Maybe just people who didn't bother to fill in the "purpose"-field when they rented some CPU time? I somehow doubt classified stuff would appear in these charts.
Yup and thanks to cryo-tech decisions will still be made by the same media-f*kheads that are in charge today so it will be 1mbit/s up, 100mbit/s down for us.
Even if I find a review using Google, the chances are that it is a bad review written by a single person who bought the product, and was stupid enough to think people care about such "reviews" and wrote one.
Why do you put "single persons" down like that?
When I'm in the process of making a buying decision I'm looking for these "single-person"-reviews in particular. I found most of these review published on private homepages or blogs much more helpful than most of the stuff I find on "professional" review-sites or - worse - these "we-pay-2cent-for-your-review" crap-stores.
Sure, often you find only one useful review out of ten. But most of the time this one useful review is written by some unbiased "single person" who just bought the thing you're looking for and took the time to publish a brief, unbiased report of the pro's and con's from a personal point of view.
Especially when looking for hardware with linux support these personal reviews have always been the premier source of actual information for me, while most "professional" articles were all about stuff that didn't matter a bit to me (like how the GFX chip performs unreal tournament and such).
The main problem is that all operating systems look and feel sufficiently different to make that damned difficult.
You have a good point. I think the main problem is that too many of the dirty details are still visible to the developer.
What is needed (imho) would be a markup language like HTML that allows the dev to describe what he needs independently of how it will be displayed.
I know there have been attempts in that direction but they apparently were too bloated and cumbersome to take off.
A simple XML DTD would probably suffice to describe most kinds of UI-forms that we encounter. IMHO *simple* is the keyword. Lazyness is what drives most coders, thus this "general" toolkit would need to be easier than programming to a particular API.
Maybe by coming up with a common GUI-API (optimally one for all relevant programming languages...)?
As I understand it, the task of popping up little windows, laying out widgets in a canvas (using a layout manager or not) is pretty well understood nowadays.
Shouldn't it be possible for someone smarter than me (ok, that would have to be a group of ppl;-) ) to take a real close look at GTK, QT, WXwindows and whatever else is out there and come up with a sane abstraction API?
This ain't a trivial task but once an open standard has been established I see no reason why we couldn't create wrappers for all relevant toolkits and finally take two significant burdens off them OSS developers shoulders:
1. Decide for a toolkit 2. Deal with the individual hassles of the toolkit
Instead we could just all program to that common API and it'd be up to the user to decide whether he prefers QT tabs over gnome tabs.
Sure, hard problems like i8n would need to be tackled, but hey, that's stuff we have to solve anyways.
You can open the lid of a car... to see how it was built and maybe improve things.
Well, when I tear open the lid of any recently built car then I see the basic stuff every car has and... SMD microchips.
I dare to claim that it does indeed take some ressources to access, reverse engineer or even modify them to any worthwhile extend.
There's a "car modding"-industry now, offering all kinds of assistance with the task but AFAIK the more interesting parts of a cars "business logic" are generally deliberately locked down.
When I can't access a HD, I will unplug the cable, blow out the areas, and plug it back in. It's one of the first things to come to mind.
Sounds like a strange approach to me. I'm sysadmin for a living (over 2yrs now) and I think I have never removed the cable, "blown" it and plugged it back in when I encounter a broken drive.
I usually only apply some pressure to the plug to make sure it hasn't loosened over time.
I don't think dust can collect inside the plug-holes to an amount where it would interfere with operation while it's plugged in. Unless you're in a *very* dusty enviroment...
So, when I made sure that it's properly plugged in and it still doesnt work then I usually connect a fresh drive to the same cable to determine whether the drive or the controller is broken. I have never had an IDE cable fail in normal operation (you can ofcourse break them by deliberate bending but that doesn't happen so often inside a closed PC case).
Well, my point is, I agree with GP that these articial test scenarios are just braindead and prove nothing except what we all know: TV shows like that are made by stupid people for an even more stupid audience.
Ah, the "catch" of anti-trust legislation. Being a monopoly isnt illegal. Doing some paticular business practices isnt illegal. But being a monopoly and engaging in some paticlar business practices is.
