I used IME under 2000 fine. Strangely enough, the options I used under 2k boxes are greyed out on the one XP machine I have access to. In any event that was present before...
Then I guess the XP look is like their getting a sports car? But I guess they reverted too far, the style looks more like Playskool...
Seriously though, It in many ways seems that they may have reached a peak and are falling from it. The 2000 products I think were the pinnacle in the professional world. Even though XP has a 'professional' edition, businesses seem to not really care about it.
Business people aren't excited about it if for no other reason than there being no 'XP Server'. While this has no technical merit, suits like to see consistancy, and feel that the best match for '2000 Servers' are '2000 Workstations', even if not always true. Plus, the new default look doesn't give an impression of 'professional', and the arrangement of the new start menu and desktop configuration can annoy them to no end. Yes all these things can be changed, but in first impressions, it really makes suits doubt the platform.
For IT people, they see that XP added shiny round windows and.... ummm..... that's just about it. They know it is an incremental update with few non-cosmetic feature enhancements. They know that while it offers little to no practical benefit, it at the same time will forever be slightly less tested and proven than Windows 2000 with all their respective updates. Additionally, though pretty efficient, the new graphics have some impact on performance, and at times the impact can be drastic if your video card isn't perfect.
Legal departments that bother to look at MS EULAs know to be scared more and more with every revision. MS is really trying to push their ground more and more, and they really haven't been giving back anything.
XP was a great thing to home users, finally going to the 2000 core for that segment. I would say XP could be the peak for the home segment, but I know full well that the home segment will buy up pretty, shiny, useless improvements endlessly. I think MS knows that too and is moving more and more into that segment (XBox, Tablets, Media Center..)
Windows 2000 offered a great deal of improvement over NT4 (mainly AD, but other stuff too). Windows XP offers next to nothing. Looking at the upcoming Windows 2003 release, there isn't that much to be excited about. Their revolutionary filesystem is the *only* feature I see that anyone cares much about, and I'm not sure how the market will ultimately view the feature.
FLAC is indeed lossless, and is also indeed large. Yes, audiophiles do insist that all lossy compression is unbearable. Even if they couldn't tell the difference, they would still make the claims I'm sure. Else how could they call themselves audiophiles?
As for myself, I think Vorbis acheives the best quality to size ratio. If there are artifiacts, they aren't as irritating as mp3 artifiacts, since I notice mp3 artifacts and don't Vorbis (until you get past 128 kbps, at which point I can't tell anything about an mp3. For listening on mediocre at best headphones in public with noise around me, I would say a 64 kbps vorbis would be good enough, not so about mp3.
First let the pitcher freeze the foam before pitching. Getting hit with a solid chunk of ice is a lot worse. Plus it seems that indications are that the heat tiles can't withstand impacts too well.
I just wish major companies (not just Apple, I've had moments with Compaq.... Not with Dell or Gateway, but that might just be pure luck) would make a bigger effort for Customer Service to be more consistant. Talking to four reps and no two agreeing on what is and is not covered by warranty really doesn't make a customer happy. I was actually happiest with the prognosis that the headphone jack was covered and the lid latch not, but if the lid latch can be fixed, might as well pursue that..
Now if they would just repair the cheap-ass plastic lid latch release button and headphone jack in my iBook.
Apple's engineering quality has declined. It seems they are trying to see what they can get away with now. The noisy power supplies are one example. Anyone else have the lid latch retainer thing break for no good reason?
And why in the latest iBook revision have they changed the metal holding a headphone plug in place to cheap plastic, and just to make *sure* it breaks, leave empty space between the plug and the more solid case. The iBook is the only device I have ever seen that saw it fit to use plastic there, and now I see why. Even a 5 dollar walkman uses metal there.
I can see why they could think that a plastic headphone jack would be sufficient without much testing, but what genious had the idea of a thin, relatively sharp lid latch fastening into rather flimsy plastic?
This recall is good, but on the whole, service is not that great nowadays. What is the price premium for if they try to weasel their way out of what they view as expensive repairs? I already drained my account to buy the damn system in the first place, don't demand 700+ dollars to fix two pieces of plastic just so I can use headphones again... I know the quote probably includes the motherboard, but still.
The battery life. temperature, and display are all top notch, but the cheap manufacturing on the outside can severely impact important functionality..
Here are the outcomes of various sporting events and here is how to bet....
Here are a list of stocks....... bail by mid 2000 for God's sake!
Lottery numbers for the biggest jackpot I can find....
