Actually there is a culture of "I am on a mac, why do I need backups" that exists because things are so easy to setup, and faily so infrequently.
Not on the server level, but on the desktop / user experience level. And if they are buying a machine that apple is selling as "easy to use as their desktop machines" with "no IT required," then yes, Apple is mis representing their product.
Of course, Apple can't sell enterprise level equipment worth shit, and knows it. 90% of their sales staff (in enterprise) would rather sell a product and never have to maintain contact with that customer until they need to buy something more again. Anything requiring more work or knowledge than what they had anticipating and deployed in house will rarely get done, unless you are buying MILLIONS of dollars of equipment.
As I've said up thread, the chances that they have an IT person at all, as in someone whose job it is to maintain the functioning and support of their IT infrastructure and services, and to secure them, is probably ZERO. They are web coders and dot commers, who just had an idea, bought an xserve, and threw their code up on it and had it run, and then collected ad revenue.
Actually, since they were using an OS X Server, this is more telling of their skills than just how pathetic their backup plan was. I admin OS X boxes, but the last thing I would imagine using them for would be a HA data store for a website. The builds of php and mysql are not kept up to date, and really the only cases I have seen an OS X box deployed for webhosting was by web coders with no experience running an actual server. They liked they could get the box up and running with multiple IP addresses quickly, and would then just go about their work in apache. Not thinking about backup, not thinking about disaster recovery.
And while the hardware may look tasty and easy to buy (hey, one click ordering at store.apple.com, unlimited user licenses!!!), you can get a lot more bang for your buck with a dell or hp linux box.
Someone who had the ability to setup a Linux box might have had a few more grains of intellect to think to atleast run rsync with cron to backup the data store offsite, if even just once a month.
I mean, how did these guys plan to do a staged update, incase an apple software update broke their apache install? Break the raid, install on drive 1, and if it works, mirror over drive 2, therefore giving them no snapshot in time to roll back to?
The logout kickstarts the windowing system again, they could rewrite it to make it switch on the fly, but that would probably break something else, and not something you would expect in a 10.5.X release.
Of course, they could write their own MacBook Pro specific windowing / graphics code, and just have that as a 'temporary' solution that supports live switching until 10.6 (where the entire graphics engine is having a major overhaul), but that defeats the whole "write once, use many" design philosophy of Apple's development cycle.
So Apple adopted a slick, bleeding edge tech, but instead of holding back for them to be able to make it work 'as it should' they introduced what could be considered the best practical work around that would involve the least amount of damage. I would hate to have to test any code that uses the graphics engine specifically on the newest macbook pro, because Apple decided to break their own development procedures to provide 'live' switching in a service pack. I'd rather just wait for the code to be in place globally, and have more incentive than "it sometimes breaks on macbook pros" to rework my code. Like being able to utilize a lot of the new low level stuff coming in 10.6.
Remember, this is flash, not magnetic bits stored on a spinning metal platter were header drift and other things would theoretically allow you to retrieve data that has long been removed.
Recovering from (intentionally overwritten flash) may be considerably harder than a traditional drive. Most flash recovery apps for cameras, etc. are really just reading the stray bits, as the formatting, etc. does not actually wipe each sector (because flash is rated in number of write operations the individual bits can support before going bad, so you want to minimize that).
Overwriting a flash storage partition on an iphone or other device also makes this harder because you can't easily pop those things open and mount the custom flash chip into some universal adapter and read its filesystem like you can do with any old hard drive (they even make forensic, read only, hard drive enclosures).
So I zero out the data on my iphone, and well, there aren't any jailbroken or app store apps that you can run on the damn thing to do a low level recovery anyway, and I don't know of any target disk raw access mode to the device when attached to a computer that is available outside of apple's developer labs.
I am part of the seed program, and there is no apple provided NDA only forum, for any of their products. This is the biggest issue folks are having with their NDA.
That is the problem. I can beta test apple's products, but since it is a one way feedback loop (that I am PAYING to be a part of), what good is it. I can't actively help in the development.
