In that case, I can't wait to see the load times on the iPhone with 3g. I wonder if they'll be more like your Voyager or more like your parents' phone. The hardware in the iPhone sure isn't shabby, but it's rendering a much more complex page.
I haven't found a consumer phone yet that has support for a QWERTY-style keyboard and isn't WM-based. I refuse to use a phone that only has a dial pad, since those things are a pain; and WM2003 and WM5 were both unstable the point of multiple "crashes" a day.
I seem to remember someone talking about the possibilities of phones and toilets and microwaves crashing in the future. Dear goodness - at least I'm there already with one of them!
I wouldn't have let the EDGE network stop me, if it weren't for the fact that the 3g iPhone is already imminent.
My own experience with Verizon has been so abysmal, I'm very excited that my current contract is up on June 6th of this year. Several of my friends who have iPhones love them, and the AT&T service in the area is excellent. Their main complaint is usually that the EDGE connection is slow
Looks like I'll be stuck without a phone from June 6th until the new iPhone comes out.
Exactly. I've done quite a bit of work out of the New Testament and we have to deal with the Critical Apparati all the time, so I've read some about manuscript transmission. That was part of my point, though not so stated. No one bothered to copy and transmit someone's daily chores list because they saw no value in it, thus the large number of people who may have made them probably did not have them preserved and thus they are not difficult for us to find in the same way that finding manuscripts of great literature is difficult, if not impossible, in almost every case.
Realize that by the time of the Roman empire quite a large number of people could read and write. There ARE shopping lists, personal letters and so forth in great abundance all over the ancient Greek and Roman worlds. You can find graffiti on buildings, etc. True, it may be hard to find modern scholarly copies of them because more people are interested in reading the old copies of the religious texts than they are someone's shopping list, but they do exist. Also, what do you do with your shopping list when you're done with it? Probably not try to save it, as the ancient world did with its literature. So thus more of those small, petty writings would have been lost, but some did survive.
Adding the tarring step to the M42.txt file to get M42.txt.tbz2 gives me 159 bytes. Using gzip alone gives me 25K (25,264 bytes) and a tgz file gives me about the same (25,335 bytes). Interesting comparison.
If the file is in decimal then I get the following comparison:
original file: 8,046,119 bytes
bzip2 file: 3,504,368 bytes
gzip file: 3,795,422 bytes
The tar files that accompany are only about 1k larger than the zip files themselves.
Funny, yes. Misinformed, also a yes. The compile would only happen once when you first creat the LiveCD, after that it is binary and you can use it for whatever system you're on (provided the CFLAGS don't break it for that architecture). Gentoo LiveCDs would have to be compiled as often as your base Gentoo install is on a normal machine (read: once).
Actually, as the previous poster mentioned, building the kernel is not that bad. If you do a Gentoo install, building the krenel takes no more time than installing a kernel designed for your hardware than it does on any other distro. The real time-consuming compiles are the major, core applicatoins. Compiling glibc can take upwards of a few hours if you compile all of the language pages, X can take a few hours also, and if you want to compile openoffice.org you're looking at probably over 10 hours (just over 12 on my 2.4GHz Intel). Going from ground up (stage1) on an automated Gentoo install (script controlled), I build a whole system, including X, in around 12 or so hours. But the kernel doesn't take too much of your time. Most certainly you would not have to compile the kernel each time you boot the CD, since the compiled kernel would go right on to your CD with all of its modules, etc.
Interesting that you should bring this up. We have a number of servers with different hardware, although they are all Intel machines. I have built a copy of Gentoo for each of the basic processors that we have (Pentium2, Pentium3, SMP, etc). Then, all I have to do is boot a new server with any LiveCD (bummer I can't get cpqarray support with the normal Gentoo LiveCD's, but it is fortunate that the FC2 CDs have support for it) that can partition the hard drive and has tar/bz2. I tar up a build of the architechture once it has reached its basic level of what we have/need/use/like and then unzip it. It is just like a Gentoo-stage3 install, but fully customized.
Every month or two I make a new tarball that has the most recent and up-to-date installs on it so that the install never gets too stale. In less time than it takes to install FC or RedHat or Mandrake I can have a Gentoo server built just the way I like it. The only exception is the few times when I have had to rebuild the kernel because the hardware in one of the servers had a strange SCSI card or something else, other than that the kernel can have the basic hardware that we use compiled as modules and run rather well.
On the other side, since I don't like my install even being one or two months stale, I recently wrote a basic shell script that sets up the chroot environment for Gentoo, downloads the stage1 and portage snapshot from our local mirror, and runs through the install of Gentoo right up until the time to choose kernel options with no need to interact other than to edit the make.conf at the beginning of the build...fully automated for our hardware and the way we like it. Once you've used Linux/Gentoo for some time, it becomes easy to set-up and install any of the secure daemons etc that you need.
