the applications will allow identifying and rectifying problems and bottlenecks within each separate process. Throwing RAM and disks at the DB will probably have a great effect, though probably not so for IIS, so why keep them together on the same box. Gigabit Ethernet will likely enhance communication between IIS and MSSQL, but most likely be useless enhancing IIS throughput if you're hooked to the internet at 1Mbs. I know, more boxes mean more MS Server licenses, but at least then you can identify and fix bottlenecks where they happen, rather than just, e.g.add more servers to a cluster. And heck, that requires more MS licenses as well.
Nope. The problem is they (AOL and now AOL/TW) have been bankrolling the development of Mozilla for years and now they've got a world class product (Netscape 7.1, based on the Mozilla engine) that they can't really sell, because it's free. That and the fact that because AOL/TW is going through some "lean" times, everyone in the corp has to cost justify the big expenditures. The Mozilla development team has got to be a big expenditure (see the posts above), and the Netscape product doesn't have a great sales future. Sadly, it's the way corporations are funded. I doubt it's much better in the movie or magazine branches of the business.
The 2 million bucks they paid toward creating the Foundation is really chicken-feed in the overall AOL/TW budget, and it's a way for them to cast off the product line without killing it outright. It's more generous than they had to be, but it's clear the Netscape browser isn't strategically important to the corporation in any way. It's probably inflammatory to say it this way, but one of the problems with "free software" is its price. I'm hoping it doesn't mean the product's demise. I'm starting to love Firebird, and hoping that pop-up free, tabbed browsing will become the internet platform of the future. My thanks and wish for good luck to all the Mozilla developers.
Beautiful. Brilliant. I bow before your awesome trolliness!