The problem really hasn't been that we are chiding Apple for(we being the OS X Server admins who support these boxes)most of us have gone and compiled our own versions of bind without issue, or being forwarding all recursion to opendns's servers, etc. (And custom installed BIND versions appear to be working fine in the server so far for most people).
It has been the total lack of response or public acknowledgement of the problem, no timeline for a fix, no patch and or updated knowledge base article on how to resolve the issue.
Apple just doesn't see it as a big enough issue to state it.
They still haven't responded to the ARDAgent.app privilege escalation exploit either.
I understand your pain. On the plus side, if you are a python / ruby developer, you have some things to look forward too, as a lot of apple's own components are being written in them, so those installs actually work most of the time. The perl one, not so much.
Of course, the biggest limitation to their serious server implementation is that there is not apple provided forum for users to be able to discuss their issues with beta release software. Let alone a publicly searchable bug tracker (right now we search by submitting bugs and waiting to see how long until the ticket is closed as either "by design" "we will get on it" or "duplicate"). So why should I bother to actually install a beta build of apples stuff to test, I can't really give any feedback, and there isn't any documentation out there floating around on how AD binding works in 10.5 vs 10.4. Which is a great example, apple's implementation of 10.5's ad plugin did not take into account that there could be multiple servers available, and that the first record returned for the ldap service may not be the same server as what was returned for the KDC. Why? probably because they only tested this with a single windows server in their test lab. So their engineers never even thought this was a problem. Of course, if we could test such things in beta, and they could see a group of folks bitching on the forums (all under NDA) about it, then they could probably even post a hot patch to the people with the issues and get faster feedback as a communication between the engineers and the people who are actually USING their code.
10.4 server made it really easy to provide recursion to the entire internet.
Also, to get your cache poisoned, all you need is an employee to visit a nice page full of LOLCats on a malicious server that will keep feeding them dns requests in the background.
The fact that it apparently isn't just shows how seriously they take their server business.
Which is a shame, because they do tend to make some good stuff, and when you want to build something to help manage and work with a group of macs, a mac server can make things a lot easier.
Of course, this is a company that didn't test their AD binding under 10.5 in anything larger than a single AD server installation (because why would apple have a multiple window servers to configure as a real AD deployment when their company doesn't actually use them)
this is related to Apple's OS X Server product, which runs DNS (bind in fact), and many mac businesses do in fact use it, if even as a local DNS cache (which a simple fix now would be to configure their boxes to us opendns).
The bigger issue is this is a pretty big deal on the security front, all of the businesses that apple has to compete with in the server space (especially in the eyes of enterprise IT), have had a fix and a public statement about it out the door. Apple is the big unix vendor missing off the list, and has not even made a public statement as such to inform it's users about the issue. Not exactly the best way to talk about how secure their products are (client and server).
Of course, they still haven't gotten around to fixing the ARDAgent.app vulnerability from a few weeks back either.
At the Angrydome (which I started out of frustration of this and other things Apple related)
The only statements we have been able to get out of apple has been from the bug reporting tool. They have stated that they are working on a fix, but it is causing problems in some instances of their deployments, but don't see it as an emergency because there isn't a targeted exploit against their user base.
They do not need to understand that this is a protocol specific issue, not a code specific issue.
Of course, we are assuming that these gateways are the only public facing ones, and there may be back end trunks or vpns bridging them to each other (so internal communications are done over private channels, while external stuff is offloaded to local internet access / backbones). Depending on how the current network is setup, compromising one of these existing 4k gateways could get you access into one of the other 3,999 networks that it is associated with.
Also, consolidating the traffic means they could track and monitor outgoing communications from their network, so instead of a researcher at one smaller agency selling their findings to the highest bidder / 'evil country' because it goes unnoticed, they get picked up on a larger gateway that is watching everything. Of course there will be the depends and pressure of monitoring such a flow of information.
