I was just installing various things to my win2k box and it still wants rebooting when:...
Generally software installs ask for reboots because the software author didn't feel like actually determining if a reboot was necessary. They just prompt for a reboot to be safe. 9 times out of 10 it isn't necessary in Win2k - just click 'No' and try.
I know this works for Nero. I don't think the ATI Catalyst drivers actually needed a reboot either for me.
I don't recall having to reboot when changing the pagefile size (or even being prompted)... in fact, Win2k often resizes it on-the-fly.
Even changing the machine's IP from static to dynamic doesn't require a reboot if you go (re)start the DHCP client service.
MS just feels safer asking you to reboot, because you might have ignorant software running that isn't aware of the changes. A reboot ensures that such software comes up in a known environment. Most programmers who write Windoze software are too lazy to detect certain changes at runtime.
I've had a lot of Sony products that piss me off in that respect. I bought a newer-model car CD player that won't read most CD-R's; yet my $25 "Lennox Sound" portable will even read CD-RW audio CDs.
OTOH, I have a Pioneer DVD player I picked up in 1999 that plays VCDs, and will read any media I can throw at it. I picked that model specifically because it mentions VCD on the front panel, and at the time I thought that was a pretty unique feature.
I personally like the 8mm CD-Rs. They hold about 24 minutes of audio or VCD, and are perfect for typical 30 minute shows (minus commercials == 22 minutes). They're much more convenient than the 12mm discs IMO.
I think Walmart is the company that could best challenge Microsoft Window's monopoly.
I doubt they are doing it to challenge MS, at least not directly. They are doing it because Walmart does whatever it takes to offer lower prices than everyone else. If Windows costs more than Linux, it only makes sense from their point of view to offer Linux. They don't care about the politics or technical merits of open-source software, etc.
They could preinstall any software they sell on the shelves.
I imagine Microtel could pre-install any software that they strike a licensing deal with. I don't think having the software on their shelf gives Walmart or Microtel any license other than to resell said software on said shelf.
If you have a problem with it, you can just bring it to a Walmart.
I seriously doubt Walmart will be offering much in the way of technical support. Hardware failures sure, you'll have your normal return-policy and warranty period, but if you trash your MBR or can't run Mozilla, I don't think they'll be much help (most likely, Microtel offers some kind of support).
If selling Lindows and Manrake increases their hardware sales, would they consider writing their own operating system? or maybe just Walmart Linux?
That's a major stretch. Again, they aren't interested in the politics, nor do I believe they are interested in becoming a tech company. They simply see that:
- PC hardware is dirt cheap right now
- Lindows/Mandrake is cheap
- As a result, they can offer a PC cheaper than the next guy
I suspect that after so many users return the machines (My AOL don't work on this computer, or Front Page won't run) they might reconsider.
Most applications worked fine if you'd click "No" when it asked you to reboot.
Also, a lot of applications would require a reboot if being installed under Win9x but not under Win2k. Rather than release two separate installers or check $WINVERSION, they just prompt for a reboot anyway.
Similar to how most installers tell you to shut down any running applications. This is only required if the installer is updating runtimes that are in use, which for most userland applications is not often.
I never shut down running programs and almost always click "No, I'll reboot later" under Win2k; very rarely do I have to actually reboot for a newly installed application to work. Even some drivers will work without a reboot...
Re:EBay and PayPal ... not my favorites
on
Ebay buys PayPal
·
· Score: 2
...but the link to verify a change to a password isn't random at all...
Hm, I checked this process out. Granted, I can't say whether or not the link URL is cryptographically secure, but I don't see any easily-identifiable pattern. Nor do I see any reason they wouldn't have a relatively secure "change password" URL generated, as it's simple to do (and a company of this size isn't going to cut corners like that, or we'd all have heard of this problem by now).
The fee where a $300 check had $24 of it shaved off the top to transfer.
I'm paid regularly via PayPal for contract work, and I've worked it out to be about 2.95% on average for me. A $600 paycheck cuts less than $18, which isn't bad in my opinion. Note that the fee appears to be the same whether my employer uses a credit card or funds already in PayPal to send (we've done it both ways), and both of us have "Business" accounts w/PayPal.
I'd have to see some evidence to support your theory on eBay's password change. Perhaps a link to a news story or article? Would that be too much to ask?
X does not crash *that* often. Just often enough to be annoying.
I run my laptop pretty much constantly, and Mozilla takes out the X server about once a week. It invariably happens after clicking a link; mind you, this happens on any site, there's nothing specific about the sites that crash it (the sites will then work fine afterword).
