Why would anyone want to stop at just building the ultimate box? Why not the ultimate home LAN, too?
I mean really... do you want your desktop box to do everything? Is it a desktop, is it a server, or is it both? Allow me to modify the "what's in your wallet" phrase in the Capital One card commercial.. "what's in your home LAN?!"
I just built two new identical systems for home.. one for my wife and one for me. Each box cost approximately $3800, a bit less than we paid for
a 500Mhz G4 Titanium Powerbook w/512mb RAM.
The quick specs:
Lian Li PC60 aluminum case w/window & light kit
300W Sparkle power supply
Tyan Tiger MP S2460 w/dual 1.2Ghz Athlon MP's
Agilent Articoolers for the CPUs
512MB ECC reg. PC2100
Matrox G550 32MB DDR AGP dual
Samsung SyncMaster 170T (x2 per machine)
Seagate Barracuda III ATA100 7200rpm (30gb)
Toshiba 16X IDE DVD-ROM
Netgear FA310-TX
SB Live 5.1
Now, we already had speakers from our previous systems (both sets are Altec Lansing, mine is the ADA880 which supports AC3, surround, etc.).
Reading over some of the comments in this article, I agree completely that SCSI is still the best choice, but for pure desktop systems, it's sheer overkill (unless you are lucky enough to have IBM donating $15k so you can afford at least one drive). And in this case, they left out fibre channel and/or solid state drives. I mean come on! Our previous desktops were all SCSI, but I chose the IDE route again this time, mostly because the type of stuff we do on our desktops is not so I/O intensive as to cause problems. ATA100 is decent as long as you tweak it, but you have to buy the right kind of drive(s). Personally, I've found that Seagate drives are reliable in this regard (others may disagree, just note this is my personal preference and I recommend you pick whatever you feel comfortable with). Also, note that the Tyan Thunder board requires the huge 450W power supply, and it's not a standard ATX connector. The Tyan Tiger MP is the same exact board as the Thunder without all the built-in stuff (for me, built-in stuff is a major no-no for a desktop box as it makes it much harder to upgrade for the future), and it uses standard ATX power connectors. You do still need to be careful about the power supply you pick, though.
So, with our new desktops, I'll tie in what I was getting at above. If you're buying/building the ultimate box to be a standalone, multi-purpose, server and desktop box.. stop reading now. I dunno about everyone else, but I can't settle for that. My desktop is for desktop tasks. I have other systems on my LAN to do the various server related tasks that compliment my desktop system (i.e I try to spread the money around)
I won't go into real specifics here, but I'm not
a Linux bigot. I'm a sysadmin, and I think every OS has a niche (well, maybe not all of them.. I'm a UNIX bigot). So with that in mind, I use FreeBSD for our dedicated firewall/router/wireless gateway, Solaris x86 for our NFS, NIS/LDAP, backup server (it's all SCSI with mirrored system and home storage disks), and Debian Linux for our media server, desktops, and home beowulf cluster. Our one Powerbook tri-boots MacOS 9.2.1, MacOS X 10.1, and Linux.
Anyway, the point is that an ultimate desktop box is all well and good, but I think anyone that goes out in search of the end-all, be-all holy grail standalone box is in for a disappointment down the road. And besides.. there is only so much fun you can have with one machine.
Speaking as the maintainer for *windowmaker.org,
it should be relatively obvious what Window Maker's focus is. All one need do is peruse our windowmaker.org.
I think we do a pretty reasonable job of keeping our users up-to-date on where the project is going, and what they can expect.
In response to a few of the comments in this thread, I'd like to point out that Window Maker concentrates more on functionality than on "features". Every window manager has its own
niche and is trying to appeal to a particular crowd. We try to appeal to those users that want a consistent, usable desktop that sits somewhere in between the minimal ones like Blackbox, and the
eyecandy ones like E. See our
comparison chart.
