The Toronto Star links to an obscene amount of ads, some of which tried to install software on my system. Visit the site several times, you'll find ads that cover 70% of the middle of the screen, and try to get you to install spyware. So much for the daily noose.
You do admit opensource software in generial is influential. Its slowly replacing AIX, Solaris, and eventually Windows. Certainly more influential than Microsoft itself.
So whats common in all Opensource software beside being opensource? Most of it is developed and run on Linux. FreeBSD's development, a few will argue, is pushed by Linux's success, not to mention the array of software that was developed for Linux and is then compiled on BSD and friends.
Linux itself as the most popular OS in two parts, the kernel and the GNU tools. The quality of these two supports ALL the applications, and had Linus done a much worse job, these applications would have suffered and the industry not paid attention to it. So the Linux kernel and GNU tools are arguable the two most important pieces of software in the IT industry, slightly more 'influential' than Windows, which would be a close third (to put things in perspective, Linux's development is influencing Windows' development more than Windows development influencing Linux's).
Between Linux and GNU, GNU is controlled by a larger number of people, each person becomes less 'influential' than he would be if there were fewer people. Linus alone finally controls what gets into the kernel, which concentrates the development process to one man. Thats what makes him critically important, but not more important than the sum of Linux's developers.
The author goes to pains to find the good in Ballmer. Keywords are man of 'action' and 'energy', and these two words are repeated, with the failure of discovery of another virtue.
Key point is Ballmer's interest in 'innovation'. Goes in line with Microsoft's PR, sounds like there was no research on this man, just interview someone at Microsoft about its CEO.. they'll just repeat the company bottom line.
When I hear 'energy', for some reason reminds me of 'developers, developers, developers'. Makes me proud of Linus' laziness.
It is very wrong to assume some Linux hobbyist will quickly put aside other things in life, to patch a bug like this, which will make companies like Redhat, novell richer.
Opensource does not guarantee quick fixes of bugs. Case in point.. many ATM cards remain in the experimental stage and crash on atmsigd 6 years on. Similar bugs with the arcnet driver will probably never be fixed.
Opensource software is 'generally' better in quality and security, but there are absolutely no guarantees...except theres no SCO code in there.
You have to first see what makes Linux popular. Its not simply because its opensource.
Linux runs on a large array of hardware, has drivers for (almost) everything under the sun, not withstanding ATM cards, has a massive binary base (for x86) now, and has many technicians who can configure it.
If Sun releases Solaris under BSD or GPL type licenses, you still cant run DOOM3 on it, using nvidia drivers, at the same speed. Instead, Solaris' benefits will bleed into Linux (many many posters are saying this), and to a smaller extent Solaris will benefit from Linux, but in its own niche market.
To think of it in a different way, Solaris is a server OS. So it might be in competition with Linux in the server market. But getting an OS for a large scale SMP server, or server farm never depends on whether its opensource. It depends on quality, capability and to a much smaller extent, price. Both OSes excel in the quality dept, Linux will be more capable with an opensource Solaris, but the current Solaris is cheap enough to be free (compare with Microsoft windows Advanced Server).
So even in the server market, Solaris might be ceding ground to Linux by being opensource. It will boost ( I suspect) Sun hardware sales, so Sun doesnt lose, Solaris will gain with benefits from Linux and the community, but will lose a little market share, and Linux will gain. All assuming Solaris is released under a BSD-like or GPL-like license.
Try any P2P system, plus bittorrent. Host it up on edonkey, DC++ kazaa etc, and hand them a link or unique name.
Many home computers have at least one of the P2P app installed. I get invited to many friends' girlfriends houses to install 'something to get music', so you can be sure it'll get delivered.
A knoppix-like live cd of solaris with leaner libraries.
A much reduced-bloat Solaris
Most important: custom compiles of Solaris kernels for speed.
So yeah people who have been using Solaris, and own tonnes of the cheap Sun hardware, will be interested in projects coming from the opensourcing of solaris. Its not about any gaps Linux/BSD left behind, the world of computers is huge, there are plenty of niche areas, not to mention Sun hardware support and some networking technologies which in the Linux kernel refuse to leave the EXPERIMENTAL stage like ATM.
Having solaris 'zones' in Linux in parallel to UML and chroot wouldnt hurt either.
