Aye, at least $50/hr and generally more like $60-$80/hr for regular customers and closer to the $100/hr mark for new business.
Because not only do you have to pay your own medical / taxes / insurance / etc. out of that gross, but also tools, training, books. Then there's the risk of lawsuits, fraud by clients, clients who won't pay until after 90 days. Plus the months out of the year that you might not have business due to slow economy, sick leave, or just wanting some vacation time.
I charge $60/hr because most of my costs are already covered by my day job. And I really only want about 5-10 hours of extra work per week. If I was doing it for my own as a primary business, I'd charge $75-$80.
Or simply allowing you to read data off of all N platters at the same time (which is probably closer to being possible) with the current design. The downside is that it would probably limit the drive structure (meaning you could have 1 head, 2 heads, 4 heads or 8 heads).
And it still wouldn't improve seek latency as much as a 2nd set of heads or simply RAID'ing together more spindles. Or increasing those RPMs again while using an even smaller diameter disk.
I've watched enough folks over the years go with the absolute cheapest parts they could find. Then I get to watch them struggle later on with either stability issues (if not outright parts failures) or upgrades (if the machine survives past a year or two). I've even done it myself on occasion, thinking that it's worth saving $10 now only to have it come around and bite me a year down the road. Upgrades are the main reason why I avoid cheaper / proprietary solutions from the pre-built folks. With commodity parts, you can usually re-use numerous components in the next-gen system.
So now I tend to go for the middle-tier parts. Motherboards from companies like Asus, Gigabyte, MSI. Power-supplies from good brands. HDs with the 3-yr or 5-yr warranty. Well-designed cases that designed for cooling and noise reduction (Antec, Lian Li, Thermaltake, etc.).
Still, it looks like I can build a pretty decent system for $300. Bottom-end performance, but definitely good enough to run a Linux distro and do word-processing / web-browsing / e-mail. That gets my geek blood pumping as I consider the possibilities. And it's a machine that I wouldn't be ashamed to give someone.
Hmm, $150 is a tough price point to hit for a regular PC if you're at all picky about components.
$45 CPU (AM2 Sempron 2800+, which means upgrade capability later)
$65 M/B (GeForce 6150 w/ integrated video)
$55 2x256 or 1x512MB (or $25 256MB single-stick)
$20 DVD-ROM
$25 Case+PSU
$45 Hard drive
===
$255
More like $300 once you buy a reasonably good case w/ PSU for $50-$75. And it would be expandable to put more memory and a more powerful CPU in it down the road.
You could still probably shave $100 off that price if you go for close-out deals, really cheap motherboards, older CPUs that are only $20ea, 256MB of RAM, and a really cheap case+PSU that will probably catch fire right after the warranty expires.
(The machines I'm building for work are around $500 for parts, but those are dual-core w/ 2GB RAM.)
GPG encryption inside regular text files (one per service) stored in a version control system. The contents of the files are encrypted using the public keys of everyone who is supposed to be able to decrypt the file contents. Unfortunately, it probably won't scale past a dozen users.
Doesn't handle keeping track of access. Although you could possibly do some fancy stuff with Subversion scripts to track that sort of thing.
I'll second GPG. Although I prefer to keep each system in a separate file (makes it easier to look at the change dates and see when things were modified).
And once you're using GPG ASCII blocks inside regular text files, you can simply dump those individual files into a version control system or central network share. Backups are also dirt-simple, e-mail everything to yourself at some other location or print out the sheets and rely on OCR in the worst-case scenario.
And yes, I'll be looking into "escrow". Our system works because there are only half a dozen of us that need to share passwords. Often I'll change a password and then e-mail the ASCII-encrypted text block to the group. I should really move towards putting these text files into our VCS.
Fortunately for AMD, they have managed to start their own 65nm manufacturing now.
There were articles about 2 weeks ago about how they're having issues with 65nm. They can't get the voltages as low as they want, which is requiring increased power requirements. Dunno if they've got it sorted out yet.
