If A, then both companies would simply duke it out until one won (namely
dirty incompatibility tricks, having even less incentive to share APIs).
Note that winning=loss of profitability for one company.
Well, you just keep doing it. Sooner or later, management would figure
out that being split repeatedly has costs and is probably not such a good
idea. If they don't figure it out, some competitor will take advantage
of their disarray.
Another approach would simply to have progressively increasing taxes depending
on company size or marketshare.
While I'd love to see MS broken up; I highly doubt that it would solve much. It would only get deregulated, given enough time. MS and Intel are in collaboration (due to mutual advantage). Why would OS / Office companies be any different.
Well, this isn't something you do once and it works forever, it's something you have to do regularly. Also, I would split them "vertically", not "horizontally": multiple competitors that each produce the OS and applications, not a single company for each product.
Remember, we're living in the same as as Enron, with accounting and inter-corporate cohabitation that puts even MS to shame. Any circumvention is possible.
The problem may be hard, but that doesn't mean we shouldn't address it. The alternative is worse.
Lastly, no government is going to break up MS any time soon; at least not until we're out of a ression. MS is a large contributor to our international exports; thereby lessoning the trade-deficit.
Foreign nations aren't going to go for that forever. Either, they are going to push Linux or domestic software heavily, or they will impose high import duties on MS software.
MS could sell the XBox for $49 without breaking a sweat... why aren't
they? They are selling their console at a price slightly higher than the
competition. Surely this shouldn't be illegal!
At issue is why Microsoft isn't charging $400, not why they aren't charging
$49. The Xbox design is too costly--it shouldn't survive in an efficient
market. There is better technology at a lower cost. The main
reason Microsoft wants Xbox is because it pushes their software into yet
other markets. Microsoft is propping up an inefficient, costly design
with subsidies from other divisions in order to drive competitors out of
business. That should very much be illegal.
When Arby's offers five roast beef sandwiches for $5, they are losing money
(i.e. selling below cost) in the hopes that people will spend money on fries
or drinks (where they make tons of money). They also hope to attract business
from their competitors... should this be illegal?
We want to prevent Arby's getting a monopoly (locally or nationally) since
we need competition on order to keep the market efficient.
Our legal system does that by punishing specific behavior. If you take
that approach, if Arby keeps selling below cost for an extended period of
time, yes, that should be illegal.
Actually, I think a better approach is to punish outcome. Let Arby's
do whatever they want, but if they manage to establish a monopoly (e.g., the
only burger joint in town), then break them up regardless of how they got
to that position. Similarly, we should stop all this bickering about
what Microsoft did and didn't do or whether they are a monopoly or almost
a monopoly; they have more than 80% of the market share in a number of sectors,
and that should be reason enough to break them up.
Either way, if we want to live in a free market and capitalist system, we
cannot tolerate the existence of unregulated monopolies--we need to curb
monopolistic tendencies somehow.
Well, I think the numbers in the article support "the bashing party", as
you call it. If you want to convince people otherwise, you need to
connect those numbers with your view of the world. We call that "making an
argument".
Probably because they had more competition in those areas 10-15 years
ago, so they wouldn't be considered a monopoly at that time. Or did you think
that Microsoft had always been the only choice?
Well, I was there 10-15 years ago, and Microsoft was pretty much as dominant
then in their market segment (small business computing, home computing) as
they are now. Apple briefly looked like a threat and another potential
monopolist, but they self-destructed.
The last time I checked, Microsoft was spending a lot of money on advertising
the Xbox. I remember seeing one billion dollars thrown around several times
for the advertising budget. They also bought out Bungie and had to pay them
to port Halo to Xbox, and they had to pay nVidia for R&D on the chipset.
They were going to lose money the first year even they didn't sell below
cost.
But the fact is that, despite all those other expenses, they are selling
below cost. And while it isn't necessarily illegal, many of those other
activities are also undesirable from the point of view of competition: if
the only reason your product succeeds is because you can outspend your competitors
in advertising, then the market isn't efficient anymore. Similarly,
buying independent software developers that otherwise would have rationally
chosen to develop for your competitor's hardware only is also highly questionable.
Hold on for just a second. A can of coke costs about a nickle to make,
can, ship and refrigerate and I just payed 0.75$ for it out of a vending machine.
That's because you mainly pay for physical delivery and distribution.
And it is because Coca Cola does have monopolies in many places (there
is only the Coke machine, nothing else, nearby).