Well, I don't pretend to understand anti-monopoly laws or anti-trust legislation indepth. But don't think I've ever heard of a case where an existing monopoly was actually broken by application of these laws.
Microsoft Windows is not an OS for games: it is an OS for Microsoft Games... It is not an OS for DBs: it is an OS for MSSQL.
All this whining... See, I hate microsoft like the next guy. But if you were in the position of MS, wouldn't you use your monopoly just the same way?
What they do is just consequently, cynically applied business logic. It's what they teach in business school and what's supposed to drive our system.
I think therein lies the true problem which is much more complex than a monopoly in the PC industry.
IMHO the only real reason why MS has stayed so long and is still doing good is because no one has offered a package that is equally "good enough", yet.
If you develop (a DB, an office package, whatever) for the MS platform you have to deal with the MS competition. If MS offers the same product as you, you lose. It's that easy - they created the playground, they make the rules.
But there might be hope. Their recent announcement makes me *hope* that apple might be heading for the mass market. And it happens that their product is actually technically better than what MS has to offer. So, let's see what darwin (sorry, pun) has to say on this matter. I for one wouldn't be too surprised if apple managed to bite off a huge chunk from MS's market share in the next decade. Don't underestimate the "cool"-factor. If apple-prices really come down to beige-box level we'll likely see some serious competition in the Joe-Sixpack-segment.
The problem with MS having the OS and the apps under one umbrella corperation is that the apps developers get inside info about the OS.... Yes you would have two virtual monopolies, (Office and OS) but at least it wouldn't be one monopoly that can squeeze out other companies.
Every time I read such arguments I wonder how anyone could ever enforce this artificial split?
There a countless ways to "smuggle" simple API details back and forth that are unknown to the competition. If the new broken up companies really wanted to suppress their competition that way, a split like this certainly wouldn't hinder them much.
Well, actually you can get quite far with polling. As you said, latency will be hell (e.g. 2 sec), but the UNIX CLI deals surprisingly well with high latency links.
Back in the day it was quite common to do stuff over 300bps links with high latency (probably rarely up to 2s but in the ballpark).
So, first of all you'll decouple Input and Output. Instead of sending a char, waiting until the server echos it and printing it you will print the typed chars immediately and probably send the whole line on each poll (these few bytes are dwarfed by http protocol overhead anyways) so the server doesn't even need to "assemble" the line but just gets the whole thing everytime.
This will basically give you "local" feel on the CLI and latency only kicks in when you actually need the server response (e.g. cycling through the history or firing the command).
For interactive things (think: editor) it gets a bit more difficult since you actually depend on the editor telling you where to put the cursor etc. Even for that stuff there are plenty ways to optimize cuz, as said, people have been tackling the "slow-link" problem very exhaustingly a long time ago.
Well, why anyone would want to spend all that effort when they can just use SSH is beyond me, tho...;-)
This is leading way offtopic but I still want to mention it:
ActiveX is the one of the worst designed pieces of software that MS ever created. I'm sure it *still* is responsible for a significant percentage of malware infections despite not being enabled by default anymore (AFAIK, might be mistaken).
BUT...
The problems that are often times "solved" with ActiveX (/Java/Flash/Javascript...) are real and I cannot grasp why they haven't been addressed by the W3C years ago. It's a shame we still have to resort to plugins or scripting for form controls (dragndrop, treeviews, cascaded dropdown menus etc) that have been the lowest common detominator of just about any UI-kit for years.
Many parts of the web have evolved in the past years (CSS, javascript, XHTML). The interactive elements available to web designers, those that actually enable you to DO something on the web, have not.
We're still stuck with the same seven basic form elements that we started with: file, input box, textarea, select box, button, checkbox, radiobox. period.
Designing an accessible and *useful* webpage would be much easier (read: HALF THE MANHOURS OR LESS!!) if we had at least the following additional form controls at our disposal:
comboselect - dropdown with optional text-input dragndrop - upload file or multiple files via dragndrop tree-menus - people script them with javascript all the time......and many more that I miss right now, you get the idea...
Can anyone explain to me why these obviously useful and both productivity- and accessiblity- (by standarization) boosting widgets have not been added to the standard, yet?
