Once it is out in the marketplace, a device which includes a fully integrated XPort can provide the following:
Interactive data communications to and from the device through a standard Internet browser (using Java applets)
System software upgrades via the Internet
Remote configuration, monitoring, and troubleshooting, including real-time device performance notification via e-mail alerts
A dedicated co-processor to optimize network activities, permitting the host microprocessor to function at maximum efficiency "
Counts as a web server, from the sense of "this is something I can jack into my network without additional hardware and get useful functions out of" Now if you have another device hooked up to it, yeah that's a useful thing.
Unless there's something in the PDF that indicates this can function standalone, I am not impressed.
Unfortunately I cannot read PDFs or I would've checked it out myself.
So what's it's memory capacity? Can you load custom software into it? Does it operate in a stand-alone mode? These are the questions that were not answered for me.
Just because it's connected to the internet and has an http stack with a few cgi scripts does not make it a web server, in my opinion. A web server is more than pushing html out to the internet. A web server is a way of interacting with users in a manner that is meaningful to your business. I'm not seeing that here.
I didn't see where this is a full webserver. The documentation seemed to indicate that it's a TCP/IP handler. You put serial data in one side and TCP/IP network packets come out the other side.
At least, that's what it's targetted at; an addition to an existing embedded system. I don't think you could just write a backdoor and stick it on a network and expect it to work. Probably not enough memory/CPU capacity for that sort of thing...
Even more disappointing was the marketing spiel. I mean if you're gonna give an interview to ZDNet on the new initiative on which your company is betting its mobile division, you'd think they'd give him someone that can answer his questions in a clear, articulated manner, and not just continually steer him back to her marketing presentation on every question and comment.
Many posts talk about proposed changes to society, government, and technology to lessen the spam problem. However, an ISP has more insight into the problem than many others, and I thought I'd ask a question to tap that insight:
Given today's society, technology and infrastructure, what can an individual do that would be effective in reducing not only the personal strain of spam, but also lessen an ISP's burden.
What kind of strategies have you seen work. For instance, in particularly bad instances I'm prone to send an e-mail to spam@isp.net, abuse@isp.net, or admin@isp.net, but usually never even get a response. Is there a better thing to do? Are there things that are absolutely the wrong thing to do (such as replying to a spam)?
In short, what would you like to see users do in response to spam today?
As I read your post, I think "wow, what a great idea!"
Then I remember the cost of pcmcia cards.
But what is a pcmcia card? It's a standard interconnect technology that's pluggable, coupled with a standard physical form factor that is external and small.
Unfortunately, the real direction the industry has taken along those lines is USB. And the form factors, while external, were not standardized and are certainly not mating. As a matter of fact, increasingly I find my desk to be a jumble of USB devices hanging on the ends of their needlessly long cables.
A case of bawls to the first slashdotter to succesfully solve my quandry, which is this:
While I enjoy the universal ease of use of USB devices, the increasing amount of clutter every perihperal creates is dismaying, therefore a standard way of organizing disparate devices in such a way that it is tidy is needed. Rubber-banding the cables doesn't do it, though cheap, easy to manufacture, and easy to retrofit cable reels (including integral winding springs) WOULD do it. The solution should cost less than $50 for my 8 devices and be removable when needed.
Quite a few of Asimov's books are based on the fact that this "zeroth law" can be derived from the rest, and that once humanity starts building sufficiently complicated, intelligent, and emotent robots they realize it independently. What do you mean quite a few? R. Daneel and what's her face (Seldon's wife) are the only ones who were written to know about the zeroth law (well, none other mentioned by name), and that was only in "Foundation and Earth" and "Forward the Foundation."
So none other mention it by name. The laws of inertia had no such name until Newton wrote it down; that doesn't change the fact that they influenced nearly every greek drama in some way.
Consider the following:
The entire foundation series only makes sense in the context of a zeroth law. R. Daneel committed murder for the good of the human race in one of the early books (as a matter of fact, I think it was his first appearance). Nearly every conflict of the First Law must be resolved by the zeroth law, and many of the robot books Asimov wrote was written about a robot detective solving an apparent First Law violation by a robot.
That last statement is not from personal experience, but from a biography I read of Asimov once that stated that he wrote books about interesting interplays between the Laws of Robotics.