They do have some technical advantages over Intel once they get 65nm working (such as SS and SOI which I don't think Intel uses).
AMD already dropped prices on Athlon64 back in July (and that was scheduled a few months in advance, maybe as early as Feb 2006). Their next scheduled price are in October, but are only on the Sempron or Turion chip lines.
Any price drops outside of those 2 months or on chips other then Sempron/Turion would be news.
I agree that having multiple pages just to increase ad impressions is a little annoying, but why do people gripe about a site trying to profit from its work/research/journalism/whatever? How else is it going to make money/pay for the bandwidth that slashdot/anyone generates? Get off your soapbox!
1) Some things make sense to be split to multiple pages. Such as cases where you have 10 images for each of the benchmarks that you ran and you put them on multiple pages to make life easier for dial-up users.
2) The article content is typically split across too many pages so that the ads overwhelm the article. If the article covered 50-70% of the page, people probably wouldn't care as much, but often it seems like we're getting down into the single paragraph per page.
3) Every time you split to a new page, the reader loses their focus. This frustrates your visitor. Especially when the page is so cluttered that you have to hunt for the next page link. Visitors are coming to your site to read the article and get information, not to be confused and bombarded by as many ads as you can throw at them.
I'm all for attempting to make money off of ad impressions, but splitting a 25 paragraph review over 25 pages is well past ridiculous. (Most of these reviews should've been done in about 1/4 of the number of pages that they're programmed for.)
I tend to buy whatever speed CPU is at the knee of the price/performance curve. I usually just calculate this based on the clock speed, which isn't perfect and somewhat naive but works quickly when all the chips are based on the same core. For example:
So for the above example, the $85/GHz is probably what I would purchase. This graph used to be a lot clearer with the Opteron 2xx series where there are 5 different single-core chips. Prices went up slowly, but the upper 2 in the group were significantly more expensive then the lower 3.
$DIETY bless the blinky bits, there's a lot of joy to be had at watching the flickering lights of a multi-disk array that you built yourself. Watching in real-time as different disks get used to service requests can be a bit mesmerizing. (Plus it points out a possible performance issue with mdadm's RAID10 implementation when disk 0 has a higher utilization then the other disks.) Of course, I used to be entranced by the lights on the front of my 14.4Kbps modem.
All in all I'm glad that Intel has decided to retake the lead in the price/performance war, AMD needs a new kick in the pants.
Indeed, prices for AMD processors have been a bit stagnant for the past few years. I think the Opterons topped out at 2.6GHz, but are very expensive. If nothing else, at least the new Intel Core Duo chips forced them to finally drop prices on the X2's. Now if we can just see some price pressure on the Opteron dual-cores I'd be a very happy geek.
For usability of said web/business applications, when dual-core only adds ~$70 to the price of a $800 system, why not add dual-core for reponsiveness?
That's been our decision. The price breaks on AMD X2 chips back in July 2006 made it an easy one ($300 chip now sells for $150, only $50 more then the $100 CPUs were were buying). For such a minor price increase, we get another few years of longevity due to the 2nd core.
Should be pretty easy for that system to save the user $50 over the course of a year because they're not waiting on all sorts of minor delays. There are even some energy-efficient X2 3800+ chips that just hit the street that could eliminate even the excess power consumption argument. Not that modern CPUs are that bad; they'll ramp their speeds back down when the power isn't needed.
I guess there's room in a SD card for multiple chips. It's just that the SD Card is the size of a postage stamp with the upper 1/4 consumed by the contact points. I guess as long as the 16Gbit chips are less then around 0.8 cm^2 then there's enough room inside the packaging for a pair of them.
Samsung has announced 32-gigabit (4GB) flash chips, not 32 megabit.
Considering that I can get a 4GB SD flash card for $60, I'm guessing that those 32 gigabit chips have been out for a while? Or do those tiny SD cards have 2 chips in them?
