Which part of the earth will it hit anyway. I dont think it will directly affect the whole world, beside the atmospheric affects, which can be dealt with, as opposed to dealing with the asteroid directly.. for example starting growing mushrooms...
I'd wanna emigrate to the country directly opposite of the impact, start a business and buy farms (critical for survival). Also important will be buying of important real estate, for example if its hitting the oceans, buy higher land areas in Bangladesh and start building apartments. Heck just buy the land, let others build apartments close to doomsday.
Shares of companies researching food sources that do not require sunlight, or low light will jump...
Theres open-source, which sometimes means you can only look at the code. Theres free software which sometimes means take the software, do whatever, but dont break it apart, take important pieces and use it to improve your own software.
But for GPL and BSD licenses, the authors really are offering their source code for 'whatever'. This includes taking pieces of their code, using in your software, and even not thanking or marking the original author of that piece of code. By releaseing software under GPL or BSD, youre really offering ALL that. If thats not OK with you, find another license.
Code's value can be measured in a few ways (1) money (2) recognition (3) effect on the market.
Most people code for money. If you take code from Windows and use it for your project, thats like taking cash. You'd have to pay cash to take that cash for proprietary companies.
Many people code for recognition. This is true for many smaller projects and a few big ones in the OSS arena. This is the major point of contention... they release the code under GPL or BSD but sometimes its not OK if someone 'steals' lines of code, and credit is not given where its due. These guys should find another license enforcing credits... or maybe just release binaries.
A few companies code for market effects. Examples are companies which invest in Linux or BSD's code to create a leverage against Microsoft, decreasing its power. One good example is Apple's OSX which creates the market for Apple hardware. They wouldnt sell the source code, nor would release an x86 version for recognition only. Another example is GPL, which in some ways forces software to be open, BSD doesnt.
Getting back to the topic... sure, if the license is GPL or BSD, take all you want. In case of GPL, make sure your license is compliant. In parts of the world, if someone offers you sweets, and you turn it down. Its offensive.
Fights between opensource projects are always sad. Part of the openness is use it however you like... recommendations are opinions. ASF isnt quite blocking PHP yet, but things can go wrong as we've seen in the case of jboss and xfree86.
Whats stopping anyone from uniting php and apache1.3 and packaging them together for each platform the way sqlite was incorporated into php? They go well together, makes alotta sense to be the same project.
And the cops could send data saying no police in sight ahead.. speed up! I wonder how that'd legally hold up.
Cars could get around crashes, congestions and the likes, the best being prevention of pileups especially in winter. Cars can detect if the wheels arent moving in unison if theres a slip, and in milliseconds this can be used to slam on the brakes of cars behind.
But sooner or later the govt will enter the race and people will get tickets mailed at home with GPS coordinates and seconds since epoch where they exceeded the speed limit by 0.1 kmph.
If an email address has not been used in a year, and someone comes around claiming the data, and can provide vital personal data, he should be allowed the data. For one, I used my first name for the first hotmail account when it was new in 1996 or something. I lost the password, and made a new one. Looks like that name is still taken and I still cant claim the login, will soon be 10 years for that account sitting dormant.
The second reason was provided by the story submitter, a dead persons emails. The dead mans switch should send a custom email to friends if the account hasnt been logged into in a year, containing all online passwords. The only problem being if two friends use each other as backup, and they die in a car crash together, the data is still lost.
I recently lost someone in the family, and saw her name come up in msn messeger, as they booted her computer, but once they logged in as someone else, that login was lost too. Microsoft will sit on that one for another 10 years.
...in protocols. In a microsoft domain, the PDC and BDC have elevated rights. Similarly in the BGP protocol, hosts are authenticated and trusted with the protocol.
All that in the legal system that exists between computers.
Similarly, the universe is completely oblivious to the human settlement on planet Earth. All laws, national or international exist on earth's surface and nowhere else. Heck its unclear which laws are implemented in orbit, outside of national boundaries... so human laws exist only for interaction between humans. Its meaningless for the lone Inuit high in Alaska, who can cross international boundaries without visas. Or for someone down in antarctica, where the lack of human interaction itself renders laws useless.
The computer-to-human interaction however is different. If the human is authorized to use the computer, (s)he has the power over the computer for any matter. If the human is not authorized, the issue really is between the human and another human who owns or is responsible for the computer.
If computers could even exist on their own, feeding off the Internet, making money to pay their rent and power bills etc, or a very sophisticated online virus which could move around, a parasite on the Internet, then it might be directly in competition with humans. In that case, a human will always be preferred over the virus/code/program/computer. We always have to watch out for ourselves first.
I just want more of the same. Faster, smaller, more stable, more ports, more porable. Would be nice to be able to fit firefox onto a floppy, as opera used to be able to do. Another version should be about speed compiled with -O4 using intels C compiler for x86, sun forte for sparc etc. As for ports, instead of more binaries I'd rather see the code being made more easily portable. I certainly wouldnt mind an SDL or SVGALIB version of firefox for linux bypassing X, or an AAlib version competing with lynx.