Therefore, it could be said, that much of Asimov's work was devoted to the zeroth law.
BTW, I just got done reading the novel (as I imagine many slashdotters are doing now) and it too appears to be a zero-law commentary...
Obviously he forgot that one. The one that says that the survival of the human species comes before the first three laws.
Quite a few of Asimov's books are based on the fact that this "zeroth law" can be derived from the rest, and that once humanity starts building sufficiently complicated, intelligent, and emotent robots they realize it independently.
For instance, a robot that commits murder because it prevents a larger attrocity, a larger amount of harm to humanity, to occur.
I surmise that the Singularity is acting in such a manner, acting to prevent the largest amount of harm that it can, and that its choice of prioritization in this is somewhat to question...
Define a moral action to be one which is ultimately preserving of an entity's own existance.
In this sense, the moral obligation of Symantec is to increase shareholder value.
This action was moral on their part, as they were providing a service to customers that were paying for that service. If you wanted the service you should have paid for it.
This was not necessarily an ETHICAL decision however; ethics is the act of applying morals to a larger body, such as a town, a state, a country, or the entire species.
Wow, thats Symantec's ONLY moral responsibility? If thats the case, then they should sell crack to kids to boost thier profit margin. After all, acording to you the only thing that matters is if they make money. Sorry, i dont buy it. Companies have an obligation to make money, but its neither thier only one, nor thier highest one.
Not selling deadly narcotics to children IS a moral responsibility; doing so would doubtless result in a loss in shareholder value.
You assert that a company has some obligation other than to increase shareholder value. I'm working on the assumption that there aren't any magical obligations that everyone just automatically has to do for no good reason (and I apparantly wasn't informed of) when I ask:
Why does a company have any obligation other than to make money? What are these obligations, and in what way has the company incurred them? I'll start off by defining an obligation as a responsibility one owes due to receiving something of value from another entity.
but Symantec has a moral responsibility to inform the public if it thinks millions will be affected.
Symantec does not have a moral responsibility to inform the public. Symantec isn't a publicly funded corporation, or a government agency.
You do not have a right to benefit for free from the hard work of others. Symantec's ONLY moral responsiblity is to increase value to their shareholders. This isn't the late 1990's where you can create a technology company based on the idea of giving things away for free and expect that to fly.
Part of that responsiblity is to treat their customers right. Given a limited timeline, and the need to provide the most value possible, they chose to send an alert to some of their (presumably) biggest and best customers. I believe that Symantec worked in a very appropriate manner in this case.
Note: I didn't read the article. I did read quite a few articles yesterday when the link was posted on hardocp.com however.
Use a small CRT. 13" or so. Not only are these cheap (used), they're what the fresnel is designed for. The bigger the screen is compared to the fresnel, the less distinct it'll be.
Turn brightness all the way up, right up to the edge of getting pixel bleed
Don't project more than a couple feet... beyond that it's too dim to be daylight readable...
I use this in my MAME arcade machine... 13" lcd projecting onto smoked glass. It works pretty good, especially with the arcade machine not having any direct light on it (it's in a corner of my living room, approximately 5 feet away from the sliding glass doors along the same wall) For the smoked glass, I started with snow frosting you can find in art stores around christmas time, but moved to acrylic sanded with 240 grit sandpaper. The higher the grit the better resolution you get...
I've also had this problem. I use an IBM Thinkpad 600 at work and the battery dies on me when I undock. It says it's just fine, but within minutes warns me that the battery is about to die, very similar to the report in the linked article.
It starts out fine with a new battery and gets progressively worse. I asked our helpdesk about it, and the helpdesk representative said that this sytem occasionally needs to be unplugged and the battery allows to run down, indicating this is a memory effect of some kind. His advice was once every 2 weeks, and he indicated that IBM suggested this to him. I'm a consultant but we have contracts both with external clients, and with clients who are sister companies. It is not uncommon therefore for me to not need to undock my laptop for months at a time.
I've gone through 2 batteries in a year. The batteries last about a month, after which I start dragging my power adapter around with me while the paperwork gets filled out for a new battery. This one died about 6 months ago, but my laptop is (and has been) scheduled for replacement since then so I haven't bothered to get it replaced.