And of course.... Date the following people who you didn't think you could, and don't date the following people who you could.....
And lastly, write all this down and go back in time when you are older and tell yourself all this.
At least, that is what I have written in my notebook, damn me for not telling me how I'm supposed to do time travel, very important detail to omit here. Now I'll create a universe destroying paradox...
Open Source is still largely deveolped as a hobby of enthusiasts. Some companies have their hands in the pie too, but even the resultant effort seems to be more in the style of the hobbiest than of the typical company effort.
Two factors. When I develop closed source apps for work, especially if it is something I have no real passion about, I tend to have messier code. No one is going to see it anyway. If I ever change jobs, a potential employer isn't going to ever see that code to review my style. If no user or the community in general will not see it, I'm more likely to take riskier shortcuts and settle for inelegant hacks. As long as no obvious runtime problems occur, then it is enough. When I submit patches for open source applications, I take more pride in the work. I want it to be clean, easy to read and follow, and free from amateurish looking code.
Secondly, even when I would like to re-evaluate approaches I use in a commercial environment, the business end of things will push deadlines. Time that I would have normally taken to go back, clean up, and rework the bits that work, but are too inelegant is denied. There is a significant amount of care with respect to market trends, customer demands, and marketing promises that interfere with quality code. With open source, you do it as you feel like it. Take as much time as you want. Sure, there are frequently deadlines in large projects (feature freeze, etc), but the penalty for not being able to meet those deadlines just means your work will be delayed to the next release cycle. There is no danger of losing your job, and even if consistantly missing feature freezes means you lose cvs write access, or are not taken as seriously, it really is no skin off your back, and you can almost always get back in through picking up the pace again...
I would assume they keyspace on the order of magnitude that MS would use would be quite significant. Additionally, it is not a simple matter of 'try combination, retry' carried out through an automated mechanism, each attempt requires a disc to be burned and a boot attempt. Maybe CDRW works (I have no idea), but that burning process is long enough to make even a weak keyspace difficult to brute force.
I would say it is more along the lines of they thought they *had* to run on PPC, Novell, i860, etc... to be considered serious for business. I could certainly understand Novell wanting to stall MS if possible, they had an iron grip on a segment now rapidly being taken over by MS. If MS had been unable to access Netware shares, NT probably would have gotten nowhere.
As far as the processors, they never dreamed businesses would ever feel that the x86 platform would be up to the task of doing 'real work', and thus they pursued architectures that would be more workstation and server quality.
I think NT is crap. But the existance of NT 3.1 cannot be outright denied, it was there. NT 3.5 was released before Windows 95. It was not because the GUI was Win3.x-like that it failed to make much inroads. NT4 happened to be the version out when the conditions were ripe for an increasing market share, and until 2000 was released, a lot of places continued to ignore NT as a toy. Home PCs really became drastically more popular while Windows 95 was standard. I think that is the reason why a lot of businesses started to think they wanted to use it in their work, so it would be like their home computer.
And saying NT was just WFW with Novell client tacked on is dumb. NT really did/still does maintain a separate, non-DOS based core. It was a truly 32 bit system. NT4 was mostly just NT 3.x with a shell upgrade, but the 3.x series needs more credit. Windows 95/98/ME were all increasingly complex hacks running on top of DOS. Really messy, really crappy. NT is crappy and messy, but not NEARLY so much the train wreck that is the 3.x/9x series.
Even though I agree that attitude is incorrect, I wouldn't go so far as to call them a monopoly in NT 3.5 days, at least not with respect to the market NT was targetting. NT was hardly a blip on the radar in professional systems.
It is so regrettable that MS has had such sucess in its inroads into professional computing. I think if the competitors (HP, IBM, Sun) had at least made a half-assed attempt at a price/performance ratio that could compete with x86 boxes, NT would continue to be seen as nothing more than a toy OS. People can say what they will about NT being a reliable system now, but it is quite clear that the system was not architected from the ground up to have the reliability of Solaris, AIX, HP-UX, etc. It is sad that a company that puts flashy, fun, marketable features before reliability and bugfixes has such a hold on the market.
First off, I must say I'm not sure if bringing software development to the end user is really important these days. Back then, there wasn't really any efficient, practically free methods to distribute software, so if you didn't write it, you had to pay at least a bit to get a copy and go to some trouble to actually go get it. For trivial apps, competent end users could write it up quickly for themselves for less time and money than it would take to find and get pre-made software. Now, computers are far more pervasive, so the competency of the average user is reduced. Also, with the internet, "real" programmes release freeware that pretty much covers the needs people once wrote custom software to do, and google, freshmeat, and versiontracker make it extremely easy to find the right tool for the job.
