To be fair, Microsoft announced a FS Search(similar technology found in Spotlight/Longhorn) to be included in NT 4. I think it was called Cairo, but could be mistaken. Microsoft has talked about this search technology in place of folder structures for a decade, which is fine, I wouldn't think machines that ran NT could run this type of indexing search, but that all comes down to implementation... they just have never found(maybe Longhorn will have) the way or ability to implement it for a public release
They can do 2 things at a time, but they are still going to the same computer components... Running a game alongside your AV scan STILL isn't going to work, because your AV software is still using the same system bus to go to the same IDE cable to go to the same harddrive. Just because your processors is duplicated, doesn't mean the rest of your system is.
Google's stated goal is to: "organizing the world's information and making it universally accessible and useful."
This is just one more check-mark on that list that spans a couple dozen different fields of information now, continuing into the new Video search technology.
"I am not yet at the point where I need my own cluster, but when calculations start to eat up hours per day or even whole days, you start to think about these things."
You may know this already but with Xgrid being built into OS X proper, buying a second Powermac could be used as a ad hoc 2-node cluster. If money isn't a problem, getting 2 of the new Powermacs, one as your workstation and both serving as grid nodes with Xgrid may be your best bet.
So what if a gnome setting is wrong, Ubuntu (and all the other major distros) patch and modify Gnome and the kernel all the time to make them customized to their system. Make a patch to fix stupid spelling mistakes that shouldn't have been made. Simple enough.
If Debian dies than it will either be remade in another Debian-influenced distrobution or the developers and users will migrate to other distrobutions... like Ubuntu. Don't be so arrogant, the world is not crashing. Free software doesn't die with its developer, remember.
Debian is where it is today because of QA and standards have crippled its ability to release stable and relatively up-to-date versions of its distribution. That is something that Ubuntu excels at.
I thought the rule of thumb was to have full speed emulation of another architecture (old atari, nintendo, etc) you needed 10x the processor speed. Is that not true?
Given the specific hardware, I wouldn't think even the fastest of the dual proc Powermacs would be able to properly play an xbox (let alone a 3-processor PS3) emulator+game in any usable or enjoyable way.
Wikipedia should have "1.0" pages... Specifically locking accurate versions as milestones for certain versions. While they would allow for new versions to be made, they could sell dvd's of the most recent milestone (a version of the article specifically checked by a real person for accuracy). I would buy a set of those.
Since this will likely be a html-based directory structure. A
Z is for bugfixes. Y is for feature additions. X is for backwards compatibility breakage.
10.3.8 was a bug fix on the 10.3 line.
10.3.0 was a major set of new features (100+) to the 10.x line.
10.0 broke(in fundamental ways, excluding Classic environment, etc) compatibility with OS 9. New kernel and all.
OS X 11.0 (not the X will likely not become XI because it doesn't sound cool) will happen when backwards compatibility is broken or their is something fundamentally big that changes to the OS itself.
PS: Apple also reserves the right to do whatever the hell they want. But they seem to follow this more or less correctly.
*Move clients(Linux, Mac, Windows, Unix, Switches, Routers, etc, etc) to a more scalable/aggressive TCP due to larger bandwidth being available with broadband users(this makes computer better tap internet connections they have available more quickly).
*Keep moving to IPv6 in the background/backbone and use DNS to hide the ugly addresses to end users. IPv6 will make random scans of hosts more difficult(theoretically) and it requires IPSEC on all connections for better authentication.
*Create more root DNS servers hosted by universities, the government, and/or interested businesses. Diversity is a good thing.
*Encourage companies to use(and sell) standards compliant tools and products. Such as the latest CSS2, XHTML, etc.
*Get rid of web-accessible default passwords for machines/appliances as quickly as possible. A first-login password change would be a start. Having virtually every switch, router and popular server appliance having a default (and sometimes unremovable) password setup is just scary.
As for spam. I think ISPs being more careful on who rogue servers they accept email from(enough to block zombie machines, not enough to block small time webmasters from hosting their own email server), and server AND client side spam filtering will fix the problem. Spam filters, white/black lists, etc can cut over 99% of spam out, even in heavy traffic environments. I don't think spam blocking needs to be a protocol or even a standard, just a common practice.