Comeon, you cant expect a huge computer company to just give out a product they worked on for years spending hundereds of millions of dollars on it.
Go ahead, donate 30% of your holdings to some charity. Lets see if you can do that.
Everyone knows why Sun is opensourcing solaris. Noone has the illusion they're so nice theyre giving out the worlds most popular UNIX. Everyone knows they're trying to ride the OSS revolution, just like any other computer company beside Microsoft and SCO. So instead of bashing them for not signing up their life away, lets be happy we've got more OSes to play with.
When Linux was relatively new in 1996, I was looking for any UNIX to install on my system, asking friends, companies universities, everywhere. I couldnt get hold of any OS beside DOS and windows95 until I discovered Linux on infomagic.com. It was awesome. We didnt care about the GPL licence, or the community or Linus' generosity, or the Cathedral and Bazaar. We WERE the bazaar. We just started using and contributing to it in many ways, and Linux is big now because of the initial hobbyists, who cared less about the political correctness of EULAs and more about hacking an OS to death.
I've been downloading Solaris ever since they allowed it for version 8. Been playing with it, learned alot and helps my job. It does have its strong points, which I'll use when I'll need it. Give me another option and I'll be happier. If its not GPL, I'll be relatively less happy but happier nevertheless
Any idea what will happen if you tell all slashdot geeks how much they could be making if they were spammers?
Sure there will always be someone spamming our mailboxes, but put out the bait to the smartest bunch, and youve just made the world a miserable place (at least online).
The govt should post a reward of $700,000 for anyone who seeks and gets enough spammers to reduce online spam by 2% or something. Being on morality's side, greedy slashdotters could then clean up the Internet, at least in western countries.
I was once playing with it, and I DDed my main XFS partition's first 2MB and last 2MB.
So for the next week, I became an expert on XFS structure, and built a fake second superblock from which the fsck copied and made the first superblock, and xfs_repair extracted ALL files. Took a LOT of reading PDFs, source code, compiling XFS and kernels etc, but I got my files back in junk-encoded filenames. I got all files except the ones in root (/) for some reason, and further grepping got me the root files, and the.tar.bz2 backup file I had been seeking.
For that reason alone I always use XFS now. I know how to extract data from it when it crashes, betten than I do for ext2 and others. More importantly, I learned the true value of backups, every seasoned IT guy should go through this.
BeOS was a great OS for its time. Its design is still that of a great OS even now. BeOS is lean and geared towards media, with opengl built in. Its threading system, filesystem were pretty advanced. Think OS X.. only simpler and faster.
Few OSes now are efficient for the desktop, linux and bsd are generalists, and windows is not efficient at all OS X itself comes close, but still not so well designed for being a desktop OS inside out (OS X doesnt have to be. it runs on awesome hardware, and has an awesome kernel+libraries, the most efficient for being a general OS, but not internally designed to be a desktop OS).
Instead of Haiku, I was hoping for changes to the Linux/BSD kernel to make it more desktop-geared like BeOS was. That, and establishing API and package standards should bring Linux/BSD out of the dark ages and onto grandma's and grandkids' computers.
10,000 windmills made a change of 2C 'locally' with its eddies. It did that by disrupting air close to ground. Trees could do that. Mountains could do that. I'm as worried about local temp change as I'm about the change in temperature in the generator of the turbine.
The article also didnt mention how many turbines will it take to cool the arctic and warm the south. Millions?
I believe 10,000 turbines are sufficient to power all Canadian homes and businesses, and will produce far less 'local' temp difference than all Canadian nuclear power plants.
I dont thing Ingres has to take attention away from Postgres and Firebird, maybe just firebird.
I've been looking for an RDBMS for running on OpenBSD and Postgresql doesnt quite cut it with the lack of certain features. Firebird, Ingres are on one level, postgres is on another, mysql and sqlite another. There is still a shortage of opensource RDBMS-scale databases out there. The opensourceness will take these DBs to OpenBSD, FreeBSD and the likes.
Mysql does have lots of attention, but just because it fits nicely on apache webservers, and thats a big market. For a newbie like me, postgres was a bit of a struggle initially (7.3) and I had to worry about tuning it and maintaining it. Mysql's default configs fits the largest market, webservers, so it gets that market regardless of merits.