All that said, of the platforms in use in homes, I would say Apple is doing the best job in the direction you want.
With MS, Visual Studio costs money, and the application and APIs are relatively difficult to use and grasp for a beginning programmer.
With Linux, the development enviornment is free, but even less newbie friendly. Sure you can get some decent IDEs and even some UI design tools for free.... *IF* you know where to look, but documentation is sparse. Adding to this, there is no well defined Linux 'platform' (or *nix platform). You can make pretty likely assumptions about the kernel and glibc... and assume X if graphical (other alternatives are there, but far from mainstream). Beyond that, things become less certain. Qt, GTK+, Motif/Lesstif, Swing, and others are possible, likely toolkits, and with less certainty you can develop with Gnome and/or KDE, or if you want to be daring, you can use a relatively exotic language binding and require that. It all *can* work together, but it is an intimidating set of choices to have to make for a novice programmer, especially one that doesn't want to alienate users by requiring some library they don't have.
OSX is just incredible. First off, compiler, headers, really nice IDE, all are free and standard (Project Builder). The platform is truly a complete platform and you can make use of any part of it without fear that an end user would lack the libraries required to run it. The API is quite elegant and straightforward. Additionally, there is one great factor that truly makes the sort of things an enduser would want to do very accessible. Interface builder. The hardest part to a novice programmer tackling simple problems is dealing with UI, and Interface Builder makes this so damn easy and the Cocoa framework really make it possible to abstract the code from the interface. Though Apple's applications are not open source, most everything an end user would want to do to them can be done through opening the.nib files in the resource bundle and going at it. My personal thing I tend to do is to make Address Book, iChat, Calculator, and Safari not use the damned Scratched Metal look. Without exposing code, users have a really easy to figure out way of modifying the Interface of Cocoa apps in drastic ways, and I think that is truly great.
iBook? Beefy? Umm.. I wouldn't say that. Typing from one right now. It is kinda a neat little machine (except too much cheap-ass plastic in its construction), but I would hardly call it beefy performance wise. I guess he means performance isn't *much* of an issue, an iBook can easily be unacceptable. There are quite a few video files that drop a lot of frames to keep up on an iBook, wheter in OSX or in Linux. And just trying to play a lot of the 3D games... painful.
That is correct, but this brings up an interesting point about 'guaranteed' throughput... Guaranteed to where?
Cable modem has pretty much no guaranteed throughput to any hop. DSL is a bit further, and while not contractually guaranteeing throughput, it does extend dedicated throughput further than cable, to the CO.
Now realistically, you can only guarantee a connection so far, and how far is it guaranteed in a T1? Is it that much more further? Obviously, the other end you want throughput to could have less throughput or be more congested, so they really can only guarantee to the next hop, maybe further than DSL, but effectively not more useful.
My experience is that the high price for a T1 is not so much for the bandwidth, but for the care and reliability. They will make 'best-effort' attempts to support DSL/cable users, but the reality is that big commercial clients come first. The bandwidth may be limited, but the equipment and cabling used is more reliable, in addition to the better support.
And his point is also correct. They really don't limit you on such lines as they tend to do in 'consumer' level connections with the TOS.
IIRC, the first mouse happened to have three buttons. And every system that runs X has had three buttons. More than two buttons is just relatively new only in mainstream PC world. The wheel is relatively new I suppose. The typical mouse is three buttons/wheel. Aside from the wheel, it is pretty much the same in function. Optical mice and mice with many buttons change things a bit, but the many buttons are somewhat fringe, and the optical mouse just makes it so much smoother and no hassle with cleaning.
Anyway, just pointing out that mice haven't really changed fundamentally that much since their conception.
This is true on any platform. If you *must* have Windows in an environment that lacks physical security (i.e. a public computer lab), then you take the precautions. BIOS configuration password and boot only from hard disk. Now boot disks are useless on standard IDE computers with no external ports. Now if they open the case and use a hard disk to boot off of, screwed again, but the presumption is that would be too conspicuous. If you had a system with bootable SCSI/firewire, one could relatively inconspicuously hook up a device (iPod...heheh) and potentially trick the boot process into using that disk. Just theoretically, of course, SCSI IDs should foil that, and I have yet to see a firewire bootable system (I think).