Damn. Your letter is much better than mine. "reprehensible" I'll have to remember that... much better than "evil" I think.
Oh well, hopefully they'll get the point. I agree with another responder of my post; we should contact the device manufacturers and let THEM know. Care to draft a letter? I think you'd do a better job at it than me... something laong the lines of "I've used your products and love them, but as you don't offer alternatives to the PCI standard I'll have to stop using them because they're evil and blah blah yadda yadda"
If only it were so simple. All motherboards, even highly integrated ones, use PCI for their interconnects. Your sound card (chip) is on the PCI bus even if its not plugged into a socket. The only things that aren't on the PCI bus are the ones plugged into the northbridge, because northbridges talk to their southbridge via PCI much of the time.
I think the nForce and maybe some of the Intel stuff is an exception.
That said I am letting them know I'm unhappy and will be avoiding PCI whenever possible.
Wanna read the letter I sent? I sent this to: Alan.Deikman@znyx.com
Here. Copy freely. Might wanna change the personal details, but the overall theme is I think quite good:
[blockquote] This is in response to the Cease & Desist sent to Jim Boelmer, maintainer of the PCI Device ID list located at http://www.yourvote.com/pci/. I'm sure your mailbox is full of mail like this today, so I'll be brief.
I have used Jim Boemler's PCI Device List in the past. In fact I have even contributed to it. I feel that it's existance does not infringe on the PCI trademark, and in fact helps PCI_SIG's brand by providing support for PCI devices in free and open operating systems.
I feel that Cease & Desist letters are dirty and underhanded, and are not the way that a responsible corporate citizen chooses to begin conflict resolution. While ethical, C&Ds are certainly not moral, and do not contribute to the universal good, except in the case where other avenues have been explored first. I believe this case is particularly evil because Mr. Boemler has contacted PCI-SIG in the past about this matter (hosting the PCI device list), and not received any response.
I am a computer consultant at a fairly large company. We specialize in systems management, e-business, and e-branding. I'm sure, in view of your already wonderfully moral way of handling matters, that you'll forgive me if I don't tell you which one. As a result of the matter with Mr. Boemler I will now be strongly advising my friends, family, associates, and clients to avoid the use of PCI devices whenever possible, and will be recommending alternate serial and bus interconnect technologies. I would do the same with my customers if I thought it appropriate. I will also recommend that anyone who does not purchase a PCI device as a result of this writes to the device manufacturer, expressing their concern with the PCI-SIG group, and advising that the device would have been purchased if it were available in another technology.
I strongly urge the PCI-SIG to seek amicable resolution in this matter with Mr. Boemler. This would convince me that PCI-SIG is not so evil that devices using PCI technology or logos should not be used by anyone, ever.
I would like to see the free PCI device list come back into being because, as I said, I believe it to be beneficial to every party involved. [/blockquote]
My #1 advice to those running MAME on older hardware:
Video card. Get a good one. 3D acceleration isn't important, except for the fact that most 3D accelerator cards build their 3D acceleration on top of a solid 2D acceleration engine.
Beware of early 3D cards, though; some of these actually made 2D speeds WORSE.
AGP is good, but not required. Other than that, any decent older 3D card should do the job. 8MB or more of RAM on the vid card is good.
FYI, my home MAME cabinet is running off a P2-450 w/ 2 64 MB sticks. I've also hooked up my VIA Eden (800 MHz, 256 MB of RAM) without much of a difference between the two...
Games that leave a little to be desired:
LightBringer (runs a little slow; I think this was a limitation of the hardware being emulated, not the emulator itself)
Mortal Kombat - load times are long. Very playable once loaded
Marvel vs. Street Fighter - load times are long. The intro scenes are a little choppy. Very playable otherwise
Most NeoGeo games - get a little choppy when large parts of the screen start getting updated, but overall pretty playable.
Built in TV-OUT (RCA & S-Video), Ethernet, and Line-out (converters to RCA available for ~ $5), as well as SPDIF if you don't need the RCA TV-OUT.
All available for ~ $100 w/proc. It's quiet. It's cheap. It's functional.