Now, if we could get email protocols that better authenticated to proper domains(spoofing and such) then we might have a something.
That is due to the way they focused on OS X. Apple focused on all the groundwork(10.0,.1, etc) and foundation and only in.2 and.3 have they focused on optimizing what is there. Each release gets faster because 1. the past release (early OS X) was quite slow) and 2. they optimize non-optimized code. There is a ceiling, one can only assume, on how long Apple can improve speed on each release on moderate/old hardware.
PS: Apple has made a release every 12-18 months on OS X (every.x release, that is), Steve Jobs has said that this pace is too break neck and they will be slowing down for 10.5 (11?) and on. Can't blame them, their release cycle has been unreal.
Just an FYI, works great in Firefox, but renders usable in Safari.. Much nicer version(when viewed through FF) then the Google page though. Nice improvement.
I don't think scalable counts for google. They have thousands of servers but each one is no more than 2 processors each. They buy lowend servers and their CUSTOM filesystem and load-balancing software makes them work so well in the high-deman world Google lives in.
If yum supported the feature to in-place upgrade distros (FC2 -> FC3, etc) without anything more than a single command, then I wouldn't mind in the slightest. But the upgrade cycle for Fedora is too awkward at the moment to enjoy a 6 month cycle.
"Anyone Remember ST:TNG episode "Pegasus" Where Starfleet had a Phased Cloaking device. Well the Federation(Capt. Picard) Came clean with the Romulans and acknowledged that They broke the no cloak Treaty. I bet some higher-ups at Google watched that same episode!"
The IBM Thinkpad X series (their 12" screen line of Thinkpads) is already manufactured in China by a compeitor to Lenovo. Their other series are (I believe) manufatured in the US.
Slashdot Mirror
To be fair, Microsoft announced a FS Search(similar technology found in Spotlight/Longhorn) to be included in NT 4. I think it was called Cairo, but could be mistaken. Microsoft has talked about this search technology in place of folder structures for a decade, which is fine, I wouldn't think machines that ran NT could run this type of indexing search, but that all comes down to implementation... they just have never found(maybe Longhorn will have) the way or ability to implement it for a public release
They can do 2 things at a time, but they are still going to the same computer components... Running a game alongside your AV scan STILL isn't going to work, because your AV software is still using the same system bus to go to the same IDE cable to go to the same harddrive. Just because your processors is duplicated, doesn't mean the rest of your system is.
Google's stated goal is to: "organizing the world's information and making it universally accessible and useful."
This is just one more check-mark on that list that spans a couple dozen different fields of information now, continuing into the new Video search technology.
"I am not yet at the point where I need my own cluster, but when calculations start to eat up hours per day or even whole days, you start to think about these things."
You may know this already but with Xgrid being built into OS X proper, buying a second Powermac could be used as a ad hoc 2-node cluster. If money isn't a problem, getting 2 of the new Powermacs, one as your workstation and both serving as grid nodes with Xgrid may be your best bet.
So what if a gnome setting is wrong, Ubuntu (and all the other major distros) patch and modify Gnome and the kernel all the time to make them customized to their system. Make a patch to fix stupid spelling mistakes that shouldn't have been made. Simple enough.
How is this modded Insightful?
If Debian dies than it will either be remade in another Debian-influenced distrobution or the developers and users will migrate to other distrobutions... like Ubuntu. Don't be so arrogant, the world is not crashing. Free software doesn't die with its developer, remember.
Debian is where it is today because of QA and standards have crippled its ability to release stable and relatively up-to-date versions of its distribution. That is something that Ubuntu excels at.
"Part One was: "How to afford two 30" Apple Monitors.""
Step one: How to find a buying on the black market for your spare kidney.
I thought the rule of thumb was to have full speed emulation of another architecture (old atari, nintendo, etc) you needed 10x the processor speed. Is that not true?
Given the specific hardware, I wouldn't think even the fastest of the dual proc Powermacs would be able to properly play an xbox (let alone a 3-processor PS3) emulator+game in any usable or enjoyable way.
