I don't think that's quite accurate. The OSS leaders and primary contributors tend to be employed, like Linus. They can't take "contracts" to do customization and lose their jobs, especially when it's often changes that would only take a few months or a year to do. The solution is simple: hire someone else to do the work.
Perhaps the big gripe is that once you invest in developing a customized branch, the realization dawns that now you have to maintain it, the same as in-house apps. Danged OSS hippies and commies won't fix it for free!;)
If half of your software purchase, deployment, and maintenance costs are consumed by the software license, then you pay your IT staff an insultingly low wage or simply don't bother with little things like disaster planning, growth projections, tuning, or anything else that makes a database efficient.
Any organization that is spending half their IT expense on licensing is a joke. The only way that claim makes any sense is if someone is rolling in service and support and claiming it's part of the capital expense instead of an ongoing run-rate service charge. Fire your accountant.
the only difference is that we're talking about something physical instead of software.
I disagree completely.
The robot is a complete assembly designed for a purpose.
Most software "patents" are on the reuse of existing techniques, algorithms, data structures, and approaches. The only thing "special" is that they identify a specific data attribute or business use.
Can you patent the use of a wrench to tighten bolts on a wheel? No. You can only patent the overall design of the wrench.
"Business process" patents are even worth. WTF -- do Americans think they're the first people in a few millenia to come up with these ideas? Half of them are taught to MBAs.
Thumbs down on the world's most ineffective and useless patent database.
Why is it lately when I see a story about IBM suing a company, it feels like good news?
Because IBM spends years trying to negotiate a deal before resorting to lawyers. They defend their patents.
Other scum sucking leeches buy up technology, kill the products, and just go around suing for infringement. They use patents to attack and rob legitimate tech producers.
How can you not cheer to see at least one company using patents the way they were meant to be used?
I wouldn't paint the entire telco community with that brush.
In regions like Canada where there are crown, private, and co-op telcos competing via copper, fiber, and cable, the billing issues for bandwidth are much clearer and cleaner than with a purely market driven mess.
Whoever builds the "last mile" network has the right to install servers and accelerators within their subnet to give their customer service. Internet traffic is only ONE aspect of what the telcos use the data backbones for.
For example, SaskTel MAX relies on data center movie and TV servers to deliver media to the home with full bitrates. You can't expect to move that kind of data across the internet backbones to do direct-to-home delivery outside the ISP/telco without someone paying for massive increases in the backbone capacity and the telco's backbone connection bandwidth.
End-user bandwidth hogs just cannot grasp the fact that they don't pay for their bandwidth. The aggregate user community of the ISP is covering the expense of their usage. Were bandwidth charges split based on usage, many high speed customers would be very annoyed at what their real access expense is.
Personally I have never understood why people try to hack utilities like XPath into database equivalents. But I guess if you learn how to use a hammer and screwdriver, you might well refuse to learn how to handle a wrench because you can "make" things work with the wrong tools.
Or perhaps it's a negative side-effect of object reuse fanatics who don't consider that reusability does not necessarily mean eliminating alternate implementations, but enforcing an appropriately designed and narrowed interface instead of specific implementations.
Imagine a rich reuse library that included not just default implementations, but specialized implementations that took advantage of particular tools and technologies to tweak performance, scalability, security, and reliability. ie. Instead of a generic JDBC connection with DbIO objects and methods, you could have the generic interface with specific implementations for Sybase ASE, DB/2 UDB, Oracle, PostgreSQL, SQLServer, etc. Some might not even use JDBC behind the interface. Some might rely on stored procs, others might rely on product-specific features like PostgreSQL table inheritance.
In any case, I'm reasonably certain such generic interface implementations would be much less vulnerable to injections, because even if the code were sloppy, the attacker would have to know which implementation they're attacking, not just which technology.
Personally I ALWAYS validate and quote data appropriately. A couple decades of database, systems, applications, and distributed software development teaches painful lessons through late night and weekend debugs.