Throw in a couple old, otherwise useless, sticks of SDRAM... ditto on the P/S... build your own OS (it uses x86, so pick your favorite distro)... make it a seldom-write OS and put it on a compact flash card. Put a reader in it somewhere (you can get them inexpensively here)... presto chango, you have a completely solid-state computer that you paid < $200 for. It's got USB so if you want a WI-FI card, add it yourself. It even has internal headers for it so you don't have to have a dongle hanging outside your case.
FYI, I have one. I put a full hard drive in mine. Plus DVD player. Plus serial-port IR remote control receiver. Total cost was ~ $350, but a lot of the components have come down in price since then. I use Windows 98 SE with Media Player 9 and Real Networks One Player v.1 and it works great. The OS license was from an old computer that I've installed Linux on, so it was free. I even wrote a custom app to boot into in VB. It took about a day, and it lets me get to most of my common functionality easily, even when running in my car. It's also convenient (and snazzy) when I have friends over. Doesn't look like Windows, it looks like a high-end av component.
IIRC, Real Networks doesn't have a Linux driver and NFS doesn't have a Windows driver, so you may be SOL as far as compatibility on both of those counts... samba may be your answer. Or one of the unix compatibility toolkits for windows. Or if WINE is working well enough, that may be the way to go.
BTW, details and fun ideas about this form factor can be had at this link
If anyone is interested in my experiences or tips about this platform, feel free to e-mail me at the slashcode-guarded e-mail address listed above. I get lots of spam so if I don't answer within a couple days, try a resend.
Way to copy that last statement out of your friend's thesis!!!
Actually I made it up on the spot out of memory. Of course, it helps that I modelled them for high school physics class... that and I read all of Niven's books...
Slashdot Mirror
The real question isn't: "Is that list worth $50 Million to the RIAA?" but "Is having my name on that list worth $12.60 to me?"
My answer was no.
Wouldn't that more appropriately be:
The article also fails to mention that Farechaser was sentenced to death by legal injunction in the same decision...
Awesome... I can build a dongle to hook up my TRS-80 to the Internet!
:)
Anyone ever write a web browser in BASIC before?
Sir I believe you would have a record on that one.
Though this is I think more geared to web serving, it should have other capabilities.
I don't see how this:
"How the XPort Can Transform Your Product
Once it is out in the marketplace, a device which includes a fully integrated XPort can provide the following:
Interactive data communications to and from the device through a standard Internet browser (using Java applets)
System software upgrades via the Internet
Remote configuration, monitoring, and troubleshooting, including real-time device performance notification via e-mail alerts
A dedicated co-processor to optimize network activities, permitting the host microprocessor to function at maximum efficiency "
Counts as a web server, from the sense of "this is something I can jack into my network without additional hardware and get useful functions out of" Now if you have another device hooked up to it, yeah that's a useful thing.
Unless there's something in the PDF that indicates this can function standalone, I am not impressed.
Unfortunately I cannot read PDFs or I would've checked it out myself.
So what's it's memory capacity? Can you load custom software into it? Does it operate in a stand-alone mode? These are the questions that were not answered for me.
Just because it's connected to the internet and has an http stack with a few cgi scripts does not make it a web server, in my opinion. A web server is more than pushing html out to the internet. A web server is a way of interacting with users in a manner that is meaningful to your business. I'm not seeing that here.
I didn't see where this is a full webserver. The documentation seemed to indicate that it's a TCP/IP handler. You put serial data in one side and TCP/IP network packets come out the other side.
At least, that's what it's targetted at; an addition to an existing embedded system. I don't think you could just write a backdoor and stick it on a network and expect it to work. Probably not enough memory/CPU capacity for that sort of thing...
http://www.allelectronics.com/
Lots of discrete components, and odd stuff. Good selection, GREAT deals.
I was disappointed by the logo...
Even more disappointing was the marketing spiel. I mean if you're gonna give an interview to ZDNet on the new initiative on which your company is betting its mobile division, you'd think they'd give him someone that can answer his questions in a clear, articulated manner, and not just continually steer him back to her marketing presentation on every question and comment.
I am not impressed.
Many posts talk about proposed changes to society, government, and technology to lessen the spam problem. However, an ISP has more insight into the problem than many others, and I thought I'd ask a question to tap that insight:
Given today's society, technology and infrastructure, what can an individual do that would be effective in reducing not only the personal strain of spam, but also lessen an ISP's burden.