Lastly I wouldnt mind an MSI version that can be easily deployed with specific configs across a microsoft network, and be made the default browser while all signs of IE are removed. Heres an even better idea, renaming firefox iexplore.exe and replacing the executable, while making firefox embeddable into explore.exe. The runtime requirements of windows suddenly becomes 32MB less ram.
Some small routers like some Linksys were based on Linux. Juniper's routers have a UNIX-like interface... so it seems apparent that a cheap router based on Linux is doable.
And I dont mean a miniITX board with flash and running a PentiumIII. I mean the newer models of MIPS CPUs, or even something like Athlon64 or Power5 for better throughput. Something to compare with higher end offerings of Cisco and Juniper.
You really mean BSD in general... or at least the 3 free BSDs. I was thinking of that a while ago about OpenBSD after I installed it on a Pentium3 to replace a much more expensive harware firewall, and a compromised windows2000 firewall(!). It was so clean and clear, still hasnt required maintenance ONCE.
I just think NetBSD is underrated precisely where it is portable.... Why in the world isnt it THE OS for embedded systems? Look at the effort going into Linux to take it anywhere. BSD follows clean design and Linux follows hacker culture, but the hacker culture must be built on strong grounds, and BSD's design impressed me, since I tried to crosscompile Linux and netbsd kernels for sparcstations on an Athlon. Kegel's crosstool scripts bombed out and I had to fix things here and there, and for netbsd, you have build.sh, which still didnt work, but the effort and design to make it possible (Linux isnt nearly as proactive about portability since its aims are general) impresses me.
I have two questions that others might have too, when shown the 54-arches NetBSD supports:
(1) Does it support ALL these arches completely, with every driver and package? I know NetBSD's driver system is awesome, where drivers are made endian-free and attached to PCI or ISA etc busses instead of arches in Linux. Sure not all devices will work with all arches, but if the electrical, performance, mechanical etc attributes work, can the NetBSD kernel drive the device in all arches that support that bus?
(2) If 54 have been conquered, how many are left? I'm more interested in knowing about 32-bit and 64-bit cpu types than architectures, since there are many architectures holding the same cpu... like IBM pSeries and iMac. Is there a 32-bit cpu that NetBSD cant be ported to because gcc/binutils do not support it well enough?
As for architectures themselves, there are just too many of them, think of all ARM evaluation boards.
What can you do? Redhat is in demand, and they have to look at the pofit curve and extract the most money. Do you blame em?
Everyone keeps hearing about this thing called Linux, too many companies are pushing it out there. Maybe your windows servers been crashing since NT 3.51, so you start looking. Redhat is the biggest Linux vendor with support. You want a big BIG company base behind your OS, and a software base, Redhat is it, with Suse coming in second regardless of price or quality of support or binaries or whatever.
So you go with the top Linux vendor. With Sun, IBM pSeries slowly defeated, and HP's HPUX platforms, well, I dont know anything about them... and Apple too vertical a market for your taste with all server apps in the wild against it, you'd head for none other than Redhat, after Microsoft, in OS sales.
For us, Redhat needs to be a rich successful company. Thats more important than the number of sales they make. Reason being their success attracts other vendors, and several competing vendors are much better than one vendor with the global supply of commercial Linux. Their success also puts them in a position to improve the Linux market itself, we've seen Redhat ads compete with Microsoft ads. Slackware couldnt do that. We've seen Domino, Oracle, and many other major server apps released in redhat packaging and supported as such. Debian couldnt do that.
So let Redhat get rich. Please. Beyond a threshold, Dell will purchase it. Below the threshold, Dell will purchase the next best thing and improve competition. If people need 'Redhat' Linux, let them pay for it until something better comes along.
Say Linux goes a bit more mainstream, Opensource apps increase in numbers, especially for powerpc, both IBM and Apple win, Intel and Microsoft lose.
This is more true of servers than desktops... for now. IBM can take the server share (cheapest pseries now is $6k, with very few under $10k, Apple the desktop share). They both have been depending more and more on opensourced apps and OSes, and have had past alliances (PReP machines), that worked. Both created successful computer lines and are confident in doing the same again. Both have been highly marginalized by Wintel Inc.
IBM is pretty much getting rid of x86 on desktops, keeping only the x86 on servers. With AMD as a good option, they really dont need Intel for anything now, havent been relying on Microsoft either for much beside xSeries OSes (online catalog shows SLES and Redhat AS as options alongside Win2003).
The whole industry, at least the bigger players are moving away from the wintel alliance, and we can expect a showdown. Wintel wants the entire market to itself, everyones threatened. Sun, SGI, Novell have seen the light at the end of the tunnel, no reason for them not to join. Apple and IBM must do something while they still have the kick.
