The HTML I type for websites sometimes has a <!-- --> comment, but usually something funny rather than my own name or address. I never leave ad advertising mark unlike many admins with my name websites and other information. Usually the webmaster or postmaster@whatever.tld points to my common but spam-filled free email (yahoo/hotmail) and people can still contact me with a quick reply latency.
I would agree totally. Some dedicated servers running Linux are less than $80 per month and that includes the bandwidth. With that budget, youre getting an industrial strength server with more bandwidth a collection of busy halflife servers will need. Just make sure you keep the root account close and that you dont have to access the server. Some dedicated services Ive seen have a web-based reboot mechanism and you can keep a redhat install CD in the cdrom drive. Then you can start playing with it.
Use them Sun Ultra machines from eBay. Theyre pretty cool, runs 64-bit although the kernel if I remember runs at 32-bit and the CPU has plenty of registers and can give you good IO. This will also introduce a non-Intel platform to your friends.
Another non-Intel option is a power macintosh G5. This beast is also 64-bit and is the most powerful desktop machine around. It could later be used for other educational stuff if people lose interest in the club or it is liquidated. I'm not sure if you can stably run Linux on it.
The bottom line is I recommend you get a non-intel platform for educational purposes.
We've all used pentium1 AT cases as routers with patchy Linux installations and uptimes of 50%. However if an x86 machine is designed as a router and used with a stable version of Linux and its applications, similar uptimes can be reached as cisco routers.
For instance, tweaking the PC bios to boot faster, with fewer options, removing legacy keyboard, PS2, audio etc ports, using a fanless CPU and chipset and using stable application versions will achieve those 9s. If a company can sit behind such a project and offer standardized interfaces to the Linux distro and the hardware, we can break free from the chains of IOS. An even better solution is to develop a free IOS-emulating software for Linux for these routers.
Remember the cisco 700 series were based on the 386 chip, only with IOS running on it. Linux 2.4.21 and FreeBSD 4.8 are no less stable.
Although I cant reach the website, its slashdotted now, Ive been waiting for such computers. I intend to replace cisco routers on many levels with these if they have available PCI slots.
I think the crashing PC prices will harm the cisco market and might spin off PC based router companies. For this reason, Cisco is focusing on management technologies that cannot be replaced by simply replacing that router. Web-frontends for management software that can manage routers and switches via SNMP and proprietary protocols, and other protocols like the CPD that will become indispensible and will make it hard to go from a $2500 router to a better $200 pc-router.
And for that reason, there is great potential for free/opensource management software as well as its cliet stubs for Linux/FreeBSD routers firewalls and other SNMP devices. Theres also great potential for an IOS emulation app for Linux/BSD.
I'm just amazed at how an operating system can run on mainframes and pdas, emulate the binaries of many OSes, have all the functions of any other OS and challenge Sun, Microsoft, Cisco and game console markets in one blow.
It would be nice to see a legacy-free Via CPU machine with say 2 free PCI slots on a mini-ITX motherboard, something thats much smaller than the ATX cases. If it had two on board ethernets to begin with, and 4 USB ports, but no keyboard, mouse, sound ports, the whole thing would be simplified much further(like some IBM aptiva cases). I think the cost could be brought lower by using Slackware or debian rather than lindows. Also ideally it would come with an external CDROM drive that wouldnt sit dormant for most of the time taking space.
Even in production environments, Ive tried to avoid cisco routers since the pentium1 system that people throw away beats the ciscos performance. It would be nice to see modular, possibly non-IDE (pcmcia flash drive) legacy-free x86 computers that are designed to run as routers using FreeBSD or Linux, hopefully in 1U form factor and with a CPU that can run under a heatsink alone (Transmeta?).
Now if they came with the server-type hot-plug PCI slots that are easily accessible with the useage of a large variety of tokenring, gigabit ethernet, voip, atm etc, we have something that can beat cisco easy in performance, price and market. There are just as many linux command-line gurus as IOS experts out there if not more and software that can emulate IOS on Linux can make its developers a fortune.
For the ones among us who do not have an XBox; why should the XBox be modded anyway? Once the CD is inserted and binaries run off it, Linux should just feel easy in an x86 surrounding and take on all the interrupts and memory ranges like in a PC but without a BIOS. Ive heard of Linux running on x86 System-on-chips with no BIOS (therefore not PC compatible).
Or is it that the most privileged level in the CPU belongs to a small program that makes SURE Linux is not running, yet doesnt give a performance hit to the games.
I become engrossed in those depending on the season. Socializing with non-geeks can also be interesting but not for long periods of time.
I play with the cat a lot, and play the piano but those interests are pretty much on and off.
I think many geeks are into photography for some reason. Photography in itself is very much an artistic endeavor and so is playing the piano or violin (not many geeks here)... nothing to do with computers except many of the techies love to put up good images on their websites and as wallpapers.
