My guess would be silicon costs and/or complexity of the spec.
Think back to the mid-late 80s and how much CPU power or silicon would've been needed to deal with character sets and interpreting or skipping over text. At the time, just reading data off of a CD at that rate, processing it and piping it through a D/A converter was considered pretty high-tech and pushing the envelope. Those early CD players were pricey.
This was also the era when programmers were still using 2-digit dates. Memory was measured in kilobytes or tens of kilobytes.
Not even sure that Unicode was a preliminary spec at that point.
The clock now has a calendar and (optionally) additional clocks in different timezones. No more opening the date/time properties just to find the date.
The clock in WinXP had a calendar as well. You could mouse-over the clock to see the current date, or expand your start bar to 2-lines which will then display the current time, day of week, and date.
Additional clocks in different timezones is probably new. Unless there were free add-ons for WinXP that served the same purpose.
As others have said, bite the bullet and go for RAID 10 (RAID0 across RAID1 pairs). Some advantages:
- Less chance of data loss in a 2-disk failure compared to RAID5.
- Rebuild rate for a RAID10 array is based on the size of the mirror pair, not the entire array size. Rebuilding a RAID5 array with more then a handful of disks is *slow*. Rebuilding a RAID10 array typically takes 3 hours... max. Maybe faster depending on the size of your mirror pairs. Which means you have a smaller window where a double-drive failure can kill your array.
- You can go paranoid (although it's a bit wasteful). Three drives in each RAID1 array (all actively mirrored) with RAID0 laid over the top allows you to handle a double-disk failure without worry. You would have to lose all 3 disks in the RAID1 set before losing the entire array. Downside is that your net capacity is going to be below 33% (including hot-spares).
- Performance of RAID10 is nice and predictable. Additional RAID1 elements in the RAID10 array increase read/write rates in a linear fashion until you hit a bottleneck. So if a RAID1 pair gives you 50MB/s read/write, a 10-disk array should hit 250 MB/s read/write (should... not always).
(I like the idea of RAID5 for large amounts of storage. But the rebuild times and risk of a 2nd drive failure during the rebuild window make me hesitant to recommend them anymore. RAID6 is pretty new and for smaller arrays of less then 8 disks, it's simpler to do RAID10. All of this is, of course, my own opinion and YMMV.)
Can anyone explain to me why SCSI drives always seem to be lagging IDE in terms of capacity? Does the increased rotational speed make them unable to discern smaller features on the disk?
SCSI drives, while using a 3.5" form factor, use smaller platters inside so that they can spin at the higher rotational speeds. Thus, lower capacity. AFAIK, SCSI drives use the same bit density per square unit of linear measure as SATA/PATA drives.
And the flip side, I've owned close to 2 dozen IBM Deskstar drives (mostly 72-80GB). No more then a handful died before their warranty period expired.
Most of those deaths were directly related to heat issues (poor cooling or poor airflow). Some were undetermined cause.
From my experience over the past decade, heat is the #1 killer. Some makes / models are better at dealing with 50C+ temperatures then others. Maxtors seemed to be a bit sensitive to anything above 50C (and Maxtor drives were a real PITA to RMA, IBM RMAs were a simple click-click-click on a web form prior to send it back).
Nowadays, I simply plan for failure (RAID1 across 3 drives or RAID10 w/ hot-spare) along with backups. I try to keep drives at or below 40C and I keep enough airflow across them that their operating them doesn't change by more then 5C between idle/active.
Which is a real shame, because it pretty much destroys the usefulness of loading pages in background tabs. I mean, the whole reason that I'm loading the page in the background is so that I can continue doing what I was doing, right?
Anyone know if Opera has a multi-threaded UI? What about IE 7?
(The poor UI responsiveness of Mozilla/Firefox is a major hassle for me.)
Wal-Mart? Come on. All they do is sell products that people want, for less money than the competition, and offer correspondingly little in the way of customer service.
The problem(s) with Wal-Mart:
- Their size makes it too easy for them to dictate conditions to suppliers. Which means that if a supplier takes on Wal-Mart as a client, it often is simply a slow-motion suicide for the supplier. Wal-Mart will demand price cuts year after year and then dump you when you can't make the mark anymore and still make a profit.
- They don't pay their employees enough, they don't offer health care (or not enough) along with other anti-union / anti-employee tactics. Again, they get away with this because of their size.
