While this is most likely a FUD troll, people may be scared about what is said.
It's quite debatable what all the terms mean and what you are allowed to do. Most of the issues have not been settled in court.
This is pure FUD on the highest level. Thus far the courts have ruled that the GPL is what it says it is and the language is plain enough. What organizations like Cisco are trying to to in the courts is limit the powers of the GPL, and are failing.
Personally, I think your advice is simplistic and wrong, though it does seem to be commonly accepted.
Everyone is entitled to their opinion, I guess.
This is like compiling a shareware disk where the programs aren't related. Once you combine the programs to create a greater program, you are no longer "merely aggregating".
Maybe you misunderstood my original post, but "combine the programs to create a greater program" is a bit ambiguous. I suspect intentionally so.
The GPL and its supporting documents describe exactly the methods by which programs can be created and how any encumbrances work. You are obscuring the facts in order to spread FUD.
If you slap a GUI on top of a GPLd program and use sockets, do you really think your program is "mere aggregation"?
As long as the GPL program runs unmodified in its own process space, absolutely, yes.
While I am not a lawyer, I have worked with several IP lawyers in creating GPL compliance guidelines.
You are going to have to do that with any database. My current client is a large publisher, and you should see the shit we have to do to Oracle to get it to perform, and that is with 6 full-time DBAs.
No one is mentioning Oracle, don't use it as a strawman. We are talking about Free/OSS replacements for commercial apps.
This is not exactly a scientific metric.
The first part was: "None of these sites are mission critical." The point was that MySQL is not up to the task.
Like I said, if we just had "pages calling Oracle", we would fry Oracle. It required extensive SQL tuning, client optimization, denormalization, caching, clustering, load-balancing, etc. to keep Oracle alive.
Like I said, we aren't talking about Oracle. That being said, equivocation is a fools game. Saying that two things are the same because both posses flaws is either lazy or cynical. Oracle has problems, as does PostgreSQL, Sybase, DB2, etc. That's why intelligent people weight the pros and cons when they evaluate technology. So, yes, it may be an amount of work to get Oracle to perform well in a specific environment, but that, in no way, is equivalent or even similar to what is need for MySQL.
Given all of the framework we have built around Oracle, I would bet that MySQL with NDB-cluster would be able to keep up, but I have no data to back that claim up with.
If you can't quantify why you think something is true, then you are not making an informed decision.
The point, of course, is moot, because we use a lot of other features in Oracle that are not present in the open source databases like MySQL and PostgreSQL.
Ahh, nice equivocation whitewash. Both MySQL and PostgreSQL lack features you use, so both are equally bad. LOL. Having worked with all these platforms, and having moved a couple systems from Oracle to PostgreSQL, I call your bluff.
It doesn't work if your kernel isn't heavyweight enough to have a process scheduler. In the case of Linux, it is, but in the case of something that runs on a handheld or embedded device with 4 MB of RAM and a single-tasking OS, it might not be. (Example: Nintendo DS)
Then that's just too bad. Pay someone to write it.
It doesn't work if the library is GPLv3 or LGPLv3 and your hardware's operating system requires all code to have been digitally signed by the hardware maker, unless you provide external hardware to run the covered code. (Example: DVRs and video game consoles)
If you have to make something as loathsome as a DRM protected anything, too f-ing bad.
I think this misses the point. The premise behind the GPL is encourage, or force, the sharing of code.
I think *you* miss the point. The GPL in no way is trying to tell you what to do with YOUR code. It is, however, insisting that you respect the rights of the people who wrote the GPL code you use.
If you link to GPL code or extend already GPL'd tools and then redistribute you must include the source code.
Yup. IF YOU'RE JUST GOING TO FIND SOME LOOPHOLE IN THE GPL (BY USING SOCKETS OR SOMETHING) IT KIND OF DEFEATS THE PURPOSE OF THE GPL ENTIRELY.