I know this is not a "favorite distro" rant, and Gentoo probably isn't my favorite to run for home applications, but as a server I have had a hard time beating it. If you use the hardened profile and only install what you want/need, it is pretty easy. And if you're going to administer a number of servers, it is best to set-up a local mirror to update portage from regularly and to download your source files once, the rest of the machines can fetch from that local mirror (in our case partial mirror, since we don't have a large enough Linux server to be a full source mirror). That severly cuts down on the time you need to do things like making tarballs of your install, etc.
If you honestly want instructions on doing any of what I have mentioned, just email me and ask, I'll be more than happy to share with you my ways. For over-all performance we have been able to cut out a few high-powered Windows servers and replace them with light-weight to medium-weight Linux boxes to do most of the same functions. Some things Microsoft obviously does better for the end-user (Exchange 2003 is loved around here) but some things OSS does better for the admins (basically stuff that's transparent to our clients).
Hmmm, we use PostgreSQL at work, running on a Linux box (most of the servers, except mail gateways, SQL and a few other miscellanies run Win2003) and it works wonderfully. PostgreSQL, exim (for a gateway setup), Samba file-sharing...they all are using either LDAP or Kerberos to authenticate either a user log-in or that mail is sent to a valid AD user. So far we have been really happy with the migration to Linux (all within the past 6 months) and some more features are starting to migrate (listserv, a CVS-clone, dial-up authentication) and we have had no problem integrating with AD running on a Win2003 box.
On a side note, I have been unable to get the latest release of the Nvidia drivers for my GeForce4 Ti to compile with gcc-3.4. No problems with the older versions of gcc, but the newest versions don't compile. Now, mind you, I have had no problems with performance, since Nvidia has awesome support for Linux, but there is a struggle somewhere in there to get the driver to compile.
If this is the case, perhaps we should look to Quantum Mechanics for a parallel to this problem. It reminds me of the statement (perhaps by Bohr) "You never really understand quantum physics, you just get used to it." It seems the same is true of the stock market. No one truly understands when or why the market takes dives and climbs, and certainly if someone knew that they were going to happen it would affect the results: just as knowing the precise location of a subatomic particle severely affects the velocity of the particle and vice-versa.
If Mendlebrot would like to start looking for patterns in the market, perhaps he should start with probability and waves at the subatomic level (in the meantime he can build a machine that overwhelms people with their tiny existence in the universe by analyzing the forces affecting a piece of cake...Douglas Adams anyone?). I wish him luck.
Slashdot Mirror
In that case, I can't wait to see the load times on the iPhone with 3g. I wonder if they'll be more like your Voyager or more like your parents' phone. The hardware in the iPhone sure isn't shabby, but it's rendering a much more complex page. I haven't found a consumer phone yet that has support for a QWERTY-style keyboard and isn't WM-based. I refuse to use a phone that only has a dial pad, since those things are a pain; and WM2003 and WM5 were both unstable the point of multiple "crashes" a day. I seem to remember someone talking about the possibilities of phones and toilets and microwaves crashing in the future. Dear goodness - at least I'm there already with one of them! I wouldn't have let the EDGE network stop me, if it weren't for the fact that the 3g iPhone is already imminent.
My own experience with Verizon has been so abysmal, I'm very excited that my current contract is up on June 6th of this year. Several of my friends who have iPhones love them, and the AT&T service in the area is excellent. Their main complaint is usually that the EDGE connection is slow Looks like I'll be stuck without a phone from June 6th until the new iPhone comes out.
Exactly. I've done quite a bit of work out of the New Testament and we have to deal with the Critical Apparati all the time, so I've read some about manuscript transmission. That was part of my point, though not so stated. No one bothered to copy and transmit someone's daily chores list because they saw no value in it, thus the large number of people who may have made them probably did not have them preserved and thus they are not difficult for us to find in the same way that finding manuscripts of great literature is difficult, if not impossible, in almost every case.
Realize that by the time of the Roman empire quite a large number of people could read and write. There ARE shopping lists, personal letters and so forth in great abundance all over the ancient Greek and Roman worlds. You can find graffiti on buildings, etc. True, it may be hard to find modern scholarly copies of them because more people are interested in reading the old copies of the religious texts than they are someone's shopping list, but they do exist. Also, what do you do with your shopping list when you're done with it? Probably not try to save it, as the ancient world did with its literature. So thus more of those small, petty writings would have been lost, but some did survive.
Adding the tarring step to the M42.txt file to get M42.txt.tbz2 gives me 159 bytes. Using gzip alone gives me 25K (25,264 bytes) and a tgz file gives me about the same (25,335 bytes). Interesting comparison. If the file is in decimal then I get the following comparison: original file: 8,046,119 bytes bzip2 file: 3,504,368 bytes gzip file: 3,795,422 bytes The tar files that accompany are only about 1k larger than the zip files themselves.
Interesting... Windows puts up a 25K compressed folder of the file, while bzip2 outputs an 85 byte file. Industry leaders lead, once again.