MS has priced their product (successfully, I'm sure) to maximise their
profit - which is NOT the cheapest price they could charge, any more than
the same is true for Coca-Cola. [...] It has nothing to do with being a monopoly.
Monopolies price their products to maximize profit as well. And
it is precisely when the prices charged become significantly higher than
the cost to produce a good that you know that the market may have become
inefficient and that you are dealing with a monopoly.
This is a feature of our modern "capitalist" society; competition only
goes so far in the face of advertising and consumer apathy.
Consumer choice and competition are essential to bring us the benefits of
free markets and capitalism: efficiency and innovation. If you are
saying that advertising and consumer apathy undermine it, then the answer
is not to lean back and say "oh, well, that's the way it is", the answer
is to figure out how to rescue our economy. Breaking up companies is
one way of doing that: if no single company dominates the market, consumers
have to make a choice.
Just because it is the law doesn't make it a good law. Removing the blindfold
from Lady Justice is far too grave a matter to justify a separate standard
for monopolies.
We have chosen as our economic model capitalism and a free market economy.
Those require competition in order to bring about the desired results
of efficiency and innovation. They don't work in the presence of monopolies,
just like they don't work in the presence of central planning, fraud, or
insider trading. That's why we have laws and policies against all of
them.
That one division carries a company is NOT an abuse of their monopoly
position. Keeping Netscape off the desktop with the threat of higher Windows
licensing costs IS. Just because a company makes a profit in 1 area and loses
in another doesn't make it abusive.
No, but it becomes abusive when the company uses profits from one division
to lower the prices of their products from another division in order to drive
out competitors.
The US itself complains about this constantly when it comes to European steel
imports and other goods. Maybe the solution to this problem is for
the Europeans simply to impose hefty import duties on Microsoft's below-cost
exports to the EU.
These numbers do not reflect the cost of MS Research. MSR is costing Microsoft
a hefty sum every year, and they actually do provide many interesting things,
especially for Windows internals.
Well, why don't you work out the numbers and let us know. So far, you
are just making wild assertions with no support.
Go back 10 years. Microsoft's main revenue drivers in 1992 were uh, Windows
3x and Office 4.3. Arguably Windows had pretty good market share but Office
was still losing to Lotus 1-2-3 and Wordperfect. [...] We're not talking monopoly
rents.
How does the fact that Microsoft has been doing this for years change the
observation that they charge monopoly rents?
We're talking about how some parts of your business become cash cows and
support other parts of your business that they believe are worth investing
in and will one day become profitable.
And what exactly is Microsoft "investing" in in those other parts? Their
losses in the Xbox division, for example, don't result from enormous activity
in development or even marketing, they result from selling the hardware below
cost in order to drive competitors out of business. It's an "investment"
if lose money on on R&D, manufacturing facilities, and advertising; for
a monopoly, it's an illegal business practice if you sell products below
cost in order to drive their competitors out of business.
With RFID tags, you should be able to implement a challenge-response protocol that's non-replayable. Not with the $0.10 tags, but with the credit card sized ones. And the cards are much harder to copy even with physical access to them.
It's 2002, and Microsoft has re-invented VNC. Not bad: this time it took them less than a decade to copy someone else's idea; usually, it takes them several decades.
If you're blazing new ground, get all the information you can and pay attention to where you're going.
The X11 designers were "blazing" enough new ground as it is, and they didn't have a cadre of foreign language experts available. Give them a break. Nobody else bothered with this until many years later either.
Where as adding an extension to the X server is meddling with something that could break non-Gnome and non-KDE usage of X
X11 applications already work against servers from many vendors, and against many different versions, often concurrently. You can't seriously think that adding an optional extension is going to break anything.
A sysadmin can download a few 10 MB tarballs, start them compiling, have lunch, install it in/usr/local/luser/linux/gnome and forget about it.
I have had many problems upgrading Gnome/Gtk+, because of changes in shared libraries. I have had no problems at all upgrading through many different versions of X11 servers.
There's no way for the international text functions to concievable handle the native languages of 20% of Earth's population. The APIs alone don't allow it. That's seriously brain-damaged. To blame the toolkit authors for working around the problem is wrong.
The X11 text handling was designed in the early 1980's--even if the designers had wanted to, there simply weren't any good standards around.
And, yes, the toolkit authors should have worked on the window system itself, not inefficient, kludgy library-based workarounds.