Well Ubuntu is arguably the fastest growing and largest of the 'minor distros' in the market... Besides which, Ubuntu already DOES run on Macs.
Him pointing out that a developer of a proprietary system owns and has control of that system doesn't make him wise, that makes him conscious.
Flame me if you want. I don't buy that software development is a religious experience.
Wikipedia should have "1.0" pages... Specifically locking accurate versions as milestones for certain versions. While they would allow for new versions to be made, they could sell dvd's of the most recent milestone (a version of the article specifically checked by a real person for accuracy). I would buy a set of those.
Since this will likely be a html-based directory structure. A
Apple uses the very sane X.Y.Z numbering system
Z is for bugfixes.
Y is for feature additions.
X is for backwards compatibility breakage.
10.3.8 was a bug fix on the 10.3 line.
10.3.0 was a major set of new features (100+) to the 10.x line.
10.0 broke(in fundamental ways, excluding Classic environment, etc) compatibility with OS 9. New kernel and all.
OS X 11.0 (not the X will likely not become XI because it doesn't sound cool) will happen when backwards compatibility is broken or their is something fundamentally big that changes to the OS itself.
PS: Apple also reserves the right to do whatever the hell they want. But they seem to follow this more or less correctly.
I would say:
*Move clients(Linux, Mac, Windows, Unix, Switches, Routers, etc, etc) to a more scalable/aggressive TCP due to larger bandwidth being available with broadband users(this makes computer better tap internet connections they have available more quickly).
*Keep moving to IPv6 in the background/backbone and use DNS to hide the ugly addresses to end users. IPv6 will make random scans of hosts more difficult(theoretically) and it requires IPSEC on all connections for better authentication.
*Create more root DNS servers hosted by universities, the government, and/or interested businesses. Diversity is a good thing.
*Encourage companies to use(and sell) standards compliant tools and products. Such as the latest CSS2, XHTML, etc.
*Get rid of web-accessible default passwords for machines/appliances as quickly as possible. A first-login password change would be a start. Having virtually every switch, router and popular server appliance having a default (and sometimes unremovable) password setup is just scary.
As for spam. I think ISPs being more careful on who rogue servers they accept email from(enough to block zombie machines, not enough to block small time webmasters from hosting their own email server), and server AND client side spam filtering will fix the problem. Spam filters, white/black lists, etc can cut over 99% of spam out, even in heavy traffic environments. I don't think spam blocking needs to be a protocol or even a standard, just a common practice.
Now, if we could get email protocols that better authenticated to proper domains(spoofing and such) then we might have a something.
That is due to the way they focused on OS X. Apple focused on all the groundwork(10.0, .1, etc) and foundation and only in .2 and .3 have they focused on optimizing what is there. Each release gets faster because 1. the past release (early OS X) was quite slow) and 2. they optimize non-optimized code. There is a ceiling, one can only assume, on how long Apple can improve speed on each release on moderate/old hardware.
.x release, that is), Steve Jobs has said that this pace is too break neck and they will be slowing down for 10.5 (11?) and on. Can't blame them, their release cycle has been unreal.
PS: Apple has made a release every 12-18 months on OS X (every
"The first one out with a working product tends to win the market, as long as their product is the best."
That should be modded Funny, not Insightful.
here is a working image for you... linkage
Just an FYI, works great in Firefox, but renders usable in Safari.. Much nicer version(when viewed through FF) then the Google page though. Nice improvement.
Be Gentle
Here is an example.
I don't think scalable counts for google. They have thousands of servers but each one is no more than 2 processors each. They buy lowend servers and their CUSTOM filesystem and load-balancing software makes them work so well in the high-deman world Google lives in.
If yum supported the feature to in-place upgrade distros (FC2 -> FC3, etc) without anything more than a single command, then I wouldn't mind in the slightest. But the upgrade cycle for Fedora is too awkward at the moment to enjoy a 6 month cycle.
Is your shift key possessed?
The IBM Thinkpad X series (their 12" screen line of Thinkpads) is already manufactured in China by a compeitor to Lenovo. Their other series are (I believe) manufatured in the US.
IBM, they are manufacturing all the processors for the consoles(Revolution, PS3, XBox2). All based on the Power5 processor line, I believe.