Now that sqlite is packaged with PHP, I expect to see it grow, being simpler than mysql, it should in theory cut into that market. Postgresql holds a niche, the spot between baby databases like mysql/sqlite and RDBMSes, and I dont think it has a competitor. If Ingres is ported around BSD, it will take a huge lead over the other commercial linux ports.
I have a similar question. I'd like to see the features list, or how advanced each database system is from sqlite to oracle.
I first thought mysql was small, until someone told me its below the level of oracle, sql2000 and sybase. Next I installed postgresql, seemed pretty advanced and learned its features. Then I found out its still one level below the mission critical databasen.
How can we compare these databases beside pure opinion and crawling the PDFs and wondering which features are more important than others? How do we not choose mysql over oracle for running a bank branch?
Zebra and Quagga already exist. They are supposed to provide BGP among other protocols. I just dont get why they dont join those projects to improve them rather than fork out a new one.
Improving the architecture of say Quagga will be more beneficial and probably welcome than forking out your own. It would also keep the code portable while supporting rip, ospf isis etc. I'd love to see a secure version of Quagga for OpenBSD, sounds much better than an all OpenBSD suite.
Articles like these really show what Solaris is all about, it is a corporate UNIX. The N1 grid, all the virtualization technologies (zones for ex), the advanced storage management, java application servers, all are about making management of large-scale technologies easier. This is where Solaris really shines, and thats why they cannot just take Linux and start using it. Linux's SMP is less scalable, threading less mature, and is a target moving too fast. It has major benefits on embedded systems, desktop systems, and a wide array of cpu architectures, but Solaris shines on its own turf.
I however disagree that firewalls are obsoleted with stronger authentication. A firewall protects a network from the 'outside world'. Such an outside world is extremely hostile and will take advantage of ANY bugs, some unknown to you. I've seen attacks on sendmail, ssh, samba, windows netbios protocols, bugs in ipv6 implementation, dictionary attacks on telnet and ssh, all kinds of buffer overflows etc. These problems arent miraculously fixed when your authentication goes 512-bits and requires a password at least 12 characters long. Software will always have bugs, and to protect them, we'll need firewalls which will present a face with far fewer possible bugs than the network inside it.
Security has two major parts, the stability of the system and additional designs to make it secure. Thats why secure OSes/distros do not head for the newest features focusing mainly on auditing the current code increasing its 'maturity'.
This is where the problems of secure Linux distros start. The Linux kernel is loaded with features, alpha and new, along with what makes it stable. Going with 2.6.9 might make a distro susceptible to bugs that will be discovered in the next couple of months, while setting up camp on 2.4.x requires them to do the same with 2.6 later on, while trailing on hardware support including the latest CPUs and chipsets.
Linux's opensourceness gives it an edge over trusted Solaris and trusted AIX, but that exists in BSD too. OpenBSD has kept the 'simple' design over additional security functions which gives it a permanent edge over any secure linux distro, but Linux enjoys the software base OpenBSD doesnt... decisions decisions.
The best thing about a secure Linux distro IMHO is its enormous potential to displace most other secure OSes. IBM has been pushing Linux and will replace trusted AIX with trusted Linux for banks and insurance markets, if Linux has the merits, Sun has already been replacing Solaris with Linux, might as well invest into its security. Windows doesnt stand a chance against any of the above, so Linux gains great market share of the security demands.
I really thing theres a major pressure to create a system that seperates the featureful 'Linux' from the stable and mature, either by version numbers, or as Linus suggested, tags in the modules or options. It might even fork the kernel, but I doubt any branch will risk breaking binary compatibility.
A rudimentary EXPERIMENTAL tag system exists in the sources, but I've had crashes because not all new features are marked so. Been tough deciding for us between Linux and OpenBSD recently, since OpenBSD doesnt really support our preferred NIC and IDE controller, and doesnt run veritas, while Linux has crapped out more than once on the few bad drivers we have to use.
Same here although Ive had no issues with XP running on my PentiumIII 800MHz. Ive played counterstrike source with my geforce4ti card, and dont think I'll need to upgrade till Athlon64 becomes cheaper.
Will I spend $999 on the new Intel chip? Only if they give me a rebate of $999
And who did ever convince you that onboard sound can be good enough for anything?