Instead of just adding a straightforward "input type=tree" to the source we still have to mess with insane javascript hacks or brain-damaging monstrosities like XUL.
Sorry this rant got longer than I intended but you somehow got me started on what bugs me every single f*ing day on my job. One day this stuff will be added and all the work many of us have spent to emulate these things for years will suddenly seem emberassingly obsolete.
Agree'd. Managing appointments in outlook, creating documents in word/excel and fancy colored "presentations" in powerpoint is what most windows-people consider "productive work".
For that stuff windows is fine, the occassional graphics job (photoshop) works, too. Even though for the latter you're better off with a mac already.
Anything that involves actually getting "stuff done" is better done on a unix OS. Don't get me started on what happens when "windows people" decide to implement business processes in scripted Excel sheets, access mdbs, lotus whatever-its-called-mess, VB or similar etch-a-sketch environments.
I pity the the poor schmocks who have to clean up the inevitable mess after a few months and "port" the pile of crap that stacked up in the meantime to something that actually works.
I used to be one of these schmocks until I learned to become very unpolite (read: say "no" very clear and insisting) early enough in the process.
I also wondered who the hell decided that the kids should be sitting in a dark room all day long. Money speaks I suppose. Who cares whether it hurts their eyesight and probably even their health (sunlight is good, remember?).
Microsoft turn out Windows and people keep on buying it in droves.
Lack of alternatives at one point in time lead to market domination. Until recently there was no serious competition in mainstream computing ("Joe Sixpack").
Slashdot Mirror
No, the real reason why they could be hacked was because the attacker was using a Mac.
Maybe just people who didn't bother to fill in the "purpose"-field when they rented some CPU time? I somehow doubt classified stuff would appear in these charts.
Yup and thanks to cryo-tech decisions will still be made by the same media-f*kheads that are in charge today so it will be 1mbit/s up, 100mbit/s down for us.
10usec?
Man, gimme a baseball bat and I'll route them packets myself!
Even if I find a review using Google, the chances are that it is a bad review written by a single person who bought the product, and was stupid enough to think people care about such "reviews" and wrote one.
Why do you put "single persons" down like that?
When I'm in the process of making a buying decision I'm looking for these "single-person"-reviews in particular. I found most of these review published on private homepages or blogs much more helpful than most of the stuff I find on "professional" review-sites or - worse - these "we-pay-2cent-for-your-review" crap-stores.
Sure, often you find only one useful review out of ten. But most of the time this one useful review is written by some unbiased "single person" who just bought the thing you're looking for and took the time to publish a brief, unbiased report of the pro's and con's from a personal point of view.
Especially when looking for hardware with linux support these personal reviews have always been the premier source of actual information for me, while most "professional" articles were all about stuff that didn't matter a bit to me (like how the GFX chip performs unreal tournament and such).
The main problem is that all operating systems look and feel sufficiently different to make that damned difficult.
You have a good point. I think the main problem is that too many of the dirty details are still visible to the developer.
What is needed (imho) would be a markup language like HTML that allows the dev to describe what he needs independently of how it will be displayed.
I know there have been attempts in that direction but they apparently were too bloated and cumbersome to take off.
A simple XML DTD would probably suffice to describe most kinds of UI-forms that we encounter. IMHO *simple* is the keyword. Lazyness is what drives most coders, thus this "general" toolkit would need to be easier than programming to a particular API.
Well, just my 2cent...
Maybe by coming up with a common GUI-API (optimally one for all relevant programming languages...)?
;-) ) to take a real close look at GTK, QT, WXwindows and whatever else is out there and come up with a sane abstraction API?
As I understand it, the task of popping up little windows, laying out widgets in a canvas (using a layout manager or not) is pretty well understood nowadays.
Shouldn't it be possible for someone smarter than me (ok, that would have to be a group of ppl
This ain't a trivial task but once an open standard has been established I see no reason why we couldn't create wrappers for all relevant toolkits and finally take two significant burdens off them OSS developers shoulders:
1. Decide for a toolkit
2. Deal with the individual hassles of the toolkit
Instead we could just all program to that common API and it'd be up to the user to decide whether he prefers QT tabs over gnome tabs.