You cannot simply tell everyone who wants to run KDE 3 or GNOME 2 to upgrade their system to the latest version of XFree86
They don't have to upgrade their X server (whatever it is), they just have to add a server-side extension as a dynamically loadable module. And that's not to much to ask if people go through all the trouble of upgrading huge, complex software systems like KDE or Gnome.
Which doesn't magically make the XRender extension or the XUtf8* functions appear in every X11 server in the world.
Yes, so what? Windows 3.1 also doesn't support many Windows XP functions.
They work great for pasting glyphs left to right; that is, for English and German and Japanese.
That's all they were intended to do: satisfy the needs of what was probably 99% of the users at the time.
But they have no provision for handling combining characters (needed for Thai), or right to left text (needed for Hebrew), or context-dependent glyphs (needed for Hindi). (Arabic needs all three.)
The new X11 text handling functions do.
Since GTK+ and Qt plan on supporting those languages, they must write their own text displaying functions.
X11 is being developed collaboratively. If there is functionality that's missing from the server, different toolkit authors shouldn't replicate it, they should create an extension. Of course, hell will freeze over before the current crop of toolkit authors can agree on anything.
Windows 2000/XP suck in a lot of ways but at least looking at my desktop
doesn't give me a headache
Well, maybe you just like the stodgy Microsoft Windows style, but Windows
desktop applications are rather inconsistent. Just take a look at the
interface hall of shame.
The "new entries" page alone has numerous different GUI and widget styles for Windows apps (plus several mutually inconsistent Mac apps).
Any redundant or emulated functionality is probably because the X11 functionality
is unportable between X implementations,
There is no other windowing environment that is as well standardized and
portable as X11.
or because it's insufficent.
Insufficient for what? X11 is a full-featured windowing environment.
It has everything you need to put bits, graphics, and text on the screen
and interact with it, efficiently and easily. Of course, you can write
fast, high-end toolkits for it. They may not look exactly like what
a Windows programmer or user might expect, but that's a different issue.
Indeed, Gtk+ was originally developed for X11 (I didn't claim otherwise, if
you read carefully). However, Gtk+ ignores many of the conventions for X11
toolkits and fails to take full advantage of the server-side facilities of
X11. Whether that's deliberate or merely means that the original designers
weren't that familiar with X11, I don't know.
The point remains: Gtk+, like Qt, and others, is not the best way to get
very small and efficient GUI applications, although it has other advantages.
Current thinking in computer science is leaning towards the idea that
assumptions should be enforced at the system level rather than at the developer
level.
But neither Windows nor MacOS enforce assumptions at the system level. The
reason why toolkits look alike on those platforms is because toolkit developers
choose to make them look alike.
According to the X mailing lists, this is the real reason people think
X is slow. Applications just aren't written for high-performance drawing.
It's not an issue with applications, it's an issue with toolkits: Gtk+, Qt,
and Swing are just not designed for maximum efficiency under X11. Since
they are more than fast enough on current machines, that doesn't bother anybody
other than people who just hate X11 for no good reason.
Applications just aren't written for high-performance drawing.
Even in your optimistic scenario, PicoGUI only speeds up the drawing of widgets,
not other application drawing. But toolkits draw the widgets. Therefore,
if you want to speed up the drawing of widgets under X11, just use a more
efficient X11 toolkit.
By putting the widget set in the server (and ideally, the entire presentation
layer) you can have one well-optimized set of code, so applications can be
implemented in a straightforward manner and still have high performance.
You can have "one well-optimized set of code" right now, by using a more
efficient widget set under X11. In fact, X11's windows are almost complete
widgets in themselves already--X11 isn't all that different from PicoGUI.
And people will not stop writing Gtk+, Qt, or wxWindows code just because
PicoGUI comes around, so those toolkits will do even worse on top of PicoGUI
than on top of X11.
Consistency, because all clients use whatever widget set the server does. By making the server modular, you can use FLTK for handhelds and GTK+ for desktops, and all apps will obey that decision.
You make the assumption that putting the widget set into the server ensures consistency, or that not doing it means GUIs become inconsistent. Experience with real-world window systems suggest otherwise.
Mac OS X ships with multiple widget sets that are consistent, and people use many more to develop on it. Windows, too, ships with multiple widget sets, and there are many additional GUI toolkits in use for Windows. Yet, most people don't even notice. That's the way GUIs work.
Performance, because by interfacing clients and the server at a high level, you reduce communication between the two by a huge amount.