And it will be a cluster pretty much how its cut, as they are migrating and shifting a huge amount of resources and management policies, and government IT is not exactly high on the list of many well experienced and intelligent IT person's places to work for.
Trying to maintain standards and practices across 4,000 gateway points vs 50. Let alone the agency bureaucracy that would be involved in doing site checks and working across various agency boundaries would be a nightmare. It would take eons to get those things in place to do consistent auditing and management to ensure standards and procedures are followed, let alone actually do them. Might as well consolidate bandwidth costs and number of checkpoints down to 50 in the process.
Because office lasers and related supplies cost more than a bulk printing center.
They could probably buy every member of congress a Kindle and still save in the end.
Actually.
Why DONT they buy every member of congress a kindle, that way they can get instant EVDO downloads of every bill that is ever submitted to congress, whenever, wherever they are? And search it.
If you want to spend the time to setup and maintain a custom LDAP directory solution to manage mac workstations, you could under linux (or use active directory, you just have to tweak the schema). The big value of OS X Server is that it integrates and manages mac workstations really well. Also, AFP, apple's native network filing protocol, isn't half bad and works best hosted on a mac server.
OS X Server licenses are 1 grand per instance (or $500 for 10 simulatenous file sharing users, no cap on web connections). No processor license, no connected user licenses. From a practical stand point, it only looks for local machines on the same subnet who have the same serial, so you could have multiple installations for testing, as long as any two were not on at the same time.
In a lot of cases I would imagine you have an xserve, or a vmware based boot loader, running OS X Server in one instance to be a lightweight file server, email server (kerio), doing open directory for the mac workstations, and then an instance of Windows 2003 Server hosting that one app you would have dedicated another box to in the past, like a web based time keeping software that requires IIS.
To reply to this, as you are pretty vehement about attacking MacFixit (not that I don't blame you, but they aren't entirely at fault for this), however Apple has announced an official response suggesting to remove APE for the bluescreen issue (http://docs.info.apple.com/article.html?artnum=306857), separate to your above login issue, so APE issue is not FUD, and considering how it works (using the 10.4 debug interface to execute arbitrary code within an application, which Leopards random library assignment is meant to prevent) no surprise it is borked in 10.5
The bigger issue is vendors like Logitech now install it on machines without notifying the user, or providing a way to update it. So you have folks who don't even know they have an old version of APE, which does not disable itself properly. APE is a very very delicate application, and should not be installed without the user knowing it. You can send some of your hate towards Logitech also.
It doesn't help that the APE developers are vehement that their program doesn't cause issues either.
it is not even that. They are following the money. Porn is dealing with the lowest common denominator and lowest barrier to entrance. Instead of a passionate love story, lets just show the 'good' parts. They went to VHS because they saw that it was allowing for lower cost films, faster turn around time and private home viewing. You had a broader audience and you got to cut out the theatre distribution chain. Since they were always a marginal aspect of society (there isn't a XXXMPAA to push laws for them) they have to meet consumer demand, instead of regulate the consumers. And to say that the industry shifted, well as a whole "porn" may changing formats and delivery mechanisms, but most of these changes are still being first to market by new comers to the industry, because each tech revolution lowered the entry point.
VHS made made amatuer porn challenge the Film market. MiniDV, DVD and the Internet made it even cheaper, so that anyone with some starting money could be a porn star. Now Vivid and other shops have to compete with 'watchmehavesexallday.com' etc, so of course they are going to adapt. They can't lobby or pass laws banning those smaller sites since they aren't exactly popular enough to get senators on their side (however I wonder if some of them backed the age verification requirement, knowing that the smaller sites would crumble under the paperwork).
To say they are broad visioned and open minded is a stretch, I'd just say that they are always on edge and have been fighting to be around, legally and financially, so they have to stay quick and adapt, or fold. A lot of porn distributors went out of business when Video hit, another group went out of business with Traci Lords underage porn too.