Funny, yes. Misinformed, also a yes. The compile would only happen once when you first creat the LiveCD, after that it is binary and you can use it for whatever system you're on (provided the CFLAGS don't break it for that architecture). Gentoo LiveCDs would have to be compiled as often as your base Gentoo install is on a normal machine (read: once).
Actually, as the previous poster mentioned, building the kernel is not that bad. If you do a Gentoo install, building the krenel takes no more time than installing a kernel designed for your hardware than it does on any other distro. The real time-consuming compiles are the major, core applicatoins. Compiling glibc can take upwards of a few hours if you compile all of the language pages, X can take a few hours also, and if you want to compile openoffice.org you're looking at probably over 10 hours (just over 12 on my 2.4GHz Intel). Going from ground up (stage1) on an automated Gentoo install (script controlled), I build a whole system, including X, in around 12 or so hours. But the kernel doesn't take too much of your time. Most certainly you would not have to compile the kernel each time you boot the CD, since the compiled kernel would go right on to your CD with all of its modules, etc.
Interesting that you should bring this up. We have a number of servers with different hardware, although they are all Intel machines. I have built a copy of Gentoo for each of the basic processors that we have (Pentium2, Pentium3, SMP, etc). Then, all I have to do is boot a new server with any LiveCD (bummer I can't get cpqarray support with the normal Gentoo LiveCD's, but it is fortunate that the FC2 CDs have support for it) that can partition the hard drive and has tar/bz2. I tar up a build of the architechture once it has reached its basic level of what we have/need/use/like and then unzip it. It is just like a Gentoo-stage3 install, but fully customized. Every month or two I make a new tarball that has the most recent and up-to-date installs on it so that the install never gets too stale. In less time than it takes to install FC or RedHat or Mandrake I can have a Gentoo server built just the way I like it. The only exception is the few times when I have had to rebuild the kernel because the hardware in one of the servers had a strange SCSI card or something else, other than that the kernel can have the basic hardware that we use compiled as modules and run rather well. On the other side, since I don't like my install even being one or two months stale, I recently wrote a basic shell script that sets up the chroot environment for Gentoo, downloads the stage1 and portage snapshot from our local mirror, and runs through the install of Gentoo right up until the time to choose kernel options with no need to interact other than to edit the make.conf at the beginning of the build...fully automated for our hardware and the way we like it. Once you've used Linux/Gentoo for some time, it becomes easy to set-up and install any of the secure daemons etc that you need. I know this is not a "favorite distro" rant, and Gentoo probably isn't my favorite to run for home applications, but as a server I have had a hard time beating it. If you use the hardened profile and only install what you want/need, it is pretty easy. And if you're going to administer a number of servers, it is best to set-up a local mirror to update portage from regularly and to download your source files once, the rest of the machines can fetch from that local mirror (in our case partial mirror, since we don't have a large enough Linux server to be a full source mirror). That severly cuts down on the time you need to do things like making tarballs of your install, etc. If you honestly want instructions on doing any of what I have mentioned, just email me and ask, I'll be more than happy to share with you my ways. For over-all performance we have been able to cut out a few high-powered Windows servers and replace them with light-weight to medium-weight Linux boxes to do most of the same functions. Some things Microsoft obviously does better for the end-user (Exchange 2003 is loved around here) but some things OSS does better for the admins (basically stuff that's transparent to our clients).
Hmmm, we use PostgreSQL at work, running on a Linux box (most of the servers, except mail gateways, SQL and a few other miscellanies run Win2003) and it works wonderfully. PostgreSQL, exim (for a gateway setup), Samba file-sharing...they all are using either LDAP or Kerberos to authenticate either a user log-in or that mail is sent to a valid AD user. So far we have been really happy with the migration to Linux (all within the past 6 months) and some more features are starting to migrate (listserv, a CVS-clone, dial-up authentication) and we have had no problem integrating with AD running on a Win2003 box.
On a side note, I have been unable to get the latest release of the Nvidia drivers for my GeForce4 Ti to compile with gcc-3.4. No problems with the older versions of gcc, but the newest versions don't compile. Now, mind you, I have had no problems with performance, since Nvidia has awesome support for Linux, but there is a struggle somewhere in there to get the driver to compile.
If this is the case, perhaps we should look to Quantum Mechanics for a parallel to this problem. It reminds me of the statement (perhaps by Bohr) "You never really understand quantum physics, you just get used to it." It seems the same is true of the stock market. No one truly understands when or why the market takes dives and climbs, and certainly if someone knew that they were going to happen it would affect the results: just as knowing the precise location of a subatomic particle severely affects the velocity of the particle and vice-versa. If Mendlebrot would like to start looking for patterns in the market, perhaps he should start with probability and waves at the subatomic level (in the meantime he can build a machine that overwhelms people with their tiny existence in the universe by analyzing the forces affecting a piece of cake...Douglas Adams anyone?). I wish him luck.
FYI: Man in room = Doctor McCoy