Ultimately, a physically insecure machine is pretty much impossible to harden against anything more than casual attacks. If an administrator of a public network fails to password protect the bios *or* fails to disable the floppy and cd boot features, then they are inviting this sort of trouble, and there is nothing any operating system can do about it. If someone suggests an encrypted filesystem that requires a passphrase to mount, you have more problems on your hands than before. You want systems in public to be able to complete a reboot without administrator intervention.
Yes, from a practical standpoint, we are pretty much doomed if we find out that in a matter of days an extinction-level impact will occur.
However, if we find out that a more localized impact will occur, there is evacuation.
Moreover, if we have say a year or more warning of an extinction-level impact, they need to make that widely known. While at the current rate of advancement, we probably still couldn't do anything, you can bet your ass that a lot of resources will be dedicated to a solution than before, and with that level of research, we could probably have a viable option with regards to avoiding impact. If a solution would take all of the world's current nuclear arsenal, you could bet they would actually sacrifice it if it meant living another day.
Re:Sun has been sunset but Java is in safe hand...
on
The Faded Sun
·
· Score: 1
Ignoring the AWT part of things, I find Swing to be quite a well structured API, though Python bindings for GTK are even easier. Any textual API will seem clunky to some degree, it isn't easy to represent GUI in terms of text. Swing is on par with as good as it gets in my opinion.
Offtopic---- Of course, I think implementing GUI elements shouldn't be in the hands of the developers. Everywhere I have worked, a human factors person will design a UI based on his experience and relative superficial knowledge of the capabilities of the toolkit, and send their mock-ups to developers. Developers then make a best-effort attempt to imitate the gui, and have that reviewed by the human factors person and explain any discrepencies.
Now that I've done some work with OSX and ProjectBuilder, I think they really have the right idea. A human factors person can learn easily to use the Interface Builder, and truly build the interface. It is easier than doing the mock up+description of 'feel' issues not captured by the mock-ups, and mostly eliminates the need for the review process unless the developer has to make a change to the UI for some reason. Send that file along to developers and tell them to do the hard backend stuff. Much more efficient flow and you end up with what best matches the human factors expert's vision, as opposed to a developer's best effort attempt to imitate mock-ups, improvising in sub-optimal ways as the need arises.
Re:Sun *is* changing already
on
The Faded Sun
·
· Score: 1
OpenOffice... They realized StarOffice wasn't profitable as a product to them. OpenOffice means cheaper, more rapid development for a viable office alternative that can run on their platform. Even when professionals have Sun workstations, they often have a PC next to it for Office because that is what they need to communicate with management and other departments. This typical work environment is a threat to Sun, as departments see the redundancy and seek to eliminate it. If OpenOffice/StarOffice was a cheap and viable alternative, they would be tempted to ditch the PCs. Otherwise, they may view the short-term cost of retraining and purchase of PC-compatible tools a good investment for the long term if they can ditch Sun hardware.
Same with Gnome, everyone who has used it sees CDE as crap. It adds no value in terms of professionalism, and it is really piss poor in terms of user friendliness. Along comes Gnome. They like the C API which gives them the greatest flexibility (compared to KDE). It actually makes some sense as to how it works (as opposed to CDE). I have never ever met a CDE fan. WindowMaker, KDE, and Gnome always were quickly used as they were available. Some people even used their PC only until someone showed them an alternative environment to CDE, then they switched to the Sun box for everything they could, as it suddenly became more convenient to do so.
Also, it means it integrates better with Linux environments. Users just coming in may be familiar with Linux and not at all with CDE. With this, there is a more familiar environment.
These moves mean they can spend less focus on CDE and StarOffice, and end up with better products... For the purpose of advancing their UltraSparc/Solaris products, not abandoning them.
Re:or just use the OSS alternative
on
The Faded Sun
·
· Score: 1
It is a Python wrapper for SDL. I've even submitted a couple of patches to it. One thing I love about python is the C API. Sometimes you need to interface with something with only a C API available, and Python is a relatively painless compared to Java for doing this.
Yes, he is a moron. Why in the world does anyone at all pay attention to him? I mean, most people who can be fooled into thinking he is intelligent could take one look at that article and realize that
a) he is ready to talk authoritatively about a subject he has no grasp on
b) he has no understanding of the scope of maintaining a kernel vs. porting their graphical system to run with a different core and maintain compatibility.
Slashdot Mirror
I used IME under 2000 fine. Strangely enough, the options I used under 2k boxes are greyed out on the one XP machine I have access to. In any event that was present before...