Actually, if you are a mac shop, you just buy one machine (intel imac + linux triple boot+ xp oem + parallels = everything you need to test a website for almost any computer) and if you are a pc shop, you buy your web designer a mac to run windows (full time if he wants) and to boot into os x also. I mean they should have one anyway, cs3 on intel macs screams.
Licensing isn't even in the same planet compared to MS.
unlimited license of server is $1000. it COMES with the xserve.
Maintenance (3 years of free upgrades, for 10.5-6, etc.) is another $1000, and entirely worth it.
So initial license purchase on top of the hardware is $1000 if you want 3 years of major versions of os x server. From past experience, that saves you $1000, because 2 more updates will happen in the next 3 years.
You are looking at $4,000 from apple vs $4313 from dell, but the dell only comes with 5 CALs (bare minimum 1u dual dual core xenon servers).
The thing is the "FCC rumour mill" would have expanded past the normal circle in the case of APPLE releasing the iPhone.
Sure, the specs on the next Sony Ericcson phone is great and all, but thats what they do, make cell phones. The press has been waiting with baited breath for the Apple iPhone for ages. All those press monkeys need is the "FCC rumour mill" to confirm the Apple iPhone, and it will be on the front page of newspapers, because you aren't going to waste time for FCC approval on a device you aren't going to sell. And the FCC approval process includes complete specifications, manual and details of the product also, so you might as well announce it prior to going in for FCC approval.
And Apple has done this before anyway. They did it with the original iBook and Airport. Built in wireless was a huge deal, but they announced it prior to getting FCC approval so they could keep Airport a secret.
Re:Prize goes to the 3D graphics provider
on
VMware Fusion goes Beta
·
· Score: 2, Informative
Nope, you are wrong. Power Mac was first introduced with the powerpc line of desktop computers.
PowerBook was a term apple was using for their laptops, which did not start with powerpc chips.
Actually there is a culture of "I am on a mac, why do I need backups" that exists because things are so easy to setup, and faily so infrequently.
Not on the server level, but on the desktop / user experience level. And if they are buying a machine that apple is selling as "easy to use as their desktop machines" with "no IT required," then yes, Apple is mis representing their product.
Of course, Apple can't sell enterprise level equipment worth shit, and knows it. 90% of their sales staff (in enterprise) would rather sell a product and never have to maintain contact with that customer until they need to buy something more again. Anything requiring more work or knowledge than what they had anticipating and deployed in house will rarely get done, unless you are buying MILLIONS of dollars of equipment.
As I've said up thread, the chances that they have an IT person at all, as in someone whose job it is to maintain the functioning and support of their IT infrastructure and services, and to secure them, is probably ZERO. They are web coders and dot commers, who just had an idea, bought an xserve, and threw their code up on it and had it run, and then collected ad revenue.
Actually, since they were using an OS X Server, this is more telling of their skills than just how pathetic their backup plan was. I admin OS X boxes, but the last thing I would imagine using them for would be a HA data store for a website. The builds of php and mysql are not kept up to date, and really the only cases I have seen an OS X box deployed for webhosting was by web coders with no experience running an actual server. They liked they could get the box up and running with multiple IP addresses quickly, and would then just go about their work in apache. Not thinking about backup, not thinking about disaster recovery.
And while the hardware may look tasty and easy to buy (hey, one click ordering at store.apple.com, unlimited user licenses!!!), you can get a lot more bang for your buck with a dell or hp linux box.
Someone who had the ability to setup a Linux box might have had a few more grains of intellect to think to atleast run rsync with cron to backup the data store offsite, if even just once a month.
I mean, how did these guys plan to do a staged update, incase an apple software update broke their apache install? Break the raid, install on drive 1, and if it works, mirror over drive 2, therefore giving them no snapshot in time to roll back to?
When developing a game for the most popular online phone game store will net you $250,000 in two months, as an independent developer:
http://toucharcade.com/2008/09/19/trism-developer-makes-250000-in-2-months/
The logout kickstarts the windowing system again, they could rewrite it to make it switch on the fly, but that would probably break something else, and not something you would expect in a 10.5.X release.