What kind of strategies have you seen work. For instance, in particularly bad instances I'm prone to send an e-mail to spam@isp.net, abuse@isp.net, or admin@isp.net, but usually never even get a response. Is there a better thing to do? Are there things that are absolutely the wrong thing to do (such as replying to a spam)?
In short, what would you like to see users do in response to spam today?
As I read your post, I think "wow, what a great idea!"
Then I remember the cost of pcmcia cards.
But what is a pcmcia card? It's a standard interconnect technology that's pluggable, coupled with a standard physical form factor that is external and small.
Unfortunately, the real direction the industry has taken along those lines is USB. And the form factors, while external, were not standardized and are certainly not mating. As a matter of fact, increasingly I find my desk to be a jumble of USB devices hanging on the ends of their needlessly long cables.
A case of bawls to the first slashdotter to succesfully solve my quandry, which is this:
While I enjoy the universal ease of use of USB devices, the increasing amount of clutter every perihperal creates is dismaying, therefore a standard way of organizing disparate devices in such a way that it is tidy is needed. Rubber-banding the cables doesn't do it, though cheap, easy to manufacture, and easy to retrofit cable reels (including integral winding springs) WOULD do it. The solution should cost less than $50 for my 8 devices and be removable when needed.
the focus seems to be the ambiguity over what counts as a human under the laws of robotics
Actually, the solution is the ambiguity over what counts as a human.
And I thought it was great at how believable the emergent behaviour was when couched in terms of the GAT...
Quite a few of Asimov's books are based on the fact that this "zeroth law" can be derived from the rest, and that once humanity starts building sufficiently complicated, intelligent, and emotent robots they realize it independently.
What do you mean quite a few? R. Daneel and what's her face (Seldon's wife) are the only ones who were written to know about the zeroth law (well, none other mentioned by name), and that was only in "Foundation and Earth" and "Forward the Foundation."
So none other mention it by name. The laws of inertia had no such name until Newton wrote it down; that doesn't change the fact that they influenced nearly every greek drama in some way.
Consider the following:
The entire foundation series only makes sense in the context of a zeroth law.
R. Daneel committed murder for the good of the human race in one of the early books (as a matter of fact, I think it was his first appearance).
Nearly every conflict of the First Law must be resolved by the zeroth law, and many of the robot books Asimov wrote was written about a robot detective solving an apparent First Law violation by a robot.
That last statement is not from personal experience, but from a biography I read of Asimov once that stated that he wrote books about interesting interplays between the Laws of Robotics.
Therefore, it could be said, that much of Asimov's work was devoted to the zeroth law.
BTW, I just got done reading the novel (as I imagine many slashdotters are doing now) and it too appears to be a zero-law commentary...
Best. Spoiler. Ever.
Seriously, I had guessed at most of this from my sporadic reading, but damn I hadn't surmised all of that...
Obviously he forgot that one. The one that says that the survival of the human species comes before the first three laws.
Quite a few of Asimov's books are based on the fact that this "zeroth law" can be derived from the rest, and that once humanity starts building sufficiently complicated, intelligent, and emotent robots they realize it independently.
For instance, a robot that commits murder because it prevents a larger attrocity, a larger amount of harm to humanity, to occur.
I surmise that the Singularity is acting in such a manner, acting to prevent the largest amount of harm that it can, and that its choice of prioritization in this is somewhat to question...
Define a moral action to be one which is ultimately preserving of an entity's own existance.
In this sense, the moral obligation of Symantec is to increase shareholder value.
This action was moral on their part, as they were providing a service to customers that were paying for that service. If you wanted the service you should have paid for it.
This was not necessarily an ETHICAL decision however; ethics is the act of applying morals to a larger body, such as a town, a state, a country, or the entire species.
Wow, thats Symantec's ONLY moral responsibility? If thats the case, then they should sell crack to kids to boost thier profit margin. After all, acording to you the only thing that matters is if they make money. Sorry, i dont buy it. Companies have an obligation to make money, but its neither thier only one, nor thier highest one.
Not selling deadly narcotics to children IS a moral responsibility; doing so would doubtless result in a loss in shareholder value.