I understand and agree with you. I enjoy programming the PICs and the 68000s sometimes for their simplicity and feel of absolute control that we had in the turboc and gwbasic days.
That post was referring to higher embedded devices where engineers would contemplate Linux vs other 'Operating Systems'. Even for digital cameras, MP3 players and watches, networking can be a useful feature but not a requirement. Companies can use linux without the networking parts in tiny devices like watches and let more rookie (read: low wage) developers make advanced applications using the already available syscontrols, binaries etc. This doesnt of course apply to $10 watches that do not need more than 4-bit logic to drive a simple LCD.
People use many types of access from various ISPS to login, using computers from a wide variety of manufacturers. Most of them use Microsoft operating systems which is a danger to the whole computing world. Microsofts software is increasingly unstable, bloated and generally low-quality. Their monopoly allows them to be lax with testing and quality control.
The same principle applies to search engines. Almost everyone uses google or yahoo to seach for anything. These two sites have become the very interface to the Internet. This also allows them to alter information (place pro-republican sites above pro-democracy for 'election' search), snoop (with the FBI they might already be doing this), and in the long run suffering the same quality control fate as Microsoft. People build a view of the world around them by exploring and communicating with the other people. The Internet allows people to talk to other people far away and share political and moral opinions which on a larger scale helps tolerance and peace itself. These are not small issues; how many politicians can anyone think of who do NOT get their information from the Internet?
More search engines will increase the diversity and break the stronghold of google. Google is a single point of failure for the Internet (the only other one is the DNS servers system) for most Internet users. Although I use it and love it, we are giving one company too much control, while knowing what the results of that are. I do doubt Microsoft can cut it as a competitor there since Ive never used MSN, and Ive seen their success with the XBox and other home-media entertainment ventures. Other skilled companies however can bring a fresh search interface to the online world.
The government is responsible for spending the least, running the most efficient and approrpriate software and keeping its systems robust. Now there are small areas where Solaris and Windows are more appropriate than Linux.
Case in point. Certain types of networking betweenn Tokenring and Ethernet do not work stably in any version of Linux or the BSDs. If such a requirement arises, the government should not let its systems become unstable or expensive when they can use Solaris. Another example: many specialized inventory management, ERP and financial applications are only sold for Windows. If such a software is required, it must be run on windows given thats the least cost route. It is in the best interest of the people and the government itself, rather than sanction a team of programmers to try and duplicate that functionality for Linux or just a team of accountants to make up for the work manually.
Of course for firewalls, other networking, high security, web and mail services and other applications, Linux or the BSDs are a better option than AIX, Irix, Solaris, Plan9, Xenix, QNX, BeOS, AtheOS, CP/M, OS/390, OS/400 or Windows 2003. Discrimiation of software licenses in the government will incur high costs that the people would not want to pay, and will harm the security, stability and efficiency of the government. Governments should therefore use free software as much as possible but not exclusively.
In fact while compiling the kernel you can remove all networks support and just use the memory manager, filesystems, security and the binaries like perl and bash.
Most embedded items DO require networking though. Linux is also a very well-known development platform now. You can whip up a new microcontroller with new USB/PCMCIA/bluetooth etc chips and make it boot linux off flash. Most of the development can be handled by anyone who has done system development on Linux which could have been on a PDA/Game console/router/PC/midrange. You no longer need someone specializing in Arm7 with detailed experience in the 8139 realtek chip and intel 802.11 chips, just someone who can make drivers for the linux kernel.
Linux has been widely ported around the town and finding a lowcost CPU that can run Linux (and includes an MMU) is easy.. so theres less need for the ucLinux or other exotic forks. Plain Linux will work well and you in one swoop have drivers for almost any networking or multimedia chip made.
I couldnt completely eliminate FreeBSD or Linux from my OS options while installing servers, each had strengths I needed.. until I ran into Solaris.
There are cards that have drivers for Solaris but not Linux or FreeBSD. There are tools and functionality in Solaris that I couldnt find in the others and had to use Solaris for some things.. such as SNAT between an ethernet and tokenring card. All linux kernels crashed after a while of doing this and so did FreeBSD with its lone Olicom driver support. Solaris held on. In other places, FreeBSDs performance just beat Solaris out of the water (no Java thrashing the harddisk like its Baghdad).
I can understand why theres little enthuisiasm for Solaris around the slashdot circles, its not free and it makes money for someone else. But just as we laugh at Microsofts lack of quality, we must laud a good product. Theres no denying the graphics capabilities of OSX and Irix, network and system admin tools of Solaris, huge features and ports of Linux and rock hard stability and efficiency of BSD.
Solaris especially makes a great package with a 64-bit Sparc CPU like the cheap used Ultra 5 systems. They still need to work on the scheduler, IDE speeds and hopefully theyll put virtual terminals back in Solaris10.