- Basically, they're not a good corporate citizen. Killing off small businesses by selling cheap crap.
I've bought maybe $100 worth of goods in the past 7 years from Wal-Mart. There are better companies to patronize with my business. Having one big retailer is not good in the long-run, I want 6-12 healthy competitors in the market.
True disaster recovery planning involves offsite storage of data IMHO, and tape is hella easier to transport than HDDs. Also, you don't have to worry about what order you stick tapes in, whereas with disk storage, re-assembling a RAID array would be a PITA, even with labelling.
The better RAID systems keep track of which drive is which in a RAID array. I know for sure that simply Linux Software RAID does it, and I'd be surprised if higher end RAID cards didn't.
For me to go from Baltimore to NYC, the train is about as expensive and as fast as driving. Plus I get to do something else with my time. Used to take about 4 hours, IIRC for the train. A bit less during off-hours a bit more during peak hours. No dealing with trying to park my car in NYC either.
Flying would be about double that cost, plus you have to deal with the TSA security theater.
Do I think Amtrak couldn't do better. You better believe it (too many stops in podunk little towns where 3 people get on/off). And they need to get more 80-100 MPH corridors like the one between Phila and NYC. Which is fast enough without needing anything fancy or expensive.
If you're going to install the NTP package anyway in order to do "ntpdate" (which is deprecated), you may as well run NTP as a daemon. Just configure it to only talk to the single server and it will work just as well as "ntpdate", except your clock won't jump around.
You can set the "iburst" option on the server line in the config file so that it synchronizes quicker at startup.
Another example would be the three prequel films in the Star Wars series.
Or [i]Snowcrash[/i] which is a book that I would *like* to enjoy, but simply can't because it's in need of a heavy edit to fix technical goofs and plot problems.
But with regards to HP, I'm not completely sure that she's crossed over into CAS-land yet. Her work in book 6 was a bit uneven, so maybe we're seeing the first symptoms of CAS.
While I agree with your sentiments... 800GB for 25 users sounds a bit overkill. (grin)
We're only allowing 4GB per user on the new server. But that's just a planning number, not a hard quota. Anyone who goes over 4GB in their maildir accounts will probably get a scripted weekly warning message. We have a few packrats and a few thrifties. I figure I'll continually adjust the warning message threshold to get looser and looser as time goes on.
(I may even pester the users via our internal Jabber server...)
The ones at work had supplied-by-the-lowest-bidder heatsinks and fans, on both the CPU and the chipset (biggest cause of motherboard death was the fan on the chipset dying and preventing air escaping).
Check out the Asus motherboards (M2N-E, M2N32-SLI) designs. They use a passive copper heatpipe / heatsink setup for the chipset cooling.
(I'm a big fan of fanless solutions on servers when I can get away with it. Such as the GeForce 6200 LE card.)
MemTest86 is no good for tracking down timing or voltage issues with "almost good" systems. It doesn't put anywhere near enough load on the CPU and other components while it does the memory testing. In general, MemTest86 will only catch completely bad memory modules.
If you *really* want to be sure of your memory (and system), run Prime95 in torture test mode. Exercising your disks and video at the same time is also recommended (while you check operating temperatures). You'll want to run for at least 4 hours, but preferably as long as 48 hours.
On the other hand, having your e-mail in a flat text file (mbox format) does wonders for interoperability. (Or being able to grep from the command line.)
Maybe they need to allow us to use maildir format mailboxes (each message goes in a separate file). The downside is disk space and whether the file system can efficiently handle thousands of small files in a directory.
I tried Google Desktop a while back. It absolutely sucks for any meaningful quantity of data.
A rough estimate is that I have around 1 million items on my system that would need to be indexed. Google Desktop started choking at the 10,000 item mark. And if you think I'm exaggerating, the subversion-user mailing list for the past year has about 19,000 messages. Multiple that by 20-30 mailing lists that I subscribe to along with all of the corporate mail, the documents on my hard drive for a few hundred projects, etc.
I believe that's what the tagging feature would be useful for...
Tagging doesn't scale. It works as long as the number of tags numbers in the low dozens. But when I look at my corporate mailbox for a single year (and I only get a few dozen messages per day), I would probably need to be creating a few hundred tags.
Much easier to edit the subject line and plug in some clarifying text.