Tell me if I'm way off my base here. I have a premonition of a tech sector mortal kombat with entire vertical architectures against each other, x86+win32 and other arches+other oses. I see IBM moving away from x86, at least from Intel... Athlon64 is too good a deal to turn down.
Am I wrong or is the Intel+Microsoft alliance just not that threatening?
There was the 386, and then, there was the 376. The 376 didnt have 16-bit parts, a different booting method, and was much simpler with the same performance as the other fatter one.Remove the silly protected mode and 16-bit silliness, and you have a lean, fast chip to which you could port Windows, and not feel bad about IBM not choosing the m68k for the PC.
Intel had a real chance there. They had already given out IP to cyrix, rise, amd, which would later compete and beat them. Restart life as 376, patented and all, and suddenly copying you becomes difficult again, and you hold all the IP.
I now wonder if the Athlon64 still has a hardware-implemented 16-bit protected mode to boot from. There must be a piece of 8086 somewhere on that huge chip, a piece we can really do without.
I think market forces will ensure that 20 years from now when we're booting the multicore,multicell quantum optical chip, somewhere deep inside an 8086 still lurks, with the complete instruction set, and Robert Browns DOS interrupt list (#21?), just to bring the monolith back to life, switch it to 32-bits, and a late-90s code to switch it to full 64-bits, and awaken the other cores/cells.
The pentiumII was a better chip than somewhere in the middle of Pentium1 and PentiumIII. The Spec Int ratings (and other ratings on the spec.org) was comparable to the PentiumIII, makes me wonder what is the PentiumIII beside having a higher clock.
Compared to the Pentium1, it was leagues ahead, similar to the Pentium's lead over the 486. Originally sold as the Pentiumpro (still looking for one as a collectors item), then as PentiumII with SSE, better clock, and weird weird packaging which AMD copied, but never took off for x86 CPUs. I still dont get the PLASTIC casing around the riser card.... bad circulation, and who will ever be impressed with ads inside the case?!?
The PentiumII-PentiumIII streak (very similar chips) began with Intel in the lead, and ended with a defeat by the Athlon. It ended really because Intel wanted a shallow pipeline and higher clock numbers... else we'd be running only a much higher clock version of the same chip now.
Somehow I dont feel the P4 will last as long. Intel needs to copy Athlon64 and quick.
Ok so the Solaris 'code' is on a NAS server, and it is run remotely. You also mentioned you recently moved to Win32.
Hmmm. It couldnt have been more unclear. Solaris most frequently runs on sparc architectures. Surely the code must be sitting one place, executing on another machine, and somehow the windows machines grab hold of the results...
So depending on your REAL situation:
(1) Run apache on the solaris box and display results.
(2) Run the code on a terminal server machine (Windows2000 terminal services or X11)
(3) Use rsync or the windows equivalent to redistribute code to all machines
(4) Use CVS
(5) Recompile the code for win32.
To get any more useful advice from slashdot, specify your problem better.
... is that OO is a complete suit, but the word processor part isnt as MSWord compliant as Abiword. Abiword is more MSWord compatible, and is standalone. They both startup slow and take more memory than good quality opensource software should.
When OO was new, I thought it was the Abiword killer.
I also dont quite get why Abiword isnt packaged as a part of OO. License incompatibility?
Lastly, I'm waiting for the firefox of word processors, something sleek and lean, fast, stable, with only the functionality I need, yet compliant with MS Word 2000. I've only needed Word and Excel, and these two applications need not be in the same office suite; only fast and compliant.
They've got alot invested in Solaris, which also drives their hardware and service markets, and wouldnt want to give it all up. It seems they're just opensourcing Solaris, as in people can look at the source. Cant copy the sources elsewhere (Linux or BSD), cant resell it, cant redistribute altered binaries, and I'm not sure if anyone can redistribute altered Solaris even with the sources.
As for taking improvements to Linux, I wonder if Linux can be forked into a more restrictive License, which doesnt go against the GPL. That way Solaris source blocks can be moved to Linux if its even worth that much.
I'd be happier to see the threading implementation and memory management of Solaris in Linux or BSD, and improvements to glibc. We dont need Solaris binary compatibility as its only really useful on the sparc platforms.
Its also kinda refreshing to see they dont intend to bring down Solaris like a kamimaze, rather plan to develop it and live on it further. Means Solaris isnt dying.
I say we fork XORP as soon as the OSFP is done, make it GPL and work to make it stable.
That way Microsoft would have funded the project and would have not received a good version. Like Zebra, XORP could stagnate if a parallel GPL version exists that ensures Microsoft can't lay their grubby hands on it.
Speaking of Microsoft, its quite possible they're trying to be back on good terms with developers around the world.... naaahh.... they're just backstabbing the networking industry giant before their interests collide through voip, webtv and the likes.