What's on the horizon? The developers recently
reworked the code that generates and maintains files which help make up the UI (menus, states, etc.). To the end user, this means there is one less library dependency to worry about, providing they're the ones compiling the code. But the focus of this question was on "features". Window Maker won't be producing breakthroughs like what the Berlin project is doing, for instance. I get the impression this is what the questionaire was driving at. In our case, we're trying to improve compatibility with existing desktop environments (GNOME, GNUSTEP, and KDE) so that Window Maker is more useful in that regard. The big one right now is better support for KDE2, and will probably be there with the next minor rev release.
My personal feeling on this subject is that while new features and new breakthroughs are great, a large majority of existing users barely make use of what's there now. This is probably because a lot of projects focus on "features" rather than on what they were originally meant to do, and most users will pick whatever their buddy is using at the time. Whether or not future technologies allow better interaction as a human<->computer interface will depend soley upon how they're implemented, and in the case of a window manager, I think most people would agree that our current idea of a window manager isn't it. Some developers will continue to concentrate on fixing the now, while others will worry about the future. Both groups are needed.
To my knowledge, the following is my own
unique wording. I did not copy this from a
website or other published source. If it's
been done before, I have no knowledge of it.
I release this into the public domain. It
is meant to be taken in a humourous context.
Microsoft is the end
Gates and Mundie it will send
Throwing all code you see
Into obscurity
Death of freedom choice
Never an open voice
Innovation's end
Never will it mend
Never
Fear
To begin the proprietary monopoly
Microsoft is the end
To begin the proprietary monopoly
Blacken our world Microsoft
Blustering of words
Free software is for the birds
Deadly game to play
Keeping geeks at bay
A pathetic license deal
Nothing left to steal
Never release code again
Microsoft is not your friend
Never
Uncertainty
To begin the proprietary monopoly
Microsoft is the end
To begin the proprietary monopoly
Blacken our world Microsoft
Microsoft
Opposition...Confrontation...Deviation...total lie
Embrace...Extend...Extinguish...standards die
Darkest hour
Proclaimed truths
True death of quality
Indication...Integration...Litigation...no justice for you and me
Humiliation...Indignation...Violation...software should be free
See our efforts
Trampled and destroyed
See our efforts die
Made the Microsoft way
Take our compilers away
Millions of our peers
Feel the grip and fears
Protesting in vain
Only corporatism remains
All is said and done
There will be no more fun
Never
Doubt
To begin the proprietary monopoly
Microsoft is the end
To begin the proprietary monopoly
FUD
Is the outcome of stupidity
Darkest plutocracy
In the exit of common sense and sanity
Blacken our world Microsoft
Microsoft
--with apologies to Metallica
I am now going to go post a copy of this outside
my door. We all know (within our community) that the whole notion of Linux and World Domination is
a fun joke. We aren't really at war with Microsoft.. we're all just trying to scratch those
programming itches by creating software that works for us (and yes, we release it for free in the hopes that it will help someone else). But out in the real-world, Microsoft is trying to dominate the world that you and I live in. This latest licensing agreement issue is just another
example of their determination to those ends.
First off, if you're not a joe average user and you're even remotely concerned about the quality
of your home Internet connection, completely and
utterly boycott cable modem. Read a story I wrote sometime ago.
Second, everyone saying you get what you pay for
is on the mark. However, there is no company anywhere on the planet that can guarantee you 24/7/365/life connectivity for any amount of money, no matter what they say. It comes down
to what level of service you can live with, and
accepting the fact that there will always
be problems at some point.
I have worked for several ISPs. I've dealt with
ILECs and CLECs. I can tell you that your best bet for DSL service is to find an ILEC or CLEC that offers DSL directly (i.e try to cut out as
many of the middlemen as possible). The trick to this is finding one that has a "fair" acceptible use policy in regards to home LANs, VPNs, personal servers, etc.
With my current DSL service in Chicago (PhoenixDSL is the ISP, NorthPoint is the CLEC, Ameritech is the ILEC), I am caught up in the middleman game. Phoenix is a decent (read: semi-responsive) provider, but when there are problems the only people I can bitch at is them. I can't go to NorthPoint or Ameritech directly, but instead have to get Phoenix to complain on my behalf. And let me just say that an ISP in their situation only has so much pull with the CLEC (given that the CLEC is what enables them to do business in the first place.) The ILECs and CLECs generally do not cooperate with one another, since they are the competition. Finding an ILEC or CLEC who offers DSL directly gives you slightly more control -- when you call them, they can't point the finger at the other guy.
