How can I be more specific than he was just wrong? The comments about compilers were wrong, as were the API, developer support, dev tools, debuggers, developer competence, sample code, and so on and so on.
"However, you do see them moving off IIS and onto Apache."
Well actually no you don't, which is reinforced by the Netcraft monthly web survey comments. The numbers of Apache sites reflect massive webhosting firms, but for people and companies who do their own dynamic web development there has been no move away from IIS. Apache is an easy migration if all you have is static content, but then if that's all you have who cares what server you use?
MS has also been increasing the value proposition with it's server pieces, especially now with the.Net development environment. It will be difficult for Linux to address this momentum as the community tends to embrace old standbys(like perl and now PHP) instead of innovating in new directions. Furthermore the recent IDC survey even shows in 2001 Linux growth was stagnant on the server side, while Microsoft had a 7% marketshare gain primarily related to customers finally understanding and adopting Windows 2000.
Ok, enough wasting my time here. I'm just always amazed how out of touch the Linux zealots are with the industry. If you think you want to beat Microsoft you have to first start being honest about where you are currently.
The fact that/. won't run an article from yesterday which discusses Linux server marketshare in unfavorable terms is more than proof enough of exactly who is "running scared."
Yes, open standards do make a case for having an open sourced reference implementation to use as a basis. You see this all the time.
But you'll also note that these open sourced implementations are not licensed under the GPL. The prime example for this is the gzip library used for HTTP 1.1 compression, they had to come up with a non-GPLed implementation specifically because of it's use as part of a standard.
I see a lot of intellectual dishonesty on the part of the GPL bigots. They frequently reference non-GPLed open source software as success stories(Apache, BSD TCP/IP, BIND, sendmail, etc), and then imply by inference that this same success will also apply to the GPLed software.
Oh, just read your last sentence. I thought you were being serious in your reply, but I guess you were just being a troll. Sorry to have wasted my time responding to a child.
The point was that Roblimo was basically admitting that/. has frequent failures with the mySQL database server and recovers from them by rebooting the box.
It appears as though you attempted to over-analyze the article as a result of some pre-programmed knee-jerk reaction.
Reminds me of a comment Roblimo made last year...
on
IBM, MS Critique MySQL
·
· Score: 5, Insightful
Back about a year ago/. experienced a major outage. Roblimo at the time noted, "By 7 a.m. it was obvious that this was not a typical, easily-fixed, reboot-the-database problem."
Can anybody imagine an SQL Server or DB/2 customer being satisfied with that solution? That's what IBM/MS is saying.
Re:At the risk of being modded redundant. . .
on
Passport vs. Plan 9
·
· Score: 2
Well, let's think about this. I actually think that the security is far more complicated than you make it out to be.
I have multiple banking, credit card and merchant accounts. I also have a whole lot of worthless accounts with places like slashdot.org, nytimes.com and so forth. It's these worthless accounts that I think are good solutions for single-signon passport type accounts. In fact this is pretty much what MS has done throughout their support websites.
As far as having multiple authorities being safer than a single one. Somewhat true, as long as you are using completely different usernames and passwords for each one of those authorities. Oh yeah, and don't write them down because that puts you at risk.
As far as credit cards, there are three well known authorities that track all of your purchases. They are named MasterCard, Visa and Discover... oh yeah and AmEx. AmEx even sends you a statement at the end of the year telling you everthing you bought.
But now how many different credit cards do you have? Hopefully you are using a different card for each merchant account you signed up for, otherwise once again you increase your risk because unfortunately these web merchants save the damn credit cards in their databases. For your convenience, of course... (weird how brick and mortar stores never do this, and it's also no coincidence they don't lose lists of thousands of valid credit card numbers either... hmm) So now we've established you've been spreading your credit card number around to dozens of websites, each one with probably questionable security.
I don't know what the answer is. I only believe Passport is a good idea for all of my less important accounts, for right now. But I also worry about my more important accounts, and I really worry about Amazon.com storing my credit card number for my convenience. It is a complicated thing, and I don't agree that saying a single signon system is dumb, because it ignores the fact that the status-quo, the way things are today is also very incredibly dumb.
We do need something better, but I'm not sure what that will be. I had some hopes of AmEx's smart card system, but they never really got that working.
"It's huge and Microsoft, in their infinite wisdom, rarely chooses to setup mirrors."
Huh? Microsoft has been using akamai for a couple years now, before that it was conxion. It's mirrored all over the world. I believe they're doing bandwidth limiting on the download in order to support more users at once, as all downloads I've seen of XPSP1 have been at exactly 10K.
For like $15 you can get a CD sent to you via Airborne Express.