Even if youre a little bit of an audiophile, you'd be looking at Audigies and Yamahas, and Grado or Sony headphones and the likes. Even the best motherboards out there have sound crappier than a measly Soundblaster Live available for $15 on eBay.
Many people and companies buy Intel because its Intel. Friends who are building computers for themselves asked me to browse motherboard combos, but it should be Intel. Asked why, they couldnt really answer beside "theyre well known".
I remember when the K5, K6 and the K6-2 had issues with certain motherboards, and running Windows95 was a pain, to an extent that you had to swap the CPU with an Intel or Cyrix. That reputation had a lasting legacy on AMD. Later the Athlon came which was awesome for the price/performance ratio but too many people used cheap fans on the 80W+ CPUs which overheated frequently. AMD makes great CPUs and I'd definitely buy AthlonXPs or 64s or Semprons over Intels unless I had a reason to do so. But its hard to argue when non-objective reasons are at play during the decisionmaking.
I'm not sure why HTML isnt sufficient for you. We build an app that had to print boxlabels, HTML fit the bill nicely. Its designed for visual presentation of data, placing of text and graphics, and thats whats happening during printing or viewing for proofing. Ours is a bad example because we didnt have XML-source data but I know HTML would work fine.
CPU performance has become a pissing contest at many places. The latest games do very well with a 2.0GHz chip, the difference in real world performance isnt so bit compared to a 3.0GHz chip. Especially if you consider the cost difference.
Cache is a big deal, think of the duron/athlon difference. Frontside bus is big, 266 and 533 FSB yield different performance metrics. And of course harddrive speeds havent changed much since the ATA100 since the Pentium2 days. Theyve changed, but not as much. Thats another performance-dragging force.
And then theres another little problem plagueing the whole tech industry. People just dont need more power anymore. I'm happily running WindowsXP on a Duron 800MHz with a SCSI Ultra160 cheetah disk and Geforce4Ti card. The games that I do play dont need anything more powerful, maybe Halflife2 will, counterstrike source doesnt.
Most of the time I'm writing, surfing the web, and telneting to the linux and solaris servers, I could do all this with a Pentium1 MMX. So why would even a company with ERP system servers buy the Pentium4 at 4GHz? Our servers are doing well with the 1.4GHz chips, the bottleneck being NICs (being upgraded to gigabit) and disks. That kills the higher-end market for chipmakers, and they will finally come back to low-heat low-cost higher-throughput chips, making laptops more affordable and computers less noisy.
We dont need 4GHz chips, and I dont think Intel plans to release chips with 2mb cache for the desktop market. Moores law can only hold for so long, market forces will bring it down.
1) Since we're getting a LOT more than the minimum, theres a queue of fresh college grads and ex-dotcom-bust employees ready to snatch the job. They have their resumes streaming to the presidents office, so he knows he can replace you without much fanfare, and negotiate better deals with the next one.
2) The environment is nice. You do not deserve a window, period. The humming of the servers are a blessing, you get fewer visitors with problems that way. Youre also relatively lonely this way with the only visits being to get more work to you.
3) The sitting down part is nice until some manager decides to switch two employees from the opposite ends of the company. Go get the trolley errand boy!
Well, I suppose sitting down is indeed nicer than a McJob, but data entry clerks arent kept standing up either. And they do get paid the minimum.
4) This is my favorite one. You dont have to swipe, and will even get a cell phone and pager, on the condition that youll always be next to it. You are also expected to put in overtime hours to keep operations running. With enough leeway to go above the 40 hour week, the employer starts piling up work and responsibility.
So you never really leave sharp at the hour, and head to your family or friends.
Despite all these issues, I still believe the original article or survey was inaccurate. I've spent some time at customer service on the phone, especially with computer technical support on the phone where you help people with everything, not just a product. There are people who cant scroll and only see files starting with A and B. There are many angry people who do not understand why their computers are slow or dont take them where they want.
I'll take managing 100 machines any day over 1000 angry users on the phone.
Slashdot Mirror
The Toronto Star links to an obscene amount of ads, some of which tried to install software on my system. Visit the site several times, you'll find ads that cover 70% of the middle of the screen, and try to get you to install spyware. So much for the daily noose.
Certianly a great number of supporting applications helped, but I wonder where the OSS movement would be today with the Linux kernel.