Pity that you posted anonymously... Seagate Momentus are 160GB, but you're right, only 5400rpm. OTOH, the older 7200.1 Momentus are 100GB and 7200rpm. So there's a good chance that as PR goes mainstream they can do a 7200rpm 160GB drive.
RPM only buys you better seek times. And on the 100GB Momentus drives, it's a difference between 8.5ms and 10.5ms. The increased density of the PR platters will give increased transfer rates even at 5400rpm.
A classic example would be the Chronicles of Narnia series, with the lion Aslan clearly representing Jesus Christ in the mythology of Narnia.
Mmm, C.S. Lewis also despised allegory from what I recall. They were close aquaintances at the time and I don't think either story was written to be allegorical. But I could be confused... I think the biographical video on the FotW extended edition talks about this a bit when discussing both authors.
Not to be a stick in the mud about this but Jackson seems to have gone out of his way to take some of Tolkien's original work and warp it. Things that he could have easily have left as written were rewritten to suit his "needs" but not the needs of the story line.
Meh, you're entitled to your opinion. Personally, I watched the extended editions and listened to the audio commentaries. PJ and crew do a good job explaining the whys and wherefores of their changes. I might not agree with all of them, but on the whole they were trying to make a good set of movies that were entertaining while simultaneously staying as true as possible to the books.
Not an easy task. And from the box office results, they at least succeeded at making something that the majority of folks enjoyed. (Even though I disliked the dwarf jokes and the elven tricks such as shield-surfing.)
Sure, it was warped in places, but usually because of the realities of film-making so that you don't bore or confuse your audience.
the law should require ANY company that keeps customers private information for any period to at least proactively make the customer aware, then divulge it at no expense to the customer.
On the flip side, it would make mail theft a more viable means of identity theft.
Right now, when you request a credit report, you'll be looking for it. If it's sent out automatically, would you realize if it didn't show up?
Thought there was a 7800 or 7900 AGP NVIDIA card? (I currently use a NVIDIA GeForce 6800 AGP and I'm thinking of keeping my 2GHz Opteron w/ 2GB of RAM for probably another 2 years now.)
Ah, the 7800GS for $270-$300. Which I believe is still a good bit faster then my old 6800 and will perform well enough to be worth the upgrade price (before I ditch the 2GHz Opteron unit for a dual-core or quad-core CPU down the road).
That reminds me, I need to go look for benchmarks...
Granted, I probably learned all of this in CS101, but if I need to remember how to do something, I typically perform a google search instead of paging thru a "missing manual". But, for those that prefer paper, this looks good.
Google is good for specifics, books are typically better at handling the why, wherefore, and "how does this fit into the big scheme" questions.
I find that most online resources are very good about the "how", but fall far short on explaining the "why" / "where" / "when" questions. Explaining why you use something, or where it fits in the overall framework, or when should I use certain features or technology is something that takes time and effort. Which is more likely to happen in a paid environment such as a book.
That and it's nice to kick back, relax, put my feet up and read hard-copy. Which looks enough like work without being anywhere near as stressful. The two foot tall towering stack of books also makes for a handy "I'm busy" look. Especially if you artfully re-arrange your books daily and sprinkle printouts all over the work surface.
As an owner of a 23" Samsung wide-screen... 24" is probabably the minimum acceptable size for a wall mount in a small to medium sized bedroom. (Defining "small" as 100 square feet, or thereabouts. And "medium" as something in the 150 sq ft range.) For a larger 200 sq ft bedroom, it might seem a bit small. Especially if you're sitting in bed and watching a TV on the far wall. OTOH, most folks don't like a big TV in the B/R so a 24" wide-screen is probably a suitable size.
I use my 23" in my home office. I sit about 48" away from it and it's a comfortable size. If I was trying to use it in a medium sized living room where I'm more like 72"-96" away it would start to feel a bit undersized. I'd find that a 28-32" display would make a more reasonable choice for that range. Still, I used to use a 19" CRT in my L/R and sat about 72" away from it and it was a decent size.