Finally, take the suggestion of another poster. Take lots of detailed notes. Be a jerk in the sense that you don't let anyone push you around. At the same time though, remember that your $40-50/mo. plan works out to roughly $1.50 a day. So unless you're down for a month or are explictly paying for better service (i.e business class with an SLA), keep in mind that the initial person you're yelling at on the other end of the phone isn't responsible for your problems.
TRGpro gives the best of Palm and Handspring
on
The new Palm VIIx
·
· Score: 2
Wider IR port, better sound, CF slot, 8mb memory plus 2megs of flash default, ability to run apps off a CF flash card (340mb IBM micro-drive anyone?), ability to store & run apps in flash. Reasonably priced... you get what you pay for.
"The mouse is also a breakthrough in technological design. Apple will be the first company to bundle an optical mouse with all its desktop systems."
The breakthough in technology is a separate sentence, and is not being disputed in this case. The article leads the reader to think that Apple is the first to use optical mice with desktops, which as a matter of history is false.
I agree with your observations regarding the limitations of Sun optical mice, but no one here is claiming they were great. Perhaps they were great in their time.. they sure beat the old DEC VAX "larger than a hockey puck" mice.
This might have been mentioned in the past, but I am curious as to why Napster could not switch to being a subscription based service, much like ISP access. For $15 to $20 a month, you would sign up with a credit card and have unlimited access.
The goal of this being a way to clear up the "gray" area of piracy issues. Obviously, Napster would need to cut deals with the record labels/RIAA and independent artists that wanted royalties. For other artists who wanted to use it as a means of distribution, they would still be able to do so.
It is clear that the RIAA wouldn't like this either, because they want to be in control of the distribution, and want to be able to make money for every song, each time it's downloaded. But it would position Napster (or similar service) in a much more favorable light. It would be reasonable for them to argue that they have this extremely large user base, an infrastructure for doing something the RIAA currently cannot do, and that now they were doing "The Right Thing" by collecting fees which (theorectically) would make its way back to the artists.
How much different is this scheme from radio? If you view radio stations as customers, they have to pay royalties to the record labels to be able to play the songs.
I would be willing to pay for such a service (the ability to download individual songs of just about anything whenever I wanted) if it meant a) it would resolve the larger issue, and b) would ensure the artist was really getting the money for their work.
I'm surprised that this didn't come up before (at least I can't find it mentioned anywhere). The Field Museum, Chicago, unveiled Sue back on May 17th. Sue is the largest, most complete T.rex specimen known to exist.
If this were a Linux or FreeBSD story, everyone would be jumping all over it saying that it was the greatest thing they'd ever heard of. But because it's Sun pushing a technology paradigm that's been around for ages, and apparently doing so effectively in ways that really count, many people seem down on it.
What's up with that? We all have our biases, and we all like to make our voices be heard when injustices are being done in the industry, but this doesn't strike me as being something anyone should be putting Sun down for. Is the hardware and software working for this school in NYC? Sounds like it. Who are any of us to rant and rave about any company that is trying to put quality hardware & software to work in our schools. Consider this at least, it's not Microsoft.
For what it's worth, here is a picture of my UDB multia. I stuffed this puppy with a 4gig Seagate SCSI drive (ignore the 2gig Quantum shown), 96mb of parity RAM (maxed), and a self-wired external SCSI connector.
The system does run pretty hot (makes a very nice hand warmer in the winter). It's about the equivalent speed of a Pentium 90/100. It's currently running Red Hat 5.2 and is being used as an NFS server for mounting home dirs on the LAN, but I'm going to convert it to a dedicated web/database intranet server. Ocassionally I run rc5 and seti on it, although it's a dog of course.:)
Has anyone gotten the PCMCIA slots to work? It would be useful if I could put in a 10/100Mbit PCMCIA ethernet card. If you have, what's your experience with performance/problems?