I havent used AIX, HPUX or OS/360/390/400. Does anyone know of their strengths and edge over other OSes?
Trolltech is a high-potential company with a bright future. The QT toolkit is the best thing around for clean fast and portable progamming. Trolltech is right to push QT to permeate across the world to reap its profits; they deserve it.
QT has given Linux alot. KDE became so big that GNOME had to be created as a free alternative before QT/X11 became GPLed. Now the Apple port will not only help apple applications, they will help Linux applications giving them more weight. Theres suddenly another big reason to shift your entire software project to QT despite any costs.
My only gripe is the really high license cost for a student. Ive built several applications in win32 but cant use them afer the 30 days. They relied heavily on printing so I couldnt port them to Linux. I even offered developers with the license to compile them for me for a small fee. I hope Trolltech sees this and if they really want to hide their code from pirates, provide a compilation service at a much lower cost for projects with low earning potential or value. I dont mind being the Toronto office manager of compilation services at all. Will even code for food(hey its 2003, not 1998)
Managing a volunteering proj is an art.
on
Paying for Volunteers?
·
· Score: 2, Insightful
Volunteering is when someone wants to support your cause. If youre profiting from the game and youre name alone is stamped on it, dont expect real Volunteering.
Now Volunteering is very much a community thing. Say you have a project for a cause that needs work and you know sincere programmers are out there. It will be best to advertise to local programmers since interacting with your neighbors and meeting them face to face increases sincerety to the cause. It can also bind them to work longer on the project. Ideally, setup a place where developers can physically come and sit on some workstations and develop. They will love the interaction with other developers from around and coke/chips offerings will complete the volunteering setup, getting the work done. Just dumping the work on someone remotely doesnt work on the sincerest of volunteers.
Now opensource software programming can be different. If it is something like gcc or the Linux kernel, the desire is globally strong enough for people to flock to it and submit patches regularly from 8 time zones away. And then, theyre working on a HUGE project with huge effects in the (computer) society. Their code will be used on tens of millions of computers. Thats the real motivation and the real itch. If you can duplicate that, or show the importance of your project to the community, you can expect help. Samba is pretty important. Everyone is using linux along with windows, and linux better look good. KDE is important. It shows the power of Linux on the desktop. Next are games projects like SDL, crystal space and mesa3d. Games are among the last remaining reasons to keep a windows partitions and is attracting a lot of volunteer programmers now. Keep a clean structure, an open environment, good responsive mailing lists, make great demos, always keep complete documentation and build 3d model/image/map import filters. Do not stamp your name all over the place (geeks are more egoistic than the rest), and that should entice a threshold number of developers.
Too much misinformation in the interview is hurting my head now. Just look at the very last line.. once code has been placed under GPL it cannot be altered! But alteration is what GPL is all about.. to take the code and change it.
They only showed PART of the code, that she recognised as Sys V 4.1 vs linux kernel 2.4. It was from no region of the source.. just here and there. Could even be a line that goes/*TODO*/ or #include<kernel.h> She also does not have experience with looking at sourcecode. She instead carried with her a tech guy. Where was he from? Who was paying him? whats his name? How much development experience does he have? Did he do a cmp or diff on the code blocks in question? Where is his interview?
Even IBMs customers I suppose are getting a gist of whats happening here. Only the most nervous and illinformed managers will be moved to Windows 2003. They will be deserving it.
Re:Do younger minds absorb quicker?
on
Ageism in IT?
·
· Score: 1
Finding a pianist who is a musician...that is a rare and precious find indeed
I dont know about down there, but up here in Toronto most pianists I know are actually musicians. Maybe all pianists who are musicians as well move to Toronto.
Re:Do younger minds absorb quicker?
on
Ageism in IT?
·
· Score: 1
How does number of years of industry experience sound? I know more about admining Linux than any admin Ive personally met, most of them working and getting paid for it. I havent gotten one interview in the past year... got very little industry experience despite the pile of certs and the fact that I'd love to goto an interview and get tested personally. Age is what matter here. Although senior technicians might feel theyre not up to speed the way they should be, theyre actually working while 6 months ago I was at a gas station during the day and was setting up IPSec VPn between Solaris and FreeBSD at home using Tokenring and Arcnet. Id much rather right now lose 3 years of my real experience to gain 3 years of paper experience. I'd be doing much better.
Yes. Releasing the patches first should alleviate the bandwidth problem. At least there should be a script in the kernel sources that downloads the latest patches with wget and applies them, also changes the directory name to reflect it. Such ease should convince people not to use 20 times the bandwidth they really need.
