You raise an interesting point. Arms races of the past have always (mostly) been about the destructive power of a warhead and the total number of warheads. It would be interesting to see an arms race who's point of escalation was merely speed. We don't need to wipe everybody out any more - contemporary scenarios are more about the speed of a tightly scoped, tactical response than they are about broadstroke devastation. Most countries seem to realize we all benefit more by working together than we do constantly jockeying for King of the Hill. Granted, there are a few pockets holding out from this way of thinking and I guess maybe that's why we need efficient "scalpels."
Ubiquity seems like a good thing in many areas of IT and I concede that Flash is orders of magnitude closer today than HTML5. However, instead of introducing (or continuing to support/invest in) new layers and proprietary standards (further complicating the stack, costing resources and making it damn impossible to secure) to address shortcomings that exist, we'd all benefit from embracing new open standards that attempt to address the issue.
Things you write in Flash do not work on all browsers. They only work on browsers that have the Flash plug-in.
Let's not gloss over that: HTML5 may support a subset of Flash today, but it could eventually encompass all of it (or, gasp- exceed Flash functionality) and will do so in all HTML5 browsers without relying on a proprietary plug-in and closed eco-system of authoring tools. I think many people prefer this approach because Adobe is neglecting their platform and also because existing authoring tools from the vendor don't provide the functionality needed at the price desired.
I emphasize the above obvious point because your post seemed to gloss over the whole point of the HTML5 vs. Flash debate.
Unfortunately it's not reasonable to expect all of the people who use IT to be able to logically evaluate a strategic choice from among the myriad IT paths available. So many follow the crowd betting that by doing so they'll minimize their chance of making a strategic error. I don't think this is to their discredit, strategic decisions are costly to make if you have to plod through all the variables on your own even when you have the expertise to apply logic.
The phenomenon can lead to an ugly place when deep pocketed (or near-monopolistic) vendors capture enough market share to get the masses on a bad path and ensuing generations of decision makers continue to follow. It is extremely difficult to make a course correction if nobody notices for 20 years- the cost of switching tracks is often prohibitive.
Delivering an implementation of a standard specification is distinctly different than "imitating.". It's one of the reason such specifications are created- to facilitate other implementations. There are a few examples of this sort of thing in IT. You might want to take a look around.
I agree. While java has made it possible to get the "same" (used in it's loosest definition possible) code to run on multiple platforms, in doing so it has turned an impossible task into a difficult one. The reality is that it can creates more work in many contexts. As examples: the equiv. of DLL hell in trying to get external dependencies coordinated (java libraries or 3rd party components) and also the accounting for sometimes nuanced sometimes not issue of talking to hardware behind the OS or other OS specific interfaces. If it were impossible you wouldn't have to do any work, now that it's possible there's an expectation that it'll happen.
Yeah- Lenovo S10e net books come like that from the manufacturer; allowing a user to either boot into a SplashTop Linux desktop or WinXP. Guess which one boots faster? Go ahead, guess.
Still- not quite the same thing as it does leave the unselected OS on the hard drive.
Here's a small dose of enlightenment: the case for servers.
Servers need OSes too and they usually need to be able to handle more simultaneous thread.s
Careful what you wish for. I for one would like my OS to make data integrity a higher priority then the user interface. Sure, I like my mouse pointer to move smoothly across the screen and for a window to close when I click on the close window control, but not if it means my bits are spooling to/dev/null.
I think programs have generally grown bloated because
it's easier to code to high level libraries than to rewrite a stack,
once everyone's addicted to high levels libraries they want more of it and the libraries grow in functional scope,
in some environments, as library interfaces evolve, multiple versions of the same library are installed side-by-side in an environment rather than doing the harder work of preserving backward compatibility or updating the calling code
If mature software architectures (such as POSIX) have taught us one thing it's that libraries (aka APIs, aka Toolboxes, aka <insert widget nomenclature du jour here>) are best kept small with well (ie. tightly) scoped functionality that adhere to some IPC standard and can be mixed & matched (vice more monolithic or Kitchen Sinkish approaches).
I agree that deleting your account may not be the best course of action if protecting your personal data is the ultimate goal. However, if MySpace does see a spike in deletes as a result of this announcement they may figure out that selling out may cost them equity they don't want to lose.
We'd make a lot more progress if everyone could support given standards rather than specific browser versions. I'm not saying we should cast disdain on those that make pragmatic decisions based on real business needs- just idealizing. Seems like we've made progress toward that ideal in recent years.
Fantastic. Can anybody think of a more effective strategy for shedding market share? Given that most of this particular segment has already formed an opinion of Windows, I'm pretty excited about their remaining options.
Will the next Mac Vs. PC commercials show the cool Mac guy caged or manacled? Can't wait.
Wow- my post got mangled. Let's try this again:
Why not switch to XPath for personal GUIDs? For example/<country>/<city|<region>/[/<address>]<name>, etc...and when it's time, prefixing/<star>/<planet>/... is easy enough.
As far as the Spam issue many have brought up, I don't think security through obscurity is ever going to be adequate protection so worrying about the public registry of IDs seems like a waste of time- the evil marketers will either derive your ID via brute force or buy the ID from somebody else. Using finer grained sender authentication at the user level (combined with cryptographic certificates for sender identity assurance) would seem to me to be the appropriate solution.