Sure, hard problems like i8n would need to be tackled, but hey, that's stuff we have to solve anyways.
You can open the lid of a car ... to see how it was built and maybe improve things.
Well, when I tear open the lid of any recently built car then I see the basic stuff every car has and... SMD microchips.
I dare to claim that it does indeed take some ressources to access, reverse engineer or even modify them to any worthwhile extend.
There's a "car modding"-industry now, offering all kinds of assistance with the task but AFAIK the more interesting parts of a cars "business logic" are generally deliberately locked down.
And next?
"Oh, sorry the buffer overflow. Bounds checking is such a pain..."
Lazyness is a pretty lame excuse for not doing it right!
Mod parent up (Insightful, not "funny").
I choked at that very phrase, too.
With that attitude the kid is up for one really very cold shower.
Welcome to the real world!
How about... phoenix?
"Oh, no, thanks. Actually I don't have a problem. But could you put me back on hold?"
When I can't access a HD, I will unplug the cable, blow out the areas, and plug it back in. It's one of the first things to come to mind.
Sounds like a strange approach to me.
I'm sysadmin for a living (over 2yrs now) and I think I have never removed the cable, "blown" it and plugged it back in when I encounter a broken drive.
I usually only apply some pressure to the plug to make sure it
hasn't loosened over time.
I don't think dust can collect inside the plug-holes to an amount where it would interfere with operation while it's plugged in. Unless you're in a *very* dusty enviroment...
So, when I made sure that it's properly plugged in and it still doesnt work then I usually connect a fresh drive to the same cable to determine whether the drive or the controller is broken. I have never had an IDE cable fail in normal operation (you can ofcourse break them by deliberate bending but that doesn't happen so often inside a closed PC case).
Well, my point is, I agree with GP that these articial test scenarios are just braindead and prove nothing except what we all know: TV shows like that are made by stupid people for an even more stupid audience.
Ah, the "catch" of anti-trust legislation. Being a monopoly isnt illegal. Doing some paticular business practices isnt illegal. But being a monopoly and engaging in some paticlar business practices is.
Well, I don't pretend to understand anti-monopoly laws or anti-trust legislation indepth. But don't think I've ever heard of a case where an existing monopoly was actually broken by application of these laws.
Have I just missed it, do you know an example?
Yup, I missed that part. I merely discussed the implementation of a secure remote shell with javascript via https.
Well, it's a braindead idea either way.
Microsoft Windows is not an OS for games: it is an OS for Microsoft Games... It is not an OS for DBs: it is an OS for MSSQL.
All this whining...
See, I hate microsoft like the next guy. But if you were in the position of MS, wouldn't you use your monopoly just the same way?
What they do is just consequently, cynically applied business logic. It's what they teach in business school and what's supposed to drive our system.
I think therein lies the true problem which is much more complex than a monopoly in the PC industry.
IMHO the only real reason why MS has stayed so long and is still doing good is because no one has offered a package that is equally "good enough", yet.
If you develop (a DB, an office package, whatever) for the MS platform you have to deal with the MS competition. If MS offers the same product as you, you lose. It's that easy - they created the playground, they make the rules.
But there might be hope.
Their recent announcement makes me *hope* that apple might be heading for the mass market. And it happens that their product is actually technically better than what MS has to offer. So, let's see what darwin (sorry, pun) has to say on this matter. I for one wouldn't be too surprised if apple managed to bite off a huge chunk from MS's market share in the next decade. Don't underestimate the "cool"-factor. If apple-prices really come down to beige-box level we'll likely see some serious competition in the Joe-Sixpack-segment.
Oh and there still is linux, too.
The problem with MS having the OS and the apps under one umbrella corperation is that the apps developers get inside info about the OS. ...
Yes you would have two virtual monopolies, (Office and OS) but at least it wouldn't be one monopoly that can squeeze out other companies.
Every time I read such arguments I wonder how anyone could ever enforce this artificial split?
There a countless ways to "smuggle" simple API details back and forth that are unknown to the competition. If the new broken up companies really wanted to suppress their competition that way, a split like this certainly wouldn't hinder them much.