If I kill and restart Mozilla every couple of days I can put off the inevitable for a bit longer at a time.
It's not hardware. I can boot into Windows on the same machine and not have any problems. I used to run Windows on the laptop full-time before a few months ago. And again, it only crashes X, the OS is fine and in fact has quite a bit of uptime right now.
I do run KDE, but same thing happens in Gnome (I switch from one to the other every couple months... I'm an odd one). The X driver is simply "trident"; no DRI or OpenGL/glx support; only video option is sw_cursor. It's a very stable machine other than the X crashes which, again, are not all that often.
FWIW, when I used Opera exclusively for about 2 months, X didn't crash once (nor did Opera).
All in all, my words may have come out more harsh than I had intended. I just consider X almost a "necessary evil", though that's probably over-stating the case as well. It still beats using Windows for my every-day surfing and email.
Replacement cost is definitely a consideration. LEDs are widely used mostly because they rarely, if ever, need replacing (anyone else remember soldering new bulbs in the front panel of your stereo receiver, trying not to burn the string for the tuning dial?;)
I love LEDs, but I don't see plastic diodes that happen to emit light when biased properly replacing other forms of artificial light. More likely, something else will come along (perhaps even these 802.11b-killers).
Personally, I run 802.11b in my house so I can use the laptop anywhere. The microwave doesn't affect my connection at all, and these lights probably wouldn't have that much effect at the short distances I typically run (it's a small apartment).
Besides, hopefully we'll all be on 802.11a before these things catch on:)
I beg to differ about X. It doesn't suck for me. But I will agree that it does mostly suck for the non-nerd; that's probably what you meant.
I run Linux/X on most of my machines. X sucks, plain and simple. I am far from a non-nerd;)
For the most part, X works okay. However, it does hog memory and crash (Mozilla 1.0 seems to crash X often for me...) My Win2k box pisses me off at how stable it has been in comparison (flies in the face of my *nix vs Windows arguments; it's hard to convince people that the OS is stable, that just the GUI crashes).
My point was simply that, in my opinion, X sucks. I use it only for lack of anything better for my Linux and BSD systems...
They light longer when powered by batteries though because their intensity doesn't drop off as much as incandescents do as the voltage drops.
LEDs are diodes. They have a constant voltage drop, generally around 1.2 volts. As you increase the supply voltage (through a resistor) you increase the current flowing through the device. It's not really a function of the voltage applied - you get the same effect by changing the resistor value. It's simply I=V/R (current = voltage / resistance), taking into account the voltage drop across the diode.
What's important is the amount of current flowing through the LED.
Not that it matters... in any case LEDs are, by comparison, not all that efficient.
I've been wondering if you really need an inverter. From my inspection of power supplies, the typical design seems to rectify the mains voltage, then "chops" it through a small isolation step-down transformer, and rectifies it to about 25 volts. Then I think this DC voltage is used to drive the voltage regulators for your 5, 12, and 3.3 volts, etc. My guess is that maybe you could disconnect the mains circuitry and the isolation transformer, and apply battery power directly to 25 volt supply node.
That wouldn't work. A typical switching power supply works by rectifying the AC input directly (hence the large 200-400V capacitors), then using an oscillator to generate a high-frequency (usually square-wave) high-voltage AC. This then goes to a multi-tap transformer that steps the voltage down to a few different levels at different taps.
The voltage regulation is handled by both the low and high-voltage sides; the switching duty-cycle is modified based on the output voltage. All of this results in much more efficiency than 60 Hz line frequency, which is why a switching supply can be so small compared to a standard transformer-based supply (think laptop power supplies). Working with sine-wave low-frequency AC results in lots of unnecessary heat dissipation etc.
Some cheaper PC supplies also use a 7905 to regulate the -5V side, and a 7912 for the -12V, as it's cheaper -- but you can only draw about an amp from these regulators, though these two voltages are rarely needed (-12V is needed for standard serial ports I believe).
So, you couldn't just stick a DC 12V supply somewhere in the mix. You might could get away with building a DC regulator for the 5V and 3.3V supplies, but you wouldn't be able to regulate the 12V side very well (you lose at least 0.7V as a result of regulation).
Believe it or not, it is many many times more efficient to use a DC-AC inverter (usually also based on a MOSFET switching supply) feeding a PC power supply than it is to try and regulate DC, which involves dissipating excess as heat. While on the surface it sounds wasteful, you'd get much more battery time with the inverter, plus you're assured to have all of the correct voltage taps needed by the system.
