Rob contributed the 1-digit spot, so I'll fill the 3-digit spot. I guess I could have registered a little earlier, but I didn't post much back then (and don't now).
Do you realize most of Windows on Alpha was running with the same old 32-bit data sizes for all the APIs? And Microsoft's applications for NT on Alpha still had most of their 32-bit limitations (SQL Server, Exchange, other Backoffice servers). They didn't get around to cleaning up their applications to run correctly in a 64-bit system until Itanium, when they started talking about the "Win64" platform. Even according to Microsoft's own press-releases, 64-bit Windows was released in 2005 (to coincide with the IA-64 releases).
I don't think Microsoft's Alpha effort compares very favorably to the Unix and Linux offerings of the same period, both of which had made a complete transition to the platform. I ran Linux on Alpha for many years, and all my applications ran natively at full speed; no emulation software or hardware required.
I tried OpenBSD once, years ago, for a corporate firewall. It was to replace a failing machine, which was configured in some sort of stupid bridging position at the head of our network (external and internal Ethernet interfaces had the same IP, and the firewall rules were based on interfaces). I knew this was a stupid way to run a network, but it was the only thing the ISP would allow, and I thought I'd give OpenBSD a shot.
I installed OpenBSD, and configured the two network cards. I assigned an IP to the external interface, and tried to use it: worked great. I assigned the same IP to the internal interface.
Kernel panic. From a user-space tool.
I grabbed a FreeBSD CD and installed that instead, which didn't crash on me when I configured the cards. A few days later we got an actual subnet between the firewall and the ISP's router, so we didn't have to do the bridge thing. The firewall has since been replaced with a new, beefier machine with five interfaces, running Linux.
I know that PostgreSQL uses write-ahead-logging so it can avoid exactly these kinds of problems. It doesn't matter how much I/O PostgreSQL is doing; all writes go to the log. If the machine crashes, it replays the log file up to its most recent write. Worst case: data that was in the process of being appended to the log when the machine crashed didn't get flushed to disk, and that last transaction is lost. No tables are corrupt. No 6+ hour delay getting back online.
You would know this to if you had read the PostgreSQL documentation.
In my opinion, the FIA needs to open up the engine rules.
Allow V8, V10, V12. Normally aspirated engines at 3.5, 3.0, and 2.5 (respecitvely). Turbo at 2, 1.5, 1 litre (respectively). These displacements are just examples, real sizes would need tweaking. The advantage to these changes would be that engine manufacturers would be able to exploit their own different strengths as applied to different designs. For example, Honda may choose a turbocharged V10, Ferrari a standard pressure V12, Ford may go back to a high torque V8. This may provide for an enjoyable shakeup in power availability.
Grab smartmontools and run them on your drive (like "smartctl -a/dev/hda" or similar). Most SCSI and most newer ATA drives will maintain a SMART error log of any defects/problems. smartmontools will also print drive attributes (for most drives) that can tell you when a drive is about to fail, before it actually does.
85 MHz? I've got a Sun 4/110 at home that still runs NetBSD pretty well. First Sparc you could fit on your desktop, and probably the slowest ever made.
Much of SourceGear's computing infrastructure is Unix-based, and free software is used for things like e-mail, DNS, backups, and mailing lists. We use this software primarily because it's reliable and efficient. These systems were mostly put in place years ago, and only need periodic software updates and hardware check-ups.
Windows and IIS were the most convenient platform for our corporate web site given our.NET product focus. You can visit Eric's Eric's personal web site, which was running Apache last time I checked.
The delay only seems to happen when using keyboard shortcuts to open new windows, although I believe the same code in librep is executed for menu-launched programs. Sometimes I'll get 2 or even 3 instances of the program I want to launch when this delay happens.
You're using sawfish, right? I am, and I tracked down the problem to a line in librep. I mailed John Harper just now about it. If you're not using sawfish, you can ignore the rest of this message.
It all has to do with the following lines from unix_processes.c:
The call to select() on kernel 2.6 causes the timeout to be fully expired every time. This code is in the parent branch after the fork() (right before it starts waitpid() on the child), so everything freezes for at least 1 full second. Dropping this timeout to 1 usec causes the problem to go away.
Considering I've been using similar models since the 1980s (and the RS/6000 workstations still seem to ship with them, and I personally own three models from the 1990s), I'd gladly pay $200 for one. What other computer component can you buy for $200, use 8 hours a day, every day, for ten years, and still have in perfect working condition? Even my Logitech mice have a life span of about 4 years.
You can still buy those really nice IBM keyboards (I own three; won't do any real work without them). The "real click" is part of the buckling spring mechanism used in the key supports. These keyboards aren't cheap, but they last forever. I just looked, and couldn't find any at commerce.www.ibm.com. I think Lexmark actually made them the last few years.
I bought one about a year ago, under the "Options from IBM" name. IBM has a notice about its limited Options Continuation Program, where it links to vendors who still carry the Options line of peripherals. I guess the Options program was discontinued by IBM sometime in the last year.
A company called I. T. Exchange is one of these Options partners, and has a keyboard/mouse inventory list, but none of the keyboards seem to be in the normal price range for one of the really nice keyboards (they were about $100.00 when I bought my last one), and I don't recognize any of the model numbers. Who knows--maybe they've got them in bulk and are trying to get rid of them at less than $20.00 each!
0%? Surely you're rounding your numbers down a bit, because I can name quite a few web servers running on AMD Athlon processors. www.lloop.com is just one example. Linux runs just fine on the Athlon; the only problems I've witnessed in Athlon systems had to do with substandard mainboard chipsets.