Not at all and I would bet that RMS himself would agree. I have not defined any such thing as a loophole. I have defined a way that is legally safe and sound for using GPL as a base for your project which is *not* GPL. It allows you to do what you feel you need too while respecting the rights of the people who have offered to share the code so generously.
RMS is about freedom, he is not about forcing anyone to do anything.
In this case its more so that Cisco/Linksys uses GPL'd software in their products and distributes it (via firmware) to its customers. They don't supply the source code for this GPL'd software as required by the GPL when distributing.
I obviously, but it is about taking someone else's work and incorporating it into a product in a way that is counter to the license through which they receive access to the source code.
They have modified someone else's code and, against the license, not shared their changes.
Yes, as the western world is quickly shedding the rights and freedoms our ancestors fought and died for, it is good to know that our leaders and would-be oppressors are idiots and smart people will be free from prosecution because a fairly well informed 14 year old will be able to hide evidence from the jack-booted inquisitors.
Cisco is being REALLY REALLY stupid here and I just don't understand why.
I've done a lot of commercial software that uses LGPL and GPL code, and its not rocket science. RMS himself even says that "mere aggregation" is not a problem.
Here are the rules:
if its LGPL, link to it, but don't modify it. If you need to modify it, make the modification in the form of a generic API extension and call it from the application. Make your extensions public.
If it is GPL, make it a service and call it through a socket.
If it is a kernel module, there seems to be some wiggle room there, otherwise make a public mini-driver and a proprietary user space app.
How hard is that? Jeez, if you screw up GPL compliance, you are not paying attention.
then how does the Shinsei Bank [linuxinsider.com] of Japan manage (US$118 million net income, first quarter 2008).
Notice too, that it is to manage their CRM product, and *NOT* their financial data.
Produce any evidence that the 'MySQL guys' faked benchmark results, produce a comparison benchmark of MySQL versus a 'commercial' database.
I never said "fake" I said "cook." The implication is that anyone can create a benchmark that makes their product look best. You should not that MySQL can't even do a STANDARD sql benchmark because it does not support enough of the standard SQL features.
There is a difference between being a troll and being opinionated.
I have a couple decades of experience with databases and have spent a very good amount of my career keeping up with the state of the art. I've used at least one version of pretty much every popular database since Aston Tate introduced dBase. To some, I am considered an expert and am paid for the advice I give. That doesn't mean much on slashdot because "MySQL is good enough."
Interesting. Tell that to Flickr, Facebook, Wikipedia, Google, Nokia and YouTube. Or, how about Slashdot and Digg - capable of bringing down moderately sized web sites with the click of a million mice?
As has been discussed many times over there are a couple points to make out about these examples.
(1) *ALL* of these sites buttress MySQL with support code. Take a look at what Slashdot has to do to enable MySQL to keep up with the sites needs.
(2) None of these sites are mission critical. Would you TRUST your bank transactions on MySQL? LOL, no way!
Now, I'm not a professional DBA. I'm just a programmer
I was a DBA, I am a software engineer. I have published articles about Windows kernel development and data acquisition. I was CTO at a dotcom, I am currently a consultant have worked directly or indirectly with AOL, Microsoft, Sun, and Yahoo.
but I was one of the maintainers of the MySQL server (I don't get to touch the Oracle servers here except on my local developers instance).
I have managed and developed on many Oracle systems. I have also done the same on Sybase, PostgresSQL, MySQL, DB2, MSSQL, mSql, SQLite, Advanced Revilation (Not SQL, but a database), and others.
I have a very good background on the subject.
I can tell you from personal experience that MySQL is easier to maintain and administer,
Than what? faster to start up, and requires far fewer system resources to keep going.
Than what?
Judging by just the performance of Wikipedia and Facebook, it seems to perform quite well under heavy load.
You are not looking at MySQL, you are looking at an aggregate system whose performance is a product of the development teams ability to work around MySQL.
None of the sites you mention simply execute a query and display the results. Every single one of them requires a lot of extra programming work to serialize and cache load because MySQL sucks.