A lot of people who are given the label "self-appointed dictator" in this realm are really just people leading by doing, but leading in a direction different than those doing the labeling would prefer.
AKA the whining masses always bitch that no one will do the work for them, for free.
OSS is about picking up your own coding skills to add in functionality you need. OSS is about hiring or recruiting a team to branch an entire project. OSS is about being able to archive source and binaries so you can access historical data years in the future.
OSS is NOT about giving lazy whiners free software, custom written for their personal wants or needs.
Cattle manure does produce significant methane. So do pig and sheep manure. Pretty much any ruminant does.
There are many examples over the past decade or two of experimental and prototype methane collection facilities using manure as a source. If I recall correctly, the issue is that keeping the slurry at the right temperature during the methane extraction limits the regions where it can be done. Doing it here in a Saskatchewan winter would burn more fuel to warm the slurry tanks than you'd get back, but farther south it ends up net-positive.
North America, Europe, the UK, and other regions actually demand that you take fiscal responsibility for delivering services. If you can't accept that, you don't run a business here, you're just waiting to be
stepped on by someone's lawyers.
Open source is a perfectly viable option for applications with mitigated risk. It is also technically viable for business. But it's your insurance company and legal department that determine if it's a viable business move -- and they like numbers. Hard numbers. Not geek speak banner waving pride.
This whole concept got me thinking about UI history overall. Let's take
a romp through history of some of the UI advancements over the decades:
Function Labels Old-school "green screen" standards such as IBM's user
interface guidelines included the use of label displays for
function keys (where supported by hardware), standardization
of keystroke actions such as "ALT-F4" closing a window, and
recommendations for font highlighting to indicate
mandatory/optional data, read/write access, primary keys, etc.
"ALT-F4" works to this day in virtually every GUI there is.
Text Menus Part of the fundamental UI models going back as far as
function keys with screen labels, if not farther.
Spreadsheet Interface Dan Briklin's Visicalc. 'nuff said. The man never did get
reasonable financial rewards for what he did.
Edit Regions Text boxes have been around as long as
green screens, as well as field validation. They're just fancier
now.
Drop-down Menus Not quite as old as the green screen, these were a display-saving
alternative to screens listing menu options. Power users would
just enter a dot-suffix navigation of menu options: 1.3.5.8 might
fire up an "Add Customer" screen, for example.
Pop-up Menus I think these started with X-11, maybe even Xerox PARC. Certainly
it was a key feature of Motif and OpenLook, which preceded MS Windows
substantially. They were also present in the Amiga UI, several years
before even Windows 3.1 was released.
Audio Feedback ANSI7 defines CTRL-G as bell. Some form of ping, alert, sound effect,
or other attention-getting audio signal has been around since the
teletypewriter. WAVs and MP3s are just fancier ways of doing the same
thing that applications have done since the Commodore PET and Apple II.
Images and Icons How far back does the BMP go? Higher resolution, compressed,
even primitive animations via GIF go back much farther than any GUI.
Once upon a time, only an image viewer displayed an image, not the UI.
Drawers Drawers of icons have been around at least since Motif.
Toolbars Tear-aside and pinnable menus have been around since at least OpenLook.
Whether icons are displayed beside, above, below, under, or to the right
of a text label, the metaphor is far from new.
Wizards I laughed myself silly when someone years ago presented
the "Wizard" as a "new" way of doing things. Ever enter a timesheet
on an old mainframe form application? GECOS email (I think that's what
it was called)? Wizards are just old fashioned step-by-step forms
prettified.
Bubble Help Green screens would display a help line to the bottom or top of
the screen. Dialogue-box help showed up with the green screen as
well. Even vi and emacs had help systems, though they weren't
triggered by the now-common F1. Pretty laughable that anyone thought
the particular shape of the dialogue box displaying the text was
important, isn't it?
Mouse Gestures The idea was around for a long time. I think I even saw prototypes
of pie menus for the Amiga or the Mac, but I'm not sure. Pie selection
is closely related to gestures -- select via stroke direction instead
of precise mouse placement. Interesting, but not comfortable for
everyone.
