Now maybe, in your opinion, that's criticism, in mine it's trolling
Bah! The issue of trolling seems stupid when you consider the title of this submission claims it's an "Linux Migration".
The linux migration part is a flop so far but people don't want to admit it and the linux crowd is riding the coat tails of OpenOffice.org and Mozilla trying to get some credit. This is a successful open source migration. It's just that while OpenOffice.org and Mozilla showed up and were the belle's of the ball, linux got drunk, started talking shit and just embarrassed itself.
Maybe instead of dropping the GNU in GNU/Linux, we should drop the other part.
I don't like bashing linux but if half the people that claim it's better than windows on the desktop put their efforts into improving it instead of talking about how great it is, things would be different.
Solar hydroscopic water extraction dates back to at least the '90's.
Yeah, stupid scientists working on solving a problem that's already been solved and people in these climates are only milking this "we don't have clean water" deal to get their less than a cup of coffee a day money from us. I mean a solar powered optical device that can view below the surface of water should have solved all these problems. Or did you mean hygroscopic?
I mean, really, I was making a light-hearted joke
If a joke falls in the forest and nobody laughs was it really a joke?
Don't be one of those guys that is wrong and then winds up looking even more foolish trying to prove that he wasn't THAT wrong. Or start a collection to buy a whole lot of cups and take your troop out to the desert and start generating the water for people to drink and grow crops because the world isn't getting any bigger and while the number of people living on it is.
Your snapshots are there and working correctly there's just no.zfs directory yet since I'm still porting that functionality. You can use 'zfs clone' to work around this for now if you'd like to browse your snapshots. You can see an example and get more info off of the Known Issues and Features page
Even the BSD folks have more than one developer working on the ZFS BSD port. Although I must commend him for not using "we" to make himself sound bigger than he is.
I think it's more than just the brand. It's also the marketing. Saying "free as in freedom, not free as in beer" turns off a lot of potential customers because their proprietary software vendors provide them with free beer and strippers.
what is getting allocated, exactly? It looks to me like that would be 10 million bytes of memory, but according to you, apparently it's not. Please enlighten me as to what it is being allocated, exactly, then?
An object that can hold that many bytes. You're not allocating anything directly. The JVM handles that.
Right. There's only references, which do... what, exactly? How do I get from the b above to the 10 million bytes of memory that I just allocated?
You don't get to that memory, you get to that object. It just happens to be stored in memory but it could be stored on stone tablets if that was possible..
or some other way to prevent pointers from pointing at something that they shouldn't be pointing at
As a java developer, you don't have pointers, which you keep going back to. You have object references. Behind the scenes there are likely pointers but you don't get access to them. You're arguing that it was done for security instead of ease for the programmer which is missing the point. Keeping your code secure by preventing memory leaks and illegal memory access is difficult and time consuming. By fixing one, you get the other. Which was the motivating factor is only something Gosling can answer.
The point is in Java applications, you can still have "memory leaks". The difference is in Java, those leaks result in poor resource management and bloated programs. What you don't get is the security risks that poor memory management in C/C++ can give you.
You don't need a garbage collector for the type of security you're talking about. Java could have had a free operator that would point all references to null and handle the security issues, but it would make things difficult for the developer.
Some things in TFA make me wonder though, like "Enterprise: no standard way of software distribution". How hard is it to set up a local repository(-ies), from where workstations get updates?
In an enterprise desktop environment, you don't want to have a repository where users can pull software updates, you want a system where you can push them onto user desktops. I think there are some now but they've only come to market recently. IBM could have helped out by coming out with Linux support for Tivoli provisioning manager much sooner.
For a majority of enterprise users, just having a good office suite, web browser and groupware would be sufficient but only 2 out of the 3 were available in a working state until recently.
The big boys in groupware were Outlook/Exchange and Notes/Domino. Exchange quickly ate into Domino's market and by the time IBM started supporting Linux with Notes clients it was too late. If Linux does get more adoption in corporate desktops the groupware they'll be using will likely be Zimbra or something other than Domino or Novell's products.