So, please tell me what basis you have to place MySQL out of the elite top-tier of database servers?
Recommending MySQL is just stupid. Not only is its development at Sun in question, it is a poor excuse as a replacement for a commercial database.
As the MySQL fanbois are used to saying, "MySQL is good enough for what I do," a commercial database is held, and rightly so, to a higher standard to which MySQL fails miserably to measure. Yea, sure, the MySQL guys can cook single instance benchmarks that look impressive, but the scalability, reliability, and feature set lack on a professional level.
I don't even need to say which is the better alternative because everyone knows what it is. Since the guy recommends MySQL, this means he didn't evaluate the application space well enough to make the recommendations he does.
The U.S.A. is going to shit. Period, end of sentence. This "zero tolerance" "get tough on [insert noun/verb]" is rediculous.
This has NEVER been a nation of laws and rules, this has historically been a nation of rule breakers.
We have all lost a sense of proportion and reason. Everyone wants everyone else screwed to the wall for the merest infraction. Everyone has become selfish and can no longer tolerate anyone else's exercise of their freedoms if it means the slightest inconvenience or offense.
Sorry everyone, this is not America. The U.S.A. of my youth was an in-your-face nation that depended on strength and a thick skin. Now the crybabies want a nice perfect little safe and oppressive police state.
People would rather feel save in a police state than enjoy the fruits of liberty and the risks that come with it.
Its a few fucking emails. Get the hell over it. Does anyone know what that suspension will cost this person in REAL dollars in tuition?
Seriously, I think the web has to become an all SSL backbone. Its fast enough through the wire, and computers are fast enough to codec that this nonsense has to stop.
ISPs should not be interfering with content. Period. This will also prevent airports and cafes from inserting their own frames around user content as well.
Over 20 years ago I would at a company called Business and Professional Software. They had a DOS presentation package called "Trumpet" (This was before Trumpet WinSock)
Anyway, I ported the DOS version to Windows. Being a Windows 2.x beta tester and developer, I had access that BPS didn't have yet, so I had the expertise and the knowledge.
I did some very cool stuff in real-mode Windows. When the program printed, I saved a new document that represented what was selected for print, and spawned a new copy of the program to print. Everyone else at the time had printing as a modal operation. I did some cool things with font mapping, kerning, color management, etc. The product was very cool. Personally, I think we were better than Hollywood (look it up) and Powerpoint was a joke. In the infinite wisdom of David Solomont, he hated Windows, and didn't push the project. He only wanted it to back up the DOS and DEC CDE software offering. Alas, so it goes. It would probably make a good base for a Linux offering today as we used AGFA font rendering and did our own layout, we did not use the Windows GUI except as an output device.
Anyway, I had a copy of the code a long time ago. (In case they need to ask me about it or need a fix) It is, however, now long lost. Most code doesn't actively get tossed when it is no longer used, but it becomes less and less likely to be backed up. In the case of companies going out of business, it exists until the last hard disk is crushed in the land-fill.
There's probably more than a million man-years work of code lost every decade.
Ever wonder how all these other people can get it working, and you can't? Every thought it might not be the technology, it might be you? Just asking...
Yes you can. Roaming User Profiles have existed in Windows for over a decade.
"Roaming Profiles" is typical of Microsoft features in Windows, its a nice little check box for IT managers. The deeper story is that "Roaming Profiles" are difficult to maintain and more work for IT departments.
Since Windows ZAK (Zero Administration Kit) it has been a bolt-on feature that does not work in anything but the strictest way.
You need 1, one, uno, system images. Multiple user directories, and a DHCP server for UNIX type systems and everything will work just like it is supposed too. In Windows, with a special server, login profiles, etc. it still does not work right. The applications don't work right, the default directories don't work, and it is, in general, unusable.
Ok, you had to change the machine name. Maybe with your method you didn't have to do anything at all in Unix. So I guess it is simpler by 15 seconds and a reboot. I now bow down to the superiority of Unix.