The "Your current 3rd party browser" is there solely as a default. The third party product needs to rewrite their installation so that they register with the new API that controls these applications.
My understanding is that the latest version of Netscape 7 does register itself properly. Opera and the others have apparently not taken the time to create new install packages.
If you take the time to actually figure out how things work you'll find that your criticism is entirely unjustified.
It's weird. I've purchased a variety of drives over the years, and I've had a variety of drives in systems I've supported. All totalled I've probably purchased 30 drives for personal use, and for work we're talking thousands.
Western Digital is the only company I've never seen a failed drive from. The next best is Seagate, where the only failure I saw there was about 7 years ago with a 170Meg and a 500Meg. Now granted, I've seen failed Seagate drives in the servers but out of the many hundreds there, the failure ratio has still been reasonable.
Maxtor I've had numerous failures from, 340Meg, 3Gig, 13Gig and so forth. I tend to avoid their product any more because the failures weren't a while ago, they just keep happening...
IBM I only had one failed 2 gig drive, years ago. I have a couple 20 and 30 gig from them now, but I otherwise haven't seen too much of their product.
Quantum was probably the worst, but that's because we had about 500 desktops with 2.5Gig bigfoot drives in them and I must have replaced half of them. Otherwise I had good luck with their older 500Meg and 1Gig Fireball drives.
Anyway, not quite sure why people bash on WD drives. Maybe you've just seen a higher volume of their drives than other brands. It would be like working in a Honda dealership and saying Honda's suck because they're always coming back in for repairs.:)
Universities began teaching Java back in '97/'98 timeframe long before it was in widespread usage. Part of this was because of a push in '98 by Sun to promote Java at the University level. Look it up, it's called the AAJC.
Every argument you could make pro-Java you can also make pro-C#. It is a well designed language, and a marketable skill for would-be graduates. The fact that you have sold out to Sun is irrelevant to this argument.
Even though DOS 6.0 was released like 10 years ago, it will still boot and run on the Pentium 4 computer I bought last month.
Wouldn't it be interesting if the hardware makers changed the specs such that new machines only ran Windows XP, and Microsoft simply claimed that they weren't doing this purposefully... they were just not going to update DOS and Windows 95 to boot on the new hardware because it is legacy code.
I must say Mac people have their head stuck in the sand far more deeply than any other advocacy group.
The Windows 2000 FTP daemon has been notoriously insecure, in contrast with open source products like MuddleFTPd and ProFTPd. Why? Because the coders who wrote this security-critical part of the system just didn't care.
Unfortunately the existence of WuFTPD pretty much invalidates your entire argument.
I'm also curious about Win2k ftp being insecure. I can only remember one exploit, do you have a list of other exploits? There have certainly been exploits of Proftpd as well, so I'm curious really how you justify your claim.
According to/. it's ok to steal music and movies, but when you steal a comic book image it's grounds for the death penalty!:)
Re:Well, I guess that's how Fascism takes root....
on
Want Freedom?
·
· Score: 2
I agree that is the way our government should work.
I also don't entirely agree with many of the other sentiments in that article or in the followups.
I would like to point out that one of the primary issues I have with the Bush administration is their unwillingess to work with the other two branches of government because it may be inconvenient. Ashcroft has been pushing for civil rights abuses despite the courts warning him otherwise, and recently Bush has been rattling the swords of War without a Declaration from Congress.
So if the Bush Administration doesn't share responsibility with the other two branches, are they not then completely at fault when they do wrong?
If you request information from the government, say like Bill Gates police report, it is going to come to you in the 5000 year old industry standard paper format.
There is a difference between regression testing where you verify that the major functionality still works, and full regression testing where you try to verify that all functionality still works and it didn't negatively impact any major products from third party vendors, etc. etc.
One takes some time, the other takes a lot of time.
Anyway, still waiting for that KDE patch from Redhat.
Patches take time to test and release. That's the reality of it.
Look at the KDE vulnerability reported last month. While the Linux media announced it took only 90 minutes to fix, it took nearly 3 weeks before a patch was actually made available.
Even then Redhat still has not made this patch available as a downloadable RPM off their site. So for dozens of Linux users out there, they are still vulnerable.
Slashdot Mirror
How can I be more specific than he was just wrong? The comments about compilers were wrong, as were the API, developer support, dev tools, debuggers, developer competence, sample code, and so on and so on.
"However, you do see them moving off IIS and onto Apache."
.Net development environment. It will be difficult for Linux to address this momentum as the community tends to embrace old standbys(like perl and now PHP) instead of innovating in new directions. Furthermore the recent IDC survey even shows in 2001 Linux growth was stagnant on the server side, while Microsoft had a 7% marketshare gain primarily related to customers finally understanding and adopting Windows 2000.