Then I guess the XP look is like their getting a sports car? But I guess they reverted too far, the style looks more like Playskool...
Seriously though, It in many ways seems that they may have reached a peak and are falling from it. The 2000 products I think were the pinnacle in the professional world. Even though XP has a 'professional' edition, businesses seem to not really care about it.
Business people aren't excited about it if for no other reason than there being no 'XP Server'. While this has no technical merit, suits like to see consistancy, and feel that the best match for '2000 Servers' are '2000 Workstations', even if not always true. Plus, the new default look doesn't give an impression of 'professional', and the arrangement of the new start menu and desktop configuration can annoy them to no end. Yes all these things can be changed, but in first impressions, it really makes suits doubt the platform.
For IT people, they see that XP added shiny round windows and.... ummm..... that's just about it. They know it is an incremental update with few non-cosmetic feature enhancements. They know that while it offers little to no practical benefit, it at the same time will forever be slightly less tested and proven than Windows 2000 with all their respective updates. Additionally, though pretty efficient, the new graphics have some impact on performance, and at times the impact can be drastic if your video card isn't perfect.
Legal departments that bother to look at MS EULAs know to be scared more and more with every revision. MS is really trying to push their ground more and more, and they really haven't been giving back anything.
XP was a great thing to home users, finally going to the 2000 core for that segment. I would say XP could be the peak for the home segment, but I know full well that the home segment will buy up pretty, shiny, useless improvements endlessly. I think MS knows that too and is moving more and more into that segment (XBox, Tablets, Media Center..)
Windows 2000 offered a great deal of improvement over NT4 (mainly AD, but other stuff too). Windows XP offers next to nothing. Looking at the upcoming Windows 2003 release, there isn't that much to be excited about. Their revolutionary filesystem is the *only* feature I see that anyone cares much about, and I'm not sure how the market will ultimately view the feature.
FLAC is indeed lossless, and is also indeed large. Yes, audiophiles do insist that all lossy compression is unbearable. Even if they couldn't tell the difference, they would still make the claims I'm sure. Else how could they call themselves audiophiles?
As for myself, I think Vorbis acheives the best quality to size ratio. If there are artifiacts, they aren't as irritating as mp3 artifiacts, since I notice mp3 artifacts and don't Vorbis (until you get past 128 kbps, at which point I can't tell anything about an mp3. For listening on mediocre at best headphones in public with noise around me, I would say a 64 kbps vorbis would be good enough, not so about mp3.
First let the pitcher freeze the foam before pitching. Getting hit with a solid chunk of ice is a lot worse. Plus it seems that indications are that the heat tiles can't withstand impacts too well.
The acronym of that would be RMS.... RMS is evil.. No wait... that *other* RMS.......
On a sidenote, I hadn't heard about MS changing their product line to 'Widows'..... Another interesting name change...
Cool, I'll give it yet another shot...
I just wish major companies (not just Apple, I've had moments with Compaq.... Not with Dell or Gateway, but that might just be pure luck) would make a bigger effort for Customer Service to be more consistant. Talking to four reps and no two agreeing on what is and is not covered by warranty really doesn't make a customer happy. I was actually happiest with the prognosis that the headphone jack was covered and the lid latch not, but if the lid latch can be fixed, might as well pursue that..
Offtopic rant...
Now if they would just repair the cheap-ass plastic lid latch release button and headphone jack in my iBook.
Apple's engineering quality has declined. It seems they are trying to see what they can get away with now. The noisy power supplies are one example. Anyone else have the lid latch retainer thing break for no good reason?
And why in the latest iBook revision have they changed the metal holding a headphone plug in place to cheap plastic, and just to make *sure* it breaks, leave empty space between the plug and the more solid case. The iBook is the only device I have ever seen that saw it fit to use plastic there, and now I see why. Even a 5 dollar walkman uses metal there.
I can see why they could think that a plastic headphone jack would be sufficient without much testing, but what genious had the idea of a thin, relatively sharp lid latch fastening into rather flimsy plastic?
This recall is good, but on the whole, service is not that great nowadays. What is the price premium for if they try to weasel their way out of what they view as expensive repairs? I already drained my account to buy the damn system in the first place, don't demand 700+ dollars to fix two pieces of plastic just so I can use headphones again... I know the quote probably includes the motherboard, but still.
The battery life. temperature, and display are all top notch, but the cheap manufacturing on the outside can severely impact important functionality..