You winbois don't get it. UNIX is designed to do this stuff from the start while Windows only has it bolted on.
In your example, all that stuff is handled by "dhcp."
The parent is a typical fanboi post, long on FUD, short on facts. I'm posting anonymously because I don't want to have people at my company know who I am.
This gives it away, of course.
But it seems to me that Linux while cheap to buy is not cheap to keep patched and secure
Please site some documentation for this statement. It is pure FUD. The nice thing about Linux however is that a very skillful and thoughtful person can plan out a very robust network and can mange the patches. But it takes effort, dicsipline and an above avegage IT guy. And if you lose that person, you are screwed. Even a new equally skilled guy probably can't get all the scripts and stuff the last guy used to manage to work.
Again, this is pure FUD. *All* of the major distributions have had large management facilities as standard for years. "apt-get update" "apt-get upgrade" is all you need to know on [K]Ubuntu. yum for centos, and rpm for RH.
So why exactly "can't I do it on Windows"? You -do- know that HKEY_CURRENT_USER can be roaming, yes?
OK, here's what you need to do to be real.
You need one, that's right, one, system image that is either replicated and maintained on all the systems or is used to netboot the clients. The image contains all the companies approved and installed applications. This is a HUGE benefit to the IT department as they only have to test and deploy one image at a time.
Any approved machine can netboot (or copy) the system image, mount a home directory and work. Try to do that with Windows. *All* applications must work.
I have set up a number of systems like this, including beowulf clusters. Windows is a complete joke. A pseudo-thin client (netboot with local disk) network made up of UNIX boxes is the simplest of things to manage. All you need are machines that can netboot, a file server with dhcp.
One of the things that truly sucks about Windows is the registry. Each windows box is its own unique little snowflake, thus impossible to replace easily.
If this is done right, all the configuration is in the user's home ditrectory, probably shared on the network, and the rest of the system is a standard image. That means any user can use any computer and have their system where they want it.
This is no surprise to us UNIX folk, but POWs "Prisoners Of Windows," will love it. Imagine being able to replace/upgrade your computer simply by dropping a new box in front of you. Your settings completely unchanged!!!
I have been doing this with Linux for so long (separate/home disk that persists), I can't believe people still put up with Windows nonsense.
Yes, most of the actual IT people hate MS, mostly because it's shitty and also because it doesn't give them options and they like options.
Ok, so you admit that IT departments generally dislike Microsoft. Good reality check.
Management, even IT management, however, thinks differently.
To a point.
For the techies, another system to support is a challenge. To management, another system to support is another FTE or two, and they don't have that in the headcount.
And there is the HUGE problem of the monoculture, and the one to which knowledgeable people object. Support is an issue when things are *not* standardized. Yea, sure, if you buy everything in microsoft's shitbag, everything works, badly, but works.
When IT departments and vendors try to reduce their support costs by getting something fundamentally less labor intensive to support, it usually backfires because Microsoft intentionally breaks standards or usually has no standard, not even in their own products. So interoperability becomes the labor intensive problem.
The only solution is to eliminate Microsoft from your vendor list and work only with standards. Currently this is difficult, but alternatives are growing. Macintosh, for instance, isn't just for kids anymore. As more and more macs come on-line, even Microsoft will be forced to support them in some way.
It is a slow process to get rid of a monopoly, especially a well funded, contemptible, spiteful, and disgusting entity like Microsoft.
What are you talking about? Common sense like this has no place in discussions of lawsuits.
Sorry, you're right. What was I thinking.
While this is most likely a FUD troll, people may be scared about what is said.
It's quite debatable what all the terms mean and what you are allowed to do. Most of the issues have not been settled in court.
This is pure FUD on the highest level. Thus far the courts have ruled that the GPL is what it says it is and the language is plain enough. What organizations like Cisco are trying to to in the courts is limit the powers of the GPL, and are failing.