(Even in MS Outlook it's not an automatic thing. You have to pull the message up and use the Edit -> Edit Message command before you can change the content. So a little "edit" icon next to the subject line would work just fine.)
Aye, even if we can't edit the message body, being able to edit the subject line for organizational purposes (without creating tags or folders out the wazoo) would be useful. Mostly because it allows you to put good searchable text into the subject line to make it easier to find the message later.
(Yet again, another thing that MS Outlook does that I sorely miss when using Thunderbird.)
Have they finally fixed the UI responsiveness issue? In Thunderbird 1.5, I find that the message pane is nigh unusable if Thunderbird is trying to retrieve mail in the background. Then there's the issue that Thunderbird gets a bit slow when dealing with folders with a few thousand messages (such as a popular mailing list where you keep a year's worth of posts for easy reference).
If you've ever used SpamBayes for MS Outlook, you'll understand why bayesian analysis engines need to have some sort of grey area instead of just a binary spam / ham bit. With SpamBayes in MS Outlook, I have (3) results after spam processing:
"ham" - Messages which scored below a rather low value (10?) and are considered non-spam. Those messages get left alone in whatever folder they were found in.
"unsure" - Anything that falls in the middle gets moved to a "Maybe Junk" folder. For the most part, this stuff is spam, but the bayesian engine isn't quite sure. So it's worth checking for false positives (which are rare, but can happen until the engine is trained).
"spam" - Stuff in the spam folder scored so high on the bayesian value that it's almost certainly spam. The odds of finding a false positive in this folder are extremely low so I never bother looking.
Now for the real magic of SpamBayes... it remembers where a message was when it was flagged as "unsure" or "spam". If you find a message that was mis-tagged, you can tell SpamBayes that it made a mistake and it will add the message to its ham corpus and move the message back where it belongs.
(That and intelligent message notification are the two things that drive me nutz with Thunderbird 1.5 and prevent me from switching over entirely.)
We stick with either Thinkpad T (or maybe X) series, Toshiba Tecras or the Apple MacBooks. We typically spend $1800-$2400 on a machine. Usually the business class products are designed / built better. Well, most days. We've had very good luck with the Tecras and Thinkpads.
OTOH, for a spare laptop that isn't going to see heavy use, it's tempting to purchase a $600 laptop. And then replace it every 2 years. Not sure which is better over the long run. Just seems like a waste. Plus the $600 machine probably doesn't have enough RAM, a big enough HD, or a hi-res screen. And you're dealing with the week of productivity loss every 2 years to migrate from the old to the new...
I guess as long as folks value price over quality, we'll continue to see the Wal-Mart effect in the market.
Why aren't any of IBM/Lenovo's Thinkpad offerings on that list? I see that computer used more often than the high-end HP's in business anyway, which is well deserved because they are rock solid and last a really long time. I had their 760L from 10 years ago working until 2 years ago when someone took the computer and smashed it to the floor. And even after that it still worked! The same went for my Thinkpad 600.
Because the Thinkpad systems last too long?
A reseller only gets revenue when you buy something. It's not in their best interests to push items that won't need to be replaced for a decade.
(Personally I think 5 years is about it for laptops that are moderately used. Mostly depends on how much time they spend in their docking station vs being toted hither and yon. Wear and tear is the typical killer for anything made after 2002, while prior to 2002 the laptop would be replaced due to being too slow after a few years.
My personal laptop is a 6 year old Tecra 9100 that averages about 3000 hours of use per year. It's just about on its last legs. Can't add a bigger HD and can't add any more RAM and the single-core CPU is a limiting factor for me.)
Slow start-up times vary between bulb types for CF. (I have a dozen or so in the house of various vintages and types.) For places where you need a quick minute (or 5 minutes of light), I make sure to put in quick-acting CFs. For the other rooms, I don't worry as much.
One thing I did a long time ago was to start paying for laundry service. I don't know what I pay per load, but it's extremely inexpensive (a small family owned business). They seem to offer the service more as a cost-saving measure (keeps the attendant busy and brings in income to offset their hourly wages) then as a profit center. Some places would charge me twice as much (especially the ones that charge based on weight).
(I rent and this apartment doesn't have room for a W/D, nor is there a unit within the building. Plus it's darn convenient to let them do it and they don't charge me an arm and a leg.)
Slashdot Mirror
My guess would be silicon costs and/or complexity of the spec.