3D User Interface SGI. 'nuff said.
Personally, I can't imagine paying royalties to use the idea of a Motif
icon/menu drawer opening sideways. It's kind of obvious.
The specific characters set selected for DNS names is a carefully selected subset that overlaps ALL character sets created by that point in time, including packed 6-bit schemes and such used by barcodes.
In order to service the widest range of computer systems and devices possible, the character set had to be pruned to include only characters that ALL those systems understood.
If it were intended to be "readable", even in English, there would have been provisions for decorative characters, case sensitivity, spacing, and the full ASCII7/LATIN1 character set.
People outside the networking infrastructures also don't realize that domain names and IP addresses comparisons and switching are implemented in hardware by the backbone hardware. This is not just a software change, any more than IPv6 is.
A viable business deals with it's business needs first. Technology is an expense to meet those needs, including risk mitigation and protection from lawsuits over failures to meet contractual obligations.
Get your head out of the sand. Business runs IT; IT does not run business.
The idea that someone has to take such a drug in order to perform their duties says a lot about the state of society, employee relations, and the treatment of contractors like robotic systems.
if you are building a program which directly competes with Word, Excel, PowerPoint, Outlook, or Access (the Microsoft applications with the new UI), you can't obtain the royalty-free license.'
So if you are doing word processing, document editing, email, calendars, diagramming, data storage/database, reporting, presentations, or anything else useful for end-users, there is no royalty-free option.
If you are doing a Mickey Mouse IM, media player, or something else that can't generate revenue due to widespread competition, feel free to implement a UI that is incompatible with any platform other than Windows. (See above.)
Read what is said, people, not what you want to hear.
Another implementation is called a "Search Engine".
It's how the vast majority of people find things. Go ahead and find the Saskatchewan Abilities Counsel website without a search engine, for example. What would that URL be?
The push for 16-bit DNS has nothing to do with whether people from other nations can view a URL because the subset of characters valid for domain names are glyphs in every multibyte character set, using codes compatible with ASCII/LATIN1.
The domain names and IP addressing issues are infrastructure technology, and government wind-emitters and spin-doctors don't belong in the meetings or plans to address those issues. They're not qualified.
Do you seriously think any CIO with a functioning brain cell is going to go with free unsupported software when they can't even find a single reference to such databases from any certified performance evaluation companies or organizations?
The downtime cost of one single failure in a five year period for a mission critical system can easily run 100 times the cost of a commercial product with support. Only bean counting fools risk their entire business without properly assessed risks and disaster recovery plans.
Not having someone to source the recovery of the smouldering crater that was your data center is a huge issue.
No kidding. The whole concept of tunneling IP packets over HTTP is thoroughly asinine. It's like mandating you load your motorcycle on a flatbed trailer in order to move it across town.
Plus now that we've got all the traffic tunneled through HTTP/HTTPS, we now have a virtual network protocol on top of a virtual network protocol (IP) on top of the actual packet and wire protocols (TCP.) Slower, and just as vulnerable to interference and bad content as IP was.
But wait!
Now, for free, you can also reform your data packets as big, fat, wordy XML documents riding on top of the HTTP stack.
With XML/RPC or SOAP/XML on top.
Layered to provide a language API transparently to the programmer.
Voila!!! It is done!!!
We've completed a full iteration of layering protocol on protocol to get right back to the exact same original functionality of shipping a packet between client and server. But now we waste WAY more processing than any mainframe OS ever could.
Microsoft is about selling a product, not providing customer satisfaction.
If Microsoft were about selling product, then there would not be the perpetual push for cascading upgrades of the entire Microsoft line. The product sold, they'd be content. But noooooo, they want you to pay repeatedly for the same "product".
But neither their product upgrades nor their service packs and updates fix the issues on the defect lists for the previous release. Instead we get fed some pablum that a survey group found a new way of screwing things up better than the old way of screwing things up, so we get a UI that lets you screw up all kinds of ways instead of fixing fundamental bugs.