The OSS movement would be exactly where it is now, since the Linux kernel exists.
Hmm I wonder what would be happening now if the pope was alive.
You do admit opensource software in generial is influential. Its slowly replacing AIX, Solaris, and eventually Windows. Certainly more influential than Microsoft itself.
So whats common in all Opensource software beside being opensource? Most of it is developed and run on Linux. FreeBSD's development, a few will argue, is pushed by Linux's success, not to mention the array of software that was developed for Linux and is then compiled on BSD and friends.
Linux itself as the most popular OS in two parts, the kernel and the GNU tools. The quality of these two supports ALL the applications, and had Linus done a much worse job, these applications would have suffered and the industry not paid attention to it. So the Linux kernel and GNU tools are arguable the two most important pieces of software in the IT industry, slightly more 'influential' than Windows, which would be a close third (to put things in perspective, Linux's development is influencing Windows' development more than Windows development influencing Linux's).
Between Linux and GNU, GNU is controlled by a larger number of people, each person becomes less 'influential' than he would be if there were fewer people. Linus alone finally controls what gets into the kernel, which concentrates the development process to one man. Thats what makes him critically important, but not more important than the sum of Linux's developers.
The author goes to pains to find the good in Ballmer. Keywords are man of 'action' and 'energy', and these two words are repeated, with the failure of discovery of another virtue.
Key point is Ballmer's interest in 'innovation'. Goes in line with Microsoft's PR, sounds like there was no research on this man, just interview someone at Microsoft about its CEO.. they'll just repeat the company bottom line.
When I hear 'energy', for some reason reminds me of 'developers, developers, developers'. Makes me proud of Linus' laziness.
It is very wrong to assume some Linux hobbyist will quickly put aside other things in life, to patch a bug like this, which will make companies like Redhat, novell richer.
..except theres no SCO code in there.
Opensource does not guarantee quick fixes of bugs. Case in point.. many ATM cards remain in the experimental stage and crash on atmsigd 6 years on. Similar bugs with the arcnet driver will probably never be fixed.
Opensource software is 'generally' better in quality and security, but there are absolutely no guarantees.
You have to first see what makes Linux popular. Its not simply because its opensource.
Linux runs on a large array of hardware, has drivers for (almost) everything under the sun, not withstanding ATM cards, has a massive binary base (for x86) now, and has many technicians who can configure it.
If Sun releases Solaris under BSD or GPL type licenses, you still cant run DOOM3 on it, using nvidia drivers, at the same speed. Instead, Solaris' benefits will bleed into Linux (many many posters are saying this), and to a smaller extent Solaris will benefit from Linux, but in its own niche market.
To think of it in a different way, Solaris is a server OS. So it might be in competition with Linux in the server market. But getting an OS for a large scale SMP server, or server farm never depends on whether its opensource. It depends on quality, capability and to a much smaller extent, price. Both OSes excel in the quality dept, Linux will be more capable with an opensource Solaris, but the current Solaris is cheap enough to be free (compare with Microsoft windows Advanced Server).
So even in the server market, Solaris might be ceding ground to Linux by being opensource. It will boost ( I suspect) Sun hardware sales, so Sun doesnt lose, Solaris will gain with benefits from Linux and the community, but will lose a little market share, and Linux will gain. All assuming Solaris is released under a BSD-like or GPL-like license.
Try any P2P system, plus bittorrent. Host it up on edonkey, DC++ kazaa etc, and hand them a link or unique name.
Many home computers have at least one of the P2P app installed. I get invited to many friends' girlfriends houses to install 'something to get music', so you can be sure it'll get delivered.
OK heres what I'm looking forward to.
Better sun hardware support in BSD.
More scalable threading for Linux/BSD
Robust ATM networking for Linux/BSD
Possibly a Solaris for my Alpha 533MHz system.
nVidia drivers for Solaris x86
A knoppix-like live cd of solaris with leaner libraries.
A much reduced-bloat Solaris
Most important: custom compiles of Solaris kernels for speed.
So yeah people who have been using Solaris, and own tonnes of the cheap Sun hardware, will be interested in projects coming from the opensourcing of solaris. Its not about any gaps Linux/BSD left behind, the world of computers is huge, there are plenty of niche areas, not to mention Sun hardware support and some networking technologies which in the Linux kernel refuse to leave the EXPERIMENTAL stage like ATM.