I experimented with this quite a bit personally, and the DC-AC-DC process is much more efficient. Think about how a UPS works...
Thankfuly (sadly?) formatting and reinstalling on either Linux OR Windows is a lot easier then fixing most major screwups.:)
Under *nix, however, you can format all partitions except/home, thus preserving your data. Assuming of course you keep your data in/home, and/home is its own partition (as it should be).
Or better yet, never log in as 'root' unless necessary, and be careful when logged in as 'root', and you won't have to reinstall (though there are situations -- a major OS upgrade, or perhaps you'd been hacked -- but for the most part a little caution goes a long way).
My solution for problem B) was to run a local web server (Apache). I have a rewrite rule to redirect everything to 'index.html', which is a zero-byte file. My hosts file on each box I use simply points offending domains to my BSD box's IP, which runs the Apache web server (I used to run my own DNS cache which made this even easier, eg edit one place to affect all boxes on the network).
This way, those silly I-frame ads and banner images will simply (and instantly, no waiting for timeouts) show empty.
You could even do this using an external web site, if you have an extra IP laying around not being used on port 80, though that may be a bit wasteful depending on the circumstances...
I think Mozilla's speed depends on a lot of factors, including platform.
In my experience, Mozilla is fast as hell on Windows and FreeBSD (XFree86 3.3.something). Under Linux (2.4.18 + low latency + preempt, XFree86 4.x, my primary system) it's not nearly as fast.
It's still faster than Opera and way way faster than Konqueror, but I often find myself running Moz on the BSD box (remotely) just to get the speed.
Galeon is a bit faster but I tend to upgrade Mozilla often, and Galeon doesn't keep up and often breaks. The only major Galeon feature is saving your session (URL for each tab you have open) in the event of a crash; but Moz hasn't crashed for me since 0.9.8 anyway...
Anyway I read somewhere once (wish I had a URL) that Mozilla's speed problems on Linux have a lot to do with the system libraries. Not sure if this is a cop-out response or not, but I have noticed major moz speed differences between Linux, Win2k and FreeBSD, even on the same physical box. My experience has been with RedHat 6.2 and 7.2, Win2k SP2, and FreeBSD 4.4-release.
I think you are confusing the term "spyware" as used here. The article discusses spyware you install on a machine to spy on other users of that machine (eg, keystroke loggers etc), not ad-ware etc...
With that in mind, there are root kits out there that install kernel modules and trojan'd binaries to hide their presense. I'm sure there exist root kits that are able to detect/disable/thwart such tools as 'chkrootkit' et al. And of course most root kits sniff passwords and keep back doors open.
Though one needs to gain root access to install, a good root kit is pretty much on par with/better (worse) than this WinWhatWhere BS, but made just for *nix:)
Hm, this invalidates my previous conclusion. I thought maybe this was being blown out of proportion, and I figured by "referer" they simply meant the HTTP_REFERER was being mucked with.
However, I know that "qksrv.net" is (IIRC) Commission Junction, which tracks affiliates. Note btw that CJ's functionality relies on your browser setting/sending cookies to a third-party server (something all non-IE browsers let you disable), but that's beside the point.
Every day I find more reasons I'm glad I stopped using Windows and MSIE...
Maybe I'm reading it wrong, but I took 'refer' to mean simply the HTTP_REFERER header:
...StreamCast can make it look like it is referring traffic...
That simply sounds like, due to a meta or Location: redirect, the browser sends their website as the HTTP_REFERER, thus appearing that they are "refering" traffic, not using an affiliate/referral program.
I could well be wrong... but my first reaction was the less paranoid conclusion. I'm sure if they were actually hijacking affiliate links there would have been a lot more fuss about it by now.
I have to agree, 2.4.17 is rock solid in my opinion. I've kept up with the latest 2.4 kernel since 2.4.0 on my home systems, and currently I'm running several systems (3 home, 2 web servers) on 2.4.17. IMO this is the best 2.4 kernel we've had yet.
With the new VM, ext3 file system, and other major changes finally stabled out, I think.17 is proving to be pretty solid. Until now, I was keeping a few servers at 2.2.19/20, but now I feel we finally have a 2.4 kernel that is ready for production use.
Though this is a neat idea that I hadn't thought of, it's not too surprising. I opened up my LinkSys 5 port 10/100 switch (I was bored), to find that inside this tiny box is more space than components. The box itself would almost fit in a standard outlet box, and if gutted and re-arranged a bit...