Here are the outcomes of various sporting events and here is how to bet....
Here are a list of stocks....... bail by mid 2000 for God's sake!
Lottery numbers for the biggest jackpot I can find....
And of course.... Date the following people who you didn't think you could, and don't date the following people who you could.....
And lastly, write all this down and go back in time when you are older and tell yourself all this.
At least, that is what I have written in my notebook, damn me for not telling me how I'm supposed to do time travel, very important detail to omit here. Now I'll create a universe destroying paradox...
Open Source is still largely deveolped as a hobby of enthusiasts. Some companies have their hands in the pie too, but even the resultant effort seems to be more in the style of the hobbiest than of the typical company effort.
Two factors. When I develop closed source apps for work, especially if it is something I have no real passion about, I tend to have messier code. No one is going to see it anyway. If I ever change jobs, a potential employer isn't going to ever see that code to review my style. If no user or the community in general will not see it, I'm more likely to take riskier shortcuts and settle for inelegant hacks. As long as no obvious runtime problems occur, then it is enough. When I submit patches for open source applications, I take more pride in the work. I want it to be clean, easy to read and follow, and free from amateurish looking code.
Secondly, even when I would like to re-evaluate approaches I use in a commercial environment, the business end of things will push deadlines. Time that I would have normally taken to go back, clean up, and rework the bits that work, but are too inelegant is denied. There is a significant amount of care with respect to market trends, customer demands, and marketing promises that interfere with quality code. With open source, you do it as you feel like it. Take as much time as you want. Sure, there are frequently deadlines in large projects (feature freeze, etc), but the penalty for not being able to meet those deadlines just means your work will be delayed to the next release cycle. There is no danger of losing your job, and even if consistantly missing feature freezes means you lose cvs write access, or are not taken as seriously, it really is no skin off your back, and you can almost always get back in through picking up the pace again...
I would assume they keyspace on the order of magnitude that MS would use would be quite significant. Additionally, it is not a simple matter of 'try combination, retry' carried out through an automated mechanism, each attempt requires a disc to be burned and a boot attempt. Maybe CDRW works (I have no idea), but that burning process is long enough to make even a weak keyspace difficult to brute force.
I would say it is more along the lines of they thought they *had* to run on PPC, Novell, i860, etc... to be considered serious for business. I could certainly understand Novell wanting to stall MS if possible, they had an iron grip on a segment now rapidly being taken over by MS. If MS had been unable to access Netware shares, NT probably would have gotten nowhere.
As far as the processors, they never dreamed businesses would ever feel that the x86 platform would be up to the task of doing 'real work', and thus they pursued architectures that would be more workstation and server quality.
Should be moderated as -1, WRONG.
I think NT is crap. But the existance of NT 3.1 cannot be outright denied, it was there. NT 3.5 was released before Windows 95. It was not because the GUI was Win3.x-like that it failed to make much inroads. NT4 happened to be the version out when the conditions were ripe for an increasing market share, and until 2000 was released, a lot of places continued to ignore NT as a toy. Home PCs really became drastically more popular while Windows 95 was standard. I think that is the reason why a lot of businesses started to think they wanted to use it in their work, so it would be like their home computer.
And saying NT was just WFW with Novell client tacked on is dumb. NT really did/still does maintain a separate, non-DOS based core. It was a truly 32 bit system. NT4 was mostly just NT 3.x with a shell upgrade, but the 3.x series needs more credit. Windows 95/98/ME were all increasingly complex hacks running on top of DOS. Really messy, really crappy. NT is crappy and messy, but not NEARLY so much the train wreck that is the 3.x/9x series.
Even though I agree that attitude is incorrect, I wouldn't go so far as to call them a monopoly in NT 3.5 days, at least not with respect to the market NT was targetting. NT was hardly a blip on the radar in professional systems.
It is so regrettable that MS has had such sucess in its inroads into professional computing. I think if the competitors (HP, IBM, Sun) had at least made a half-assed attempt at a price/performance ratio that could compete with x86 boxes, NT would continue to be seen as nothing more than a toy OS. People can say what they will about NT being a reliable system now, but it is quite clear that the system was not architected from the ground up to have the reliability of Solaris, AIX, HP-UX, etc. It is sad that a company that puts flashy, fun, marketable features before reliability and bugfixes has such a hold on the market.