In exchange, you get to buy more memory, a faster CPU, and upgrade a whole boatload of other software that now proves incompatible with the MS suite upgrades you installed.
*Heavy Exasperated Sigh*
Inevitably you run into it. The critical package whose vendor is null, void, and gone. No upgrades. No patches. No enhancements.
And it won't run with the latest "updates".
Look at it this way. A company that has been around for a couple decades using Microsoft Word has now paid for Word 1..n, 95, 98, 2000, 2Kn, and XP. $300-500 for the first go-round, and $150-300 for at least five upgrades.
Is your word processor worth $1050-2000?!?!?!?!
Oh, yes, but wait! There are the OS Upgrades as well. Win 3.1, WFW, Win95, Win98, WinNT, Win2K, WinXP. Let's just ballpark those at $150 each. Another $1050.
Is your glorified typewriter worth $2100-3050?!?!?!
Now imagine an office that has loyally purchased the entire suite of Microsoft servers and tools.
I fail to see any significant saving over the "bad old days" of centralized computing, other than the drop in hardware costs. And there is still someone trying to tell me what I can and cannot do with the machine, only now it's not the admin who works with me -- it's some third party in Washington that doesn't even know who I am except as a license validation string.
Instead of changing the fundamental DNS which is a programmer's and administrator's tool, not an advertising medium. It is founded, like programming languages, on a fundamental 7-bit ASCII character set, and is not intended to be used for NLS text.
A far better solution is some form of VDNS that translates NLS text names into the proper domain name at the system level. That also allows the same domain to have multiple language translations to reflect localized product and service names.
We seriously need to kick the general political community in the arse. They keep trying to impose technical decisions, and it fails as miserably as any corporate PHB's uninformed decisions. ASK the techies to propose solutions instead of shoving ill-conceived ideas down our throats.
For example -- once you mandate multibyte domains, you implicitly mandate multibyte URL components. Goodbye direct mapping of names to the directories, file systems, and servers.
This is the most common mistake people make. If you need to do it today, you will need to do it again in the future.
The business owner who asked for the one-off will want changes. Or it will turn out to be a good idea that upper management wants to make part of the monthly, weekly, or daily reports.
The "one-off" application is an excuse to kick something out the door without design, documentation, or maintainability concerns. I would hazard a guess that fully a third of production systems are comprised of "one-off" code that was deemed too important to let go.
No matter how much of a hurry you're in, remember that someone (hopefully you) will have to step into that steaming "one-off" pile you're creating today...
Yep. Microsoft's commitment to their "Plays for Sure" campaign with the Zune really instills confidence in their backwards compatability.
At least with OpenOffice I can legally archive the source code and install images needed to access the data for that period (say, every year or six months.) Sort of like dropping a copy of TrueCrypt on a DVD full of crypto archives.
With the new DRM keys and license enforcement policies, I dread someday trying to resurrect an old image so I can access data archives, only to find it wants to register with a DRM verification service that no longer runs or is no longer compatible with a 4-5 year old install image.
It would be the responsibility of the OpenGL rendering implementation to take advantage of a fused CPU/GPU core, much as the drivers took advantage of SSE implementations. The application should see no difference under OpenGL or DirectX unless it's using vendor-specific extensions.
What I find more interesting is the possibility of applying "graphics" operations to general data types, not just the presumed integral axis references. Imagine feeding currency into a bump-map transform to texture a global sales map, or temperature to texture a thermal relief map. Additional axis could be conveyed as color, and key references like "average" or "mean" could display as "sea level".
Date projections act the same as Z-buffer clipping/projections, except for the data type. Hardware accelerated time-based history data could define a "current" plane feeding the above visual displays, allowing you to literally scroll through years of data.
Algorithms used for motion detection in video processing, temporal cleaners, static removers, etc. could be very interesting for detecting and correcting data anomolies.
Data is data, regardless of whether the end result is pixels.