The 3 requirements of any job is: experience, degree and certs.
Communication skills, personality and motivation are a distant fourth, and hard to measure.
So If you have a crap degree (I.E. compared to an ivy league student) 2 additional years of experience, especially with well-known companies will make you equal. Having additional certs, especially relevant ones will help as well. For security, database, networking jobs, certs are critical, sometimes meaning more than the degree if you browse joblists, and will do you more good than a masters program. So the right certs will land you the job over the same ivy league student without them.
Basically if you lack anything from the above 3 requirements, getting jobs will be difficult. However all else being equal, you do get a better edge getting a better known degree in CS circles, so if you can help it, and if it doesnt affect anything else, transfer.
Microsoft is very slowly finding its direction. The success of any OS platform depends in a large part on its developer community.
With the development tools free, a developer and application base forms naturally, that can better sustain any given company. After all Linux started with gcc.
The cost of VisualC has been obscene, with Microsoft assuming win32 developers have no other option. Nowadays we've got wxwindows, QT, the bcc and intel compilers, all free (except QT) and of better quality, and ticked off developers can easily switch to OSX and Linux. Gates has acknowledged Microsoft made a mistake in not rallying a developer base around it.
Free VisualC... hmmm if they release such a thing it would be the culmination of the 'developers,developers,developers' we've been hearding of...
In some ways you guys are right. BSD is dying and has been dying for over a decade. BSD started dying when AIX, SCO, Solaris starting grabbing the growing market.
But see thats the beauty of free operating systems, and thats what scares Microsoft, and scared IBM, Novell, Sun into joining what they cant defeat. An opensourced OS can never really die.. they just live on with a smaller number of supporters until developers elsewhere feel their favorite OS is insufficient for their purposes, and join forces. In that regard Linux, BSD and the smaller/less known opensourced OSes feed off each other, exchanging work and developers depending on the politics. Linux has to be grateful to BSD for being a great template of an OS, for its TCPIP stack, for the UNIX philosophy which was lost among the commercial OSes. BSD must be grateful to Linux for the new hacker culture, and boom in the free OS market from which customers, developers and applications (drivers even) trickle back to it from Linux.
BSD is dying, but it will never be dead. Same for Solaris if they 'really' opensource it. Then it will not become a SysV, or a Unixware.
Many contractors/consultants work at their pace, whereby 1 hour of continuous work is done against 4-8 hours of relaxation, taking a break, whatever. Try working on any problem for more than several hours straight and the quality of your work falls exponentially.
Hours are quoted by consultants as a way to measure the cost of work, not that he'd be spending exactly that many hours of work, its far too difficult to anticipate the time needed. So the work worth a consultants 80 hours, will require 20 hours of work with many continuous breaks if it is to run within the warranty period without a major crash.
I hope as a result of this, more than anything else, the next Civ will have the top-down view option that the first Civ had. I've always had trouble with the side view Civ2 onwards had. The north-south directio n is twice as squished as the east-west and moving diagonally is terrible without the grid.
The other 'features' newer civs have are realistic and animated units, which go through pains to move/attack/fortify, whereas I like to build huge civilizations and move around units real fast, no sounds, animation, messages or delay.
Another feature I'm hoping for is grouping up units, like you do in Command n Conquer and Tiberian Sun, where CTRL-num after selecting a bunch of units bunches them under the number, and ALT-num selects them immediately. Usually I'm using more than one unit attacking a city, transporting them elsewhere, even with workers, flashmobbing a city zone. When you have hundereds of units, civ becomes a pain.
Lastly, the simple queueing of production in cities in Civ3 is terrific, but I'd like more, maybe through scripting. For example NEVER let a city goto civil disorder at the cost of causing starvation, or NEVER let shields fall below at the cost of civil disorder etc.
PS a civ that doesnt end in 2040, and smoothly translates into an Alpha Centauri game wouldnt hurt. At that point it would be civs on planets rather than continents, where the cost of transportation would be very high.
I'd still want Sid Meyer to be part of the development team though.
Reminds me of the mid-90s websites. Everything was as flashed and shockwaved, and javascripted as possible. Those were the days when big was good. I remember plenty of highly unreadable websites which took 4 minutes to load on 28.8k dialup, but had a coolfactor.
The same effort that was put into building this interface, couldve been used to put it into sgml, or at least xhtml. They want the layout preserved? PDF! Even a series of GIFs like some online books would be easier.
Slashdot Mirror
Which part of the earth will it hit anyway. I dont think it will directly affect the whole world, beside the atmospheric affects, which can be dealt with, as opposed to dealing with the asteroid directly.. for example starting growing mushrooms...
I'd wanna emigrate to the country directly opposite of the impact, start a business and buy farms (critical for survival). Also important will be buying of important real estate, for example if its hitting the oceans, buy higher land areas in Bangladesh and start building apartments. Heck just buy the land, let others build apartments close to doomsday.