First off, I must say I'm not sure if bringing software development to the end user is really important these days. Back then, there wasn't really any efficient, practically free methods to distribute software, so if you didn't write it, you had to pay at least a bit to get a copy and go to some trouble to actually go get it. For trivial apps, competent end users could write it up quickly for themselves for less time and money than it would take to find and get pre-made software. Now, computers are far more pervasive, so the competency of the average user is reduced. Also, with the internet, "real" programmes release freeware that pretty much covers the needs people once wrote custom software to do, and google, freshmeat, and versiontracker make it extremely easy to find the right tool for the job.
.nib files in the resource bundle and going at it. My personal thing I tend to do is to make Address Book, iChat, Calculator, and Safari not use the damned Scratched Metal look. Without exposing code, users have a really easy to figure out way of modifying the Interface of Cocoa apps in drastic ways, and I think that is truly great.
All that said, of the platforms in use in homes, I would say Apple is doing the best job in the direction you want.
With MS, Visual Studio costs money, and the application and APIs are relatively difficult to use and grasp for a beginning programmer.
With Linux, the development enviornment is free, but even less newbie friendly. Sure you can get some decent IDEs and even some UI design tools for free.... *IF* you know where to look, but documentation is sparse. Adding to this, there is no well defined Linux 'platform' (or *nix platform). You can make pretty likely assumptions about the kernel and glibc... and assume X if graphical (other alternatives are there, but far from mainstream). Beyond that, things become less certain. Qt, GTK+, Motif/Lesstif, Swing, and others are possible, likely toolkits, and with less certainty you can develop with Gnome and/or KDE, or if you want to be daring, you can use a relatively exotic language binding and require that. It all *can* work together, but it is an intimidating set of choices to have to make for a novice programmer, especially one that doesn't want to alienate users by requiring some library they don't have.
OSX is just incredible. First off, compiler, headers, really nice IDE, all are free and standard (Project Builder). The platform is truly a complete platform and you can make use of any part of it without fear that an end user would lack the libraries required to run it. The API is quite elegant and straightforward. Additionally, there is one great factor that truly makes the sort of things an enduser would want to do very accessible. Interface builder. The hardest part to a novice programmer tackling simple problems is dealing with UI, and Interface Builder makes this so damn easy and the Cocoa framework really make it possible to abstract the code from the interface. Though Apple's applications are not open source, most everything an end user would want to do to them can be done through opening the
Only *one* meeting in a day? That sounds like a much more fun place to work than most, they hiring?
iBook? Beefy? Umm.. I wouldn't say that. Typing from one right now. It is kinda a neat little machine (except too much cheap-ass plastic in its construction), but I would hardly call it beefy performance wise. I guess he means performance isn't *much* of an issue, an iBook can easily be unacceptable. There are quite a few video files that drop a lot of frames to keep up on an iBook, wheter in OSX or in Linux. And just trying to play a lot of the 3D games... painful.
That is correct, but this brings up an interesting point about 'guaranteed' throughput... Guaranteed to where?
Cable modem has pretty much no guaranteed throughput to any hop. DSL is a bit further, and while not contractually guaranteeing throughput, it does extend dedicated throughput further than cable, to the CO.
Now realistically, you can only guarantee a connection so far, and how far is it guaranteed in a T1? Is it that much more further? Obviously, the other end you want throughput to could have less throughput or be more congested, so they really can only guarantee to the next hop, maybe further than DSL, but effectively not more useful.
My experience is that the high price for a T1 is not so much for the bandwidth, but for the care and reliability. They will make 'best-effort' attempts to support DSL/cable users, but the reality is that big commercial clients come first. The bandwidth may be limited, but the equipment and cabling used is more reliable, in addition to the better support.
And his point is also correct. They really don't limit you on such lines as they tend to do in 'consumer' level connections with the TOS.
IIRC, the first mouse happened to have three buttons. And every system that runs X has had three buttons. More than two buttons is just relatively new only in mainstream PC world. The wheel is relatively new I suppose. The typical mouse is three buttons/wheel. Aside from the wheel, it is pretty much the same in function. Optical mice and mice with many buttons change things a bit, but the many buttons are somewhat fringe, and the optical mouse just makes it so much smoother and no hassle with cleaning.
Anyway, just pointing out that mice haven't really changed fundamentally that much since their conception.
Windows 2000 recovery console is only available at boot time from the CD. It can't run once the system is booted.