Think back to the mid-late 80s and how much CPU power or silicon would've been needed to deal with character sets and interpreting or skipping over text. At the time, just reading data off of a CD at that rate, processing it and piping it through a D/A converter was considered pretty high-tech and pushing the envelope. Those early CD players were pricey.
This was also the era when programmers were still using 2-digit dates. Memory was measured in kilobytes or tens of kilobytes.
Not even sure that Unicode was a preliminary spec at that point.
The clock now has a calendar and (optionally) additional clocks in different timezones. No more opening the date/time properties just to find the date.
The clock in WinXP had a calendar as well. You could mouse-over the clock to see the current date, or expand your start bar to 2-lines which will then display the current time, day of week, and date.
Additional clocks in different timezones is probably new. Unless there were free add-ons for WinXP that served the same purpose.
As others have said, bite the bullet and go for RAID 10 (RAID0 across RAID1 pairs). Some advantages:
- Less chance of data loss in a 2-disk failure compared to RAID5.
- Rebuild rate for a RAID10 array is based on the size of the mirror pair, not the entire array size. Rebuilding a RAID5 array with more then a handful of disks is *slow*. Rebuilding a RAID10 array typically takes 3 hours... max. Maybe faster depending on the size of your mirror pairs. Which means you have a smaller window where a double-drive failure can kill your array.
- You can go paranoid (although it's a bit wasteful). Three drives in each RAID1 array (all actively mirrored) with RAID0 laid over the top allows you to handle a double-disk failure without worry. You would have to lose all 3 disks in the RAID1 set before losing the entire array. Downside is that your net capacity is going to be below 33% (including hot-spares).
- Performance of RAID10 is nice and predictable. Additional RAID1 elements in the RAID10 array increase read/write rates in a linear fashion until you hit a bottleneck. So if a RAID1 pair gives you 50MB/s read/write, a 10-disk array should hit 250 MB/s read/write (should... not always).
(I like the idea of RAID5 for large amounts of storage. But the rebuild times and risk of a 2nd drive failure during the rebuild window make me hesitant to recommend them anymore. RAID6 is pretty new and for smaller arrays of less then 8 disks, it's simpler to do RAID10. All of this is, of course, my own opinion and YMMV.)
Can anyone explain to me why SCSI drives always seem to be lagging IDE in terms of capacity? Does the increased rotational speed make them unable to discern smaller features on the disk?
SCSI drives, while using a 3.5" form factor, use smaller platters inside so that they can spin at the higher rotational speeds. Thus, lower capacity. AFAIK, SCSI drives use the same bit density per square unit of linear measure as SATA/PATA drives.
And the flip side, I've owned close to 2 dozen IBM Deskstar drives (mostly 72-80GB). No more then a handful died before their warranty period expired.
Most of those deaths were directly related to heat issues (poor cooling or poor airflow). Some were undetermined cause.
From my experience over the past decade, heat is the #1 killer. Some makes / models are better at dealing with 50C+ temperatures then others. Maxtors seemed to be a bit sensitive to anything above 50C (and Maxtor drives were a real PITA to RMA, IBM RMAs were a simple click-click-click on a web form prior to send it back).
Nowadays, I simply plan for failure (RAID1 across 3 drives or RAID10 w/ hot-spare) along with backups. I try to keep drives at or below 40C and I keep enough airflow across them that their operating them doesn't change by more then 5C between idle/active.
Which is a real shame, because it pretty much destroys the usefulness of loading pages in background tabs. I mean, the whole reason that I'm loading the page in the background is so that I can continue doing what I was doing, right?
Anyone know if Opera has a multi-threaded UI? What about IE 7?
(The poor UI responsiveness of Mozilla/Firefox is a major hassle for me.)
Wal-Mart? Come on. All they do is sell products that people want, for less money than the competition, and offer correspondingly little in the way of customer service.
The problem(s) with Wal-Mart:
- Their size makes it too easy for them to dictate conditions to suppliers. Which means that if a supplier takes on Wal-Mart as a client, it often is simply a slow-motion suicide for the supplier. Wal-Mart will demand price cuts year after year and then dump you when you can't make the mark anymore and still make a profit.
- They don't pay their employees enough, they don't offer health care (or not enough) along with other anti-union / anti-employee tactics. Again, they get away with this because of their size.
- Basically, they're not a good corporate citizen. Killing off small businesses by selling cheap crap.