To be fair anyway, I admit I'm downloading the full sources:)
I agree with you completely. IBM has spent over 1 billion dollars on Linux. Theyre a company of executives. But sun is a company of geeks and theyve carried UNIX itself forward with enormous help to BSD and Linux. On one hand you have the feel of knoppix and redhat. Nice slick and easy. On the other hand, its a very different feeling using Solaris on the Ultra 5. Simple, robust, standard, with a heavy and dedicated company behind it. Not to mention you get more jobs after learning Solaris. I hated Sun a while ago because everyone else in the Linux community hated them, but then, I tried it. Now I'm typing from an Ultra 5 and loving the feeling and stability.
All the more reason to get Sun certifications eh? The demand for Sun experts is greater than Linux at present and too many companies dont want to tear away from one *nix to another.
And what if all the unemployed Linux techies get Sun certs and work as their admins? There will be enough Sun techies going about to support the stations, no need for switching to Linux in the first place.
This balance is called market momentum and Sun has it. By giving out J2EE specs, free Java, cheap Solaris and other stuff, Sun knows it is increasing its own market momentum.
OK I understand your position very well. I have less than 2 years of Solaris experience myself, and been using Linux since 95.
Here in Toronto, jobs for Solaris beat jobs for Linux, so I had to learn it. I downloaded Solaris 8 for x86 and it looked crap. no VTs, not much software, slow and cumbersome with packages, management software and its X server.
But internally the technology is very well designed. I bought an Ultra 5 for running Solaris 9 SPARC and it is really rock stable. There are things you will see in Solaris that beat Linux, like standards. Every company out there knows theres only one standard Solaris distro, and packages (and binary drivers) for it always install with no problems. Solaris has drivers for products linux doesnt, since Suns a proprietary company and can communicate with the other companies. Software makers have known clients who need UNIX and are willing to pay for it, are using Solaris, so theres a great deal of good proprietary software for Solaris (in many places they beat the free offerings in Linux/BSD). Suns compiler beats GCC 3.2 too.
Now I know the performance of FreeBSD and gentoo beats solaris clear. I personally use FreeBSD in all of my installed servers for clients. But I have gained new respect in the past year trying out various features of Solaris that seem advanced, standard-based and very robust compared to Linux.. such as IPSec VPNs, SNAT between ethernet, tokenring and arcnet. Sun is not as omnipresent on the Net as Linux. It is omnipresent in the corporate.
Re:Same thing was said about Apple
on
Sun's Last Stand
·
· Score: 1
Unlike Unisys, Sun offers a certain kind of good feeling with their OS and hardware.. you know.. the feeling that this product is rock stable, designed very well for UNIX and has good value. Linux on x86 is a good contender but depending on the whims of many IT directors, Sun will prevail for its name alone. Pointy haired bosses will (hopefully) at least know this was the most popular UNIX company, created Java and J2EE, spun off veritas, has awesome 64-bit cheap platforms and is run by true geeks who love the technology. Sun has pioneered so much technology used now in Linux/BSD/M$ and the rest of the world. That knowledge itself is good enough to depend on Sun (on its excellence, not survival. Thats being debated). Sun hardware is cheap nowadays and jobs for Solaris technicians exceed jobs for Linux. So unemployed Linux technicians will keep getting Solaris certs, and later as decision makers will choose a heavier company (opposed to RedHat or SuSE) to make the management happy. Solaris is also a very standard platform in itself, while Linux is struggling with standards. RedHat isnt even LSB compliant and many companies are packing products as redhat rpms only. Sun has a good market mass like Apple only better, and will stick around although in a bad shape.
Youre very right about that idea. Look at the sales of Ultra workstations on ebay. People need cheap machines will definitely buy them,,, look at el cheapo ECS and eMachines computers out there selling like hotcakes.
Sun needs to license out their UltraSPARC motherboards to low-cost Indian and Taiwanese companies who will sell combo motherboards for ATX or miniATX form factors. Sun should really add a good SCSI chip and possibly 802.11 or gigabit ethernet chips if they can get some cheap and add those to the mobos. The result might be Ultrasparcs with good SCSI and networking at less than $500. What geek would NOT want such little beasts for their home systems, not to mention Linux desktop companies might prefer them over x86 (possibly, but I dont know).
At the very least Sun should release specs to the Ultrasparc to the OpenBSD and other teams. Like Java, Sun needs to get people more involved with their products even against the risk of people making it off with cheap products instead of combos. IBM is making money off J2EE, but that only improves Suns position being in the center of gravity of J2EE and therefore always being compared with websphere and weblogic. The same should be applied to software, OS and hardware by Sun.
If I were the CEO of Sun, I would release all of Solaris and Forte under the GPL, try to incorporate their good features into the Linux kernel, and use the linux kernel in either x86 based machines or ultrasparcs as cheap reliable solutions. This is better than just dying out or getting bought out by someone who will just shelve these technologies and put Suns engineers to other use.