Sun's Java Enterprise System could have worked since it was web based, but they didn't have the resources to come into the market compete with two already large players. Plus there's a general anti-sun sentiment in the Linux community.
That's just for regular office workers. When you mix in others that might need other special software, you can run into problems.
Most people don't care about the OS, they only care about the applications they interact with. If the applications aren't there it doesn't matter if you have the most perfect OS that has ever or will ever be written.
Linux became popular on edge servers because it could provide services through applications (Apache server and other ASF stuff, MySQL, sendmail, Samba, ftp, etc.) that run on it
Those applications brought open source into the business world and Linux came along for the ride. Then once Linux was in the corporate data center, other ISVs started releasing version of their enterprise server software for it.
There hasn't been enough ISV support for desktop Linux. It's also more difficult because desktops tend to run a variety of software, unlike servers that are commonly deployed with a limited software stack for one application.
I'd like to see it happen, but it's not quite there yet. It's good to see that open source desktop software is making it, but Linux hasn't been invited to this party yet.
No, I see it, and your bolding, but I don't see the focus of the question being one of a site that has specifically done a migration.
I don't know how much clearer it could be.
'It' can be read as a migration, the thing they just mentioned, or everything, as in run a different OS, run different apps, migrate clients from one to the other, etc.
Yeah... why assume they're asking for what they mentioned? Why not just assume they mean something other than what they said.
No, because people are interpreting that as one site that is 11k+ desktop computers, using OpenOffice, which all migrated from Window NT.
Doesn't have to be 11+k could be anywhere around that number. How about 5k? Can you find any Win/Office to Linux/OO.o migrations for regular government office workers at that time?
Sure, the flat answer is no. But that's sort of lazy.
The email was a follow up question to the study. The guy wasn't being hired as a consultant or analyst to create a detailed analysis to find answers to his question.
. The better answer is, here's a list of sites doing things that are like you'd have to do and a bit about what you could learn from examining each.
Would you mind posting examples of your OO.o network diagrams? I was looking for examples a while ago and couldn't find any good ones. Also, do you just use generic shapes or have you found a set of images that work together?
He could have. But I don't know of many people who type into their Office application of choice rather than just their browser or a lightweight pad when posting.
Ugh. That just reminded me of all the times I'd open up a word document that was sent as an email attachment that just said something like "Project meeting today at 2:00."
It's more like "Google has to pay for the privilege of displaying content creators freshly created content next to Google ads."
There are no ads on the Google News homepage or the Google home page or even the iGoogle homepage so I don't see how they are using ads with other people's content in your case.
Without you using Google, those news sites wouldn't get the 10% of clicks you generate.
If newspapers don't like it they can use their robots.txt file to block googlebot. Even worse, Google News has become more of an opt-in crawl where you have to request it and meet certain crtieria. You even need to include a unique numerical id in your urls for google to include you in the news index.
Newspapers could opt out of google news but it would be the equivalent of providing newsstands with front pages that contained no headlines or stories. People walking by wouldn't see the attention grabbing headlines that might cause them to buy the paper and see the advertisements contained.
Before that, they would just crawl news sites and display headlines and summaries, just like in their normal search.
It seems odd. Google has to pay for the privilege of sending them traffic. I wish I could get a deal like that.
If I were Google, the next time the traditional news outlets came to me with their hands out I'd tell them I've decided that I'd be more than happy to remove all their content from my index and no longer "steal" their business. Thew newspaper execs wouldn't like that too much.
I didn't get the emphasis on the 'do the migration', to me it appeared they wanted to see any significant OSS site.
I tried to emphasis the part of the email that asked for migrations but maybe your handheld browser can't see bold so here it is again.
We definitely do not want to embark in a migration without having verified that others have done it successfully before us, and that the benefits would exceed the disadvantages. In this spirit, we would like to visit 2 or 3 successful sites, if any exist.
It does seem like there should have been SOME medium-to-large scale OSS sites in Europe in 2005.
Nobody has been able to name 2-3 sites that could satisfy this guy's request. But that doesn't stop people from whining about how gartner and microsoft are conspiring to keep people away from linux, even though half of those people (likely more) are running windows.