Prices on LCD glass keep dropping. The 23" was the sweet spot last year ($800-$900) when I bought, now I'd imagine that you can get a 23" for $400-$600. I'd be more likely to go with a Mac Mini chained to a VESA-mountable 23" LCD screen that is easily wall mountable.
It's the difference between 0.22mm / 0.255mm / 0.285mm dot pitches (too lazy to conver to ppi numbers).
What I find is that for older users, a 0.285mm dot pitch is a better fit even though you get fewer pixels per inch. Computer technology and web / UI design is still stuck in the old mentality that every display has the same pixels/inch value (within +/- 10%).
My 15" Toshiba Tecra has a 1400x1050 display which works out to around 125ppi (0.21mm dot pitch). I have to use large fonts even with a good pair of reading glasses to keep myself from leaning forward to get 16" away from the display. (Proper posture puts my eyes 24" away.)
The downside is that some programs don't deal with the dot pitch change properly so you get dialogs that are unreadable or images that are 40% smaller then normal.
The only real objection I'd agree with from that list is the "no video card" one. Notebook drives are now up to 160GB and 7200rpm, which is competitive with desktop drives.
That it now allows 2GB and uses a dual-core CPU actually makes it much more attractive then the older Mac Mini (which I felt was underpowered and too limited in RAM).
Wouldn't use it for a power-user, but might be reasonable for a business user. Price on the Mac Mini w/ 2GB, keyboard+mouse and the 3-year warranty is roughly $1275. That's not terribly expensive, but I could build a white-box PC for about $900 that would match it (figure an extra $200 for 3-year warranty).
Is OS X worth $675? Maybe. At least the Mac Mini is small enough to attach to my KVM without taking up oodles of space. I still might pick up a Mac at some point to play with (I'll sell it to the bean-counters as a way of better supporting our Mac users).
Slashdot Mirror
Aye, at least $50/hr and generally more like $60-$80/hr for regular customers and closer to the $100/hr mark for new business.
Because not only do you have to pay your own medical / taxes / insurance / etc. out of that gross, but also tools, training, books. Then there's the risk of lawsuits, fraud by clients, clients who won't pay until after 90 days. Plus the months out of the year that you might not have business due to slow economy, sick leave, or just wanting some vacation time.
I charge $60/hr because most of my costs are already covered by my day job. And I really only want about 5-10 hours of extra work per week. If I was doing it for my own as a primary business, I'd charge $75-$80.
Hopefully it will get caught in meta-moderation then... (something I've been remiss about doing lately).
Or simply allowing you to read data off of all N platters at the same time (which is probably closer to being possible) with the current design. The downside is that it would probably limit the drive structure (meaning you could have 1 head, 2 heads, 4 heads or 8 heads).
And it still wouldn't improve seek latency as much as a 2nd set of heads or simply RAID'ing together more spindles. Or increasing those RPMs again while using an even smaller diameter disk.
I've watched enough folks over the years go with the absolute cheapest parts they could find. Then I get to watch them struggle later on with either stability issues (if not outright parts failures) or upgrades (if the machine survives past a year or two). I've even done it myself on occasion, thinking that it's worth saving $10 now only to have it come around and bite me a year down the road. Upgrades are the main reason why I avoid cheaper / proprietary solutions from the pre-built folks. With commodity parts, you can usually re-use numerous components in the next-gen system.
So now I tend to go for the middle-tier parts. Motherboards from companies like Asus, Gigabyte, MSI. Power-supplies from good brands. HDs with the 3-yr or 5-yr warranty. Well-designed cases that designed for cooling and noise reduction (Antec, Lian Li, Thermaltake, etc.).
Still, it looks like I can build a pretty decent system for $300. Bottom-end performance, but definitely good enough to run a Linux distro and do word-processing / web-browsing / e-mail. That gets my geek blood pumping as I consider the possibilities. And it's a machine that I wouldn't be ashamed to give someone.
Hmm, $150 is a tough price point to hit for a regular PC if you're at all picky about components.