Shares of companies researching food sources that do not require sunlight, or low light will jump...
Theres open-source, which sometimes means you can only look at the code. Theres free software which sometimes means take the software, do whatever, but dont break it apart, take important pieces and use it to improve your own software.
But for GPL and BSD licenses, the authors really are offering their source code for 'whatever'. This includes taking pieces of their code, using in your software, and even not thanking or marking the original author of that piece of code. By releaseing software under GPL or BSD, youre really offering ALL that. If thats not OK with you, find another license.
Code's value can be measured in a few ways (1) money (2) recognition (3) effect on the market.
Most people code for money. If you take code from Windows and use it for your project, thats like taking cash. You'd have to pay cash to take that cash for proprietary companies.
Many people code for recognition. This is true for many smaller projects and a few big ones in the OSS arena. This is the major point of contention... they release the code under GPL or BSD but sometimes its not OK if someone 'steals' lines of code, and credit is not given where its due. These guys should find another license enforcing credits... or maybe just release binaries.
A few companies code for market effects. Examples are companies which invest in Linux or BSD's code to create a leverage against Microsoft, decreasing its power. One good example is Apple's OSX which creates the market for Apple hardware. They wouldnt sell the source code, nor would release an x86 version for recognition only. Another example is GPL, which in some ways forces software to be open, BSD doesnt.
Getting back to the topic... sure, if the license is GPL or BSD, take all you want. In case of GPL, make sure your license is compliant. In parts of the world, if someone offers you sweets, and you turn it down. Its offensive.
Fights between opensource projects are always sad. Part of the openness is use it however you like... recommendations are opinions. ASF isnt quite blocking PHP yet, but things can go wrong as we've seen in the case of jboss and xfree86.
Whats stopping anyone from uniting php and apache1.3 and packaging them together for each platform the way sqlite was incorporated into php? They go well together, makes alotta sense to be the same project.
And the cops could send data saying no police in sight ahead.. speed up! I wonder how that'd legally hold up.
Cars could get around crashes, congestions and the likes, the best being prevention of pileups especially in winter. Cars can detect if the wheels arent moving in unison if theres a slip, and in milliseconds this can be used to slam on the brakes of cars behind.
But sooner or later the govt will enter the race and people will get tickets mailed at home with GPS coordinates and seconds since epoch where they exceeded the speed limit by 0.1 kmph.
If an email address has not been used in a year, and someone comes around claiming the data, and can provide vital personal data, he should be allowed the data.
For one, I used my first name for the first hotmail account when it was new in 1996 or something. I lost the password, and made a new one. Looks like that name is still taken and I still cant claim the login, will soon be 10 years for that account sitting dormant.
The second reason was provided by the story submitter, a dead persons emails. The dead mans switch should send a custom email to friends if the account hasnt been logged into in a year, containing all online passwords. The only problem being if two friends use each other as backup, and they die in a car crash together, the data is still lost.
I recently lost someone in the family, and saw her name come up in msn messeger, as they booted her computer, but once they logged in as someone else, that login was lost too. Microsoft will sit on that one for another 10 years.
...in protocols. In a microsoft domain, the PDC and BDC have elevated rights. Similarly in the BGP protocol, hosts are authenticated and trusted with the protocol.
All that in the legal system that exists between computers.
Similarly, the universe is completely oblivious to the human settlement on planet Earth. All laws, national or international exist on earth's surface and nowhere else. Heck its unclear which laws are implemented in orbit, outside of national boundaries... so human laws exist only for interaction between humans. Its meaningless for the lone Inuit high in Alaska, who can cross international boundaries without visas. Or for someone down in antarctica, where the lack of human interaction itself renders laws useless.
The computer-to-human interaction however is different. If the human is authorized to use the computer, (s)he has the power over the computer for any matter. If the human is not authorized, the issue really is between the human and another human who owns or is responsible for the computer.
If computers could even exist on their own, feeding off the Internet, making money to pay their rent and power bills etc, or a very sophisticated online virus which could move around, a parasite on the Internet, then it might be directly in competition with humans. In that case, a human will always be preferred over the virus/code/program/computer. We always have to watch out for ourselves first.
I just want more of the same. Faster, smaller, more stable, more ports, more porable.
Would be nice to be able to fit firefox onto a floppy, as opera used to be able to do. Another version should be about speed compiled with -O4 using intels C compiler for x86, sun forte for sparc etc. As for ports, instead of more binaries I'd rather see the code being made more easily portable. I certainly wouldnt mind an SDL or SVGALIB version of firefox for linux bypassing X, or an AAlib version competing with lynx.
Lastly I wouldnt mind an MSI version that can be easily deployed with specific configs across a microsoft network, and be made the default browser while all signs of IE are removed. Heres an even better idea, renaming firefox iexplore.exe and replacing the executable, while making firefox embeddable into explore.exe. The runtime requirements of windows suddenly becomes 32MB less ram.