Personally, I think your advice is simplistic and wrong, though it does seem to be commonly accepted.
Everyone is entitled to their opinion, I guess.
This is like compiling a shareware disk where the programs aren't related. Once you combine the programs to create a greater program, you are no longer "merely aggregating".
Maybe you misunderstood my original post, but "combine the programs to create a greater program" is a bit ambiguous. I suspect intentionally so.
The GPL and its supporting documents describe exactly the methods by which programs can be created and how any encumbrances work. You are obscuring the facts in order to spread FUD.
If you slap a GUI on top of a GPLd program and use sockets, do you really think your program is "mere aggregation"?
As long as the GPL program runs unmodified in its own process space, absolutely, yes.
While I am not a lawyer, I have worked with several IP lawyers in creating GPL compliance guidelines.
You're ignoring the circumstances of the situation.
What the fact that someone is too stupid to ensure they don't let go of a remote control while they are playing with a video game?
I have a Wii. Breaking your T.V. with a remote is the act of pure stupidity!.
What's next, if you kill yourself in a car accident, it's your fault?
Hardly the same thing.
You are going to have to do that with any database. My current client is a large publisher, and you should see the shit we have to do to Oracle to get it to perform, and that is with 6 full-time DBAs.
No one is mentioning Oracle, don't use it as a strawman. We are talking about Free/OSS replacements for commercial apps.
This is not exactly a scientific metric.
The first part was: "None of these sites are mission critical." The point was that MySQL is not up to the task.
Like I said, if we just had "pages calling Oracle", we would fry Oracle. It required extensive SQL tuning, client optimization, denormalization, caching, clustering, load-balancing, etc. to keep Oracle alive.
Like I said, we aren't talking about Oracle. That being said, equivocation is a fools game. Saying that two things are the same because both posses flaws is either lazy or cynical. Oracle has problems, as does PostgreSQL, Sybase, DB2, etc. That's why intelligent people weight the pros and cons when they evaluate technology. So, yes, it may be an amount of work to get Oracle to perform well in a specific environment, but that, in no way, is equivalent or even similar to what is need for MySQL.
Given all of the framework we have built around Oracle, I would bet that MySQL with NDB-cluster would be able to keep up, but I have no data to back that claim up with.
If you can't quantify why you think something is true, then you are not making an informed decision.
The point, of course, is moot, because we use a lot of other features in Oracle that are not present in the open source databases like MySQL and PostgreSQL.
Ahh, nice equivocation whitewash. Both MySQL and PostgreSQL lack features you use, so both are equally bad. LOL. Having worked with all these platforms, and having moved a couple systems from Oracle to PostgreSQL, I call your bluff.
Seriously, if you break your TV with a remote, its your fault.
It doesn't work if your kernel isn't heavyweight enough to have a process scheduler. In the case of Linux, it is, but in the case of something that runs on a handheld or embedded device with 4 MB of RAM and a single-tasking OS, it might not be. (Example: Nintendo DS)
Then that's just too bad. Pay someone to write it.
It doesn't work if the library is GPLv3 or LGPLv3 and your hardware's operating system requires all code to have been digitally signed by the hardware maker, unless you provide external hardware to run the covered code. (Example: DVRs and video game consoles)
If you have to make something as loathsome as a DRM protected anything, too f-ing bad.
I think this misses the point. The premise behind the GPL is encourage, or force, the sharing of code.
I think *you* miss the point. The GPL in no way is trying to tell you what to do with YOUR code. It is, however, insisting that you respect the rights of the people who wrote the GPL code you use.
If you link to GPL code or extend already GPL'd tools and then redistribute you must include the source code.
Yup.
IF YOU'RE JUST GOING TO FIND SOME LOOPHOLE IN THE GPL (BY USING SOCKETS OR SOMETHING) IT KIND OF DEFEATS THE PURPOSE OF THE GPL ENTIRELY.