$45 CPU (AM2 Sempron 2800+, which means upgrade capability later)
$65 M/B (GeForce 6150 w/ integrated video)
$55 2x256 or 1x512MB (or $25 256MB single-stick)
$20 DVD-ROM
$25 Case+PSU
$45 Hard drive
===
$255
More like $300 once you buy a reasonably good case w/ PSU for $50-$75. And it would be expandable to put more memory and a more powerful CPU in it down the road.
You could still probably shave $100 off that price if you go for close-out deals, really cheap motherboards, older CPUs that are only $20ea, 256MB of RAM, and a really cheap case+PSU that will probably catch fire right after the warranty expires.
(The machines I'm building for work are around $500 for parts, but those are dual-core w/ 2GB RAM.)
GPG encryption inside regular text files (one per service) stored in a version control system. The contents of the files are encrypted using the public keys of everyone who is supposed to be able to decrypt the file contents. Unfortunately, it probably won't scale past a dozen users.
Doesn't handle keeping track of access. Although you could possibly do some fancy stuff with Subversion scripts to track that sort of thing.
I'll second GPG. Although I prefer to keep each system in a separate file (makes it easier to look at the change dates and see when things were modified).
And once you're using GPG ASCII blocks inside regular text files, you can simply dump those individual files into a version control system or central network share. Backups are also dirt-simple, e-mail everything to yourself at some other location or print out the sheets and rely on OCR in the worst-case scenario.
And yes, I'll be looking into "escrow". Our system works because there are only half a dozen of us that need to share passwords. Often I'll change a password and then e-mail the ASCII-encrypted text block to the group. I should really move towards putting these text files into our VCS.
Fortunately for AMD, they have managed to start their own 65nm manufacturing now.
There were articles about 2 weeks ago about how they're having issues with 65nm. They can't get the voltages as low as they want, which is requiring increased power requirements. Dunno if they've got it sorted out yet.
They do have some technical advantages over Intel once they get 65nm working (such as SS and SOI which I don't think Intel uses).
AMD already dropped prices on Athlon64 back in July (and that was scheduled a few months in advance, maybe as early as Feb 2006). Their next scheduled price are in October, but are only on the Sempron or Turion chip lines.
Any price drops outside of those 2 months or on chips other then Sempron/Turion would be news.
I agree that having multiple pages just to increase ad impressions is a little annoying, but why do people gripe about a site trying to profit from its work/research/journalism/whatever? How else is it going to make money/pay for the bandwidth that slashdot/anyone generates? Get off your soapbox!
1) Some things make sense to be split to multiple pages. Such as cases where you have 10 images for each of the benchmarks that you ran and you put them on multiple pages to make life easier for dial-up users.
2) The article content is typically split across too many pages so that the ads overwhelm the article. If the article covered 50-70% of the page, people probably wouldn't care as much, but often it seems like we're getting down into the single paragraph per page.
3) Every time you split to a new page, the reader loses their focus. This frustrates your visitor. Especially when the page is so cluttered that you have to hunt for the next page link. Visitors are coming to your site to read the article and get information, not to be confused and bombarded by as many ads as you can throw at them.
I'm all for attempting to make money off of ad impressions, but splitting a 25 paragraph review over 25 pages is well past ridiculous. (Most of these reviews should've been done in about 1/4 of the number of pages that they're programmed for.)
Well, there's cheap and then there's inexpensive.
I tend to buy whatever speed CPU is at the knee of the price/performance curve. I usually just calculate this based on the clock speed, which isn't perfect and somewhat naive but works quickly when all the chips are based on the same core. For example:
$0150 $75/GHz Athlon64 X2 3800+ 2.0GHz
$0187 $85/GHz Athlon64 X2 4200+ 2.2GHz
$0240 $100/GHz Athlon64 X2 4600+ 2.4GHz
So for the above example, the $85/GHz is probably what I would purchase. This graph used to be a lot clearer with the Opteron 2xx series where there are 5 different single-core chips. Prices went up slowly, but the upper 2 in the group were significantly more expensive then the lower 3.