This is true on any platform. If you *must* have Windows in an environment that lacks physical security (i.e. a public computer lab), then you take the precautions. BIOS configuration password and boot only from hard disk. Now boot disks are useless on standard IDE computers with no external ports. Now if they open the case and use a hard disk to boot off of, screwed again, but the presumption is that would be too conspicuous. If you had a system with bootable SCSI/firewire, one could relatively inconspicuously hook up a device (iPod...heheh) and potentially trick the boot process into using that disk. Just theoretically, of course, SCSI IDs should foil that, and I have yet to see a firewire bootable system (I think).
Ultimately, a physically insecure machine is pretty much impossible to harden against anything more than casual attacks. If an administrator of a public network fails to password protect the bios *or* fails to disable the floppy and cd boot features, then they are inviting this sort of trouble, and there is nothing any operating system can do about it. If someone suggests an encrypted filesystem that requires a passphrase to mount, you have more problems on your hands than before. You want systems in public to be able to complete a reboot without administrator intervention.
Yes, from a practical standpoint, we are pretty much doomed if we find out that in a matter of days an extinction-level impact will occur.
However, if we find out that a more localized impact will occur, there is evacuation.
Moreover, if we have say a year or more warning of an extinction-level impact, they need to make that widely known. While at the current rate of advancement, we probably still couldn't do anything, you can bet your ass that a lot of resources will be dedicated to a solution than before, and with that level of research, we could probably have a viable option with regards to avoiding impact. If a solution would take all of the world's current nuclear arsenal, you could bet they would actually sacrifice it if it meant living another day.
Ignoring the AWT part of things, I find Swing to be quite a well structured API, though Python bindings for GTK are even easier. Any textual API will seem clunky to some degree, it isn't easy to represent GUI in terms of text. Swing is on par with as good as it gets in my opinion.
Offtopic----
Of course, I think implementing GUI elements shouldn't be in the hands of the developers. Everywhere I have worked, a human factors person will design a UI based on his experience and relative superficial knowledge of the capabilities of the toolkit, and send their mock-ups to developers. Developers then make a best-effort attempt to imitate the gui, and have that reviewed by the human factors person and explain any discrepencies.
Now that I've done some work with OSX and ProjectBuilder, I think they really have the right idea. A human factors person can learn easily to use the Interface Builder, and truly build the interface. It is easier than doing the mock up+description of 'feel' issues not captured by the mock-ups, and mostly eliminates the need for the review process unless the developer has to make a change to the UI for some reason. Send that file along to developers and tell them to do the hard backend stuff. Much more efficient flow and you end up with what best matches the human factors expert's vision, as opposed to a developer's best effort attempt to imitate mock-ups, improvising in sub-optimal ways as the need arises.
OpenOffice... They realized StarOffice wasn't profitable as a product to them. OpenOffice means cheaper, more rapid development for a viable office alternative that can run on their platform. Even when professionals have Sun workstations, they often have a PC next to it for Office because that is what they need to communicate with management and other departments. This typical work environment is a threat to Sun, as departments see the redundancy and seek to eliminate it. If OpenOffice/StarOffice was a cheap and viable alternative, they would be tempted to ditch the PCs. Otherwise, they may view the short-term cost of retraining and purchase of PC-compatible tools a good investment for the long term if they can ditch Sun hardware.
Same with Gnome, everyone who has used it sees CDE as crap. It adds no value in terms of professionalism, and it is really piss poor in terms of user friendliness. Along comes Gnome. They like the C API which gives them the greatest flexibility (compared to KDE). It actually makes some sense as to how it works (as opposed to CDE). I have never ever met a CDE fan. WindowMaker, KDE, and Gnome always were quickly used as they were available. Some people even used their PC only until someone showed them an alternative environment to CDE, then they switched to the Sun box for everything they could, as it suddenly became more convenient to do so.
Also, it means it integrates better with Linux environments. Users just coming in may be familiar with Linux and not at all with CDE. With this, there is a more familiar environment.
These moves mean they can spend less focus on CDE and StarOffice, and end up with better products... For the purpose of advancing their UltraSparc/Solaris products, not abandoning them.
It is a Python wrapper for SDL. I've even submitted a couple of patches to it. One thing I love about python is the C API. Sometimes you need to interface with something with only a C API available, and Python is a relatively painless compared to Java for doing this.
Yes, he is a moron. Why in the world does anyone at all pay attention to him? I mean, most people who can be fooled into thinking he is intelligent could take one look at that article and realize that
a) he is ready to talk authoritatively about a subject he has no grasp on
b) he has no understanding of the scope of maintaining a kernel vs. porting their graphical system to run with a different core and maintain compatibility.