Of course the price tag is a bit steep considering $60 buys the LinkSys, which is small enough not to worry about, but it's still a neat idea. I'm sure they'll get cheaper as time goes on.
Re:So, what's the best way to upgrade?
on
Linux 2.4.16 Released
·
· Score: 4, Informative
From my understanding the bug affects all filesystem types.
I patched my kernel to 2.4.16-pre1 yesterday in light of this bug, and here's what I did:
1) Compile kernel using my normal procedure
2) Switch to single user mode ('init 1')
3) 'sync' and 'umount' each partition (except/)
4) sync
5) shutdown -r -F now
No corruption, no problems (I'm on ext3 so the forced check wasn't even noticable).
You might be tempted to remount / read-only first, but if you do, first create '/forcefsck', which is exactly what the -F flag on 'shutdown' would do, but of course only if / was writable.
I agree, Windows isn't the problem in the case of DLLs. It really is stupid for an uninstall routine to ask the user whether to delete a DLL. It seems it should either know that it's not needed by any other program, or leave it alone. Asking the user (and really think about your typical Windows user) about deleting system files is a mistake. I've walked more friends and family through reinstalls after having uninstalled crappy shareware...
Unfortunately this practice is common thanks to InstallShield being used by so many programs, as InstallShield always asks before deleting a so-called "shared" DLL. Keep in mind, half of the time the DLL is program-specific (ie not shared), and other times it's something the program itself did not install in the first place (was already there). I don't think Windows itself is to blame here...
Win2k still suffers from this, but if you do delete a DLL it almost always magically reappears. It's part of some scheme to protect the system from its users I believe, but it is a real pain when you actually want to remove a DLL...
As for the Unix side, I've always wondered about the organisation (or lack thereof) of programs. Many tools do IMO belong in central locations (cat, grep, ls...) but anything larger should have its own directory. I long for the day when I can say:
export PATH=$PATH:/usr/programs/*/bin
or something to that effect...
Most of your larger packages do attempt to install into their own locations; Apache by default ends up with/usr/local/apache/* though it does tend to scatter a few things around. MySQL, Qmail, and a few others generally create subdirectories for most of their files. Not perfect, but it's a step in the right direction anyway.
I personally hate RPM, and I generally snag a tarball over an.rpm any day. I do like *BSD's ports collection quite a bit, but on RedHat RPM is about the best we've got. RPM is fine for the initial install, and even for adding some system-level tools/packages/upgrades, but any major software instalation after that I prefer to install manually; and of course, this doesn't help the issue at hand one bit...
Unfortunately, I have my complaints about filesystem standards, but I don't have any solutions either, really. Too much software exists that depends upon our current system, though a proposed future standard might be nice. Maybe a new POSIX recommendation is in order... and once some years go by, software vendors will slowly migrate to the new standard... of course I don't know what that standard might be...
I don't have any one favorite programming language, and I tend to choose the best tool for the job. I don't know Java, and never really cared for C++, but I use C, Perl, PHP, various shell scripting, and even VB...
If I'm writing an app for *n[ui]x, I usually stick with C for anything serious (where a shell script won't do). I may hand off certain tasks to other languages though.
When doing Windows programming (though I mostly use Linux, I do find myself working with Windows sometimes), I always choose VB for the interface, and any computational stuff goes in a DLL written in plain C. As much as I hate to admit it here, VB is the quickest way to build a usable interface on Windows IMO (and I have lots of BASIC background). It works well with DLLs written in any language (with the right calling convention), and this is a perfect example of using the best tool for the job.
When I'm asked what lanugage do you use, my response is simply whatever language(s) is/are best for the task at hand. Would you ask a mechanic what tool do you use, a wrench or a ratchet? Though it's not exactly the same thing, we do have many tools, some of which are more appropriate for a particular task. I wouldn't use bash to decode MP3 data, and I wouldn't use C to write a startup or installation script, though many times the choice isn't that obvious.
I like the X where it's at, but what would be nice is some feedback. It's not really a button, and sometimes it takes >1 second to actually close; using a laptop (touch pad) it's not always clear if you actually clicked it or not...
However, I'm glad Mozilla is at the point that I'm complaining about such simple things as this; it's been my primary browser since 0.9.2.
I did notice (as others have mentioned) how much faster it is on Windows. It may be my imagination but I could swear 0.9.6 is a lot faster than.5 was at startup and loading pages... but still has much room for improvement on the Linux side of things.
Slashdot Mirror
I was just installing various things to my win2k box and it still wants rebooting when: ...