Well actually no you don't, which is reinforced by the Netcraft monthly web survey comments. The numbers of Apache sites reflect massive webhosting firms, but for people and companies who do their own dynamic web development there has been no move away from IIS. Apache is an easy migration if all you have is static content, but then if that's all you have who cares what server you use?
MS has also been increasing the value proposition with it's server pieces, especially now with the
Ok, enough wasting my time here. I'm just always amazed how out of touch the Linux zealots are with the industry. If you think you want to beat Microsoft you have to first start being honest about where you are currently.
The fact that /. won't run an article from yesterday which discusses Linux server marketshare in unfavorable terms is more than proof enough of exactly who is "running scared."
http://news.com.com/2100-1001-959049.html
It would be most helpful in the future if you would only comment on those things which you have some knowledge of.
Obviously developing on the Microsoft platform is beyond your capability.
Yes, open standards do make a case for having an open sourced reference implementation to use as a basis. You see this all the time.
But you'll also note that these open sourced implementations are not licensed under the GPL. The prime example for this is the gzip library used for HTTP 1.1 compression, they had to come up with a non-GPLed implementation specifically because of it's use as part of a standard.
I see a lot of intellectual dishonesty on the part of the GPL bigots. They frequently reference non-GPLed open source software as success stories(Apache, BSD TCP/IP, BIND, sendmail, etc), and then imply by inference that this same success will also apply to the GPLed software.
Oh, just read your last sentence. I thought you were being serious in your reply, but I guess you were just being a troll. Sorry to have wasted my time responding to a child.
The answer is: Reading comprehension problems.
/. has frequent failures with the mySQL database server and recovers from them by rebooting the box.
The point was that Roblimo was basically admitting that
It appears as though you attempted to over-analyze the article as a result of some pre-programmed knee-jerk reaction.
Back about a year ago /. experienced a major outage. Roblimo at the time noted, "By 7 a.m. it was obvious that this was not a typical, easily-fixed, reboot-the-database problem."
Can anybody imagine an SQL Server or DB/2 customer being satisfied with that solution? That's what IBM/MS is saying.
Well, let's think about this. I actually think that the security is far more complicated than you make it out to be.
I have multiple banking, credit card and merchant accounts. I also have a whole lot of worthless accounts with places like slashdot.org, nytimes.com and so forth. It's these worthless accounts that I think are good solutions for single-signon passport type accounts. In fact this is pretty much what MS has done throughout their support websites.
As far as having multiple authorities being safer than a single one. Somewhat true, as long as you are using completely different usernames and passwords for each one of those authorities. Oh yeah, and don't write them down because that puts you at risk.
As far as credit cards, there are three well known authorities that track all of your purchases. They are named MasterCard, Visa and Discover... oh yeah and AmEx. AmEx even sends you a statement at the end of the year telling you everthing you bought.
But now how many different credit cards do you have? Hopefully you are using a different card for each merchant account you signed up for, otherwise once again you increase your risk because unfortunately these web merchants save the damn credit cards in their databases. For your convenience, of course... (weird how brick and mortar stores never do this, and it's also no coincidence they don't lose lists of thousands of valid credit card numbers either... hmm) So now we've established you've been spreading your credit card number around to dozens of websites, each one with probably questionable security.
I don't know what the answer is. I only believe Passport is a good idea for all of my less important accounts, for right now. But I also worry about my more important accounts, and I really worry about Amazon.com storing my credit card number for my convenience. It is a complicated thing, and I don't agree that saying a single signon system is dumb, because it ignores the fact that the status-quo, the way things are today is also very incredibly dumb.
We do need something better, but I'm not sure what that will be. I had some hopes of AmEx's smart card system, but they never really got that working.
Hard to find? I didn't know where it was myself, so I just went to the MSDN site and did a search for "Program Access and Defaults".
It was the second link, the first described how to use the new tool from a user's perspective.
Definately not hard to find at all.
"It's huge and Microsoft, in their infinite wisdom, rarely chooses to setup mirrors."
Huh? Microsoft has been using akamai for a couple years now, before that it was conxion. It's mirrored all over the world. I believe they're doing bandwidth limiting on the download in order to support more users at once, as all downloads I've seen of XPSP1 have been at exactly 10K.
For like $15 you can get a CD sent to you via Airborne Express.
The "Your current 3rd party browser" is there solely as a default. The third party product needs to rewrite their installation so that they register with the new API that controls these applications.
It's all documented here.
My understanding is that the latest version of Netscape 7 does register itself properly. Opera and the others have apparently not taken the time to create new install packages.