$DIETY bless the blinky bits, there's a lot of joy to be had at watching the flickering lights of a multi-disk array that you built yourself. Watching in real-time as different disks get used to service requests can be a bit mesmerizing. (Plus it points out a possible performance issue with mdadm's RAID10 implementation when disk 0 has a higher utilization then the other disks.) Of course, I used to be entranced by the lights on the front of my 14.4Kbps modem.
All in all I'm glad that Intel has decided to retake the lead in the price/performance war, AMD needs a new kick in the pants.
Indeed, prices for AMD processors have been a bit stagnant for the past few years. I think the Opterons topped out at 2.6GHz, but are very expensive. If nothing else, at least the new Intel Core Duo chips forced them to finally drop prices on the X2's. Now if we can just see some price pressure on the Opteron dual-cores I'd be a very happy geek.
For usability of said web/business applications, when dual-core only adds ~$70 to the price of a $800 system, why not add dual-core for reponsiveness?
That's been our decision. The price breaks on AMD X2 chips back in July 2006 made it an easy one ($300 chip now sells for $150, only $50 more then the $100 CPUs were were buying). For such a minor price increase, we get another few years of longevity due to the 2nd core.
Should be pretty easy for that system to save the user $50 over the course of a year because they're not waiting on all sorts of minor delays. There are even some energy-efficient X2 3800+ chips that just hit the street that could eliminate even the excess power consumption argument. Not that modern CPUs are that bad; they'll ramp their speeds back down when the power isn't needed.
I guess there's room in a SD card for multiple chips. It's just that the SD Card is the size of a postage stamp with the upper 1/4 consumed by the contact points. I guess as long as the 16Gbit chips are less then around 0.8 cm^2 then there's enough room inside the packaging for a pair of them.
Samsung has announced 32-gigabit (4GB) flash chips, not 32 megabit.
Considering that I can get a 4GB SD flash card for $60, I'm guessing that those 32 gigabit chips have been out for a while? Or do those tiny SD cards have 2 chips in them?
Pity that you posted anonymously... Seagate Momentus are 160GB, but you're right, only 5400rpm. OTOH, the older 7200.1 Momentus are 100GB and 7200rpm. So there's a good chance that as PR goes mainstream they can do a 7200rpm 160GB drive.
RPM only buys you better seek times. And on the 100GB Momentus drives, it's a difference between 8.5ms and 10.5ms. The increased density of the PR platters will give increased transfer rates even at 5400rpm.
A classic example would be the Chronicles of Narnia series, with the lion Aslan clearly representing Jesus Christ in the mythology of Narnia.
Mmm, C.S. Lewis also despised allegory from what I recall. They were close aquaintances at the time and I don't think either story was written to be allegorical. But I could be confused... I think the biographical video on the FotW extended edition talks about this a bit when discussing both authors.
Not to be a stick in the mud about this but Jackson seems to have gone out of his way to take some of Tolkien's original work and warp it. Things that he could have easily have left as written were rewritten to suit his "needs" but not the needs of the story line.
Meh, you're entitled to your opinion. Personally, I watched the extended editions and listened to the audio commentaries. PJ and crew do a good job explaining the whys and wherefores of their changes. I might not agree with all of them, but on the whole they were trying to make a good set of movies that were entertaining while simultaneously staying as true as possible to the books.
Not an easy task. And from the box office results, they at least succeeded at making something that the majority of folks enjoyed. (Even though I disliked the dwarf jokes and the elven tricks such as shield-surfing.)
Sure, it was warped in places, but usually because of the realities of film-making so that you don't bore or confuse your audience.
the law should require ANY company that keeps customers private information for any period to at least proactively make the customer aware, then divulge it at no expense to the customer.
On the flip side, it would make mail theft a more viable means of identity theft.
Right now, when you request a credit report, you'll be looking for it. If it's sent out automatically, would you realize if it didn't show up?
Good job on the formatting (IMO). It's legible and slashdot didn't kill it.