Some small routers like some Linksys were based on Linux. Juniper's routers have a UNIX-like interface... so it seems apparent that a cheap router based on Linux is doable.
And I dont mean a miniITX board with flash and running a PentiumIII. I mean the newer models of MIPS CPUs, or even something like Athlon64 or Power5 for better throughput. Something to compare with higher end offerings of Cisco and Juniper.
You must mean prep. CHRP is the newer IBM machines, prep was the alliance with apple, and those run 5.1L at best.
You really mean BSD in general... or at least the 3 free BSDs. I was thinking of that a while ago about OpenBSD after I installed it on a Pentium3 to replace a much more expensive harware firewall, and a compromised windows2000 firewall(!). It was so clean and clear, still hasnt required maintenance ONCE.
I just think NetBSD is underrated precisely where it is portable.... Why in the world isnt it THE OS for embedded systems? Look at the effort going into Linux to take it anywhere. BSD follows clean design and Linux follows hacker culture, but the hacker culture must be built on strong grounds, and BSD's design impressed me, since I tried to crosscompile Linux and netbsd kernels for sparcstations on an Athlon. Kegel's crosstool scripts bombed out and I had to fix things here and there, and for netbsd, you have build.sh, which still didnt work, but the effort and design to make it possible (Linux isnt nearly as proactive about portability since its aims are general) impresses me.
I have two questions that others might have too, when shown the 54-arches NetBSD supports:
(1) Does it support ALL these arches completely, with every driver and package? I know NetBSD's driver system is awesome, where drivers are made endian-free and attached to PCI or ISA etc busses instead of arches in Linux. Sure not all devices will work with all arches, but if the electrical, performance, mechanical etc attributes work, can the NetBSD kernel drive the device in all arches that support that bus?
(2) If 54 have been conquered, how many are left? I'm more interested in knowing about 32-bit and 64-bit cpu types than architectures, since there are many architectures holding the same cpu... like IBM pSeries and iMac. Is there a 32-bit cpu that NetBSD cant be ported to because gcc/binutils do not support it well enough?
As for architectures themselves, there are just too many of them, think of all ARM evaluation boards.
What can you do? Redhat is in demand, and they have to look at the pofit curve and extract the most money. Do you blame em?
Everyone keeps hearing about this thing called Linux, too many companies are pushing it out there. Maybe your windows servers been crashing since NT 3.51, so you start looking. Redhat is the biggest Linux vendor with support. You want a big BIG company base behind your OS, and a software base, Redhat is it, with Suse coming in second regardless of price or quality of support or binaries or whatever.
So you go with the top Linux vendor. With Sun, IBM pSeries slowly defeated, and HP's HPUX platforms, well, I dont know anything about them... and Apple too vertical a market for your taste with all server apps in the wild against it, you'd head for none other than Redhat, after Microsoft, in OS sales.
For us, Redhat needs to be a rich successful company. Thats more important than the number of sales they make. Reason being their success attracts other vendors, and several competing vendors are much better than one vendor with the global supply of commercial Linux. Their success also puts them in a position to improve the Linux market itself, we've seen Redhat ads compete with Microsoft ads. Slackware couldnt do that. We've seen Domino, Oracle, and many other major server apps released in redhat packaging and supported as such. Debian couldnt do that.
So let Redhat get rich. Please. Beyond a threshold, Dell will purchase it. Below the threshold, Dell will purchase the next best thing and improve competition. If people need 'Redhat' Linux, let them pay for it until something better comes along.
Your comment suddenly made me realize something.
Add up these facts:
Apple is powerpc
IBM is powerpc
Apple is OSX based on FreeBSD
IBM spent $1 bil on Linux last year
Apple competes with Intel and Microsoft
IBM competes with Intel and Microsoft
Microsofts apps depend on Intel (Wintel)
Intel Sales depend on Microsoft OS and apps
Intel is a monopoly (they still are, declining)
Microsoft is a monopoly
IBM and Apple are losers in these monopolies
IBM has been releasing apps for Linux on pSeries
Apple has been pushing apps for UNIX on PPC
IBM supports OSS community, increasing free apps
Free apps can be compiled on any arch.
Making sense now?
No?
Say Linux goes a bit more mainstream, Opensource apps increase in numbers, especially for powerpc, both IBM and Apple win, Intel and Microsoft lose.
This is more true of servers than desktops... for now. IBM can take the server share (cheapest pseries now is $6k, with very few under $10k, Apple the desktop share). They both have been depending more and more on opensourced apps and OSes, and have had past alliances (PReP machines), that worked. Both created successful computer lines and are confident in doing the same again. Both have been highly marginalized by Wintel Inc.