Wonder what took them to epoch
on
Sun Opens Java.net
·
· Score: 4, Interesting
... to provide community sites for java? And java.net sounds like java.NET. I think there should be a seperate community site for J2EE systems with a code repository, forums and recommendations, or am I ignorant of such a site?
Slashdot Mirror
The HTML I type for websites sometimes has a <!-- --> comment, but usually something funny rather than my own name or address. I never leave ad advertising mark unlike many admins with my name websites and other information. Usually the webmaster or postmaster@whatever.tld points to my common but spam-filled free email (yahoo/hotmail) and people can still contact me with a quick reply latency.
Oh yeah, C++ source code is a different game...
I would agree totally. Some dedicated servers running Linux are less than $80 per month and that includes the bandwidth. With that budget, youre getting an industrial strength server with more bandwidth a collection of busy halflife servers will need. Just make sure you keep the root account close and that you dont have to access the server. Some dedicated services Ive seen have a web-based reboot mechanism and you can keep a redhat install CD in the cdrom drive. Then you can start playing with it.
Use them Sun Ultra machines from eBay. Theyre pretty cool, runs 64-bit although the kernel if I remember runs at 32-bit and the CPU has plenty of registers and can give you good IO. This will also introduce a non-Intel platform to your friends.
Another non-Intel option is a power macintosh G5. This beast is also 64-bit and is the most powerful desktop machine around. It could later be used for other educational stuff if people lose interest in the club or it is liquidated. I'm not sure if you can stably run Linux on it.
The bottom line is I recommend you get a non-intel platform for educational purposes.
We've all used pentium1 AT cases as routers with patchy Linux installations and uptimes of 50%. However if an x86 machine is designed as a router and used with a stable version of Linux and its applications, similar uptimes can be reached as cisco routers.
For instance, tweaking the PC bios to boot faster, with fewer options, removing legacy keyboard, PS2, audio etc ports, using a fanless CPU and chipset and using stable application versions will achieve those 9s. If a company can sit behind such a project and offer standardized interfaces to the Linux distro and the hardware, we can break free from the chains of IOS. An even better solution is to develop a free IOS-emulating software for Linux for these routers.
Remember the cisco 700 series were based on the 386 chip, only with IOS running on it. Linux 2.4.21 and FreeBSD 4.8 are no less stable.
Although I cant reach the website, its slashdotted now, Ive been waiting for such computers. I intend to replace cisco routers on many levels with these if they have available PCI slots.
I think the crashing PC prices will harm the cisco market and might spin off PC based router companies. For this reason, Cisco is focusing on management technologies that cannot be replaced by simply replacing that router. Web-frontends for management software that can manage routers and switches via SNMP and proprietary protocols, and other protocols like the CPD that will become indispensible and will make it hard to go from a $2500 router to a better $200 pc-router.
And for that reason, there is great potential for free/opensource management software as well as its cliet stubs for Linux/FreeBSD routers firewalls and other SNMP devices. Theres also great potential for an IOS emulation app for Linux/BSD.
I'm just amazed at how an operating system can run on mainframes and pdas, emulate the binaries of many OSes, have all the functions of any other OS and challenge Sun, Microsoft, Cisco and game console markets in one blow.
It would be nice to see a legacy-free Via CPU machine with say 2 free PCI slots on a mini-ITX motherboard, something thats much smaller than the ATX cases. If it had two on board ethernets to begin with, and 4 USB ports, but no keyboard, mouse, sound ports, the whole thing would be simplified much further(like some IBM aptiva cases). I think the cost could be brought lower by using Slackware or debian rather than lindows. Also ideally it would come with an external CDROM drive that wouldnt sit dormant for most of the time taking space.
Even in production environments, Ive tried to avoid cisco routers since the pentium1 system that people throw away beats the ciscos performance. It would be nice to see modular, possibly non-IDE (pcmcia flash drive) legacy-free x86 computers that are designed to run as routers using FreeBSD or Linux, hopefully in 1U form factor and with a CPU that can run under a heatsink alone (Transmeta?).
Now if they came with the server-type hot-plug PCI slots that are easily accessible with the useage of a large variety of tokenring, gigabit ethernet, voip, atm etc, we have something that can beat cisco easy in performance, price and market. There are just as many linux command-line gurus as IOS experts out there if not more and software that can emulate IOS on Linux can make its developers a fortune.
For the ones among us who do not have an XBox; why should the XBox be modded anyway? Once the CD is inserted and binaries run off it, Linux should just feel easy in an x86 surrounding and take on all the interrupts and memory ranges like in a PC but without a BIOS. Ive heard of Linux running on x86 System-on-chips with no BIOS (therefore not PC compatible).
Or is it that the most privileged level in the CPU belongs to a small program that makes SURE Linux is not running, yet doesnt give a performance hit to the games.
Modders please enlighten us.
I become engrossed in those depending on the season. Socializing with non-geeks can also be interesting but not for long periods of time.