Look at Linux? This text was originally posted by Shoeboy, who has stated he works for Microsoft. The original comment was "Sorry to do this you CmdrTaco", and can be found archived in this thread (Just Say No to Reading About Drugs).
Slashdot Mirror
My second question was the more important one. Any answer?
Who told you that? Why did you believe it? Please cite your sources.
Rob contributed the 1-digit spot, so I'll fill the 3-digit spot. I guess I could have registered a little earlier, but I didn't post much back then (and don't now).
Hello!
Do you realize most of Windows on Alpha was running with the same old 32-bit data sizes for all the APIs? And Microsoft's applications for NT on Alpha still had most of their 32-bit limitations (SQL Server, Exchange, other Backoffice servers). They didn't get around to cleaning up their applications to run correctly in a 64-bit system until Itanium, when they started talking about the "Win64" platform. Even according to Microsoft's own press-releases, 64-bit Windows was released in 2005 (to coincide with the IA-64 releases).
I don't think Microsoft's Alpha effort compares very favorably to the Unix and Linux offerings of the same period, both of which had made a complete transition to the platform. I ran Linux on Alpha for many years, and all my applications ran natively at full speed; no emulation software or hardware required.
How would you describe the software you write?
I tried OpenBSD once, years ago, for a corporate firewall. It was to replace a failing machine, which was configured in some sort of stupid bridging position at the head of our network (external and internal Ethernet interfaces had the same IP, and the firewall rules were based on interfaces). I knew this was a stupid way to run a network, but it was the only thing the ISP would allow, and I thought I'd give OpenBSD a shot.
I installed OpenBSD, and configured the two network cards. I assigned an IP to the external interface, and tried to use it: worked great. I assigned the same IP to the internal interface.
Kernel panic. From a user-space tool.
I grabbed a FreeBSD CD and installed that instead, which didn't crash on me when I configured the cards. A few days later we got an actual subnet between the firewall and the ISP's router, so we didn't have to do the bridge thing. The firewall has since been replaced with a new, beefier machine with five interfaces, running Linux.
Actually, that would be a trilemma.
You would know this to if you had read the PostgreSQL documentation.
V12s are gone. F1 cars all use V10s now.
In my opinion, the FIA needs to open up the engine rules.
Allow V8, V10, V12. Normally aspirated engines at 3.5, 3.0, and 2.5 (respecitvely). Turbo at 2, 1.5, 1 litre (respectively). These displacements are just examples, real sizes would need tweaking. The advantage to these changes would be that engine manufacturers would be able to exploit their own different strengths as applied to different designs. For example, Honda may choose a turbocharged V10, Ferrari a standard pressure V12, Ford may go back to a high torque V8. This may provide for an enjoyable shakeup in power availability.
Grab smartmontools and run them on your drive (like "smartctl -a /dev/hda" or similar). Most SCSI and most newer ATA drives will maintain a SMART error log of any defects/problems. smartmontools will also print drive attributes (for most drives) that can tell you when a drive is about to fail, before it actually does.
85 MHz? I've got a Sun 4/110 at home that still runs NetBSD pretty well. First Sparc you could fit on your desktop, and probably the slowest ever made.
Windows and IIS were the most convenient platform for our corporate web site given our .NET product focus. You can visit Eric's Eric's personal web site, which was running Apache last time I checked.
I've edited 40 GB files in XEmacs (raw disk devices) to recover a few password hashes. Works fine.
Oh, I forgot:
The delay only seems to happen when using keyboard shortcuts to open new windows, although I believe the same code in librep is executed for menu-launched programs. Sometimes I'll get 2 or even 3 instances of the program I want to launch when this delay happens.
You're using sawfish, right? I am, and I tracked down the problem to a line in librep. I mailed John Harper just now about it. If you're not using sawfish, you can ignore the rest of this message.
It all has to do with the following lines from unix_processes.c:
The call to select() on kernel 2.6 causes the timeout to be fully expired every time. This code is in the parent branch after the fork() (right before it starts waitpid() on the child), so everything freezes for at least 1 full second. Dropping this timeout to 1 usec causes the problem to go away.
I'm sitting on the same T1 as the AbiSource server and things are snappy for me.
Most cell phones these days have an "off" button.
--
Considering I've been using similar models since the 1980s (and the RS/6000 workstations still seem to ship with them, and I personally own three models from the 1990s), I'd gladly pay $200 for one. What other computer component can you buy for $200, use 8 hours a day, every day, for ten years, and still have in perfect working condition? Even my Logitech mice have a life span of about 4 years.
--
I bought one about a year ago, under the "Options from IBM" name. IBM has a notice about its limited Options Continuation Program, where it links to vendors who still carry the Options line of peripherals. I guess the Options program was discontinued by IBM sometime in the last year.
A company called I. T. Exchange is one of these Options partners, and has a keyboard/mouse inventory list, but none of the keyboards seem to be in the normal price range for one of the really nice keyboards (they were about $100.00 when I bought my last one), and I don't recognize any of the model numbers. Who knows--maybe they've got them in bulk and are trying to get rid of them at less than $20.00 each!
--
0%? Surely you're rounding your numbers down a bit, because I can name quite a few web servers running on AMD Athlon processors. www.lloop.com is just one example. Linux runs just fine on the Athlon; the only problems I've witnessed in Athlon systems had to do with substandard mainboard chipsets.
--
Yes, just like that.
--
I agree--personally, I think the web browser thing is just too much. Call me traditional.
--
Look at Linux? This text was originally posted by Shoeboy, who has stated he works for Microsoft. The original comment was "Sorry to do this you CmdrTaco", and can be found archived in this thread (Just Say No to Reading About Drugs).
--
MacOS: Dogcow says moof!
--