Slashdot Mirror
I don't think that's quite accurate. The OSS leaders and primary contributors tend to be employed, like Linus. They can't take "contracts" to do customization and lose their jobs, especially when it's often changes that would only take a few months or a year to do. The solution is simple: hire someone else to do the work.
Perhaps the big gripe is that once you invest in developing a customized branch, the realization dawns that now you have to maintain it, the same as in-house apps. Danged OSS hippies and commies won't fix it for free! ;)
If half of your software purchase, deployment, and maintenance costs are consumed by the software license, then you pay your IT staff an insultingly low wage or simply don't bother with little things like disaster planning, growth projections, tuning, or anything else that makes a database efficient.
Any organization that is spending half their IT expense on licensing is a joke. The only way that claim makes any sense is if someone is rolling in service and support and claiming it's part of the capital expense instead of an ongoing run-rate service charge. Fire your accountant.
I disagree completely.
The robot is a complete assembly designed for a purpose.
Most software "patents" are on the reuse of existing techniques, algorithms, data structures, and approaches. The only thing "special" is that they identify a specific data attribute or business use.
Can you patent the use of a wrench to tighten bolts on a wheel? No. You can only patent the overall design of the wrench.
"Business process" patents are even worth. WTF -- do Americans think they're the first people in a few millenia to come up with these ideas? Half of them are taught to MBAs.
Thumbs down on the world's most ineffective and useless patent database.
*snerk*
Good flamebait. Irrelevant, but amusing.
Because IBM spends years trying to negotiate a deal before resorting to lawyers. They defend their patents.
Other scum sucking leeches buy up technology, kill the products, and just go around suing for infringement. They use patents to attack and rob legitimate tech producers.
How can you not cheer to see at least one company using patents the way they were meant to be used?
I wouldn't paint the entire telco community with that brush.
In regions like Canada where there are crown, private, and co-op telcos competing via copper, fiber, and cable, the billing issues for bandwidth are much clearer and cleaner than with a purely market driven mess.
Whoever builds the "last mile" network has the right to install servers and accelerators within their subnet to give their customer service. Internet traffic is only ONE aspect of what the telcos use the data backbones for.
For example, SaskTel MAX relies on data center movie and TV servers to deliver media to the home with full bitrates. You can't expect to move that kind of data across the internet backbones to do direct-to-home delivery outside the ISP/telco without someone paying for massive increases in the backbone capacity and the telco's backbone connection bandwidth.
End-user bandwidth hogs just cannot grasp the fact that they don't pay for their bandwidth. The aggregate user community of the ISP is covering the expense of their usage. Were bandwidth charges split based on usage, many high speed customers would be very annoyed at what their real access expense is.
Personally I have never understood why people try to hack utilities like XPath into database equivalents. But I guess if you learn how to use a hammer and screwdriver, you might well refuse to learn how to handle a wrench because you can "make" things work with the wrong tools.
Or perhaps it's a negative side-effect of object reuse fanatics who don't consider that reusability does not necessarily mean eliminating alternate implementations, but enforcing an appropriately designed and narrowed interface instead of specific implementations.
Imagine a rich reuse library that included not just default implementations, but specialized implementations that took advantage of particular tools and technologies to tweak performance, scalability, security, and reliability. ie. Instead of a generic JDBC connection with DbIO objects and methods, you could have the generic interface with specific implementations for Sybase ASE, DB/2 UDB, Oracle, PostgreSQL, SQLServer, etc. Some might not even use JDBC behind the interface. Some might rely on stored procs, others might rely on product-specific features like PostgreSQL table inheritance.
In any case, I'm reasonably certain such generic interface implementations would be much less vulnerable to injections, because even if the code were sloppy, the attacker would have to know which implementation they're attacking, not just which technology.
Personally I ALWAYS validate and quote data appropriately. A couple decades of database, systems, applications, and distributed software development teaches painful lessons through late night and weekend debugs.
AKA the whining masses always bitch that no one will do the work for them, for free.