I play with the cat a lot, and play the piano but those interests are pretty much on and off.
I think many geeks are into photography for some reason. Photography in itself is very much an artistic endeavor and so is playing the piano or violin (not many geeks here)... nothing to do with computers except many of the techies love to put up good images on their websites and as wallpapers.
I understand and agree with you. I enjoy programming the PICs and the 68000s sometimes for their simplicity and feel of absolute control that we had in the turboc and gwbasic days.
That post was referring to higher embedded devices where engineers would contemplate Linux vs other 'Operating Systems'. Even for digital cameras, MP3 players and watches, networking can be a useful feature but not a requirement. Companies can use linux without the networking parts in tiny devices like watches and let more rookie (read: low wage) developers make advanced applications using the already available syscontrols, binaries etc. This doesnt of course apply to $10 watches that do not need more than 4-bit logic to drive a simple LCD.
People use many types of access from various ISPS to login, using computers from a wide variety of manufacturers. Most of them use Microsoft operating systems which is a danger to the whole computing world. Microsofts software is increasingly unstable, bloated and generally low-quality. Their monopoly allows them to be lax with testing and quality control.
The same principle applies to search engines. Almost everyone uses google or yahoo to seach for anything. These two sites have become the very interface to the Internet. This also allows them to alter information (place pro-republican sites above pro-democracy for 'election' search), snoop (with the FBI they might already be doing this), and in the long run suffering the same quality control fate as Microsoft. People build a view of the world around them by exploring and communicating with the other people. The Internet allows people to talk to other people far away and share political and moral opinions which on a larger scale helps tolerance and peace itself. These are not small issues; how many politicians can anyone think of who do NOT get their information from the Internet?
More search engines will increase the diversity and break the stronghold of google. Google is a single point of failure for the Internet (the only other one is the DNS servers system) for most Internet users. Although I use it and love it, we are giving one company too much control, while knowing what the results of that are. I do doubt Microsoft can cut it as a competitor there since Ive never used MSN, and Ive seen their success with the XBox and other home-media entertainment ventures. Other skilled companies however can bring a fresh search interface to the online world.
The government is responsible for spending the least, running the most efficient and approrpriate software and keeping its systems robust. Now there are small areas where Solaris and Windows are more appropriate than Linux.
Case in point. Certain types of networking betweenn Tokenring and Ethernet do not work stably in any version of Linux or the BSDs. If such a requirement arises, the government should not let its systems become unstable or expensive when they can use Solaris. Another example: many specialized inventory management, ERP and financial applications are only sold for Windows. If such a software is required, it must be run on windows given thats the least cost route. It is in the best interest of the people and the government itself, rather than sanction a team of programmers to try and duplicate that functionality for Linux or just a team of accountants to make up for the work manually.
Of course for firewalls, other networking, high security, web and mail services and other applications, Linux or the BSDs are a better option than AIX, Irix, Solaris, Plan9, Xenix, QNX, BeOS, AtheOS, CP/M, OS/390, OS/400 or Windows 2003. Discrimiation of software licenses in the government will incur high costs that the people would not want to pay, and will harm the security, stability and efficiency of the government. Governments should therefore use free software as much as possible but not exclusively.
In fact while compiling the kernel you can remove all networks support and just use the memory manager, filesystems, security and the binaries like perl and bash.
Most embedded items DO require networking though. Linux is also a very well-known development platform now. You can whip up a new microcontroller with new USB/PCMCIA/bluetooth etc chips and make it boot linux off flash. Most of the development can be handled by anyone who has done system development on Linux which could have been on a PDA/Game console/router/PC/midrange. You no longer need someone specializing in Arm7 with detailed experience in the 8139 realtek chip and intel 802.11 chips, just someone who can make drivers for the linux kernel.
Linux has been widely ported around the town and finding a lowcost CPU that can run Linux (and includes an MMU) is easy.. so theres less need for the ucLinux or other exotic forks. Plain Linux will work well and you in one swoop have drivers for almost any networking or multimedia chip made.
I couldnt completely eliminate FreeBSD or Linux from my OS options while installing servers, each had strengths I needed.. until I ran into Solaris.
There are cards that have drivers for Solaris but not Linux or FreeBSD. There are tools and functionality in Solaris that I couldnt find in the others and had to use Solaris for some things.. such as SNAT between an ethernet and tokenring card. All linux kernels crashed after a while of doing this and so did FreeBSD with its lone Olicom driver support. Solaris held on. In other places, FreeBSDs performance just beat Solaris out of the water (no Java thrashing the harddisk like its Baghdad).
I can understand why theres little enthuisiasm for Solaris around the slashdot circles, its not free and it makes money for someone else. But just as we laugh at Microsofts lack of quality, we must laud a good product. Theres no denying the graphics capabilities of OSX and Irix, network and system admin tools of Solaris, huge features and ports of Linux and rock hard stability and efficiency of BSD.