You assert that a company has some obligation other than to increase shareholder value. I'm working on the assumption that there aren't any magical obligations that everyone just automatically has to do for no good reason (and I apparantly wasn't informed of) when I ask:
Why does a company have any obligation other than to make money? What are these obligations, and in what way has the company incurred them? I'll start off by defining an obligation as a responsibility one owes due to receiving something of value from another entity.
but Symantec has a moral responsibility to inform the public if it thinks millions will be affected.
Symantec does not have a moral responsibility to inform the public. Symantec isn't a publicly funded corporation, or a government agency.
You do not have a right to benefit for free from the hard work of others. Symantec's ONLY moral responsiblity is to increase value to their shareholders. This isn't the late 1990's where you can create a technology company based on the idea of giving things away for free and expect that to fly.
Part of that responsiblity is to treat their customers right. Given a limited timeline, and the need to provide the most value possible, they chose to send an alert to some of their (presumably) biggest and best customers. I believe that Symantec worked in a very appropriate manner in this case.
Note: I didn't read the article. I did read quite a few articles yesterday when the link was posted on hardocp.com however.
I've used a fresnel lens. A couple notes:
Use a small CRT. 13" or so. Not only are these cheap (used), they're what the fresnel is designed for. The bigger the screen is compared to the fresnel, the less distinct it'll be.
Turn brightness all the way up, right up to the edge of getting pixel bleed
Don't project more than a couple feet... beyond that it's too dim to be daylight readable...
I use this in my MAME arcade machine... 13" lcd projecting onto smoked glass. It works pretty good, especially with the arcade machine not having any direct light on it (it's in a corner of my living room, approximately 5 feet away from the sliding glass doors along the same wall) For the smoked glass, I started with snow frosting you can find in art stores around christmas time, but moved to acrylic sanded with 240 grit sandpaper. The higher the grit the better resolution you get...
I've also had this problem. I use an IBM Thinkpad 600 at work and the battery dies on me when I undock. It says it's just fine, but within minutes warns me that the battery is about to die, very similar to the report in the linked article.
It starts out fine with a new battery and gets progressively worse. I asked our helpdesk about it, and the helpdesk representative said that this sytem occasionally needs to be unplugged and the battery allows to run down, indicating this is a memory effect of some kind. His advice was once every 2 weeks, and he indicated that IBM suggested this to him. I'm a consultant but we have contracts both with external clients, and with clients who are sister companies. It is not uncommon therefore for me to not need to undock my laptop for months at a time.
I've gone through 2 batteries in a year. The batteries last about a month, after which I start dragging my power adapter around with me while the paperwork gets filled out for a new battery. This one died about 6 months ago, but my laptop is (and has been) scheduled for replacement since then so I haven't bothered to get it replaced.
Damn. Your letter is much better than mine. "reprehensible" I'll have to remember that... much better than "evil" I think.
Oh well, hopefully they'll get the point. I agree with another responder of my post; we should contact the device manufacturers and let THEM know. Care to draft a letter? I think you'd do a better job at it than me... something laong the lines of "I've used your products and love them, but as you don't offer alternatives to the PCI standard I'll have to stop using them because they're evil and blah blah yadda yadda"
If only it were so simple. All motherboards, even highly integrated ones, use PCI for their interconnects. Your sound card (chip) is on the PCI bus even if its not plugged into a socket. The only things that aren't on the PCI bus are the ones plugged into the northbridge, because northbridges talk to their southbridge via PCI much of the time.
I think the nForce and maybe some of the Intel stuff is an exception.
That said I am letting them know I'm unhappy and will be avoiding PCI whenever possible.
Wanna read the letter I sent? I sent this to: Alan.Deikman@znyx.com
Here. Copy freely. Might wanna change the personal details, but the overall theme is I think quite good:
[blockquote]
This is in response to the Cease & Desist sent to Jim Boelmer, maintainer of the PCI Device ID list located at http://www.yourvote.com/pci/. I'm sure your mailbox is full of mail like this today, so I'll be brief.
I have used Jim Boemler's PCI Device List in the past. In fact I have even contributed to it. I feel that it's existance does not infringe on the PCI trademark, and in fact helps PCI_SIG's brand by providing support for PCI devices in free and open operating systems.