Of course, they could write their own MacBook Pro specific windowing / graphics code, and just have that as a 'temporary' solution that supports live switching until 10.6 (where the entire graphics engine is having a major overhaul), but that defeats the whole "write once, use many" design philosophy of Apple's development cycle.
So Apple adopted a slick, bleeding edge tech, but instead of holding back for them to be able to make it work 'as it should' they introduced what could be considered the best practical work around that would involve the least amount of damage. I would hate to have to test any code that uses the graphics engine specifically on the newest macbook pro, because Apple decided to break their own development procedures to provide 'live' switching in a service pack. I'd rather just wait for the code to be in place globally, and have more incentive than "it sometimes breaks on macbook pros" to rework my code. Like being able to utilize a lot of the new low level stuff coming in 10.6.
Remember, this is flash, not magnetic bits stored on a spinning metal platter were header drift and other things would theoretically allow you to retrieve data that has long been removed.
Recovering from (intentionally overwritten flash) may be considerably harder than a traditional drive. Most flash recovery apps for cameras, etc. are really just reading the stray bits, as the formatting, etc. does not actually wipe each sector (because flash is rated in number of write operations the individual bits can support before going bad, so you want to minimize that).
Overwriting a flash storage partition on an iphone or other device also makes this harder because you can't easily pop those things open and mount the custom flash chip into some universal adapter and read its filesystem like you can do with any old hard drive (they even make forensic, read only, hard drive enclosures).
So I zero out the data on my iphone, and well, there aren't any jailbroken or app store apps that you can run on the damn thing to do a low level recovery anyway, and I don't know of any target disk raw access mode to the device when attached to a computer that is available outside of apple's developer labs.
I am part of the seed program, and there is no apple provided NDA only forum, for any of their products. This is the biggest issue folks are having with their NDA.
That is the problem. I can beta test apple's products, but since it is a one way feedback loop (that I am PAYING to be a part of), what good is it. I can't actively help in the development.
The problem really hasn't been that we are chiding Apple for(we being the OS X Server admins who support these boxes)most of us have gone and compiled our own versions of bind without issue, or being forwarding all recursion to opendns's servers, etc. (And custom installed BIND versions appear to be working fine in the server so far for most people).
It has been the total lack of response or public acknowledgement of the problem, no timeline for a fix, no patch and or updated knowledge base article on how to resolve the issue.
Apple just doesn't see it as a big enough issue to state it.
They still haven't responded to the ARDAgent.app privilege escalation exploit either.
I understand your pain. On the plus side, if you are a python / ruby developer, you have some things to look forward too, as a lot of apple's own components are being written in them, so those installs actually work most of the time. The perl one, not so much.
Of course, the biggest limitation to their serious server implementation is that there is not apple provided forum for users to be able to discuss their issues with beta release software. Let alone a publicly searchable bug tracker (right now we search by submitting bugs and waiting to see how long until the ticket is closed as either "by design" "we will get on it" or "duplicate"). So why should I bother to actually install a beta build of apples stuff to test, I can't really give any feedback, and there isn't any documentation out there floating around on how AD binding works in 10.5 vs 10.4. Which is a great example, apple's implementation of 10.5's ad plugin did not take into account that there could be multiple servers available, and that the first record returned for the ldap service may not be the same server as what was returned for the KDC. Why? probably because they only tested this with a single windows server in their test lab. So their engineers never even thought this was a problem. Of course, if we could test such things in beta, and they could see a group of folks bitching on the forums (all under NDA) about it, then they could probably even post a hot patch to the people with the issues and get faster feedback as a communication between the engineers and the people who are actually USING their code.
10.4 server made it really easy to provide recursion to the entire internet.
Also, to get your cache poisoned, all you need is an employee to visit a nice page full of LOLCats on a malicious server that will keep feeding them dns requests in the background.
The fact that it apparently isn't just shows how seriously they take their server business.