Solaris especially makes a great package with a 64-bit Sparc CPU like the cheap used Ultra 5 systems. They still need to work on the scheduler, IDE speeds and hopefully theyll put virtual terminals back in Solaris10.
I havent used AIX, HPUX or OS/360/390/400. Does anyone know of their strengths and edge over other OSes?
Trolltech is a high-potential company with a bright future. The QT toolkit is the best thing around for clean fast and portable progamming. Trolltech is right to push QT to permeate across the world to reap its profits; they deserve it.
QT has given Linux alot. KDE became so big that GNOME had to be created as a free alternative before QT/X11 became GPLed. Now the Apple port will not only help apple applications, they will help Linux applications giving them more weight. Theres suddenly another big reason to shift your entire software project to QT despite any costs.
My only gripe is the really high license cost for a student. Ive built several applications in win32 but cant use them afer the 30 days. They relied heavily on printing so I couldnt port them to Linux. I even offered developers with the license to compile them for me for a small fee. I hope Trolltech sees this and if they really want to hide their code from pirates, provide a compilation service at a much lower cost for projects with low earning potential or value. I dont mind being the Toronto office manager of compilation services at all. Will even code for food(hey its 2003, not 1998)
Volunteering is when someone wants to support your cause. If youre profiting from the game and youre name alone is stamped on it, dont expect real Volunteering.
Now Volunteering is very much a community thing. Say you have a project for a cause that needs work and you know sincere programmers are out there. It will be best to advertise to local programmers since interacting with your neighbors and meeting them face to face increases sincerety to the cause. It can also bind them to work longer on the project. Ideally, setup a place where developers can physically come and sit on some workstations and develop. They will love the interaction with other developers from around and coke/chips offerings will complete the volunteering setup, getting the work done. Just dumping the work on someone remotely doesnt work on the sincerest of volunteers.
Now opensource software programming can be different. If it is something like gcc or the Linux kernel, the desire is globally strong enough for people to flock to it and submit patches regularly from 8 time zones away. And then, theyre working on a HUGE project with huge effects in the (computer) society. Their code will be used on tens of millions of computers. Thats the real motivation and the real itch. If you can duplicate that, or show the importance of your project to the community, you can expect help. Samba is pretty important. Everyone is using linux along with windows, and linux better look good. KDE is important. It shows the power of Linux on the desktop. Next are games projects like SDL, crystal space and mesa3d. Games are among the last remaining reasons to keep a windows partitions and is attracting a lot of volunteer programmers now. Keep a clean structure, an open environment, good responsive mailing lists, make great demos, always keep complete documentation and build 3d model/image/map import filters. Do not stamp your name all over the place (geeks are more egoistic than the rest), and that should entice a threshold number of developers.
Too much misinformation in the interview is hurting my head now. Just look at the very last line.. once code has been placed under GPL it cannot be altered! But alteration is what GPL is all about.. to take the code and change it.
/*TODO*/ or #include<kernel.h>
They only showed PART of the code, that she recognised as Sys V 4.1 vs linux kernel 2.4. It was from no region of the source.. just here and there. Could even be a line that goes
She also does not have experience with looking at sourcecode. She instead carried with her a tech guy. Where was he from? Who was paying him? whats his name? How much development experience does he have? Did he do a cmp or diff on the code blocks in question? Where is his interview?
Even IBMs customers I suppose are getting a gist of whats happening here. Only the most nervous and illinformed managers will be moved to Windows 2003. They will be deserving it.
Finding a pianist who is a musician...that is a rare and precious find indeed
I dont know about down there, but up here in Toronto most pianists I know are actually musicians. Maybe all pianists who are musicians as well move to Toronto.
How does number of years of industry experience sound? I know more about admining Linux than any admin Ive personally met, most of them working and getting paid for it. I havent gotten one interview in the past year... got very little industry experience despite the pile of certs and the fact that I'd love to goto an interview and get tested personally. Age is what matter here. Although senior technicians might feel theyre not up to speed the way they should be, theyre actually working while 6 months ago I was at a gas station during the day and was setting up IPSec VPn between Solaris and FreeBSD at home using Tokenring and Arcnet. Id much rather right now lose 3 years of my real experience to gain 3 years of paper experience. I'd be doing much better.
Yes. Releasing the patches first should alleviate the bandwidth problem. At least there should be a script in the kernel sources that downloads the latest patches with wget and applies them, also changes the directory name to reflect it. Such ease should convince people not to use 20 times the bandwidth they really need.