While not directly related to this person's unfortunate incident, I have been keeping a log of my @Home service and have it published here. If you already have a cable-modem, or are planning on purchasing cable-modem service, chances are good that you want to read this. I have submitted this to/. as a frontpage story... it will be interesting to see if it makes it.
Slashdot Mirror
I just built two new identical systems for home.. one for my wife and one for me. Each box cost approximately $3800, a bit less than we paid for a 500Mhz G4 Titanium Powerbook w/512mb RAM.
The quick specs:
Lian Li PC60 aluminum case w/window & light kit
300W Sparkle power supply
Tyan Tiger MP S2460 w/dual 1.2Ghz Athlon MP's
Agilent Articoolers for the CPUs
512MB ECC reg. PC2100
Matrox G550 32MB DDR AGP dual
Samsung SyncMaster 170T (x2 per machine)
Seagate Barracuda III ATA100 7200rpm (30gb)
Toshiba 16X IDE DVD-ROM
Netgear FA310-TX
SB Live 5.1
Now, we already had speakers from our previous systems (both sets are Altec Lansing, mine is the ADA880 which supports AC3, surround, etc.).
Reading over some of the comments in this article, I agree completely that SCSI is still the best choice, but for pure desktop systems, it's sheer overkill (unless you are lucky enough to have IBM donating $15k so you can afford at least one drive). And in this case, they left out fibre channel and/or solid state drives. I mean come on! Our previous desktops were all SCSI, but I chose the IDE route again this time, mostly because the type of stuff we do on our desktops is not so I/O intensive as to cause problems. ATA100 is decent as long as you tweak it, but you have to buy the right kind of drive(s). Personally, I've found that Seagate drives are reliable in this regard (others may disagree, just note this is my personal preference and I recommend you pick whatever you feel comfortable with). Also, note that the Tyan Thunder board requires the huge 450W power supply, and it's not a standard ATX connector. The Tyan Tiger MP is the same exact board as the Thunder without all the built-in stuff (for me, built-in stuff is a major no-no for a desktop box as it makes it much harder to upgrade for the future), and it uses standard ATX power connectors. You do still need to be careful about the power supply you pick, though.
So, with our new desktops, I'll tie in what I was getting at above. If you're buying/building the ultimate box to be a standalone, multi-purpose, server and desktop box.. stop reading now. I dunno about everyone else, but I can't settle for that. My desktop is for desktop tasks. I have other systems on my LAN to do the various server related tasks that compliment my desktop system (i.e I try to spread the money around)
I won't go into real specifics here, but I'm not a Linux bigot. I'm a sysadmin, and I think every OS has a niche (well, maybe not all of them.. I'm a UNIX bigot). So with that in mind, I use FreeBSD for our dedicated firewall/router/wireless gateway, Solaris x86 for our NFS, NIS/LDAP, backup server (it's all SCSI with mirrored system and home storage disks), and Debian Linux for our media server, desktops, and home beowulf cluster. Our one Powerbook tri-boots MacOS 9.2.1, MacOS X 10.1, and Linux.
Anyway, the point is that an ultimate desktop box is all well and good, but I think anyone that goes out in search of the end-all, be-all holy grail standalone box is in for a disappointment down the road. And besides.. there is only so much fun you can have with one machine.
Check out our image gallery for pics and stuff.
In response to a few of the comments in this thread, I'd like to point out that Window Maker concentrates more on functionality than on "features". Every window manager has its own niche and is trying to appeal to a particular crowd. We try to appeal to those users that want a consistent, usable desktop that sits somewhere in between the minimal ones like Blackbox, and the eyecandy ones like E. See our comparison chart.