I've bought maybe $100 worth of goods in the past 7 years from Wal-Mart. There are better companies to patronize with my business. Having one big retailer is not good in the long-run, I want 6-12 healthy competitors in the market.
True disaster recovery planning involves offsite storage of data IMHO, and tape is hella easier to transport than HDDs. Also, you don't have to worry about what order you stick tapes in, whereas with disk storage, re-assembling a RAID array would be a PITA, even with labelling.
The better RAID systems keep track of which drive is which in a RAID array. I know for sure that simply Linux Software RAID does it, and I'd be surprised if higher end RAID cards didn't.
TV on your own schedule, on your own terms.
Except for this little pesky thing called the "broadcast flag".
For me to go from Baltimore to NYC, the train is about as expensive and as fast as driving. Plus I get to do something else with my time. Used to take about 4 hours, IIRC for the train. A bit less during off-hours a bit more during peak hours. No dealing with trying to park my car in NYC either.
Flying would be about double that cost, plus you have to deal with the TSA security theater.
Do I think Amtrak couldn't do better. You better believe it (too many stops in podunk little towns where 3 people get on/off). And they need to get more 80-100 MPH corridors like the one between Phila and NYC. Which is fast enough without needing anything fancy or expensive.
See also, Perfect Blue, which features some web pages as a minor part of the plot.
(A very well written psycho thriller.)
If you're going to install the NTP package anyway in order to do "ntpdate" (which is deprecated), you may as well run NTP as a daemon. Just configure it to only talk to the single server and it will work just as well as "ntpdate", except your clock won't jump around.
You can set the "iburst" option on the server line in the config file so that it synchronizes quicker at startup.
Goodness yes...
Another example would be the three prequel films in the Star Wars series.
Or [i]Snowcrash[/i] which is a book that I would *like* to enjoy, but simply can't because it's in need of a heavy edit to fix technical goofs and plot problems.
But with regards to HP, I'm not completely sure that she's crossed over into CAS-land yet. Her work in book 6 was a bit uneven, so maybe we're seeing the first symptoms of CAS.
While I agree with your sentiments... 800GB for 25 users sounds a bit overkill. (grin)
We're only allowing 4GB per user on the new server. But that's just a planning number, not a hard quota. Anyone who goes over 4GB in their maildir accounts will probably get a scripted weekly warning message. We have a few packrats and a few thrifties. I figure I'll continually adjust the warning message threshold to get looser and looser as time goes on.
(I may even pester the users via our internal Jabber server...)
The ones at work had supplied-by-the-lowest-bidder heatsinks and fans, on both the CPU and the chipset (biggest cause of motherboard death was the fan on the chipset dying and preventing air escaping).
Check out the Asus motherboards (M2N-E, M2N32-SLI) designs. They use a passive copper heatpipe / heatsink setup for the chipset cooling.
(I'm a big fan of fanless solutions on servers when I can get away with it. Such as the GeForce 6200 LE card.)
MemTest86 is no good for tracking down timing or voltage issues with "almost good" systems. It doesn't put anywhere near enough load on the CPU and other components while it does the memory testing. In general, MemTest86 will only catch completely bad memory modules.
If you *really* want to be sure of your memory (and system), run Prime95 in torture test mode. Exercising your disks and video at the same time is also recommended (while you check operating temperatures). You'll want to run for at least 4 hours, but preferably as long as 48 hours.
On the other hand, having your e-mail in a flat text file (mbox format) does wonders for interoperability. (Or being able to grep from the command line.)
Maybe they need to allow us to use maildir format mailboxes (each message goes in a separate file). The downside is disk space and whether the file system can efficiently handle thousands of small files in a directory.
I installed Google Desktop a while back.
I tried Google Desktop a while back. It absolutely sucks for any meaningful quantity of data.
A rough estimate is that I have around 1 million items on my system that would need to be indexed. Google Desktop started choking at the 10,000 item mark. And if you think I'm exaggerating, the subversion-user mailing list for the past year has about 19,000 messages. Multiple that by 20-30 mailing lists that I subscribe to along with all of the corporate mail, the documents on my hard drive for a few hundred projects, etc.
I believe that's what the tagging feature would be useful for...
Tagging doesn't scale. It works as long as the number of tags numbers in the low dozens. But when I look at my corporate mailbox for a single year (and I only get a few dozen messages per day), I would probably need to be creating a few hundred tags.