I have seen no practical indication that client/server communication is a bottleneck for X11. Why "optimize" something that doesn't need optimizing? Furthermore, X11 already takes care of many widget-related issues, like geometry management, bit-blitting of retained pixels, and event dispatch. Basically, in X11, if you open a subwindow and draw some text on it, you already have a widget, and you can eliminate almost all client/server communications through backing store.
or your desktop looks like an example of applied chaos theory
Gtk+, Qt, Motif, Fltk+, and Tk applications are nearly indistiguishable visually and behaviorally. And the situation is the same on other major desktops as well: your average Windows and Macintosh desktop probably has applications written in half a dozen different toolkits.
If pico is reasonably modular and well designed, you could have different versions (different in complexity, hardware abstraction, your own ultimate widget collection, etc) and still use *every* app *native* with the different flavours of picoGUI
That's the wrong kind of generality as far as I'm concerned. I like programming to different widget sets because the APIs of different widget sets are good for different things. What you suggest is just what I don't want: a single, uniform API with different appearances.
My only problem with XWindows is that it isn't as fast as it could be
And how fast "could it be"? X11 seems to be faster than both GDI and Cocoa/Quartz on comparable hardware in my experience.
(any tweaking hints are welcome)
Most supposed performance problems with X11 seem to be due to the toolkits and desktop environment used. Gnome or KDE on X11 can be a bit sluggish on low-end machines Tweaking hint: use a different desktop environment.
Of course, you can also enable backing store by default in your X11 server, which is how a lot of other window systems achieve their slick look; backing store is disabled in XFree86 by default because well-behaved X11 clients shouldn't rely on it.
Cross-platform GUI wrappers impose enormous overhead because they don't take full advantage of the features available in X11. Toolkits like Gtk+, Qt, or wxWindows duplicate a lot of the functionality already provided by the X11 server, and they try to emulate behavior that would be more naturally provided differently under X11. What makes things worse is that those cross-platform toolkits are usually optimized for Windows and ported to X11 as an afterthought. The enormous size and performance problems of systems like Gnome, KDE, and Java's Swing are due to this.
If you can live with that overhead, fine, use cross-platform toolkits. I use wxWindows myself and think it's a pretty nice toolkit. But you have no basis on which to complain about X11 performance or functionality if you hamstring it by using cross-platform GUI wrappers with it constantly.
The major difference between PicoGUI and systems like X11 and Quartz seems to be that PicoGUI defines widgets and layout in the display server.
I find that a dubious design decision. I like the fact that I can have many different kinds of widgets and graphics under traditional window systems. For example, for handhelds, I can use something like FLTK, which is very light and does not have much in the way of geometry management, while on desktops, I can use something heavy-weight like Gtk+ or wxX11. Also, there is no widespread agreement of how widget layout should be handled in GUIs.
If you want a widget server like PicoGUI provides, you could easily add that between client applications and an X11 (or Quartz) server.
Eventually, X11 will get replaced by something, but among the current contenders (Windows GDI, Quartz, PicoGUI, Berlin, etc.), I don't see anything that has compelling advantages.
Slashdot Mirror
If A, then both companies would simply duke it out until one won (namely dirty incompatibility tricks, having even less incentive to share APIs). Note that winning=loss of profitability for one company.
Well, you just keep doing it. Sooner or later, management would figure out that being split repeatedly has costs and is probably not such a good idea. If they don't figure it out, some competitor will take advantage of their disarray.
Another approach would simply to have progressively increasing taxes depending on company size or marketshare.
Well, this isn't something you do once and it works forever, it's something you have to do regularly. Also, I would split them "vertically", not "horizontally": multiple competitors that each produce the OS and applications, not a single company for each product.
Remember, we're living in the same as as Enron, with accounting and inter-corporate cohabitation that puts even MS to shame. Any circumvention is possible.
The problem may be hard, but that doesn't mean we shouldn't address it. The alternative is worse.
Lastly, no government is going to break up MS any time soon; at least not until we're out of a ression. MS is a large contributor to our international exports; thereby lessoning the trade-deficit.
Foreign nations aren't going to go for that forever. Either, they are going to push Linux or domestic software heavily, or they will impose high import duties on MS software.
MS could sell the XBox for $49 without breaking a sweat... why aren't they? They are selling their console at a price slightly higher than the competition. Surely this shouldn't be illegal!