Thought there was a 7800 or 7900 AGP NVIDIA card? (I currently use a NVIDIA GeForce 6800 AGP and I'm thinking of keeping my 2GHz Opteron w/ 2GB of RAM for probably another 2 years now.)
Ah, the 7800GS for $270-$300. Which I believe is still a good bit faster then my old 6800 and will perform well enough to be worth the upgrade price (before I ditch the 2GHz Opteron unit for a dual-core or quad-core CPU down the road).
That reminds me, I need to go look for benchmarks...
Granted, I probably learned all of this in CS101, but if I need to remember how to do something, I typically perform a google search instead of paging thru a "missing manual". But, for those that prefer paper, this looks good.
Google is good for specifics, books are typically better at handling the why, wherefore, and "how does this fit into the big scheme" questions.
I find that most online resources are very good about the "how", but fall far short on explaining the "why" / "where" / "when" questions. Explaining why you use something, or where it fits in the overall framework, or when should I use certain features or technology is something that takes time and effort. Which is more likely to happen in a paid environment such as a book.
That and it's nice to kick back, relax, put my feet up and read hard-copy. Which looks enough like work without being anywhere near as stressful. The two foot tall towering stack of books also makes for a handy "I'm busy" look. Especially if you artfully re-arrange your books daily and sprinkle printouts all over the work surface.
As an owner of a 23" Samsung wide-screen... 24" is probabably the minimum acceptable size for a wall mount in a small to medium sized bedroom. (Defining "small" as 100 square feet, or thereabouts. And "medium" as something in the 150 sq ft range.) For a larger 200 sq ft bedroom, it might seem a bit small. Especially if you're sitting in bed and watching a TV on the far wall. OTOH, most folks don't like a big TV in the B/R so a 24" wide-screen is probably a suitable size.
I use my 23" in my home office. I sit about 48" away from it and it's a comfortable size. If I was trying to use it in a medium sized living room where I'm more like 72"-96" away it would start to feel a bit undersized. I'd find that a 28-32" display would make a more reasonable choice for that range. Still, I used to use a 19" CRT in my L/R and sat about 72" away from it and it was a decent size.
Prices on LCD glass keep dropping. The 23" was the sweet spot last year ($800-$900) when I bought, now I'd imagine that you can get a 23" for $400-$600. I'd be more likely to go with a Mac Mini chained to a VESA-mountable 23" LCD screen that is easily wall mountable.
It's the difference between 0.22mm / 0.255mm / 0.285mm dot pitches (too lazy to conver to ppi numbers).
What I find is that for older users, a 0.285mm dot pitch is a better fit even though you get fewer pixels per inch. Computer technology and web / UI design is still stuck in the old mentality that every display has the same pixels/inch value (within +/- 10%).
My 15" Toshiba Tecra has a 1400x1050 display which works out to around 125ppi (0.21mm dot pitch). I have to use large fonts even with a good pair of reading glasses to keep myself from leaning forward to get 16" away from the display. (Proper posture puts my eyes 24" away.)
The downside is that some programs don't deal with the dot pitch change properly so you get dialogs that are unreadable or images that are 40% smaller then normal.
The only real objection I'd agree with from that list is the "no video card" one. Notebook drives are now up to 160GB and 7200rpm, which is competitive with desktop drives.
That it now allows 2GB and uses a dual-core CPU actually makes it much more attractive then the older Mac Mini (which I felt was underpowered and too limited in RAM).
Wouldn't use it for a power-user, but might be reasonable for a business user. Price on the Mac Mini w/ 2GB, keyboard+mouse and the 3-year warranty is roughly $1275. That's not terribly expensive, but I could build a white-box PC for about $900 that would match it (figure an extra $200 for 3-year warranty).
Is OS X worth $675? Maybe. At least the Mac Mini is small enough to attach to my KVM without taking up oodles of space. I still might pick up a Mac at some point to play with (I'll sell it to the bean-counters as a way of better supporting our Mac users).