I feel that Cease & Desist letters are dirty and underhanded, and are not the way that a responsible corporate citizen chooses to begin conflict resolution. While ethical, C&Ds are certainly not moral, and do not contribute to the universal good, except in the case where other avenues have been explored first. I believe this case is particularly evil because Mr. Boemler has contacted PCI-SIG in the past about this matter (hosting the PCI device list), and not received any response.
I am a computer consultant at a fairly large company. We specialize in systems management, e-business, and e-branding. I'm sure, in view of your already wonderfully moral way of handling matters, that you'll forgive me if I don't tell you which one. As a result of the matter with Mr. Boemler I will now be strongly advising my friends, family, associates, and clients to avoid the use of PCI devices whenever possible, and will be recommending alternate serial and bus interconnect technologies. I would do the same with my customers if I thought it appropriate. I will also recommend that anyone who does not purchase a PCI device as a result of this writes to the device manufacturer, expressing their concern with the PCI-SIG group, and advising that the device would have been purchased if it were available in another technology.
I strongly urge the PCI-SIG to seek amicable resolution in this matter with Mr. Boemler. This would convince me that PCI-SIG is not so evil that devices using PCI technology or logos should not be used by anyone, ever.
I would like to see the free PCI device list come back into being because, as I said, I believe it to be beneficial to every party involved.
[/blockquote]
My #1 advice to those running MAME on older hardware:
Video card. Get a good one. 3D acceleration isn't important, except for the fact that most 3D accelerator cards build their 3D acceleration on top of a solid 2D acceleration engine.
Beware of early 3D cards, though; some of these actually made 2D speeds WORSE.
AGP is good, but not required. Other than that, any decent older 3D card should do the job. 8MB or more of RAM on the vid card is good.
FYI, my home MAME cabinet is running off a P2-450 w/ 2 64 MB sticks. I've also hooked up my VIA Eden (800 MHz, 256 MB of RAM) without much of a difference between the two...
Games that leave a little to be desired:
LightBringer (runs a little slow; I think this was a limitation of the hardware being emulated, not the emulator itself)
Mortal Kombat - load times are long. Very playable once loaded
Marvel vs. Street Fighter - load times are long. The intro scenes are a little choppy. Very playable otherwise
Most NeoGeo games - get a little choppy when large parts of the screen start getting updated, but overall pretty playable.
VIA Eden my friend.
Built in TV-OUT (RCA & S-Video), Ethernet, and Line-out (converters to RCA available for ~ $5), as well as SPDIF if you don't need the RCA TV-OUT.
All available for ~ $100 w/proc. It's quiet. It's cheap. It's functional.
Throw in a couple old, otherwise useless, sticks of SDRAM... ditto on the P/S... build your own OS (it uses x86, so pick your favorite distro)... make it a seldom-write OS and put it on a compact flash card. Put a reader in it somewhere (you can get them inexpensively here)... presto chango, you have a completely solid-state computer that you paid < $200 for. It's got USB so if you want a WI-FI card, add it yourself. It even has internal headers for it so you don't have to have a dongle hanging outside your case.
FYI, I have one. I put a full hard drive in mine. Plus DVD player. Plus serial-port IR remote control receiver. Total cost was ~ $350, but a lot of the components have come down in price since then. I use Windows 98 SE with Media Player 9 and Real Networks One Player v.1 and it works great. The OS license was from an old computer that I've installed Linux on, so it was free. I even wrote a custom app to boot into in VB. It took about a day, and it lets me get to most of my common functionality easily, even when running in my car. It's also convenient (and snazzy) when I have friends over. Doesn't look like Windows, it looks like a high-end av component.
IIRC, Real Networks doesn't have a Linux driver and NFS doesn't have a Windows driver, so you may be SOL as far as compatibility on both of those counts... samba may be your answer. Or one of the unix compatibility toolkits for windows. Or if WINE is working well enough, that may be the way to go.
BTW, details and fun ideas about this form factor can be had at this link
If anyone is interested in my experiences or tips about this platform, feel free to e-mail me at the slashcode-guarded e-mail address listed above. I get lots of spam so if I don't answer within a couple days, try a resend.
Way to copy that last statement out of your friend's thesis!!!
Actually I made it up on the spot out of memory. Of course, it helps that I modelled them for high school physics class... that and I read all of Niven's books...