By the way, the number of nodes and the hardware in the nodes for this Hadoop cluster is -optimized- for this contest.
The number of nodes was reduced to run the 100TB benchmark but I don't see anything that backs up your comment that the hardware was optimized for this contest. The cluster hardware doesn't look like anything special. Maybe it's optimized for Hadoop which is different than being optimized for the contest.
Yahoo's cluster had 3,800 nodes with 4 disks per node giving it roughly 15,200 drives plus or minus the dead nodes/drives in the cluster.
The Google cluster had 4,000 nodes with 48,000 hard drives. 12 drives per node doesn't sound like the typical Google servers I've seen. That one looks like 2-4 drives. This other video seems to show the storage node which looks like it has 5-10 drives.
The reason I bring up drives is that sorting 1PB likely involves hd access the more drives, the higher I/O throughput.
Whatever the case, the nodes seem to be vastly different and making a comparison based on the number of nodes doesn't seem appropriate.
Slashdot Mirror
Now maybe, in your opinion, that's criticism, in mine it's trolling
Bah! The issue of trolling seems stupid when you consider the title of this submission claims it's an "Linux Migration".
The linux migration part is a flop so far but people don't want to admit it and the linux crowd is riding the coat tails of OpenOffice.org and Mozilla trying to get some credit. This is a successful open source migration. It's just that while OpenOffice.org and Mozilla showed up and were the belle's of the ball, linux got drunk, started talking shit and just embarrassed itself.
Maybe instead of dropping the GNU in GNU/Linux, we should drop the other part.
I don't like bashing linux but if half the people that claim it's better than windows on the desktop put their efforts into improving it instead of talking about how great it is, things would be different.
It's not the fans.
You don't mess with people that blow stuff up for a living.
Even if they're nerds.
Solar hydroscopic water extraction dates back to at least the '90's.
Yeah, stupid scientists working on solving a problem that's already been solved and people in these climates are only milking this "we don't have clean water" deal to get their less than a cup of coffee a day money from us. I mean a solar powered optical device that can view below the surface of water should have solved all these problems. Or did you mean hygroscopic?
I mean, really, I was making a light-hearted joke
If a joke falls in the forest and nobody laughs was it really a joke?
Don't be one of those guys that is wrong and then winds up looking even more foolish trying to prove that he wasn't THAT wrong. Or start a collection to buy a whole lot of cups and take your troop out to the desert and start generating the water for people to drink and grow crops because the world isn't getting any bigger and while the number of people living on it is.
I'm thinking they didn't put enough resources into ZFS?
From http://zfs.macosforge.org/trac/wiki
Where is .zfs? I can't find my snapshots
Your snapshots are there and working correctly there's just no .zfs directory yet since I'm still porting that functionality. You can use 'zfs clone' to work around this for now if you'd like to browse your snapshots. You can see an example and get more info off of the Known Issues and Features page
Even the BSD folks have more than one developer working on the ZFS BSD port. Although I must commend him for not using "we" to make himself sound bigger than he is.
We thank you for reading this comment.
I'm glad that people are focusing on answers for people in underprivileged parts of the world, but it's not some sort of magical discovery.
You must have read the wrong article. They never claimed it was magic.
P.S. Claiming you haven't read the article doesn't absolve you if you make a mistake.
I think it's more than just the brand. It's also the marketing. Saying "free as in freedom, not free as in beer" turns off a lot of potential customers because their proprietary software vendors provide them with free beer and strippers.
what is getting allocated, exactly? It looks to me like that would be 10 million bytes of memory, but according to you, apparently it's not. Please enlighten me as to what it is being allocated, exactly, then?
An object that can hold that many bytes. You're not allocating anything directly. The JVM handles that.
Right. There's only references, which do... what, exactly? How do I get from the b above to the 10 million bytes of memory that I just allocated?