:)
To be fair anyway, I admit I'm downloading the full sources
I agree with you completely. IBM has spent over 1 billion dollars on Linux. Theyre a company of executives. But sun is a company of geeks and theyve carried UNIX itself forward with enormous help to BSD and Linux. On one hand you have the feel of knoppix and redhat. Nice slick and easy. On the other hand, its a very different feeling using Solaris on the Ultra 5. Simple, robust, standard, with a heavy and dedicated company behind it. Not to mention you get more jobs after learning Solaris. I hated Sun a while ago because everyone else in the Linux community hated them, but then, I tried it. Now I'm typing from an Ultra 5 and loving the feeling and stability.
All the more reason to get Sun certifications eh? The demand for Sun experts is greater than Linux at present and too many companies dont want to tear away from one *nix to another.
And what if all the unemployed Linux techies get Sun certs and work as their admins? There will be enough Sun techies going about to support the stations, no need for switching to Linux in the first place.
This balance is called market momentum and Sun has it. By giving out J2EE specs, free Java, cheap Solaris and other stuff, Sun knows it is increasing its own market momentum.
OK I understand your position very well. I have less than 2 years of Solaris experience myself, and been using Linux since 95.
Here in Toronto, jobs for Solaris beat jobs for Linux, so I had to learn it. I downloaded Solaris 8 for x86 and it looked crap. no VTs, not much software, slow and cumbersome with packages, management software and its X server.
But internally the technology is very well designed. I bought an Ultra 5 for running Solaris 9 SPARC and it is really rock stable. There are things you will see in Solaris that beat Linux, like standards. Every company out there knows theres only one standard Solaris distro, and packages (and binary drivers) for it always install with no problems. Solaris has drivers for products linux doesnt, since Suns a proprietary company and can communicate with the other companies. Software makers have known clients who need UNIX and are willing to pay for it, are using Solaris, so theres a great deal of good proprietary software for Solaris (in many places they beat the free offerings in Linux/BSD). Suns compiler beats GCC 3.2 too.
Now I know the performance of FreeBSD and gentoo beats solaris clear. I personally use FreeBSD in all of my installed servers for clients. But I have gained new respect in the past year trying out various features of Solaris that seem advanced, standard-based and very robust compared to Linux.. such as IPSec VPNs, SNAT between ethernet, tokenring and arcnet. Sun is not as omnipresent on the Net as Linux. It is omnipresent in the corporate.
Unlike Unisys, Sun offers a certain kind of good feeling with their OS and hardware.. you know.. the feeling that this product is rock stable, designed very well for UNIX and has good value. Linux on x86 is a good contender but depending on the whims of many IT directors, Sun will prevail for its name alone. Pointy haired bosses will (hopefully) at least know this was the most popular UNIX company, created Java and J2EE, spun off veritas, has awesome 64-bit cheap platforms and is run by true geeks who love the technology. Sun has pioneered so much technology used now in Linux/BSD/M$ and the rest of the world. That knowledge itself is good enough to depend on Sun (on its excellence, not survival. Thats being debated). Sun hardware is cheap nowadays and jobs for Solaris technicians exceed jobs for Linux. So unemployed Linux technicians will keep getting Solaris certs, and later as decision makers will choose a heavier company (opposed to RedHat or SuSE) to make the management happy. Solaris is also a very standard platform in itself, while Linux is struggling with standards. RedHat isnt even LSB compliant and many companies are packing products as redhat rpms only. Sun has a good market mass like Apple only better, and will stick around although in a bad shape.
Youre very right about that idea. Look at the sales of Ultra workstations on ebay. People need cheap machines will definitely buy them,,, look at el cheapo ECS and eMachines computers out there selling like hotcakes.
Sun needs to license out their UltraSPARC motherboards to low-cost Indian and Taiwanese companies who will sell combo motherboards for ATX or miniATX form factors. Sun should really add a good SCSI chip and possibly 802.11 or gigabit ethernet chips if they can get some cheap and add those to the mobos. The result might be Ultrasparcs with good SCSI and networking at less than $500. What geek would NOT want such little beasts for their home systems, not to mention Linux desktop companies might prefer them over x86 (possibly, but I dont know).
At the very least Sun should release specs to the Ultrasparc to the OpenBSD and other teams. Like Java, Sun needs to get people more involved with their products even against the risk of people making it off with cheap products instead of combos. IBM is making money off J2EE, but that only improves Suns position being in the center of gravity of J2EE and therefore always being compared with websphere and weblogic. The same should be applied to software, OS and hardware by Sun.
If I were the CEO of Sun, I would release all of Solaris and Forte under the GPL, try to incorporate their good features into the Linux kernel, and use the linux kernel in either x86 based machines or ultrasparcs as cheap reliable solutions. This is better than just dying out or getting bought out by someone who will just shelve these technologies and put Suns engineers to other use.
... to provide community sites for java? And java.net sounds like java.NET. I think there should be a seperate community site for J2EE systems with a code repository, forums and recommendations, or am I ignorant of such a site?