At issue is why Microsoft isn't charging $400, not why they aren't charging $49. The Xbox design is too costly--it shouldn't survive in an efficient market. There is better technology at a lower cost. The main reason Microsoft wants Xbox is because it pushes their software into yet other markets. Microsoft is propping up an inefficient, costly design with subsidies from other divisions in order to drive competitors out of business. That should very much be illegal.
When Arby's offers five roast beef sandwiches for $5, they are losing money (i.e. selling below cost) in the hopes that people will spend money on fries or drinks (where they make tons of money). They also hope to attract business from their competitors... should this be illegal?
We want to prevent Arby's getting a monopoly (locally or nationally) since we need competition on order to keep the market efficient.
Our legal system does that by punishing specific behavior. If you take that approach, if Arby keeps selling below cost for an extended period of time, yes, that should be illegal.
Actually, I think a better approach is to punish outcome. Let Arby's do whatever they want, but if they manage to establish a monopoly (e.g., the only burger joint in town), then break them up regardless of how they got to that position. Similarly, we should stop all this bickering about what Microsoft did and didn't do or whether they are a monopoly or almost a monopoly; they have more than 80% of the market share in a number of sectors, and that should be reason enough to break them up.
Either way, if we want to live in a free market and capitalist system, we cannot tolerate the existence of unregulated monopolies--we need to curb monopolistic tendencies somehow.
Well, and breaking up big monopolies is a good first step towards addressing that problem and restoring a free market.
Well, I think the numbers in the article support "the bashing party", as you call it. If you want to convince people otherwise, you need to connect those numbers with your view of the world. We call that "making an argument".
Well, I was there 10-15 years ago, and Microsoft was pretty much as dominant then in their market segment (small business computing, home computing) as they are now. Apple briefly looked like a threat and another potential monopolist, but they self-destructed.
The last time I checked, Microsoft was spending a lot of money on advertising the Xbox. I remember seeing one billion dollars thrown around several times for the advertising budget. They also bought out Bungie and had to pay them to port Halo to Xbox, and they had to pay nVidia for R&D on the chipset. They were going to lose money the first year even they didn't sell below cost.But the fact is that, despite all those other expenses, they are selling below cost. And while it isn't necessarily illegal, many of those other activities are also undesirable from the point of view of competition: if the only reason your product succeeds is because you can outspend your competitors in advertising, then the market isn't efficient anymore. Similarly, buying independent software developers that otherwise would have rationally chosen to develop for your competitor's hardware only is also highly questionable.
Hold on for just a second. A can of coke costs about a nickle to make, can, ship and refrigerate and I just payed 0.75$ for it out of a vending machine.
That's because you mainly pay for physical delivery and distribution. And it is because Coca Cola does have monopolies in many places (there is only the Coke machine, nothing else, nearby).
MS has priced their product (successfully, I'm sure) to maximise their profit - which is NOT the cheapest price they could charge, any more than the same is true for Coca-Cola. [...] It has nothing to do with being a monopoly.
Monopolies price their products to maximize profit as well. And it is precisely when the prices charged become significantly higher than the cost to produce a good that you know that the market may have become inefficient and that you are dealing with a monopoly.
This is a feature of our modern "capitalist" society; competition only goes so far in the face of advertising and consumer apathy.
Consumer choice and competition are essential to bring us the benefits of free markets and capitalism: efficiency and innovation. If you are saying that advertising and consumer apathy undermine it, then the answer is not to lean back and say "oh, well, that's the way it is", the answer is to figure out how to rescue our economy. Breaking up companies is one way of doing that: if no single company dominates the market, consumers have to make a choice.
Just because it is the law doesn't make it a good law. Removing the blindfold from Lady Justice is far too grave a matter to justify a separate standard for monopolies.
We have chosen as our economic model capitalism and a free market economy. Those require competition in order to bring about the desired results of efficiency and innovation. They don't work in the presence of monopolies, just like they don't work in the presence of central planning, fraud, or insider trading. That's why we have laws and policies against all of them.
That one division carries a company is NOT an abuse of their monopoly position. Keeping Netscape off the desktop with the threat of higher Windows licensing costs IS. Just because a company makes a profit in 1 area and loses in another doesn't make it abusive.
No, but it becomes abusive when the company uses profits from one division to lower the prices of their products from another division in order to drive out competitors.
The US itself complains about this constantly when it comes to European steel imports and other goods. Maybe the solution to this problem is for the Europeans simply to impose hefty import duties on Microsoft's below-cost exports to the EU.