Not at all and I would bet that RMS himself would agree. I have not defined any such thing as a loophole. I have defined a way that is legally safe and sound for using GPL as a base for your project which is *not* GPL. It allows you to do what you feel you need too while respecting the rights of the people who have offered to share the code so generously.
RMS is about freedom, he is not about forcing anyone to do anything.
In this case its more so that Cisco/Linksys uses GPL'd software in their products and distributes it (via firmware) to its customers. They don't supply the source code for this GPL'd software as required by the GPL when distributing.
I obviously, but it is about taking someone else's work and incorporating it into a product in a way that is counter to the license through which they receive access to the source code.
They have modified someone else's code and, against the license, not shared their changes.
Good call, I remember that discussion.
You give them the key to the ..
In the U.S.A. you don't have to give your keys. The 4th amendment is a wonderful thing.
Yes, as the western world is quickly shedding the rights and freedoms our ancestors fought and died for, it is good to know that our leaders and would-be oppressors are idiots and smart people will be free from prosecution because a fairly well informed 14 year old will be able to hide evidence from the jack-booted inquisitors.
Cisco is being REALLY REALLY stupid here and I just don't understand why.
I've done a lot of commercial software that uses LGPL and GPL code, and its not rocket science. RMS himself even says that "mere aggregation" is not a problem.
Here are the rules:
if its LGPL, link to it, but don't modify it. If you need to modify it, make the modification in the form of a generic API extension and call it from the application. Make your extensions public.
If it is GPL, make it a service and call it through a socket.
If it is a kernel module, there seems to be some wiggle room there, otherwise make a public mini-driver and a proprietary user space app.
How hard is that? Jeez, if you screw up GPL compliance, you are not paying attention.
then how does the Shinsei Bank [linuxinsider.com] of Japan manage (US$118 million net income, first quarter 2008).
Notice too, that it is to manage their CRM product, and *NOT* their financial data.
Produce any evidence that the 'MySQL guys' faked benchmark results, produce a comparison benchmark of MySQL versus a 'commercial' database.
I never said "fake" I said "cook." The implication is that anyone can create a benchmark that makes their product look best. You should not that MySQL can't even do a STANDARD sql benchmark because it does not support enough of the standard SQL features.
troll detect score 5+, keywords fanbois, cook, benchmark
There is a difference between being a troll and being opinionated.
I have a couple decades of experience with databases and have spent a very good amount of my career keeping up with the state of the art. I've used at least one version of pretty much every popular database since Aston Tate introduced dBase. To some, I am considered an expert and am paid for the advice I give. That doesn't mean much on slashdot because "MySQL is good enough."
Interesting. Tell that to Flickr, Facebook, Wikipedia, Google, Nokia and YouTube. Or, how about Slashdot and Digg - capable of bringing down moderately sized web sites with the click of a million mice?
As has been discussed many times over there are a couple points to make out about these examples.
(1) *ALL* of these sites buttress MySQL with support code. Take a look at what Slashdot has to do to enable MySQL to keep up with the sites needs.
(2) None of these sites are mission critical. Would you TRUST your bank transactions on MySQL? LOL, no way!
Now, I'm not a professional DBA. I'm just a programmer
I was a DBA, I am a software engineer. I have published articles about Windows kernel development and data acquisition. I was CTO at a dotcom, I am currently a consultant have worked directly or indirectly with AOL, Microsoft, Sun, and Yahoo.
but I was one of the maintainers of the MySQL server (I don't get to touch the Oracle servers here except on my local developers instance).
I have managed and developed on many Oracle systems. I have also done the same on Sybase, PostgresSQL, MySQL, DB2, MSSQL, mSql, SQLite, Advanced Revilation (Not SQL, but a database), and others.
I have a very good background on the subject.
I can tell you from personal experience that MySQL is easier to maintain and administer,
Than what?
faster to start up, and requires far fewer system resources to keep going.
Than what?
Judging by just the performance of Wikipedia and Facebook, it seems to perform quite well under heavy load.