What's on the horizon? The developers recently reworked the code that generates and maintains files which help make up the UI (menus, states, etc.). To the end user, this means there is one less library dependency to worry about, providing they're the ones compiling the code. But the focus of this question was on "features". Window Maker won't be producing breakthroughs like what the Berlin project is doing, for instance. I get the impression this is what the questionaire was driving at. In our case, we're trying to improve compatibility with existing desktop environments (GNOME, GNUSTEP, and KDE) so that Window Maker is more useful in that regard. The big one right now is better support for KDE2, and will probably be there with the next minor rev release.
My personal feeling on this subject is that while new features and new breakthroughs are great, a large majority of existing users barely make use of what's there now. This is probably because a lot of projects focus on "features" rather than on what they were originally meant to do, and most users will pick whatever their buddy is using at the time. Whether or not future technologies allow better interaction as a human<->computer interface will depend soley upon how they're implemented, and in the case of a window manager, I think most people would agree that our current idea of a window manager isn't it. Some developers will continue to concentrate on fixing the now, while others will worry about the future. Both groups are needed.
Disclaimer:
Microsoft is the end
Gates and Mundie it will send
Throwing all code you see
Into obscurity
Death of freedom choice
Never an open voice
Innovation's end
Never will it mend
Never
Fear
To begin the proprietary monopoly
Microsoft is the end
To begin the proprietary monopoly
Blacken our world Microsoft
Blustering of words
Free software is for the birds
Deadly game to play
Keeping geeks at bay
A pathetic license deal
Nothing left to steal
Never release code again
Microsoft is not your friend
Never
Uncertainty
To begin the proprietary monopoly
Microsoft is the end
To begin the proprietary monopoly
Blacken our world Microsoft
Microsoft
Opposition...Confrontation...Deviation...total lie
Embrace...Extend...Extinguish...standards die
Darkest hour
Proclaimed truths
True death of quality
Indication...Integration...Litigation...no justice for you and me
Humiliation...Indignation...Violation...software should be free
See our efforts
Trampled and destroyed
See our efforts die
Made the Microsoft way
Take our compilers away
Millions of our peers
Feel the grip and fears
Protesting in vain
Only corporatism remains
All is said and done
There will be no more fun
Never
Doubt
To begin the proprietary monopoly
Microsoft is the end
To begin the proprietary monopoly
FUD
Is the outcome of stupidity
Darkest plutocracy
In the exit of common sense and sanity
Blacken our world Microsoft
Microsoft
I am now going to go post a copy of this outside my door. We all know (within our community) that the whole notion of Linux and World Domination is a fun joke. We aren't really at war with Microsoft.. we're all just trying to scratch those programming itches by creating software that works for us (and yes, we release it for free in the hopes that it will help someone else). But out in the real-world, Microsoft is trying to dominate the world that you and I live in. This latest licensing agreement issue is just another example of their determination to those ends.
So, in the spirt of the joke, enjoy.
First off, if you're not a joe average user and you're even remotely concerned about the quality of your home Internet connection, completely and utterly boycott cable modem. Read a story I wrote sometime ago.
Second, everyone saying you get what you pay for is on the mark. However, there is no company anywhere on the planet that can guarantee you 24/7/365/life connectivity for any amount of money, no matter what they say. It comes down to what level of service you can live with, and accepting the fact that there will always be problems at some point.
I have worked for several ISPs. I've dealt with ILECs and CLECs. I can tell you that your best bet for DSL service is to find an ILEC or CLEC that offers DSL directly (i.e try to cut out as many of the middlemen as possible). The trick to this is finding one that has a "fair" acceptible use policy in regards to home LANs, VPNs, personal servers, etc.
With my current DSL service in Chicago (PhoenixDSL is the ISP, NorthPoint is the CLEC, Ameritech is the ILEC), I am caught up in the middleman game. Phoenix is a decent (read: semi-responsive) provider, but when there are problems the only people I can bitch at is them. I can't go to NorthPoint or Ameritech directly, but instead have to get Phoenix to complain on my behalf. And let me just say that an ISP in their situation only has so much pull with the CLEC (given that the CLEC is what enables them to do business in the first place.) The ILECs and CLECs generally do not cooperate with one another, since they are the competition. Finding an ILEC or CLEC who offers DSL directly gives you slightly more control -- when you call them, they can't point the finger at the other guy.