Well, actually you can get quite far with polling.
;-)
As you said, latency will be hell (e.g. 2 sec), but the UNIX CLI deals surprisingly well with high latency links.
Back in the day it was quite common to do stuff over 300bps links with high latency (probably rarely up to 2s but in the ballpark).
So, first of all you'll decouple Input and Output.
Instead of sending a char, waiting until the server echos it and printing it you will print the typed chars immediately and probably send the whole line on each poll (these few bytes are dwarfed by http protocol overhead anyways) so the server doesn't even need to "assemble" the line but just gets the whole thing everytime.
This will basically give you "local" feel on the CLI and latency only kicks in when you actually need the server response (e.g. cycling through the history or firing the command).
For interactive things (think: editor) it gets a bit more difficult since you actually depend on the editor telling you where to put the cursor etc.
Even for that stuff there are plenty ways to optimize cuz, as said, people have been tackling the "slow-link" problem very exhaustingly a long time ago.
Well, why anyone would want to spend all that effort when they can just use SSH is beyond me, tho...
This is leading way offtopic but I still want to mention it:
...and many more that I miss right now, you get the idea...
ActiveX is the one of the worst designed pieces of software that MS ever created. I'm sure it *still* is responsible for a significant percentage of malware infections despite not being enabled by default anymore (AFAIK, might be mistaken).
BUT...
The problems that are often times "solved" with ActiveX (/Java/Flash/Javascript...) are real and I cannot grasp why they haven't been addressed by the W3C years ago. It's a shame we still have to resort to plugins or scripting for form controls (dragndrop, treeviews, cascaded dropdown menus etc) that have been the lowest common detominator of just about any UI-kit for years.
Many parts of the web have evolved in the past years (CSS, javascript, XHTML).
The interactive elements available to web designers, those that actually enable you to DO something on the web, have not.
We're still stuck with the same seven basic form elements that we started with: file, input box, textarea, select box, button, checkbox, radiobox. period.
Designing an accessible and *useful* webpage would be much easier (read: HALF THE MANHOURS OR LESS!!) if we had at least the following additional form controls at our disposal:
comboselect - dropdown with optional text-input
dragndrop - upload file or multiple files via dragndrop
tree-menus - people script them with javascript all the time...
Can anyone explain to me why these obviously useful and both productivity- and accessiblity- (by standarization) boosting widgets have not been added to the standard, yet?
Instead of just adding a straightforward "input type=tree" to the source we still have to mess with insane javascript hacks or brain-damaging monstrosities like XUL.
Sorry this rant got longer than I intended but you somehow got me started on what bugs me every single f*ing day on my job. One day this stuff will be added and all the work many of us have spent to emulate these things for years will suddenly seem emberassingly obsolete.
Agree'd. Managing appointments in outlook, creating documents in word/excel and fancy colored "presentations" in powerpoint is what most windows-people consider "productive work".
For that stuff windows is fine, the occassional graphics job (photoshop) works, too. Even though for the latter you're better off with a mac already.
Anything that involves actually getting "stuff done" is better done on a unix OS. Don't get me started on what happens when "windows people" decide to implement business processes in scripted Excel sheets, access mdbs, lotus whatever-its-called-mess, VB or similar etch-a-sketch environments.
I pity the the poor schmocks who have to clean up the inevitable mess after a few months and "port" the pile of crap that stacked up in the meantime to something that actually works.
I used to be one of these schmocks until I learned to become very unpolite (read: say "no" very clear and insisting) early enough in the process.
Did NetCraft confirm, yet?
Great point!
I also wondered who the hell decided that the kids should be sitting in a dark room all day long. Money speaks I suppose. Who cares whether it hurts their eyesight and probably even their health (sunlight is good, remember?).
I don't mean to feed the trolls but he has a point.
Most things people do with applescript today could be done with rexx on my old amiga500 in 1990.
I'll leave the comments on who they grow up to be to someone else.
My product manager.
Microsoft turn out Windows and people keep on buying it in droves.
Lack of alternatives at one point in time lead to market domination.
Until recently there was no serious competition in mainstream computing ("Joe Sixpack").
Apple is catching up. There is hope.