These numbers do not reflect the cost of MS Research. MSR is costing Microsoft a hefty sum every year, and they actually do provide many interesting things, especially for Windows internals.
Well, why don't you work out the numbers and let us know. So far, you are just making wild assertions with no support.
Go back 10 years. Microsoft's main revenue drivers in 1992 were uh, Windows 3x and Office 4.3. Arguably Windows had pretty good market share but Office was still losing to Lotus 1-2-3 and Wordperfect. [...] We're not talking monopoly rents.
How does the fact that Microsoft has been doing this for years change the observation that they charge monopoly rents?
We're talking about how some parts of your business become cash cows and support other parts of your business that they believe are worth investing in and will one day become profitable.
And what exactly is Microsoft "investing" in in those other parts? Their losses in the Xbox division, for example, don't result from enormous activity in development or even marketing, they result from selling the hardware below cost in order to drive competitors out of business. It's an "investment" if lose money on on R&D, manufacturing facilities, and advertising; for a monopoly, it's an illegal business practice if you sell products below cost in order to drive their competitors out of business.
With RFID tags, you should be able to implement a challenge-response protocol that's non-replayable. Not with the $0.10 tags, but with the credit card sized ones. And the cards are much harder to copy even with physical access to them.
It's 2002, and Microsoft has re-invented VNC. Not bad: this time it took them less than a decade to copy someone else's idea; usually, it takes them several decades.
The X11 designers were "blazing" enough new ground as it is, and they didn't have a cadre of foreign language experts available. Give them a break. Nobody else bothered with this until many years later either.
Where as adding an extension to the X server is meddling with something that could break non-Gnome and non-KDE usage of X
X11 applications already work against servers from many vendors, and against many different versions, often concurrently. You can't seriously think that adding an optional extension is going to break anything.
A sysadmin can download a few 10 MB tarballs, start them compiling, have lunch, install it in /usr/local/luser/linux/gnome and forget about it.
I have had many problems upgrading Gnome/Gtk+, because of changes in shared libraries. I have had no problems at all upgrading through many different versions of X11 servers.
The X11 text handling was designed in the early 1980's--even if the designers had wanted to, there simply weren't any good standards around.
And, yes, the toolkit authors should have worked on the window system itself, not inefficient, kludgy library-based workarounds.
You cannot simply tell everyone who wants to run KDE 3 or GNOME 2 to upgrade their system to the latest version of XFree86
They don't have to upgrade their X server (whatever it is), they just have to add a server-side extension as a dynamically loadable module. And that's not to much to ask if people go through all the trouble of upgrading huge, complex software systems like KDE or Gnome.
Yes, so what? Windows 3.1 also doesn't support many Windows XP functions.
They work great for pasting glyphs left to right; that is, for English and German and Japanese.
That's all they were intended to do: satisfy the needs of what was probably 99% of the users at the time.
But they have no provision for handling combining characters (needed for Thai), or right to left text (needed for Hebrew), or context-dependent glyphs (needed for Hindi). (Arabic needs all three.)
The new X11 text handling functions do.
Since GTK+ and Qt plan on supporting those languages, they must write their own text displaying functions.
X11 is being developed collaboratively. If there is functionality that's missing from the server, different toolkit authors shouldn't replicate it, they should create an extension. Of course, hell will freeze over before the current crop of toolkit authors can agree on anything.
Well, maybe you just like the stodgy Microsoft Windows style, but Windows desktop applications are rather inconsistent. Just take a look at the interface hall of shame. The "new entries" page alone has numerous different GUI and widget styles for Windows apps (plus several mutually inconsistent Mac apps).
(and my 2d performance is more than a bit better)
Doubtful.Any redundant or emulated functionality is probably because the X11 functionality is unportable between X implementations,
There is no other windowing environment that is as well standardized and portable as X11.
or because it's insufficent.
Insufficient for what? X11 is a full-featured windowing environment. It has everything you need to put bits, graphics, and text on the screen and interact with it, efficiently and easily. Of course, you can write fast, high-end toolkits for it. They may not look exactly like what a Windows programmer or user might expect, but that's a different issue.
Indeed, Gtk+ was originally developed for X11 (I didn't claim otherwise, if you read carefully). However, Gtk+ ignores many of the conventions for X11 toolkits and fails to take full advantage of the server-side facilities of X11. Whether that's deliberate or merely means that the original designers weren't that familiar with X11, I don't know.