OSS is about picking up your own coding skills to add in functionality you need. OSS is about hiring or recruiting a team to branch an entire project. OSS is about being able to archive source and binaries so you can access historical data years in the future.
OSS is NOT about giving lazy whiners free software, custom written for their personal wants or needs.
Cattle manure does produce significant methane. So do pig and sheep manure. Pretty much any ruminant does.
There are many examples over the past decade or two of experimental and prototype methane collection facilities using manure as a source. If I recall correctly, the issue is that keeping the slurry at the right temperature during the methane extraction limits the regions where it can be done. Doing it here in a Saskatchewan winter would burn more fuel to warm the slurry tanks than you'd get back, but farther south it ends up net-positive.
North America, Europe, the UK, and other regions actually demand that you take fiscal responsibility for delivering services. If you can't accept that, you don't run a business here, you're just waiting to be stepped on by someone's lawyers.
Open source is a perfectly viable option for applications with mitigated risk. It is also technically viable for business. But it's your insurance company and legal department that determine if it's a viable business move -- and they like numbers. Hard numbers. Not geek speak banner waving pride.
This whole concept got me thinking about UI history overall. Let's take a romp through history of some of the UI advancements over the decades:
Function Labels
Old-school "green screen" standards such as IBM's user interface guidelines included the use of label displays for function keys (where supported by hardware), standardization of keystroke actions such as "ALT-F4" closing a window, and recommendations for font highlighting to indicate mandatory/optional data, read/write access, primary keys, etc. "ALT-F4" works to this day in virtually every GUI there is.
Text Menus
Part of the fundamental UI models going back as far as function keys with screen labels, if not farther.
Spreadsheet Interface
Dan Briklin's Visicalc. 'nuff said. The man never did get reasonable financial rewards for what he did.
Edit Regions
Text boxes have been around as long as green screens, as well as field validation. They're just fancier now.
Drop-down Menus
Not quite as old as the green screen, these were a display-saving alternative to screens listing menu options. Power users would just enter a dot-suffix navigation of menu options: 1.3.5.8 might fire up an "Add Customer" screen, for example.
Pop-up Menus
I think these started with X-11, maybe even Xerox PARC. Certainly it was a key feature of Motif and OpenLook, which preceded MS Windows substantially. They were also present in the Amiga UI, several years before even Windows 3.1 was released.
Audio Feedback
ANSI7 defines CTRL-G as bell. Some form of ping, alert, sound effect, or other attention-getting audio signal has been around since the teletypewriter. WAVs and MP3s are just fancier ways of doing the same thing that applications have done since the Commodore PET and Apple II.
Images and Icons
How far back does the BMP go? Higher resolution, compressed, even primitive animations via GIF go back much farther than any GUI. Once upon a time, only an image viewer displayed an image, not the UI.
Drawers
Drawers of icons have been around at least since Motif.
Toolbars
Tear-aside and pinnable menus have been around since at least OpenLook. Whether icons are displayed beside, above, below, under, or to the right of a text label, the metaphor is far from new.
Wizards
I laughed myself silly when someone years ago presented the "Wizard" as a "new" way of doing things. Ever enter a timesheet on an old mainframe form application? GECOS email (I think that's what it was called)? Wizards are just old fashioned step-by-step forms prettified.
Bubble Help
Green screens would display a help line to the bottom or top of the screen. Dialogue-box help showed up with the green screen as well. Even vi and emacs had help systems, though they weren't triggered by the now-common F1. Pretty laughable that anyone thought the particular shape of the dialogue box displaying the text was important, isn't it?
Mouse Gestures
The idea was around for a long time. I think I even saw prototypes of pie menus for the Amiga or the Mac, but I'm not sure. Pie selection is closely related to gestures -- select via stroke direction instead of precise mouse placement. Interesting, but not comfortable for everyone.
3D User Interface
SGI. 'nuff said.
Personally, I can't imagine paying royalties to use the idea of a Motif icon/menu drawer opening sideways. It's kind of obvious.
Incorrect.