Having solaris 'zones' in Linux in parallel to UML and chroot wouldnt hurt either.
Comeon, you cant expect a huge computer company to just give out a product they worked on for years spending hundereds of millions of dollars on it.
Go ahead, donate 30% of your holdings to some charity. Lets see if you can do that.
Everyone knows why Sun is opensourcing solaris. Noone has the illusion they're so nice theyre giving out the worlds most popular UNIX. Everyone knows they're trying to ride the OSS revolution, just like any other computer company beside Microsoft and SCO. So instead of bashing them for not signing up their life away, lets be happy we've got more OSes to play with.
When Linux was relatively new in 1996, I was looking for any UNIX to install on my system, asking friends, companies universities, everywhere. I couldnt get hold of any OS beside DOS and windows95 until I discovered Linux on infomagic.com. It was awesome. We didnt care about the GPL licence, or the community or Linus' generosity, or the Cathedral and Bazaar. We WERE the bazaar. We just started using and contributing to it in many ways, and Linux is big now because of the initial hobbyists, who cared less about the political correctness of EULAs and more about hacking an OS to death.
I've been downloading Solaris ever since they allowed it for version 8. Been playing with it, learned alot and helps my job. It does have its strong points, which I'll use when I'll need it. Give me another option and I'll be happier. If its not GPL, I'll be relatively less happy but happier nevertheless
Any idea what will happen if you tell all slashdot geeks how much they could be making if they were spammers?
Sure there will always be someone spamming our mailboxes, but put out the bait to the smartest bunch, and youve just made the world a miserable place (at least online).
The govt should post a reward of $700,000 for anyone who seeks and gets enough spammers to reduce online spam by 2% or something. Being on morality's side, greedy slashdotters could then clean up the Internet, at least in western countries.
dd is dangerous.
.tar.bz2 backup file I had been seeking.
I was once playing with it, and I DDed my main XFS partition's first 2MB and last 2MB.
So for the next week, I became an expert on XFS structure, and built a fake second superblock from which the fsck copied and made the first superblock, and xfs_repair extracted ALL files. Took a LOT of reading PDFs, source code, compiling XFS and kernels etc, but I got my files back in junk-encoded filenames. I got all files except the ones in root (/) for some reason, and further grepping got me the root files, and the
For that reason alone I always use XFS now. I know how to extract data from it when it crashes, betten than I do for ext2 and others. More importantly, I learned the true value of backups, every seasoned IT guy should go through this.
They can?
BeOS was a great OS for its time. Its design is still that of a great OS even now. BeOS is lean and geared towards media, with opengl built in. Its threading system, filesystem were pretty advanced. Think OS X.. only simpler and faster.
Few OSes now are efficient for the desktop, linux and bsd are generalists, and windows is not efficient at all OS X itself comes close, but still not so well designed for being a desktop OS inside out (OS X doesnt have to be. it runs on awesome hardware, and has an awesome kernel+libraries, the most efficient for being a general OS, but not internally designed to be a desktop OS).
Instead of Haiku, I was hoping for changes to the Linux/BSD kernel to make it more desktop-geared like BeOS was. That, and establishing API and package standards should bring Linux/BSD out of the dark ages and onto grandma's and grandkids' computers.
10,000 windmills made a change of 2C 'locally' with its eddies. It did that by disrupting air close to ground. Trees could do that. Mountains could do that. I'm as worried about local temp change as I'm about the change in temperature in the generator of the turbine.
The article also didnt mention how many turbines will it take to cool the arctic and warm the south. Millions?
I believe 10,000 turbines are sufficient to power all Canadian homes and businesses, and will produce far less 'local' temp difference than all Canadian nuclear power plants.
I dont thing Ingres has to take attention away from Postgres and Firebird, maybe just firebird.
I've been looking for an RDBMS for running on OpenBSD and Postgresql doesnt quite cut it with the lack of certain features. Firebird, Ingres are on one level, postgres is on another, mysql and sqlite another. There is still a shortage of opensource RDBMS-scale databases out there. The opensourceness will take these DBs to OpenBSD, FreeBSD and the likes.
