I ran into the ISA slot spacing thing too. Had an old XT case that I crammed a 386 into. I widened one of the case's ISA slots with a hacksaw and passed the KVM cables through the opening. The mobo was set about 4" forward in the case and secured with nylon standoffs and lots of electrical tape. I was probably 15 at the time.
Another one I can't take credit for: I got a discarded PC from the local university around 1999. It was some oddball HP that someone had retrofitted a standard AT power supply into. They custom fabricated a power switch bracket out of scrap aluminum and neatly dremeled out the plastic cooling slots in the back to allow the power cord to be plugged in. It was actually quite well executed, I'll shake the hand of the guy who did it if I ever meet him.
I think many people are more "verbal" in email communication because of the pace. Face-to-face, dead air is a sin, so you're pressured to blurt half-cooked thoughts before they're done baking. In an email, you have time to think about how to word things so they sound somewhat intelligent. You can also edit things if you don't like the way it came out the first time, something that's impossible in voice (or IM, for that matter).
Asperger's or whatever, I think many people just don't think quite as fast on their feet. At least, I don't.:)
"...oil change advice. Based on nothing more than folk lore and a profit motive"
Actually, the 3K miles change is for hard-service applications, like frequent short trips and city driving. If you car does not spend most of its time at normal operating temperature (around 200*F), moisture and other things don't separate out of your engine oil and escape via the PCV system. Over time, your oil turns acidic and starts to eat away at your bearings. Cold oil is also thicker and does not reach all the little nooks and crannies of your engine.
If you do mostly highway driving, with few cool down / warm up cycles and constant rpm, then the 6K mile oil changes are entirely reasonable.
"I had to spend the money on a massive YAGI UHF antenna and some signal amps"
Actually, amps don't do a lot for improving signals; they can't improve the S/N ratio much because they tend to amplify the Noise as much as they do the Signal. You generally only use an amp to compensate for losses introduced by a long cable run (which are significant at UHF). It's better skip the amp and use coax that's less lossy, like RG-11U, if at all possible.
That's just a pedantic nitpick though, I agree with the rest of your post.:)
Windows steals focus from me all the time. I can't count how many times I'll be typing in an email (or on a web page like/.) when suddenly I'm somewhere else and windows and dialogs are flashing wildly. Another pet peeve is entering keys for Microsoft products. I'm concentrating on reading the key and typing on the keyboard, then when I look up I discover I have to type the whole thing all over again because something stole the freaking focus while I was typing.
If I were to write a Window Manager, any program with input focus and has had keystrokes in the last 3 seconds *cannot* lose focus to *anything* short of something actually crashing.
Also, any confirmation dialogs would have no default button selected, so hitting enter or space would do nothing.
"While it is true that there are only three distinct, non-overlapping, slices of spectrum"
Agreed. Wardriving around my neighborhood, I found a surprising large number of instances where two or even three neighbors had APs on the same channel. Obviously it doesn't bother them enough for anyone to change channels.
Actually, while 1, 6, and 11 are the "primary" channels, you can often sqeeze in two more at 3|4 and at 9 with only a slight degredation in performance. Most of the bandwidth is concentrated in the middle of the channel; off to the sides (above and below), only a little bandwidth is used so if these areas are overlapped it's not too big a deal.
At first I thought this was kind of stupid since the federal gevernment doesn't have that power. Then I remembered that the federal goverment has ruled that marijuana grown in California, sold in California, and consumed in California constitutes interstate commerce and can therefore be regulated or banned by the federal gov't.
"It puts packages where the originators expects them to be -- this means that I can download a source tarball, build it, and have it actually work."
This is a *major* advantage to Slackware. Why do the other distros put things in weird places for all of their packages? Drives me nuts.
To be fair though, I have had troubles with Slackware's Berkely DB files being "in the wrong place" or missing libraries or being the wrong version, most notably with OpenLDAP and Cyrus SASL. It could just be OpenLDAP and Cyrus that are the ones messed up, I'm not sure... but other than BDB issues, I can't recall any other programs that won't compile from source and just work.