Generally software installs ask for reboots because the software author didn't feel like actually determining if a reboot was necessary. They just prompt for a reboot to be safe. 9 times out of 10 it isn't necessary in Win2k - just click 'No' and try.
I know this works for Nero. I don't think the ATI Catalyst drivers actually needed a reboot either for me.
I don't recall having to reboot when changing the pagefile size (or even being prompted)... in fact, Win2k often resizes it on-the-fly.
Even changing the machine's IP from static to dynamic doesn't require a reboot if you go (re)start the DHCP client service.
MS just feels safer asking you to reboot, because you might have ignorant software running that isn't aware of the changes. A reboot ensures that such software comes up in a known environment. Most programmers who write Windoze software are too lazy to detect certain changes at runtime.
I'd never heard of TurboLinux until I bought a Linksys switch a couple of years ago. Also got another copy with my Linksys PCMCIA ethernet adaptor.
Never installed it, but I did browse the CD. Looked very much like RedHat 6.x to me, so I never bothered with it...
I've had a lot of Sony products that piss me off in that respect. I bought a newer-model car CD player that won't read most CD-R's; yet my $25 "Lennox Sound" portable will even read CD-RW audio CDs.
OTOH, I have a Pioneer DVD player I picked up in 1999 that plays VCDs, and will read any media I can throw at it. I picked that model specifically because it mentions VCD on the front panel, and at the time I thought that was a pretty unique feature.
I personally like the 8mm CD-Rs. They hold about 24 minutes of audio or VCD, and are perfect for typical 30 minute shows (minus commercials == 22 minutes). They're much more convenient than the 12mm discs IMO.
I think Walmart is the company that could best challenge Microsoft Window's monopoly.
I doubt they are doing it to challenge MS, at least not directly. They are doing it because Walmart does whatever it takes to offer lower prices than everyone else. If Windows costs more than Linux, it only makes sense from their point of view to offer Linux. They don't care about the politics or technical merits of open-source software, etc.
They could preinstall any software they sell on the shelves.
I imagine Microtel could pre-install any software that they strike a licensing deal with. I don't think having the software on their shelf gives Walmart or Microtel any license other than to resell said software on said shelf.
If you have a problem with it, you can just bring it to a Walmart.
I seriously doubt Walmart will be offering much in the way of technical support. Hardware failures sure, you'll have your normal return-policy and warranty period, but if you trash your MBR or can't run Mozilla, I don't think they'll be much help (most likely, Microtel offers some kind of support).
If selling Lindows and Manrake increases their hardware sales, would they consider writing their own operating system? or maybe just Walmart Linux?
That's a major stretch. Again, they aren't interested in the politics, nor do I believe they are interested in becoming a tech company. They simply see that:
- PC hardware is dirt cheap right now
- Lindows/Mandrake is cheap
- As a result, they can offer a PC cheaper than the next guy
I suspect that after so many users return the machines (My AOL don't work on this computer, or Front Page won't run) they might reconsider.
Most applications worked fine if you'd click "No" when it asked you to reboot.
Also, a lot of applications would require a reboot if being installed under Win9x but not under Win2k. Rather than release two separate installers or check $WINVERSION, they just prompt for a reboot anyway.
Similar to how most installers tell you to shut down any running applications. This is only required if the installer is updating runtimes that are in use, which for most userland applications is not often.
I never shut down running programs and almost always click "No, I'll reboot later" under Win2k; very rarely do I have to actually reboot for a newly installed application to work. Even some drivers will work without a reboot...
...but the link to verify a change to a password isn't random at all...
Hm, I checked this process out. Granted, I can't say whether or not the link URL is cryptographically secure, but I don't see any easily-identifiable pattern. Nor do I see any reason they wouldn't have a relatively secure "change password" URL generated, as it's simple to do (and a company of this size isn't going to cut corners like that, or we'd all have heard of this problem by now).
The fee where a $300 check had $24 of it shaved off the top to transfer.
I'm paid regularly via PayPal for contract work, and I've worked it out to be about 2.95% on average for me. A $600 paycheck cuts less than $18, which isn't bad in my opinion. Note that the fee appears to be the same whether my employer uses a credit card or funds already in PayPal to send (we've done it both ways), and both of us have "Business" accounts w/PayPal.
I'd have to see some evidence to support your theory on eBay's password change. Perhaps a link to a news story or article? Would that be too much to ask?
IDG: How about just the idea of having an open source browser, the Opera Web browser for instance...
:p
Anyone have the tarball handy?
X does not crash *that* often. Just often enough to be annoying.