Mysql does have lots of attention, but just because it fits nicely on apache webservers, and thats a big market. For a newbie like me, postgres was a bit of a struggle initially (7.3) and I had to worry about tuning it and maintaining it. Mysql's default configs fits the largest market, webservers, so it gets that market regardless of merits.
Now that sqlite is packaged with PHP, I expect to see it grow, being simpler than mysql, it should in theory cut into that market. Postgresql holds a niche, the spot between baby databases like mysql/sqlite and RDBMSes, and I dont think it has a competitor. If Ingres is ported around BSD, it will take a huge lead over the other commercial linux ports.
I have a similar question. I'd like to see the features list, or how advanced each database system is from sqlite to oracle.
I first thought mysql was small, until someone told me its below the level of oracle, sql2000 and sybase. Next I installed postgresql, seemed pretty advanced and learned its features. Then I found out its still one level below the mission critical databasen.
How can we compare these databases beside pure opinion and crawling the PDFs and wondering which features are more important than others? How do we not choose mysql over oracle for running a bank branch?
Zebra and Quagga already exist. They are supposed to provide BGP among other protocols. I just dont get why they dont join those projects to improve them rather than fork out a new one.
Improving the architecture of say Quagga will be more beneficial and probably welcome than forking out your own. It would also keep the code portable while supporting rip, ospf isis etc. I'd love to see a secure version of Quagga for OpenBSD, sounds much better than an all OpenBSD suite.
Articles like these really show what Solaris is all about, it is a corporate UNIX. The N1 grid, all the virtualization technologies (zones for ex), the advanced storage management, java application servers, all are about making management of large-scale technologies easier. This is where Solaris really shines, and thats why they cannot just take Linux and start using it. Linux's SMP is less scalable, threading less mature, and is a target moving too fast. It has major benefits on embedded systems, desktop systems, and a wide array of cpu architectures, but Solaris shines on its own turf.
I however disagree that firewalls are obsoleted with stronger authentication. A firewall protects a network from the 'outside world'. Such an outside world is extremely hostile and will take advantage of ANY bugs, some unknown to you. I've seen attacks on sendmail, ssh, samba, windows netbios protocols, bugs in ipv6 implementation, dictionary attacks on telnet and ssh, all kinds of buffer overflows etc. These problems arent miraculously fixed when your authentication goes 512-bits and requires a password at least 12 characters long. Software will always have bugs, and to protect them, we'll need firewalls which will present a face with far fewer possible bugs than the network inside it.
Security has two major parts, the stability of the system and additional designs to make it secure. Thats why secure OSes/distros do not head for the newest features focusing mainly on auditing the current code increasing its 'maturity'.
This is where the problems of secure Linux distros start. The Linux kernel is loaded with features, alpha and new, along with what makes it stable. Going with 2.6.9 might make a distro susceptible to bugs that will be discovered in the next couple of months, while setting up camp on 2.4.x requires them to do the same with 2.6 later on, while trailing on hardware support including the latest CPUs and chipsets.
Linux's opensourceness gives it an edge over trusted Solaris and trusted AIX, but that exists in BSD too. OpenBSD has kept the 'simple' design over additional security functions which gives it a permanent edge over any secure linux distro, but Linux enjoys the software base OpenBSD doesnt... decisions decisions.
The best thing about a secure Linux distro IMHO is its enormous potential to displace most other secure OSes. IBM has been pushing Linux and will replace trusted AIX with trusted Linux for banks and insurance markets, if Linux has the merits, Sun has already been replacing Solaris with Linux, might as well invest into its security. Windows doesnt stand a chance against any of the above, so Linux gains great market share of the security demands.
I really thing theres a major pressure to create a system that seperates the featureful 'Linux' from the stable and mature, either by version numbers, or as Linus suggested, tags in the modules or options. It might even fork the kernel, but I doubt any branch will risk breaking binary compatibility.
A rudimentary EXPERIMENTAL tag system exists in the sources, but I've had crashes because not all new features are marked so. Been tough deciding for us between Linux and OpenBSD recently, since OpenBSD doesnt really support our preferred NIC and IDE controller, and doesnt run veritas, while Linux has crapped out more than once on the few bad drivers we have to use.
Same here although Ive had no issues with XP running on my PentiumIII 800MHz. Ive played counterstrike source with my geforce4ti card, and dont think I'll need to upgrade till Athlon64 becomes cheaper.