Finally, take the suggestion of another poster. Take lots of detailed notes. Be a jerk in the sense that you don't let anyone push you around. At the same time though, remember that your $40-50/mo. plan works out to roughly $1.50 a day. So unless you're down for a month or are explictly paying for better service (i.e business class with an SLA), keep in mind that the initial person you're yelling at on the other end of the phone isn't responsible for your problems.
www.trgpro.com
Wider IR port, better sound, CF slot, 8mb memory plus 2megs of flash default, ability to run apps off a CF flash card (340mb IBM micro-drive anyone?), ability to store & run apps in flash. Reasonably priced... you get what you pay for.
Ok, that's enough.
"The mouse is also a breakthrough in technological design. Apple will be the first company to bundle an optical mouse with all its desktop systems."
The breakthough in technology is a separate sentence, and is not being disputed in this case. The article leads the reader to think that Apple is the first to use optical mice with desktops, which as a matter of history is false.
I agree with your observations regarding the limitations of Sun optical mice, but no one here is claiming they were great. Perhaps they were great in their time.. they sure beat the old DEC VAX "larger than a hockey puck" mice.
The goal of this being a way to clear up the "gray" area of piracy issues. Obviously, Napster would need to cut deals with the record labels/RIAA and independent artists that wanted royalties. For other artists who wanted to use it as a means of distribution, they would still be able to do so.
It is clear that the RIAA wouldn't like this either, because they want to be in control of the distribution, and want to be able to make money for every song, each time it's downloaded. But it would position Napster (or similar service) in a much more favorable light. It would be reasonable for them to argue that they have this extremely large user base, an infrastructure for doing something the RIAA currently cannot do, and that now they were doing "The Right Thing" by collecting fees which (theorectically) would make its way back to the artists.
How much different is this scheme from radio? If you view radio stations as customers, they have to pay royalties to the record labels to be able to play the songs.
I would be willing to pay for such a service (the ability to download individual songs of just about anything whenever I wanted) if it meant a) it would resolve the larger issue, and b) would ensure the artist was really getting the money for their work.
TANSTAFL
I'm surprised that this didn't come up before (at least I can't find it mentioned anywhere). The Field Museum, Chicago, unveiled Sue back on May 17th. Sue is the largest, most complete T.rex specimen known to exist.
Full details and additional information can be found at http://www.fieldmuseum.org/sue/
If this were a Linux or FreeBSD story, everyone would be jumping all over it saying that it was the greatest thing they'd ever heard of. But because it's Sun pushing a technology paradigm that's been around for ages, and apparently doing so effectively in ways that really count, many people seem down on it.
What's up with that? We all have our biases, and we all like to make our voices be heard when injustices are being done in the industry, but this doesn't strike me as being something anyone should be putting Sun down for. Is the hardware and software working for this school in NYC? Sounds like it. Who are any of us to rant and rave about any company that is trying to put
quality hardware & software to work in our schools. Consider this at least, it's not Microsoft.
For what it's worth, here is a picture of my UDB multia. I stuffed this puppy with a 4gig Seagate SCSI drive (ignore the 2gig Quantum shown), 96mb of parity RAM (maxed), and a self-wired external SCSI connector.
The system does run pretty hot (makes a very nice hand warmer in the winter). It's about the equivalent speed of a Pentium 90/100. It's currently running Red Hat 5.2 and is being used as an NFS server for mounting home dirs on the LAN, but I'm going to convert it to a dedicated web/database intranet server. Ocassionally I run rc5 and seti on it, although it's a dog of course. :)
Has anyone gotten the PCMCIA slots to work? It would be useful if I could put in a 10/100Mbit PCMCIA ethernet card. If you have, what's your experience with performance/problems?
While not directly related to this person's unfortunate incident, I have been keeping a log of my @Home service and have it published here. If you already have a cable-modem, or are planning on purchasing cable-modem service, chances are good that you want to read this. I have submitted this to /. as a frontpage story... it will be interesting to see if it makes it.