I run my laptop pretty much constantly, and Mozilla takes out the X server about once a week. It invariably happens after clicking a link; mind you, this happens on any site, there's nothing specific about the sites that crash it (the sites will then work fine afterword).
If I kill and restart Mozilla every couple of days I can put off the inevitable for a bit longer at a time.
It's not hardware. I can boot into Windows on the same machine and not have any problems. I used to run Windows on the laptop full-time before a few months ago. And again, it only crashes X, the OS is fine and in fact has quite a bit of uptime right now.
I do run KDE, but same thing happens in Gnome (I switch from one to the other every couple months... I'm an odd one). The X driver is simply "trident"; no DRI or OpenGL/glx support; only video option is sw_cursor. It's a very stable machine other than the X crashes which, again, are not all that often.
FWIW, when I used Opera exclusively for about 2 months, X didn't crash once (nor did Opera).
All in all, my words may have come out more harsh than I had intended. I just consider X almost a "necessary evil", though that's probably over-stating the case as well. It still beats using Windows for my every-day surfing and email.
Replacement cost is definitely a consideration. LEDs are widely used mostly because they rarely, if ever, need replacing (anyone else remember soldering new bulbs in the front panel of your stereo receiver, trying not to burn the string for the tuning dial? ;)
:)
I love LEDs, but I don't see plastic diodes that happen to emit light when biased properly replacing other forms of artificial light. More likely, something else will come along (perhaps even these 802.11b-killers).
Personally, I run 802.11b in my house so I can use the laptop anywhere. The microwave doesn't affect my connection at all, and these lights probably wouldn't have that much effect at the short distances I typically run (it's a small apartment).
Besides, hopefully we'll all be on 802.11a before these things catch on
I beg to differ about X. It doesn't suck for me. But I will agree that it does mostly suck for the non-nerd; that's probably what you meant.
;)
I run Linux/X on most of my machines. X sucks, plain and simple. I am far from a non-nerd
For the most part, X works okay. However, it does hog memory and crash (Mozilla 1.0 seems to crash X often for me...) My Win2k box pisses me off at how stable it has been in comparison (flies in the face of my *nix vs Windows arguments; it's hard to convince people that the OS is stable, that just the GUI crashes).
My point was simply that, in my opinion, X sucks. I use it only for lack of anything better for my Linux and BSD systems...
They light longer when powered by batteries though because their intensity doesn't drop off as much as incandescents do as the voltage drops.
LEDs are diodes. They have a constant voltage drop, generally around 1.2 volts. As you increase the supply voltage (through a resistor) you increase the current flowing through the device. It's not really a function of the voltage applied - you get the same effect by changing the resistor value. It's simply I=V/R (current = voltage / resistance), taking into account the voltage drop across the diode.
What's important is the amount of current flowing through the LED.
Not that it matters... in any case LEDs are, by comparison, not all that efficient.
Actually, it is 0 bits...
This information is invaluable. More specifically, this equates to [FPE Exception: Divide by Zero] in USD per byte!
I've been wondering if you really need an inverter. From my inspection of power supplies, the typical design seems to rectify the mains voltage, then "chops" it through a small isolation step-down transformer, and rectifies it to about 25 volts. Then I think this DC voltage is used to drive the voltage regulators for your 5, 12, and 3.3 volts, etc. My guess is that maybe you could disconnect the mains circuitry and the isolation transformer, and apply battery power directly to 25 volt supply node.
That wouldn't work. A typical switching power supply works by rectifying the AC input directly (hence the large 200-400V capacitors), then using an oscillator to generate a high-frequency (usually square-wave) high-voltage AC. This then goes to a multi-tap transformer that steps the voltage down to a few different levels at different taps.
The voltage regulation is handled by both the low and high-voltage sides; the switching duty-cycle is modified based on the output voltage. All of this results in much more efficiency than 60 Hz line frequency, which is why a switching supply can be so small compared to a standard transformer-based supply (think laptop power supplies). Working with sine-wave low-frequency AC results in lots of unnecessary heat dissipation etc.
Some cheaper PC supplies also use a 7905 to regulate the -5V side, and a 7912 for the -12V, as it's cheaper -- but you can only draw about an amp from these regulators, though these two voltages are rarely needed (-12V is needed for standard serial ports I believe).
So, you couldn't just stick a DC 12V supply somewhere in the mix. You might could get away with building a DC regulator for the 5V and 3.3V supplies, but you wouldn't be able to regulate the 12V side very well (you lose at least 0.7V as a result of regulation).