Will I spend $999 on the new Intel chip? Only if they give me a rebate of $999
Why do you want onboard sound.
And who did ever convince you that onboard sound can be good enough for anything?
Even if youre a little bit of an audiophile, you'd be looking at Audigies and Yamahas, and Grado or Sony headphones and the likes. Even the best motherboards out there have sound crappier than a measly Soundblaster Live available for $15 on eBay.
Go figure.
Many people and companies buy Intel because its Intel. Friends who are building computers for themselves asked me to browse motherboard combos, but it should be Intel. Asked why, they couldnt really answer beside "theyre well known".
I remember when the K5, K6 and the K6-2 had issues with certain motherboards, and running Windows95 was a pain, to an extent that you had to swap the CPU with an Intel or Cyrix. That reputation had a lasting legacy on AMD. Later the Athlon came which was awesome for the price/performance ratio but too many people used cheap fans on the 80W+ CPUs which overheated frequently. AMD makes great CPUs and I'd definitely buy AthlonXPs or 64s or Semprons over Intels unless I had a reason to do so. But its hard to argue when non-objective reasons are at play during the decisionmaking.
I'm not sure why HTML isnt sufficient for you. We build an app that had to print boxlabels, HTML fit the bill nicely. Its designed for visual presentation of data, placing of text and graphics, and thats whats happening during printing or viewing for proofing. Ours is a bad example because we didnt have XML-source data but I know HTML would work fine.
CPU performance has become a pissing contest at many places. The latest games do very well with a 2.0GHz chip, the difference in real world performance isnt so bit compared to a 3.0GHz chip. Especially if you consider the cost difference.
Cache is a big deal, think of the duron/athlon difference. Frontside bus is big, 266 and 533 FSB yield different performance metrics. And of course harddrive speeds havent changed much since the ATA100 since the Pentium2 days. Theyve changed, but not as much. Thats another performance-dragging force.
And then theres another little problem plagueing the whole tech industry. People just dont need more power anymore. I'm happily running WindowsXP on a Duron 800MHz with a SCSI Ultra160 cheetah disk and Geforce4Ti card. The games that I do play dont need anything more powerful, maybe Halflife2 will, counterstrike source doesnt.
Most of the time I'm writing, surfing the web, and telneting to the linux and solaris servers, I could do all this with a Pentium1 MMX. So why would even a company with ERP system servers buy the Pentium4 at 4GHz? Our servers are doing well with the 1.4GHz chips, the bottleneck being NICs (being upgraded to gigabit) and disks. That kills the higher-end market for chipmakers, and they will finally come back to low-heat low-cost higher-throughput chips, making laptops more affordable and computers less noisy.
We dont need 4GHz chips, and I dont think Intel plans to release chips with 2mb cache for the desktop market. Moores law can only hold for so long, market forces will bring it down.
Nice answer.
1) Since we're getting a LOT more than the minimum, theres a queue of fresh college grads and ex-dotcom-bust employees ready to snatch the job. They have their resumes streaming to the presidents office, so he knows he can replace you without much fanfare, and negotiate better deals with the next one.
2) The environment is nice. You do not deserve a window, period. The humming of the servers are a blessing, you get fewer visitors with problems that way. Youre also relatively lonely this way with the only visits being to get more work to you.
3) The sitting down part is nice until some manager decides to switch two employees from the opposite ends of the company. Go get the trolley errand boy!
Well, I suppose sitting down is indeed nicer than a McJob, but data entry clerks arent kept standing up either. And they do get paid the minimum.
4) This is my favorite one. You dont have to swipe, and will even get a cell phone and pager, on the condition that youll always be next to it. You are also expected to put in overtime hours to keep operations running. With enough leeway to go above the 40 hour week, the employer starts piling up work and responsibility.
So you never really leave sharp at the hour, and head to your family or friends.
Despite all these issues, I still believe the original article or survey was inaccurate. I've spent some time at customer service on the phone, especially with computer technical support on the phone where you help people with everything, not just a product. There are people who cant scroll and only see files starting with A and B. There are many angry people who do not understand why their computers are slow or dont take them where they want.
I'll take managing 100 machines any day over 1000 angry users on the phone.