If you take the time to actually figure out how things work you'll find that your criticism is entirely unjustified.
Interesting... according to my stats I'm only getting around 5-8 hits per day on port 80. Far far less than I was receiving at this time last year.
You didn't mention what size WD, or from which line.(Caviar?)
"In one server, we lost 4/4 western digital drives over the course of 3 months."
Hmm. Western Digital hasn't made SCSI drives for years, and even then it wasn't common to see them in servers.
It's weird. I've purchased a variety of drives over the years, and I've had a variety of drives in systems I've supported. All totalled I've probably purchased 30 drives for personal use, and for work we're talking thousands.
:)
Western Digital is the only company I've never seen a failed drive from. The next best is Seagate, where the only failure I saw there was about 7 years ago with a 170Meg and a 500Meg. Now granted, I've seen failed Seagate drives in the servers but out of the many hundreds there, the failure ratio has still been reasonable.
Maxtor I've had numerous failures from, 340Meg, 3Gig, 13Gig and so forth. I tend to avoid their product any more because the failures weren't a while ago, they just keep happening...
IBM I only had one failed 2 gig drive, years ago. I have a couple 20 and 30 gig from them now, but I otherwise haven't seen too much of their product.
Quantum was probably the worst, but that's because we had about 500 desktops with 2.5Gig bigfoot drives in them and I must have replaced half of them. Otherwise I had good luck with their older 500Meg and 1Gig Fireball drives.
Anyway, not quite sure why people bash on WD drives. Maybe you've just seen a higher volume of their drives than other brands. It would be like working in a Honda dealership and saying Honda's suck because they're always coming back in for repairs.
Universities began teaching Java back in '97/'98 timeframe long before it was in widespread usage. Part of this was because of a push in '98 by Sun to promote Java at the University level. Look it up, it's called the AAJC.
Every argument you could make pro-Java you can also make pro-C#. It is a well designed language, and a marketable skill for would-be graduates. The fact that you have sold out to Sun is irrelevant to this argument.
Even though DOS 6.0 was released like 10 years ago, it will still boot and run on the Pentium 4 computer I bought last month.
Wouldn't it be interesting if the hardware makers changed the specs such that new machines only ran Windows XP, and Microsoft simply claimed that they weren't doing this purposefully... they were just not going to update DOS and Windows 95 to boot on the new hardware because it is legacy code.
I must say Mac people have their head stuck in the sand far more deeply than any other advocacy group.
The Windows 2000 FTP daemon has been notoriously insecure, in contrast with open source products like MuddleFTPd and ProFTPd. Why? Because the coders who wrote this security-critical part of the system just didn't care.
Unfortunately the existence of WuFTPD pretty much invalidates your entire argument.
I'm also curious about Win2k ftp being insecure. I can only remember one exploit, do you have a list of other exploits? There have certainly been exploits of Proftpd as well, so I'm curious really how you justify your claim.
According to /. it's ok to steal music and movies, but when you steal a comic book image it's grounds for the death penalty! :)
I agree that is the way our government should work.
I also don't entirely agree with many of the other sentiments in that article or in the followups.
I would like to point out that one of the primary issues I have with the Bush administration is their unwillingess to work with the other two branches of government because it may be inconvenient. Ashcroft has been pushing for civil rights abuses despite the courts warning him otherwise, and recently Bush has been rattling the swords of War without a Declaration from Congress.
So if the Bush Administration doesn't share responsibility with the other two branches, are they not then completely at fault when they do wrong?
Differentiating between innovative creations and innovate uses of technology is semantic nonsense.
No, actually it's not semantic nonsense.
But thanks for trying to participate in an intelligent discussion.
If you request information from the government, say like Bill Gates police report, it is going to come to you in the 5000 year old industry standard paper format.
I believe he was asking for examples of innovative creations of technology, not uses.
If Redhat adds lag that's unacceptable to you then why are you using redhat?
I just want to run the OS, I don't want to have to compile software just to fix a security hole.
Is that really so hard to ask?
If Linux is to be profitable on the service model, then people better be providing the service.
There is a difference between regression testing where you verify that the major functionality still works, and full regression testing where you try to verify that all functionality still works and it didn't negatively impact any major products from third party vendors, etc. etc.
One takes some time, the other takes a lot of time.
Anyway, still waiting for that KDE patch from Redhat.
Patches take time to test and release. That's the reality of it.
Look at the KDE vulnerability reported last month. While the Linux media announced it took only 90 minutes to fix, it took nearly 3 weeks before a patch was actually made available.
Even then Redhat still has not made this patch available as a downloadable RPM off their site. So for dozens of Linux users out there, they are still vulnerable.