The specific characters set selected for DNS names is a carefully selected subset that overlaps ALL character sets created by that point in time, including packed 6-bit schemes and such used by barcodes.
In order to service the widest range of computer systems and devices possible, the character set had to be pruned to include only characters that ALL those systems understood.
If it were intended to be "readable", even in English, there would have been provisions for decorative characters, case sensitivity, spacing, and the full ASCII7/LATIN1 character set.
People outside the networking infrastructures also don't realize that domain names and IP addresses comparisons and switching are implemented in hardware by the backbone hardware. This is not just a software change, any more than IPv6 is.
Databases run on computer systems.
Businesses run on legal and financial systems.
A viable business deals with it's business needs first. Technology is an expense to meet those needs, including risk mitigation and protection from lawsuits over failures to meet contractual obligations.
Get your head out of the sand. Business runs IT; IT does not run business.
The idea that someone has to take such a drug in order to perform their duties says a lot about the state of society, employee relations, and the treatment of contractors like robotic systems.
And none of what it says is "good".
So if you are doing word processing, document editing, email, calendars, diagramming, data storage/database, reporting, presentations, or anything else useful for end-users, there is no royalty-free option.
If you are doing a Mickey Mouse IM, media player, or something else that can't generate revenue due to widespread competition, feel free to implement a UI that is incompatible with any platform other than Windows. (See above.)
Read what is said, people, not what you want to hear.
Another implementation is called a "Search Engine".
It's how the vast majority of people find things. Go ahead and find the Saskatchewan Abilities Counsel website without a search engine, for example. What would that URL be?
The push for 16-bit DNS has nothing to do with whether people from other nations can view a URL because the subset of characters valid for domain names are glyphs in every multibyte character set, using codes compatible with ASCII/LATIN1.
The domain names and IP addressing issues are infrastructure technology, and government wind-emitters and spin-doctors don't belong in the meetings or plans to address those issues. They're not qualified.
Do you seriously think any CIO with a functioning brain cell is going to go with free unsupported software when they can't even find a single reference to such databases from any certified performance evaluation companies or organizations?
The downtime cost of one single failure in a five year period for a mission critical system can easily run 100 times the cost of a commercial product with support. Only bean counting fools risk their entire business without properly assessed risks and disaster recovery plans.
Not having someone to source the recovery of the smouldering crater that was your data center is a huge issue.
No kidding. The whole concept of tunneling IP packets over HTTP is thoroughly asinine. It's like mandating you load your motorcycle on a flatbed trailer in order to move it across town.
Plus now that we've got all the traffic tunneled through HTTP/HTTPS, we now have a virtual network protocol on top of a virtual network protocol (IP) on top of the actual packet and wire protocols (TCP.) Slower, and just as vulnerable to interference and bad content as IP was.
But wait!
Now, for free, you can also reform your data packets as big, fat, wordy XML documents riding on top of the HTTP stack.
With XML/RPC or SOAP/XML on top.
Layered to provide a language API transparently to the programmer.
Voila!!! It is done!!!
We've completed a full iteration of layering protocol on protocol to get right back to the exact same original functionality of shipping a packet between client and server. But now we waste WAY more processing than any mainframe OS ever could.
If Microsoft were about selling product, then there would not be the perpetual push for cascading upgrades of the entire Microsoft line. The product sold, they'd be content. But noooooo, they want you to pay repeatedly for the same "product".
But neither their product upgrades nor their service packs and updates fix the issues on the defect lists for the previous release. Instead we get fed some pablum that a survey group found a new way of screwing things up better than the old way of screwing things up, so we get a UI that lets you screw up all kinds of ways instead of fixing fundamental bugs.
In exchange, you get to buy more memory, a faster CPU, and upgrade a whole boatload of other software that now proves incompatible with the MS suite upgrades you installed.
*Heavy Exasperated Sigh*
Inevitably you run into it. The critical package whose vendor is null, void, and gone. No upgrades. No patches. No enhancements.
And it won't run with the latest "updates".