You are not looking at MySQL, you are looking at an aggregate system whose performance is a product of the development teams ability to work around MySQL.
None of the sites you mention simply execute a query and display the results. Every single one of them requires a lot of extra programming work to serialize and cache load because MySQL sucks.
So, please tell me what basis you have to place MySQL out of the elite top-tier of database servers?
I think I made my point.
Recommending MySQL is just stupid. Not only is its development at Sun in question, it is a poor excuse as a replacement for a commercial database.
As the MySQL fanbois are used to saying, "MySQL is good enough for what I do," a commercial database is held, and rightly so, to a higher standard to which MySQL fails miserably to measure. Yea, sure, the MySQL guys can cook single instance benchmarks that look impressive, but the scalability, reliability, and feature set lack on a professional level.
I don't even need to say which is the better alternative because everyone knows what it is. Since the guy recommends MySQL, this means he didn't evaluate the application space well enough to make the recommendations he does.
The U.S.A. is going to shit. Period, end of sentence. This "zero tolerance" "get tough on [insert noun/verb]" is rediculous.
This has NEVER been a nation of laws and rules, this has historically been a nation of rule breakers.
We have all lost a sense of proportion and reason. Everyone wants everyone else screwed to the wall for the merest infraction. Everyone has become selfish and can no longer tolerate anyone else's exercise of their freedoms if it means the slightest inconvenience or offense.
Sorry everyone, this is not America. The U.S.A. of my youth was an in-your-face nation that depended on strength and a thick skin. Now the crybabies want a nice perfect little safe and oppressive police state.
People would rather feel save in a police state than enjoy the fruits of liberty and the risks that come with it.
Its a few fucking emails. Get the hell over it. Does anyone know what that suspension will cost this person in REAL dollars in tuition?
OK, that does it!!!
Seriously, I think the web has to become an all SSL backbone. Its fast enough through the wire, and computers are fast enough to codec that this nonsense has to stop.
ISPs should not be interfering with content. Period. This will also prevent airports and cafes from inserting their own frames around user content as well.
Over 20 years ago I would at a company called Business and Professional Software. They had a DOS presentation package called "Trumpet" (This was before Trumpet WinSock)
Anyway, I ported the DOS version to Windows. Being a Windows 2.x beta tester and developer, I had access that BPS didn't have yet, so I had the expertise and the knowledge.
I did some very cool stuff in real-mode Windows. When the program printed, I saved a new document that represented what was selected for print, and spawned a new copy of the program to print. Everyone else at the time had printing as a modal operation. I did some cool things with font mapping, kerning, color management, etc. The product was very cool. Personally, I think we were better than Hollywood (look it up) and Powerpoint was a joke. In the infinite wisdom of David Solomont, he hated Windows, and didn't push the project. He only wanted it to back up the DOS and DEC CDE software offering. Alas, so it goes. It would probably make a good base for a Linux offering today as we used AGFA font rendering and did our own layout, we did not use the Windows GUI except as an output device.
Anyway, I had a copy of the code a long time ago. (In case they need to ask me about it or need a fix) It is, however, now long lost. Most code doesn't actively get tossed when it is no longer used, but it becomes less and less likely to be backed up. In the case of companies going out of business, it exists until the last hard disk is crushed in the land-fill.
There's probably more than a million man-years work of code lost every decade.
Ever wonder how all these other people can get it working, and you can't? Every thought it might not be the technology, it might be you? Just asking...
I'm just assuming they are astroturfers.
Yes you can. Roaming User Profiles have existed in Windows for over a decade.
"Roaming Profiles" is typical of Microsoft features in Windows, its a nice little check box for IT managers. The deeper story is that "Roaming Profiles" are difficult to maintain and more work for IT departments.
Since Windows ZAK (Zero Administration Kit) it has been a bolt-on feature that does not work in anything but the strictest way.