Believe it or not, it is many many times more efficient to use a DC-AC inverter (usually also based on a MOSFET switching supply) feeding a PC power supply than it is to try and regulate DC, which involves dissipating excess as heat. While on the surface it sounds wasteful, you'd get much more battery time with the inverter, plus you're assured to have all of the correct voltage taps needed by the system.
I experimented with this quite a bit personally, and the DC-AC-DC process is much more efficient. Think about how a UPS works...
Thankfuly (sadly?) formatting and reinstalling on either Linux OR Windows is a lot easier then fixing most major screwups. :)
/home, thus preserving your data. Assuming of course you keep your data in /home, and /home is its own partition (as it should be).
Under *nix, however, you can format all partitions except
Or better yet, never log in as 'root' unless necessary, and be careful when logged in as 'root', and you won't have to reinstall (though there are situations -- a major OS upgrade, or perhaps you'd been hacked -- but for the most part a little caution goes a long way).
My solution for problem B) was to run a local web server (Apache). I have a rewrite rule to redirect everything to 'index.html', which is a zero-byte file. My hosts file on each box I use simply points offending domains to my BSD box's IP, which runs the Apache web server (I used to run my own DNS cache which made this even easier, eg edit one place to affect all boxes on the network).
This way, those silly I-frame ads and banner images will simply (and instantly, no waiting for timeouts) show empty.
You could even do this using an external web site, if you have an extra IP laying around not being used on port 80, though that may be a bit wasteful depending on the circumstances...
I think Mozilla's speed depends on a lot of factors, including platform.
In my experience, Mozilla is fast as hell on Windows and FreeBSD (XFree86 3.3.something). Under Linux (2.4.18 + low latency + preempt, XFree86 4.x, my primary system) it's not nearly as fast.
It's still faster than Opera and way way faster than Konqueror, but I often find myself running Moz on the BSD box (remotely) just to get the speed.
Galeon is a bit faster but I tend to upgrade Mozilla often, and Galeon doesn't keep up and often breaks. The only major Galeon feature is saving your session (URL for each tab you have open) in the event of a crash; but Moz hasn't crashed for me since 0.9.8 anyway...
Anyway I read somewhere once (wish I had a URL) that Mozilla's speed problems on Linux have a lot to do with the system libraries. Not sure if this is a cop-out response or not, but I have noticed major moz speed differences between Linux, Win2k and FreeBSD, even on the same physical box. My experience has been with RedHat 6.2 and 7.2, Win2k SP2, and FreeBSD 4.4-release.
I think you are confusing the term "spyware" as used here. The article discusses spyware you install on a machine to spy on other users of that machine (eg, keystroke loggers etc), not ad-ware etc...
:)
With that in mind, there are root kits out there that install kernel modules and trojan'd binaries to hide their presense. I'm sure there exist root kits that are able to detect/disable/thwart such tools as 'chkrootkit' et al. And of course most root kits sniff passwords and keep back doors open.
Though one needs to gain root access to install, a good root kit is pretty much on par with/better (worse) than this WinWhatWhere BS, but made just for *nix
Hm, this invalidates my previous conclusion. I thought maybe this was being blown out of proportion, and I figured by "referer" they simply meant the HTTP_REFERER was being mucked with.
However, I know that "qksrv.net" is (IIRC) Commission Junction, which tracks affiliates. Note btw that CJ's functionality relies on your browser setting/sending cookies to a third-party server (something all non-IE browsers let you disable), but that's beside the point.
Every day I find more reasons I'm glad I stopped using Windows and MSIE...
Maybe I'm reading it wrong, but I took 'refer' to mean simply the HTTP_REFERER header:
...StreamCast can make it look like it is referring traffic...
That simply sounds like, due to a meta or Location: redirect, the browser sends their website as the HTTP_REFERER, thus appearing that they are "refering" traffic, not using an affiliate/referral program.
I could well be wrong... but my first reaction was the less paranoid conclusion. I'm sure if they were actually hijacking affiliate links there would have been a lot more fuss about it by now.
I have to agree, 2.4.17 is rock solid in my opinion. I've kept up with the latest 2.4 kernel since 2.4.0 on my home systems, and currently I'm running several systems (3 home, 2 web servers) on 2.4.17. IMO this is the best 2.4 kernel we've had yet.
.17 is proving to be pretty solid. Until now, I was keeping a few servers at 2.2.19/20, but now I feel we finally have a 2.4 kernel that is ready for production use.