You don't get to that memory, you get to that object. It just happens to be stored in memory but it could be stored on stone tablets if that was possible..
or some other way to prevent pointers from pointing at something that they shouldn't be pointing at
As a java developer, you don't have pointers, which you keep going back to. You have object references. Behind the scenes there are likely pointers but you don't get access to them. You're arguing that it was done for security instead of ease for the programmer which is missing the point. Keeping your code secure by preventing memory leaks and illegal memory access is difficult and time consuming. By fixing one, you get the other. Which was the motivating factor is only something Gosling can answer.
The point is in Java applications, you can still have "memory leaks". The difference is in Java, those leaks result in poor resource management and bloated programs. What you don't get is the security risks that poor memory management in C/C++ can give you.
You don't need a garbage collector for the type of security you're talking about. Java could have had a free operator that would point all references to null and handle the security issues, but it would make things difficult for the developer.
I mean, how hard is to reformat a pc?
And how was he going to get windows back on it? Just because he's a thief, doesn't mean he's a pirate. :)
A customer is someone that pays you for goods and services. Not someone that sends you an email.
If you want to ignore everything this "customer" asked and just answer what you think he asked you can make up anything you want. Just like you did.
Pfft. You need to keep up. The future of data center cooling is oil.
Personally, I prefer bacon fat for the extra flavor.
It doesn't matter. Biden was probably wrong.
Some things in TFA make me wonder though, like "Enterprise: no standard way of software distribution". How hard is it to set up a local repository(-ies), from where workstations get updates?
In an enterprise desktop environment, you don't want to have a repository where users can pull software updates, you want a system where you can push them onto user desktops. I think there are some now but they've only come to market recently. IBM could have helped out by coming out with Linux support for Tivoli provisioning manager much sooner.
For a majority of enterprise users, just having a good office suite, web browser and groupware would be sufficient but only 2 out of the 3 were available in a working state until recently.
The big boys in groupware were Outlook/Exchange and Notes/Domino. Exchange quickly ate into Domino's market and by the time IBM started supporting Linux with Notes clients it was too late. If Linux does get more adoption in corporate desktops the groupware they'll be using will likely be Zimbra or something other than Domino or Novell's products.
Sun's Java Enterprise System could have worked since it was web based, but they didn't have the resources to come into the market compete with two already large players. Plus there's a general anti-sun sentiment in the Linux community.
That's just for regular office workers. When you mix in others that might need other special software, you can run into problems.
Most people don't care about the OS, they only care about the applications they interact with. If the applications aren't there it doesn't matter if you have the most perfect OS that has ever or will ever be written.
Linux became popular on edge servers because it could provide services through applications (Apache server and other ASF stuff, MySQL, sendmail, Samba, ftp, etc.) that run on it
Those applications brought open source into the business world and Linux came along for the ride. Then once Linux was in the corporate data center, other ISVs started releasing version of their enterprise server software for it.
There hasn't been enough ISV support for desktop Linux. It's also more difficult because desktops tend to run a variety of software, unlike servers that are commonly deployed with a limited software stack for one application.
I'd like to see it happen, but it's not quite there yet. It's good to see that open source desktop software is making it, but Linux hasn't been invited to this party yet.
No, I see it, and your bolding, but I don't see the focus of the question being one of a site that has specifically done a migration.
I don't know how much clearer it could be.
'It' can be read as a migration, the thing they just mentioned, or everything, as in run a different OS, run different apps, migrate clients from one to the other, etc.
Yeah... why assume they're asking for what they mentioned? Why not just assume they mean something other than what they said.
No, because people are interpreting that as one site that is 11k+ desktop computers, using OpenOffice, which all migrated from Window NT.
Doesn't have to be 11+k could be anywhere around that number. How about 5k? Can you find any Win/Office to Linux/OO.o migrations for regular government office workers at that time?
Sure, the flat answer is no. But that's sort of lazy.
The email was a follow up question to the study. The guy wasn't being hired as a consultant or analyst to create a detailed analysis to find answers to his question.
. The better answer is, here's a list of sites doing things that are like you'd have to do and a bit about what you could learn from examining each.
And that's what he did.
he perfect example of how MS bloat comes about
You can't blame people's stupidity on MS.