You need 1, one, uno, system images. Multiple user directories, and a DHCP server for UNIX type systems and everything will work just like it is supposed too. In Windows, with a special server, login profiles, etc. it still does not work right. The applications don't work right, the default directories don't work, and it is, in general, unusable.
Ok, you had to change the machine name. Maybe with your method you didn't have to do anything at all in Unix. So I guess it is simpler by 15 seconds and a reboot. I now bow down to the superiority of Unix.
You winbois don't get it. UNIX is designed to do this stuff from the start while Windows only has it bolted on.
In your example, all that stuff is handled by "dhcp."
The parent is a typical fanboi post, long on FUD, short on facts.
I'm posting anonymously because I don't want to have people at my company know who I am.
This gives it away, of course.
But it seems to me that Linux while cheap to buy is not cheap to keep patched and secure
Please site some documentation for this statement. It is pure FUD.
The nice thing about Linux however is that a very skillful and thoughtful person can plan out a very robust network and can mange the patches. But it takes effort, dicsipline and an above avegage IT guy. And if you lose that person, you are screwed. Even a new equally skilled guy probably can't get all the scripts and stuff the last guy used to manage to work.
Again, this is pure FUD. *All* of the major distributions have had large management facilities as standard for years. "apt-get update" "apt-get upgrade" is all you need to know on [K]Ubuntu. yum for centos, and rpm for RH.
So why exactly "can't I do it on Windows"? You -do- know that HKEY_CURRENT_USER can be roaming, yes?
OK, here's what you need to do to be real.
You need one, that's right, one, system image that is either replicated and maintained on all the systems or is used to netboot the clients. The image contains all the companies approved and installed applications. This is a HUGE benefit to the IT department as they only have to test and deploy one image at a time.
Any approved machine can netboot (or copy) the system image, mount a home directory and work. Try to do that with Windows. *All* applications must work.
I have set up a number of systems like this, including beowulf clusters. Windows is a complete joke. A pseudo-thin client (netboot with local disk) network made up of UNIX boxes is the simplest of things to manage. All you need are machines that can netboot, a file server with dhcp.
On home network its a little more awkward, but in corporate environment, this is common and easy to do with Windows too
LOL, you can "say" this, but it isn't true.
Its not auto-magical as it is with Linux,
So, you can't do it on Windows.
One of the things that truly sucks about Windows is the registry. Each windows box is its own unique little snowflake, thus impossible to replace easily.
If this is done right, all the configuration is in the user's home ditrectory, probably shared on the network, and the rest of the system is a standard image. That means any user can use any computer and have their system where they want it.
This is no surprise to us UNIX folk, but POWs "Prisoners Of Windows," will love it. Imagine being able to replace/upgrade your computer simply by dropping a new box in front of you. Your settings completely unchanged!!!
I have been doing this with Linux for so long (separate /home disk that persists), I can't believe people still put up with Windows nonsense.
Yes, most of the actual IT people hate MS, mostly because it's shitty and also because it doesn't give them options and they like options.
Ok, so you admit that IT departments generally dislike Microsoft. Good reality check.
Management, even IT management, however, thinks differently.
To a point.
For the techies, another system to support is a challenge. To management, another system to support is another FTE or two, and they don't have that in the headcount.
And there is the HUGE problem of the monoculture, and the one to which knowledgeable people object. Support is an issue when things are *not* standardized. Yea, sure, if you buy everything in microsoft's shitbag, everything works, badly, but works.
When IT departments and vendors try to reduce their support costs by getting something fundamentally less labor intensive to support, it usually backfires because Microsoft intentionally breaks standards or usually has no standard, not even in their own products. So interoperability becomes the labor intensive problem.
The only solution is to eliminate Microsoft from your vendor list and work only with standards. Currently this is difficult, but alternatives are growing. Macintosh, for instance, isn't just for kids anymore. As more and more macs come on-line, even Microsoft will be forced to support them in some way.
It is a slow process to get rid of a monopoly, especially a well funded, contemptible, spiteful, and disgusting entity like Microsoft.