Look at it this way. A company that has been around for a couple decades using Microsoft Word has now paid for Word 1..n, 95, 98, 2000, 2Kn, and XP. $300-500 for the first go-round, and $150-300 for at least five upgrades.
Is your word processor worth $1050-2000?!?!?!?!
Oh, yes, but wait! There are the OS Upgrades as well. Win 3.1, WFW, Win95, Win98, WinNT, Win2K, WinXP. Let's just ballpark those at $150 each. Another $1050.
Is your glorified typewriter worth $2100-3050?!?!?!
Now imagine an office that has loyally purchased the entire suite of Microsoft servers and tools.
I fail to see any significant saving over the "bad old days" of centralized computing, other than the drop in hardware costs. And there is still someone trying to tell me what I can and cannot do with the machine, only now it's not the admin who works with me -- it's some third party in Washington that doesn't even know who I am except as a license validation string.
Instead of changing the fundamental DNS which is a programmer's and administrator's tool, not an advertising medium. It is founded, like programming languages, on a fundamental 7-bit ASCII character set, and is not intended to be used for NLS text.
A far better solution is some form of VDNS that translates NLS text names into the proper domain name at the system level. That also allows the same domain to have multiple language translations to reflect localized product and service names.
We seriously need to kick the general political community in the arse. They keep trying to impose technical decisions, and it fails as miserably as any corporate PHB's uninformed decisions. ASK the techies to propose solutions instead of shoving ill-conceived ideas down our throats.
For example -- once you mandate multibyte domains, you implicitly mandate multibyte URL components. Goodbye direct mapping of names to the directories, file systems, and servers.
Bad idea. Very bad idea.
The agreement says nothing about real issues of patent violation or the validity of such potential lawsuits.
It just says it's a lot cheaper to just swap money now and cut the lawyers out of the loop.
Distributed IIOP services via IDL, RMI, or SOAP/XML.
The browser is reduced to a layout engine for active widgets or applets, or some variant on XUL or XForms.
"Web" is a misconception derived from the assumption that IP traffic must be HTTP/HTTPS based. There are another 32000+ ports aside from 80 and 88...
This is the most common mistake people make. If you need to do it today, you will need to do it again in the future.
The business owner who asked for the one-off will want changes. Or it will turn out to be a good idea that upper management wants to make part of the monthly, weekly, or daily reports.
The "one-off" application is an excuse to kick something out the door without design, documentation, or maintainability concerns. I would hazard a guess that fully a third of production systems are comprised of "one-off" code that was deemed too important to let go.
No matter how much of a hurry you're in, remember that someone (hopefully you) will have to step into that steaming "one-off" pile you're creating today...
Yep. Microsoft's commitment to their "Plays for Sure" campaign with the Zune really instills confidence in their backwards compatability.
At least with OpenOffice I can legally archive the source code and install images needed to access the data for that period (say, every year or six months.) Sort of like dropping a copy of TrueCrypt on a DVD full of crypto archives.
With the new DRM keys and license enforcement policies, I dread someday trying to resurrect an old image so I can access data archives, only to find it wants to register with a DRM verification service that no longer runs or is no longer compatible with a 4-5 year old install image.
It would be the responsibility of the OpenGL rendering implementation to take advantage of a fused CPU/GPU core, much as the drivers took advantage of SSE implementations. The application should see no difference under OpenGL or DirectX unless it's using vendor-specific extensions.
What I find more interesting is the possibility of applying "graphics" operations to general data types, not just the presumed integral axis references. Imagine feeding currency into a bump-map transform to texture a global sales map, or temperature to texture a thermal relief map. Additional axis could be conveyed as color, and key references like "average" or "mean" could display as "sea level".
Date projections act the same as Z-buffer clipping/projections, except for the data type. Hardware accelerated time-based history data could define a "current" plane feeding the above visual displays, allowing you to literally scroll through years of data.
Algorithms used for motion detection in video processing, temporal cleaners, static removers, etc. could be very interesting for detecting and correcting data anomolies.
Data is data, regardless of whether the end result is pixels.