The point remains: Gtk+, like Qt, and others, is not the best way to get very small and efficient GUI applications, although it has other advantages.
Current thinking in computer science is leaning towards the idea that assumptions should be enforced at the system level rather than at the developer level.
But neither Windows nor MacOS enforce assumptions at the system level. The reason why toolkits look alike on those platforms is because toolkit developers choose to make them look alike.
According to the X mailing lists, this is the real reason people think X is slow. Applications just aren't written for high-performance drawing.
It's not an issue with applications, it's an issue with toolkits: Gtk+, Qt, and Swing are just not designed for maximum efficiency under X11. Since they are more than fast enough on current machines, that doesn't bother anybody other than people who just hate X11 for no good reason.
Applications just aren't written for high-performance drawing.
Even in your optimistic scenario, PicoGUI only speeds up the drawing of widgets, not other application drawing. But toolkits draw the widgets. Therefore, if you want to speed up the drawing of widgets under X11, just use a more efficient X11 toolkit.
By putting the widget set in the server (and ideally, the entire presentation layer) you can have one well-optimized set of code, so applications can be implemented in a straightforward manner and still have high performance.
You can have "one well-optimized set of code" right now, by using a more efficient widget set under X11. In fact, X11's windows are almost complete widgets in themselves already--X11 isn't all that different from PicoGUI.
And people will not stop writing Gtk+, Qt, or wxWindows code just because PicoGUI comes around, so those toolkits will do even worse on top of PicoGUI than on top of X11.
You make the assumption that putting the widget set into the server ensures consistency, or that not doing it means GUIs become inconsistent. Experience with real-world window systems suggest otherwise.
Mac OS X ships with multiple widget sets that are consistent, and people use many more to develop on it. Windows, too, ships with multiple widget sets, and there are many additional GUI toolkits in use for Windows. Yet, most people don't even notice. That's the way GUIs work.
Performance, because by interfacing clients and the server at a high level, you reduce communication between the two by a huge amount.
I have seen no practical indication that client/server communication is a bottleneck for X11. Why "optimize" something that doesn't need optimizing? Furthermore, X11 already takes care of many widget-related issues, like geometry management, bit-blitting of retained pixels, and event dispatch. Basically, in X11, if you open a subwindow and draw some text on it, you already have a widget, and you can eliminate almost all client/server communications through backing store.
Gtk+, Qt, Motif, Fltk+, and Tk applications are nearly indistiguishable visually and behaviorally. And the situation is the same on other major desktops as well: your average Windows and Macintosh desktop probably has applications written in half a dozen different toolkits.
If pico is reasonably modular and well designed, you could have different versions (different in complexity, hardware abstraction, your own ultimate widget collection, etc) and still use *every* app *native* with the different flavours of picoGUI
That's the wrong kind of generality as far as I'm concerned. I like programming to different widget sets because the APIs of different widget sets are good for different things. What you suggest is just what I don't want: a single, uniform API with different appearances.
And how fast "could it be"? X11 seems to be faster than both GDI and Cocoa/Quartz on comparable hardware in my experience.
(any tweaking hints are welcome)
Most supposed performance problems with X11 seem to be due to the toolkits and desktop environment used. Gnome or KDE on X11 can be a bit sluggish on low-end machines Tweaking hint: use a different desktop environment.
Of course, you can also enable backing store by default in your X11 server, which is how a lot of other window systems achieve their slick look; backing store is disabled in XFree86 by default because well-behaved X11 clients shouldn't rely on it.
If you can live with that overhead, fine, use cross-platform toolkits. I use wxWindows myself and think it's a pretty nice toolkit. But you have no basis on which to complain about X11 performance or functionality if you hamstring it by using cross-platform GUI wrappers with it constantly.
I find that a dubious design decision. I like the fact that I can have many different kinds of widgets and graphics under traditional window systems. For example, for handhelds, I can use something like FLTK, which is very light and does not have much in the way of geometry management, while on desktops, I can use something heavy-weight like Gtk+ or wxX11. Also, there is no widespread agreement of how widget layout should be handled in GUIs.
If you want a widget server like PicoGUI provides, you could easily add that between client applications and an X11 (or Quartz) server.
Eventually, X11 will get replaced by something, but among the current contenders (Windows GDI, Quartz, PicoGUI, Berlin, etc.), I don't see anything that has compelling advantages.