Would you mind posting examples of your OO.o network diagrams? I was looking for examples a while ago and couldn't find any good ones. Also, do you just use generic shapes or have you found a set of images that work together?
On that note, would you mind telling me what it is that Visio does?
Visio network diagrams are the currency used in many IT procurement departments.
He could have. But I don't know of many people who type into their Office application of choice rather than just their browser or a lightweight pad when posting.
Ugh. That just reminded me of all the times I'd open up a word document that was sent as an email attachment that just said something like "Project meeting today at 2:00."
Have you tried Dia as an alternative to Visio? I've used Visio myself in the past, but it seems that Dia does just as much as I ever did with Visio.
It would be nice if there was a visio replacement in openoffice.org.
Dia comes pretty close in functionality but it's biggest drawback is the lack of great looking shapes for different lines of work.
Conservative media brings on opposing sides, they just tell the opposing side to shut-up.
It's more like "Google has to pay for the privilege of displaying content creators freshly created content next to Google ads."
There are no ads on the Google News homepage or the Google home page or even the iGoogle homepage so I don't see how they are using ads with other people's content in your case.
Without you using Google, those news sites wouldn't get the 10% of clicks you generate.
If newspapers don't like it they can use their robots.txt file to block googlebot. Even worse, Google News has become more of an opt-in crawl where you have to request it and meet certain crtieria. You even need to include a unique numerical id in your urls for google to include you in the news index.
Newspapers could opt out of google news but it would be the equivalent of providing newsstands with front pages that contained no headlines or stories. People walking by wouldn't see the attention grabbing headlines that might cause them to buy the paper and see the advertisements contained.
Another group of fantasists speculate about ways of extorting money from Google, which they portray as a parasitic feeder on their hallowed produce.
From what I understand Google licenses news from the big news wires as well as from some of the big newspapers. Some of that has been forced through lawsuits.
Before that, they would just crawl news sites and display headlines and summaries, just like in their normal search.
It seems odd. Google has to pay for the privilege of sending them traffic. I wish I could get a deal like that.
If I were Google, the next time the traditional news outlets came to me with their hands out I'd tell them I've decided that I'd be more than happy to remove all their content from my index and no longer "steal" their business. Thew newspaper execs wouldn't like that too much.
I didn't get the emphasis on the 'do the migration', to me it appeared they wanted to see any significant OSS site.
I tried to emphasis the part of the email that asked for migrations but maybe your handheld browser can't see bold so here it is again.
We definitely do not want to embark in a migration without having verified that others have done it successfully before us, and that the benefits would exceed the disadvantages. In this spirit, we would like to visit 2 or 3 successful sites, if any exist.
It does seem like there should have been SOME medium-to-large scale OSS sites in Europe in 2005.
Nobody has been able to name 2-3 sites that could satisfy this guy's request. But that doesn't stop people from whining about how gartner and microsoft are conspiring to keep people away from linux, even though half of those people (likely more) are running windows.
By the way, the number of nodes and the hardware in the nodes for this Hadoop cluster is -optimized- for this contest.
The number of nodes was reduced to run the 100TB benchmark but I don't see anything that backs up your comment that the hardware was optimized for this contest. The cluster hardware doesn't look like anything special. Maybe it's optimized for Hadoop which is different than being optimized for the contest.
Something doesn't seem right.
Yahoo's cluster had 3,800 nodes with 4 disks per node giving it roughly 15,200 drives plus or minus the dead nodes/drives in the cluster.
The Google cluster had 4,000 nodes with 48,000 hard drives. 12 drives per node doesn't sound like the typical Google servers I've seen. That one looks like 2-4 drives. This other video seems to show the storage node which looks like it has 5-10 drives.
The reason I bring up drives is that sorting 1PB likely involves hd access the more drives, the higher I/O throughput.
Whatever the case, the nodes seem to be vastly different and making a comparison based on the number of nodes doesn't seem appropriate.
But has anyone patented it yet? Patents trump copyright after all.
There are a number of patent applications related to MapReduce from Google and Yahoo.