Much easier to edit the subject line and plug in some clarifying text.
(Even in MS Outlook it's not an automatic thing. You have to pull the message up and use the Edit -> Edit Message command before you can change the content. So a little "edit" icon next to the subject line would work just fine.)
Aye, even if we can't edit the message body, being able to edit the subject line for organizational purposes (without creating tags or folders out the wazoo) would be useful. Mostly because it allows you to put good searchable text into the subject line to make it easier to find the message later.
(Yet again, another thing that MS Outlook does that I sorely miss when using Thunderbird.)
Have they finally fixed the UI responsiveness issue? In Thunderbird 1.5, I find that the message pane is nigh unusable if Thunderbird is trying to retrieve mail in the background. Then there's the issue that Thunderbird gets a bit slow when dealing with folders with a few thousand messages (such as a popular mailing list where you keep a year's worth of posts for easy reference).
If you've ever used SpamBayes for MS Outlook, you'll understand why bayesian analysis engines need to have some sort of grey area instead of just a binary spam / ham bit. With SpamBayes in MS Outlook, I have (3) results after spam processing:
"ham" - Messages which scored below a rather low value (10?) and are considered non-spam. Those messages get left alone in whatever folder they were found in.
"unsure" - Anything that falls in the middle gets moved to a "Maybe Junk" folder. For the most part, this stuff is spam, but the bayesian engine isn't quite sure. So it's worth checking for false positives (which are rare, but can happen until the engine is trained).
"spam" - Stuff in the spam folder scored so high on the bayesian value that it's almost certainly spam. The odds of finding a false positive in this folder are extremely low so I never bother looking.
Now for the real magic of SpamBayes... it remembers where a message was when it was flagged as "unsure" or "spam". If you find a message that was mis-tagged, you can tell SpamBayes that it made a mistake and it will add the message to its ham corpus and move the message back where it belongs.
(That and intelligent message notification are the two things that drive me nutz with Thunderbird 1.5 and prevent me from switching over entirely.)
We stick with either Thinkpad T (or maybe X) series, Toshiba Tecras or the Apple MacBooks. We typically spend $1800-$2400 on a machine. Usually the business class products are designed / built better. Well, most days. We've had very good luck with the Tecras and Thinkpads.
OTOH, for a spare laptop that isn't going to see heavy use, it's tempting to purchase a $600 laptop. And then replace it every 2 years. Not sure which is better over the long run. Just seems like a waste. Plus the $600 machine probably doesn't have enough RAM, a big enough HD, or a hi-res screen. And you're dealing with the week of productivity loss every 2 years to migrate from the old to the new...
I guess as long as folks value price over quality, we'll continue to see the Wal-Mart effect in the market.
Why aren't any of IBM/Lenovo's Thinkpad offerings on that list? I see that computer used more often than the high-end HP's in business anyway, which is well deserved because they are rock solid and last a really long time. I had their 760L from 10 years ago working until 2 years ago when someone took the computer and smashed it to the floor. And even after that it still worked! The same went for my Thinkpad 600.
Because the Thinkpad systems last too long?
A reseller only gets revenue when you buy something. It's not in their best interests to push items that won't need to be replaced for a decade.
(Personally I think 5 years is about it for laptops that are moderately used. Mostly depends on how much time they spend in their docking station vs being toted hither and yon. Wear and tear is the typical killer for anything made after 2002, while prior to 2002 the laptop would be replaced due to being too slow after a few years.
My personal laptop is a 6 year old Tecra 9100 that averages about 3000 hours of use per year. It's just about on its last legs. Can't add a bigger HD and can't add any more RAM and the single-core CPU is a limiting factor for me.)
Slow start-up times vary between bulb types for CF. (I have a dozen or so in the house of various vintages and types.) For places where you need a quick minute (or 5 minutes of light), I make sure to put in quick-acting CFs. For the other rooms, I don't worry as much.
One thing I did a long time ago was to start paying for laundry service. I don't know what I pay per load, but it's extremely inexpensive (a small family owned business). They seem to offer the service more as a cost-saving measure (keeps the attendant busy and brings in income to offset their hourly wages) then as a profit center. Some places would charge me twice as much (especially the ones that charge based on weight).
(I rent and this apartment doesn't have room for a W/D, nor is there a unit within the building. Plus it's darn convenient to let them do it and they don't charge me an arm and a leg.)