That sounds a lot like OTEC, Ocean Thermal Energy Conversion. There are a lot of pumping losses to pull cool seawater from the depths of the ocean, so it's not terribly efficient but it does work (there's been a small one operating off the coast of Hawaii for a number of years now). Some of the same principals could be applied to datacenter waste heat. Energy conversion is often more cost effective on a large scale, so if the idea even is worthwhile, you'd probably need a major datacenter to do it with.
Actually that's not such a bad idea. Google or wiki the Rankine Cycle. You will have a net loss of energy, but you will reduce the total power consumption since you're no longer throwing away heat energy.
You have a point there... also, I was told this a number of years ago when clock speeds were measured in MHz rather than GHz so things were an order of magnatude slower "back then".
I was always taught that two 1GHz CPUs are slower than one 2GHz CPU, because of the extra overhead of the OS managing the 2nd CPU. On all the servers at work, HT CPUs show up as two virtual CPUs to Windows... so yeah, I would fully expect HT to be slower on heavily loaded systems -- no surprises there.
"Tech support guys know that they are zeros and will most likely always be zeros.... employed at chump wages... never have the social status... working... in dead soulless drab cubicles."
Slashdot Mirror
Indeed. Dune was the first thing I thought of when I read the headline too.
I ran into the ISA slot spacing thing too. Had an old XT case that I crammed a 386 into. I widened one of the case's ISA slots with a hacksaw and passed the KVM cables through the opening. The mobo was set about 4" forward in the case and secured with nylon standoffs and lots of electrical tape. I was probably 15 at the time.
Another one I can't take credit for: I got a discarded PC from the local university around 1999. It was some oddball HP that someone had retrofitted a standard AT power supply into. They custom fabricated a power switch bracket out of scrap aluminum and neatly dremeled out the plastic cooling slots in the back to allow the power cord to be plugged in. It was actually quite well executed, I'll shake the hand of the guy who did it if I ever meet him.
I think many people are more "verbal" in email communication because of the pace. Face-to-face, dead air is a sin, so you're pressured to blurt half-cooked thoughts before they're done baking. In an email, you have time to think about how to word things so they sound somewhat intelligent. You can also edit things if you don't like the way it came out the first time, something that's impossible in voice (or IM, for that matter).
:)
Asperger's or whatever, I think many people just don't think quite as fast on their feet. At least, I don't.
"...oil change advice. Based on nothing more than folk lore and a profit motive"
Actually, the 3K miles change is for hard-service applications, like frequent short trips and city driving. If you car does not spend most of its time at normal operating temperature (around 200*F), moisture and other things don't separate out of your engine oil and escape via the PCV system. Over time, your oil turns acidic and starts to eat away at your bearings. Cold oil is also thicker and does not reach all the little nooks and crannies of your engine.
If you do mostly highway driving, with few cool down / warm up cycles and constant rpm, then the 6K mile oil changes are entirely reasonable.
"I had to spend the money on a massive YAGI UHF antenna and some signal amps"
:)
Actually, amps don't do a lot for improving signals; they can't improve the S/N ratio much because they tend to amplify the Noise as much as they do the Signal. You generally only use an amp to compensate for losses introduced by a long cable run (which are significant at UHF). It's better skip the amp and use coax that's less lossy, like RG-11U, if at all possible.
That's just a pedantic nitpick though, I agree with the rest of your post.
"People play the sims to create death traps, set families on fire, and haunt kids with the ghosts of their parents."
And to lock people in rooms with no doors until they pee on the floor. Don't forget that part.
"Why buy Cosmo ... Why buy Marth Stewart Living? ... Why buy O?"
/me ducks
So, what does that say about people who buy Slashdot subscriptions?
Thanks for that second link. Reminds me of George Orwell (of "Big Brother" 1984 and Animal Farm fame).