Slashdot Mirror
You raise an interesting point. Arms races of the past have always (mostly) been about the destructive power of a warhead and the total number of warheads. It would be interesting to see an arms race who's point of escalation was merely speed. We don't need to wipe everybody out any more - contemporary scenarios are more about the speed of a tightly scoped, tactical response than they are about broadstroke devastation. Most countries seem to realize we all benefit more by working together than we do constantly jockeying for King of the Hill. Granted, there are a few pockets holding out from this way of thinking and I guess maybe that's why we need efficient "scalpels."
Yes, but the sand is on the beach. Moving the sand to the fab facility and filtering out all the seagull poop can't be cheap.
You just watched Charlie & the Chocolate factory, didn't you?
Ubiquity seems like a good thing in many areas of IT and I concede that Flash is orders of magnitude closer today than HTML5. However, instead of introducing (or continuing to support/invest in) new layers and proprietary standards (further complicating the stack, costing resources and making it damn impossible to secure) to address shortcomings that exist, we'd all benefit from embracing new open standards that attempt to address the issue.
Things you write in Flash do not work on all browsers. They only work on browsers that have the Flash plug-in.
Let's not gloss over that: HTML5 may support a subset of Flash today, but it could eventually encompass all of it (or, gasp- exceed Flash functionality) and will do so in all HTML5 browsers without relying on a proprietary plug-in and closed eco-system of authoring tools. I think many people prefer this approach because Adobe is neglecting their platform and also because existing authoring tools from the vendor don't provide the functionality needed at the price desired.
I emphasize the above obvious point because your post seemed to gloss over the whole point of the HTML5 vs. Flash debate.
Your list is incomplete.
Unfortunately it's not reasonable to expect all of the people who use IT to be able to logically evaluate a strategic choice from among the myriad IT paths available. So many follow the crowd betting that by doing so they'll minimize their chance of making a strategic error. I don't think this is to their discredit, strategic decisions are costly to make if you have to plod through all the variables on your own even when you have the expertise to apply logic.
The phenomenon can lead to an ugly place when deep pocketed (or near-monopolistic) vendors capture enough market share to get the masses on a bad path and ensuing generations of decision makers continue to follow. It is extremely difficult to make a course correction if nobody notices for 20 years- the cost of switching tracks is often prohibitive.
Delivering an implementation of a standard specification is distinctly different than "imitating.". It's one of the reason such specifications are created- to facilitate other implementations. There are a few examples of this sort of thing in IT. You might want to take a look around.
I agree. While java has made it possible to get the "same" (used in it's loosest definition possible) code to run on multiple platforms, in doing so it has turned an impossible task into a difficult one. The reality is that it can creates more work in many contexts. As examples: the equiv. of DLL hell in trying to get external dependencies coordinated (java libraries or 3rd party components) and also the accounting for sometimes nuanced sometimes not issue of talking to hardware behind the OS or other OS specific interfaces. If it were impossible you wouldn't have to do any work, now that it's possible there's an expectation that it'll happen.
Why can't people use words the way they want to use them instead of the way you'd prefer they use them?
Insert MSlam/Reactor safety quip
Not all file (or network) operations are able to be slowed down. That's why we have hardware interrupts.
In other news, "Easier Classes Raise GPAs"
Yeah- Lenovo S10e net books come like that from the manufacturer; allowing a user to either boot into a SplashTop Linux desktop or WinXP. Guess which one boots faster? Go ahead, guess.
Still- not quite the same thing as it does leave the unselected OS on the hard drive.
Here's a small dose of enlightenment: the case for servers.
Servers need OSes too and they usually need to be able to handle more simultaneous thread.s
Careful what you wish for. I for one would like my OS to make data integrity a higher priority then the user interface. Sure, I like my mouse pointer to move smoothly across the screen and for a window to close when I click on the close window control, but not if it means my bits are spooling to /dev/null.
Now if there could just be a "Pick your OS" pull-down on first boot ...
If mature software architectures (such as POSIX) have taught us one thing it's that libraries (aka APIs, aka Toolboxes, aka <insert widget nomenclature du jour here>) are best kept small with well (ie. tightly) scoped functionality that adhere to some IPC standard and can be mixed & matched (vice more monolithic or Kitchen Sinkish approaches).
I'll get right on that as soon as you define "juicy" in a machine-readable format.
Boss, is that you?
I agree that deleting your account may not be the best course of action if protecting your personal data is the ultimate goal. However, if MySpace does see a spike in deletes as a result of this announcement they may figure out that selling out may cost them equity they don't want to lose.
We'd make a lot more progress if everyone could support given standards rather than specific browser versions. I'm not saying we should cast disdain on those that make pragmatic decisions based on real business needs- just idealizing. Seems like we've made progress toward that ideal in recent years.
Fantastic. Can anybody think of a more effective strategy for shedding market share? Given that most of this particular segment has already formed an opinion of Windows, I'm pretty excited about their remaining options.
Will the next Mac Vs. PC commercials show the cool Mac guy caged or manacled? Can't wait.
Sure it can- Wine. I've had surprisingly good luck running Windows apps natively on Linux (ie. not in a virtual machine or emulator).
Wow- my post got mangled. Let's try this again: Why not switch to XPath for personal GUIDs? For example /<country>/<city|<region>/[/<address>]<name>, etc...and when it's time, prefixing /<star>/<planet>/... is easy enough.
As far as the Spam issue many have brought up, I don't think security through obscurity is ever going to be adequate protection so worrying about the public registry of IDs seems like a waste of time- the evil marketers will either derive your ID via brute force or buy the ID from somebody else. Using finer grained sender authentication at the user level (combined with cryptographic certificates for sender identity assurance) would seem to me to be the appropriate solution.