IBM is pretty much getting rid of x86 on desktops, keeping only the x86 on servers. With AMD as a good option, they really dont need Intel for anything now, havent been relying on Microsoft either for much beside xSeries OSes (online catalog shows SLES and Redhat AS as options alongside Win2003).
The whole industry, at least the bigger players are moving away from the wintel alliance, and we can expect a showdown. Wintel wants the entire market to itself, everyones threatened. Sun, SGI, Novell have seen the light at the end of the tunnel, no reason for them not to join. Apple and IBM must do something while they still have the kick.
Tell me if I'm way off my base here. I have a premonition of a tech sector mortal kombat with entire vertical architectures against each other, x86+win32 and other arches+other oses. I see IBM moving away from x86, at least from Intel... Athlon64 is too good a deal to turn down.
Am I wrong or is the Intel+Microsoft alliance just not that threatening?
There was the 386, and then, there was the 376. The 376 didnt have 16-bit parts, a different booting method, and was much simpler with the same performance as the other fatter one.Remove the silly protected mode and 16-bit silliness, and you have a lean, fast chip to which you could port Windows, and not feel bad about IBM not choosing the m68k for the PC.
Intel had a real chance there. They had already given out IP to cyrix, rise, amd, which would later compete and beat them. Restart life as 376, patented and all, and suddenly copying you becomes difficult again, and you hold all the IP.
I now wonder if the Athlon64 still has a hardware-implemented 16-bit protected mode to boot from. There must be a piece of 8086 somewhere on that huge chip, a piece we can really do without.
I think market forces will ensure that 20 years from now when we're booting the multicore,multicell quantum optical chip, somewhere deep inside an 8086 still lurks, with the complete instruction set, and Robert Browns DOS interrupt list (#21?), just to bring the monolith back to life, switch it to 32-bits, and a late-90s code to switch it to full 64-bits, and awaken the other cores/cells.
The pentiumII was a better chip than somewhere in the middle of Pentium1 and PentiumIII. The Spec Int ratings (and other ratings on the spec.org) was comparable to the PentiumIII, makes me wonder what is the PentiumIII beside having a higher clock.
Compared to the Pentium1, it was leagues ahead, similar to the Pentium's lead over the 486. Originally sold as the Pentiumpro (still looking for one as a collectors item), then as PentiumII with SSE, better clock, and weird weird packaging which AMD copied, but never took off for x86 CPUs. I still dont get the PLASTIC casing around the riser card.... bad circulation, and who will ever be impressed with ads inside the case?!?
The PentiumII-PentiumIII streak (very similar chips) began with Intel in the lead, and ended with a defeat by the Athlon. It ended really because Intel wanted a shallow pipeline and higher clock numbers... else we'd be running only a much higher clock version of the same chip now.
Somehow I dont feel the P4 will last as long. Intel needs to copy Athlon64 and quick.
Ok so the Solaris 'code' is on a NAS server, and it is run remotely. You also mentioned you recently moved to Win32.
Hmmm. It couldnt have been more unclear. Solaris most frequently runs on sparc architectures. Surely the code must be sitting one place, executing on another machine, and somehow the windows machines grab hold of the results...
So depending on your REAL situation:
(1) Run apache on the solaris box and display results.
(2) Run the code on a terminal server machine (Windows2000 terminal services or X11)
(3) Use rsync or the windows equivalent to redistribute code to all machines
(4) Use CVS
(5) Recompile the code for win32.
To get any more useful advice from slashdot, specify your problem better.
... is that OO is a complete suit, but the word processor part isnt as MSWord compliant as Abiword. Abiword is more MSWord compatible, and is standalone. They both startup slow and take more memory than good quality opensource software should.
When OO was new, I thought it was the Abiword killer.
I also dont quite get why Abiword isnt packaged as a part of OO. License incompatibility?
Lastly, I'm waiting for the firefox of word processors, something sleek and lean, fast, stable, with only the functionality I need, yet compliant with MS Word 2000. I've only needed Word and Excel, and these two applications need not be in the same office suite; only fast and compliant.
They've got alot invested in Solaris, which also drives their hardware and service markets, and wouldnt want to give it all up. It seems they're just opensourcing Solaris, as in people can look at the source. Cant copy the sources elsewhere (Linux or BSD), cant resell it, cant redistribute altered binaries, and I'm not sure if anyone can redistribute altered Solaris even with the sources.
As for taking improvements to Linux, I wonder if Linux can be forked into a more restrictive License, which doesnt go against the GPL. That way Solaris source blocks can be moved to Linux if its even worth that much.
I'd be happier to see the threading implementation and memory management of Solaris in Linux or BSD, and improvements to glibc. We dont need Solaris binary compatibility as its only really useful on the sparc platforms.
Its also kinda refreshing to see they dont intend to bring down Solaris like a kamimaze, rather plan to develop it and live on it further. Means Solaris isnt dying.