Windows steals focus from me all the time. I can't count how many times I'll be typing in an email (or on a web page like /.) when suddenly I'm somewhere else and windows and dialogs are flashing wildly. Another pet peeve is entering keys for Microsoft products. I'm concentrating on reading the key and typing on the keyboard, then when I look up I discover I have to type the whole thing all over again because something stole the freaking focus while I was typing.
If I were to write a Window Manager, any program with input focus and has had keystrokes in the last 3 seconds *cannot* lose focus to *anything* short of something actually crashing.
Also, any confirmation dialogs would have no default button selected, so hitting enter or space would do nothing.
Ok, so you can't copyright a game. Cool. Only problem is, I bet RISK and Carmen Sandiego are patented.
Nah, just enable to evil bit in the TCP headers.
I do run Blackbox (a lightweight WM) on the aforementioned Slack 10 box w/ 32MB, but let me tell you right now that it just *crawls*.
There's also sbootmgr, or Smart Boot Manager: http://btmgr.webframe.org/index.php3?body=download .html.
Heck I've run Slackware 10 on 32MB of RAM, so Puppy should be able to handle it no sweat.
;)
Just don't run X.
"While it is true that there are only three distinct, non-overlapping, slices of spectrum"
Agreed. Wardriving around my neighborhood, I found a surprising large number of instances where two or even three neighbors had APs on the same channel. Obviously it doesn't bother them enough for anyone to change channels.
Actually, while 1, 6, and 11 are the "primary" channels, you can often sqeeze in two more at 3|4 and at 9 with only a slight degredation in performance. Most of the bandwidth is concentrated in the middle of the channel; off to the sides (above and below), only a little bandwidth is used so if these areas are overlapped it's not too big a deal.
"passport and use it for interstate travel"
At first I thought this was kind of stupid since the federal gevernment doesn't have that power. Then I remembered that the federal goverment has ruled that marijuana grown in California, sold in California, and consumed in California constitutes interstate commerce and can therefore be regulated or banned by the federal gov't.
Yeah, we're screwed.
I think you have a while to wait. Usually lots of people have to get killed before a revolution is born.
"It puts packages where the originators expects them to be -- this means that I can download a source tarball, build it, and have it actually work."
This is a *major* advantage to Slackware. Why do the other distros put things in weird places for all of their packages? Drives me nuts.
To be fair though, I have had troubles with Slackware's Berkely DB files being "in the wrong place" or missing libraries or being the wrong version, most notably with OpenLDAP and Cyrus SASL. It could just be OpenLDAP and Cyrus that are the ones messed up, I'm not sure... but other than BDB issues, I can't recall any other programs that won't compile from source and just work.
At least it was better than Episode III.
That sounds a lot like OTEC, Ocean Thermal Energy Conversion. There are a lot of pumping losses to pull cool seawater from the depths of the ocean, so it's not terribly efficient but it does work (there's been a small one operating off the coast of Hawaii for a number of years now). Some of the same principals could be applied to datacenter waste heat. Energy conversion is often more cost effective on a large scale, so if the idea even is worthwhile, you'd probably need a major datacenter to do it with.
Actually that's not such a bad idea. Google or wiki the Rankine Cycle. You will have a net loss of energy, but you will reduce the total power consumption since you're no longer throwing away heat energy.
I love how that article states "No one know exactly why the grenade exploded."
IT WAS A GRENADE.
I'm not usually the sort to say "mod parent up", but I think he deserves it in this case.
You have a point there... also, I was told this a number of years ago when clock speeds were measured in MHz rather than GHz so things were an order of magnatude slower "back then".
I was always taught that two 1GHz CPUs are slower than one 2GHz CPU, because of the extra overhead of the OS managing the 2nd CPU. On all the servers at work, HT CPUs show up as two virtual CPUs to Windows... so yeah, I would fully expect HT to be slower on heavily loaded systems -- no surprises there.
"Tech support guys know that they are zeros and will most likely always be zeros. ... employed at chump wages. .. never have the social status ... working ... in dead soulless drab cubicles."
Hm, you sound like a tech support guy.