Which is a shame, because they do tend to make some good stuff, and when you want to build something to help manage and work with a group of macs, a mac server can make things a lot easier.
Of course, this is a company that didn't test their AD binding under 10.5 in anything larger than a single AD server installation (because why would apple have a multiple window servers to configure as a real AD deployment when their company doesn't actually use them)
this is related to Apple's OS X Server product, which runs DNS (bind in fact), and many mac businesses do in fact use it, if even as a local DNS cache (which a simple fix now would be to configure their boxes to us opendns).
The bigger issue is this is a pretty big deal on the security front, all of the businesses that apple has to compete with in the server space (especially in the eyes of enterprise IT), have had a fix and a public statement about it out the door. Apple is the big unix vendor missing off the list, and has not even made a public statement as such to inform it's users about the issue. Not exactly the best way to talk about how secure their products are (client and server).
Of course, they still haven't gotten around to fixing the ARDAgent.app vulnerability from a few weeks back either.
need to lay off the coffee right now.
At the Angrydome (which I started out of frustration of this and other things Apple related)
The only statements we have been able to get out of apple has been from the bug reporting tool. They have stated that they are working on a fix, but it is causing problems in some instances of their deployments, but don't see it as an emergency because there isn't a targeted exploit against their user base.
They do not need to understand that this is a protocol specific issue, not a code specific issue.
Of course, we are assuming that these gateways are the only public facing ones, and there may be back end trunks or vpns bridging them to each other (so internal communications are done over private channels, while external stuff is offloaded to local internet access / backbones). Depending on how the current network is setup, compromising one of these existing 4k gateways could get you access into one of the other 3,999 networks that it is associated with.
Also, consolidating the traffic means they could track and monitor outgoing communications from their network, so instead of a researcher at one smaller agency selling their findings to the highest bidder / 'evil country' because it goes unnoticed, they get picked up on a larger gateway that is watching everything. Of course there will be the depends and pressure of monitoring such a flow of information.
And it will be a cluster pretty much how its cut, as they are migrating and shifting a huge amount of resources and management policies, and government IT is not exactly high on the list of many well experienced and intelligent IT person's places to work for.
Are you kidding?
Trying to maintain standards and practices across 4,000 gateway points vs 50. Let alone the agency bureaucracy that would be involved in doing site checks and working across various agency boundaries would be a nightmare. It would take eons to get those things in place to do consistent auditing and management to ensure standards and procedures are followed, let alone actually do them. Might as well consolidate bandwidth costs and number of checkpoints down to 50 in the process.
Question- Was it the approved budget or the proposed budget for 2008?
I can understand that an approved document has to be put online, but this is bushes Proposal for 2009, which has yet to be approved IIRC.
Actually more like -$[some amount of money]
Because office lasers and related supplies cost more than a bulk printing center.
They could probably buy every member of congress a Kindle and still save in the end.
Actually.
Why DONT they buy every member of congress a kindle, that way they can get instant EVDO downloads of every bill that is ever submitted to congress, whenever, wherever they are? And search it.
If you want to spend the time to setup and maintain a custom LDAP directory solution to manage mac workstations, you could under linux (or use active directory, you just have to tweak the schema). The big value of OS X Server is that it integrates and manages mac workstations really well. Also, AFP, apple's native network filing protocol, isn't half bad and works best hosted on a mac server.
OS X Server licenses are 1 grand per instance (or $500 for 10 simulatenous file sharing users, no cap on web connections). No processor license, no connected user licenses. From a practical stand point, it only looks for local machines on the same subnet who have the same serial, so you could have multiple installations for testing, as long as any two were not on at the same time.
In a lot of cases I would imagine you have an xserve, or a vmware based boot loader, running OS X Server in one instance to be a lightweight file server, email server (kerio), doing open directory for the mac workstations, and then an instance of Windows 2003 Server hosting that one app you would have dedicated another box to in the past, like a web based time keeping software that requires IIS.