I say we fork XORP as soon as the OSFP is done, make it GPL and work to make it stable.
That way Microsoft would have funded the project and would have not received a good version. Like Zebra, XORP could stagnate if a parallel GPL version exists that ensures Microsoft can't lay their grubby hands on it.
Speaking of Microsoft, its quite possible they're trying to be back on good terms with developers around the world.... naaahh.... they're just backstabbing the networking industry giant before their interests collide through voip, webtv and the likes.
The 3 requirements of any job is: experience, degree and certs.
Communication skills, personality and motivation are a distant fourth, and hard to measure.
So If you have a crap degree (I.E. compared to an ivy league student) 2 additional years of experience, especially with well-known companies will make you equal. Having additional certs, especially relevant ones will help as well. For security, database, networking jobs, certs are critical, sometimes meaning more than the degree if you browse joblists, and will do you more good than a masters program. So the right certs will land you the job over the same ivy league student without them.
Basically if you lack anything from the above 3 requirements, getting jobs will be difficult. However all else being equal, you do get a better edge getting a better known degree in CS circles, so if you can help it, and if it doesnt affect anything else, transfer.
Microsoft is very slowly finding its direction. The success of any OS platform depends in a large part on its developer community.
With the development tools free, a developer and application base forms naturally, that can better sustain any given company. After all Linux started with gcc.
The cost of VisualC has been obscene, with Microsoft assuming win32 developers have no other option. Nowadays we've got wxwindows, QT, the bcc and intel compilers, all free (except QT) and of better quality, and ticked off developers can easily switch to OSX and Linux. Gates has acknowledged Microsoft made a mistake in not rallying a developer base around it.
Free VisualC... hmmm if they release such a thing it would be the culmination of the 'developers,developers,developers' we've been hearding of...
In some ways you guys are right. BSD is dying and has been dying for over a decade. BSD started dying when AIX, SCO, Solaris starting grabbing the growing market.
But see thats the beauty of free operating systems, and thats what scares Microsoft, and scared IBM, Novell, Sun into joining what they cant defeat. An opensourced OS can never really die.. they just live on with a smaller number of supporters until developers elsewhere feel their favorite OS is insufficient for their purposes, and join forces. In that regard Linux, BSD and the smaller/less known opensourced OSes feed off each other, exchanging work and developers depending on the politics. Linux has to be grateful to BSD for being a great template of an OS, for its TCPIP stack, for the UNIX philosophy which was lost among the commercial OSes. BSD must be grateful to Linux for the new hacker culture, and boom in the free OS market from which customers, developers and applications (drivers even) trickle back to it from Linux.
BSD is dying, but it will never be dead. Same for Solaris if they 'really' opensource it. Then it will not become a SysV, or a Unixware.
Many contractors/consultants work at their pace, whereby 1 hour of continuous work is done against 4-8 hours of relaxation, taking a break, whatever. Try working on any problem for more than several hours straight and the quality of your work falls exponentially.
Hours are quoted by consultants as a way to measure the cost of work, not that he'd be spending exactly that many hours of work, its far too difficult to anticipate the time needed. So the work worth a consultants 80 hours, will require 20 hours of work with many continuous breaks if it is to run within the warranty period without a major crash.
I hope as a result of this, more than anything else, the next Civ will have the top-down view option that the first Civ had. I've always had trouble with the side view Civ2 onwards had. The north-south directio n is twice as squished as the east-west and moving diagonally is terrible without the grid.
The other 'features' newer civs have are realistic and animated units, which go through pains to move/attack/fortify, whereas I like to build huge civilizations and move around units real fast, no sounds, animation, messages or delay.
Another feature I'm hoping for is grouping up units, like you do in Command n Conquer and Tiberian Sun, where CTRL-num after selecting a bunch of units bunches them under the number, and ALT-num selects them immediately. Usually I'm using more than one unit attacking a city, transporting them elsewhere, even with workers, flashmobbing a city zone. When you have hundereds of units, civ becomes a pain.
Lastly, the simple queueing of production in cities in Civ3 is terrific, but I'd like more, maybe through scripting. For example NEVER let a city goto civil disorder at the cost of causing starvation, or NEVER let shields fall below at the cost of civil disorder etc.
PS a civ that doesnt end in 2040, and smoothly translates into an Alpha Centauri game wouldnt hurt. At that point it would be civs on planets rather than continents, where the cost of transportation would be very high.
I'd still want Sid Meyer to be part of the development team though.
Reminds me of the mid-90s websites. Everything was as flashed and shockwaved, and javascripted as possible. Those were the days when big was good. I remember plenty of highly unreadable websites which took 4 minutes to load on 28.8k dialup, but had a coolfactor.
The same effort that was put into building this interface, couldve been used to put it into sgml, or at least xhtml. They want the layout preserved? PDF! Even a series of GIFs like some online books would be easier.