With the new VM, ext3 file system, and other major changes finally stabled out, I think
Though this is a neat idea that I hadn't thought of, it's not too surprising. I opened up my LinkSys 5 port 10/100 switch (I was bored), to find that inside this tiny box is more space than components. The box itself would almost fit in a standard outlet box, and if gutted and re-arranged a bit...
Of course the price tag is a bit steep considering $60 buys the LinkSys, which is small enough not to worry about, but it's still a neat idea. I'm sure they'll get cheaper as time goes on.
From my understanding the bug affects all filesystem types.
/)
I patched my kernel to 2.4.16-pre1 yesterday in light of this bug, and here's what I did:
1) Compile kernel using my normal procedure
2) Switch to single user mode ('init 1')
3) 'sync' and 'umount' each partition (except
4) sync
5) shutdown -r -F now
No corruption, no problems (I'm on ext3 so the forced check wasn't even noticable).
You might be tempted to remount / read-only first, but if you do, first create '/forcefsck', which is exactly what the -F flag on 'shutdown' would do, but of course only if / was writable.
I agree, Windows isn't the problem in the case of DLLs. It really is stupid for an uninstall routine to ask the user whether to delete a DLL. It seems it should either know that it's not needed by any other program, or leave it alone. Asking the user (and really think about your typical Windows user) about deleting system files is a mistake. I've walked more friends and family through reinstalls after having uninstalled crappy shareware...
/usr/local/apache/* though it does tend to scatter a few things around. MySQL, Qmail, and a few others generally create subdirectories for most of their files. Not perfect, but it's a step in the right direction anyway.
.rpm any day. I do like *BSD's ports collection quite a bit, but on RedHat RPM is about the best we've got. RPM is fine for the initial install, and even for adding some system-level tools/packages/upgrades, but any major software instalation after that I prefer to install manually; and of course, this doesn't help the issue at hand one bit...
Unfortunately this practice is common thanks to InstallShield being used by so many programs, as InstallShield always asks before deleting a so-called "shared" DLL. Keep in mind, half of the time the DLL is program-specific (ie not shared), and other times it's something the program itself did not install in the first place (was already there). I don't think Windows itself is to blame here...
Win2k still suffers from this, but if you do delete a DLL it almost always magically reappears. It's part of some scheme to protect the system from its users I believe, but it is a real pain when you actually want to remove a DLL...
As for the Unix side, I've always wondered about the organisation (or lack thereof) of programs. Many tools do IMO belong in central locations (cat, grep, ls...) but anything larger should have its own directory. I long for the day when I can say:
export PATH=$PATH:/usr/programs/*/bin
or something to that effect...
Most of your larger packages do attempt to install into their own locations; Apache by default ends up with
I personally hate RPM, and I generally snag a tarball over an
Unfortunately, I have my complaints about filesystem standards, but I don't have any solutions either, really. Too much software exists that depends upon our current system, though a proposed future standard might be nice. Maybe a new POSIX recommendation is in order... and once some years go by, software vendors will slowly migrate to the new standard... of course I don't know what that standard might be...
I don't have any one favorite programming language, and I tend to choose the best tool for the job. I don't know Java, and never really cared for C++, but I use C, Perl, PHP, various shell scripting, and even VB...
If I'm writing an app for *n[ui]x, I usually stick with C for anything serious (where a shell script won't do). I may hand off certain tasks to other languages though.
When doing Windows programming (though I mostly use Linux, I do find myself working with Windows sometimes), I always choose VB for the interface, and any computational stuff goes in a DLL written in plain C. As much as I hate to admit it here, VB is the quickest way to build a usable interface on Windows IMO (and I have lots of BASIC background). It works well with DLLs written in any language (with the right calling convention), and this is a perfect example of using the best tool for the job.
When I'm asked what lanugage do you use, my response is simply whatever language(s) is/are best for the task at hand. Would you ask a mechanic what tool do you use, a wrench or a ratchet? Though it's not exactly the same thing, we do have many tools, some of which are more appropriate for a particular task. I wouldn't use bash to decode MP3 data, and I wouldn't use C to write a startup or installation script, though many times the choice isn't that obvious.
I like the X where it's at, but what would be nice is some feedback. It's not really a button, and sometimes it takes >1 second to actually close; using a laptop (touch pad) it's not always clear if you actually clicked it or not...
.5 was at startup and loading pages... but still has much room for improvement on the Linux side of things.
However, I'm glad Mozilla is at the point that I'm complaining about such simple things as this; it's been my primary browser since 0.9.2.
I did notice (as others have mentioned) how much faster it is on Windows. It may be my imagination but I could swear 0.9.6 is a lot faster than