To reply to this, as you are pretty vehement about attacking MacFixit (not that I don't blame you, but they aren't entirely at fault for this), however Apple has announced an official response suggesting to remove APE for the bluescreen issue (http://docs.info.apple.com/article.html?artnum=306857), separate to your above login issue, so APE issue is not FUD, and considering how it works (using the 10.4 debug interface to execute arbitrary code within an application, which Leopards random library assignment is meant to prevent) no surprise it is borked in 10.5
The bigger issue is vendors like Logitech now install it on machines without notifying the user, or providing a way to update it. So you have folks who don't even know they have an old version of APE, which does not disable itself properly. APE is a very very delicate application, and should not be installed without the user knowing it. You can send some of your hate towards Logitech also.
It doesn't help that the APE developers are vehement that their program doesn't cause issues either.
it is not even that. They are following the money. Porn is dealing with the lowest common denominator and lowest barrier to entrance. Instead of a passionate love story, lets just show the 'good' parts. They went to VHS because they saw that it was allowing for lower cost films, faster turn around time and private home viewing. You had a broader audience and you got to cut out the theatre distribution chain. Since they were always a marginal aspect of society (there isn't a XXXMPAA to push laws for them) they have to meet consumer demand, instead of regulate the consumers. And to say that the industry shifted, well as a whole "porn" may changing formats and delivery mechanisms, but most of these changes are still being first to market by new comers to the industry, because each tech revolution lowered the entry point.
VHS made made amatuer porn challenge the Film market.
MiniDV, DVD and the Internet made it even cheaper, so that anyone with some starting money could be a porn star. Now Vivid and other shops have to compete with 'watchmehavesexallday.com' etc, so of course they are going to adapt. They can't lobby or pass laws banning those smaller sites since they aren't exactly popular enough to get senators on their side (however I wonder if some of them backed the age verification requirement, knowing that the smaller sites would crumble under the paperwork).
To say they are broad visioned and open minded is a stretch, I'd just say that they are always on edge and have been fighting to be around, legally and financially, so they have to stay quick and adapt, or fold. A lot of porn distributors went out of business when Video hit, another group went out of business with Traci Lords underage porn too.
Actually, if you are a mac shop, you just buy one machine (intel imac + linux triple boot+ xp oem + parallels = everything you need to test a website for almost any computer) and if you are a pc shop, you buy your web designer a mac to run windows (full time if he wants) and to boot into os x also. I mean they should have one anyway, cs3 on intel macs screams.
Licensing isn't even in the same planet compared to MS.
unlimited license of server is $1000. it COMES with the xserve.
Maintenance (3 years of free upgrades, for 10.5-6, etc.) is another $1000, and entirely worth it.
So initial license purchase on top of the hardware is $1000 if you want 3 years of major versions of os x server. From past experience, that saves you $1000, because 2 more updates will happen in the next 3 years.
You are looking at $4,000 from apple vs $4313 from dell, but the dell only comes with 5 CALs (bare minimum 1u dual dual core xenon servers).
Why couldn't she be the mermaid with the fish part on top and the lady part on the bottom!? / Fry
The thing is the "FCC rumour mill" would have expanded past the normal circle in the case of APPLE releasing the iPhone.
Sure, the specs on the next Sony Ericcson phone is great and all, but thats what they do, make cell phones. The press has been waiting with baited breath for the Apple iPhone for ages. All those press monkeys need is the "FCC rumour mill" to confirm the Apple iPhone, and it will be on the front page of newspapers, because you aren't going to waste time for FCC approval on a device you aren't going to sell. And the FCC approval process includes complete specifications, manual and details of the product also, so you might as well announce it prior to going in for FCC approval.
And Apple has done this before anyway. They did it with the original iBook and Airport. Built in wireless was a huge deal, but they announced it prior to getting FCC approval so they could keep Airport a secret.
Nope, you are wrong. Power Mac was first introduced with the powerpc line of desktop computers.
PowerBook was a term apple was using for their laptops, which did not start with powerpc chips.