Can there be room for another system?

[Waccoon]

UNIX has a sound philosophy, but it's full of warts and rust.  A major update should be considered, especially where security and localization are concerned.  Unfortunately, such things are invisible to end-users and only really affect programmers.  There's not much incentive to make a better kernel/OS when you can just put more gloss and eye candy over such old designs.

Irishmike:   My feeling is that innovation is nothing more than stealing an idea from an open-source project and mainly I too feel that the computer industry is making more or less an appliance for the masses.

Concerning the "theft" of ideas, I usually see the opposite.  Just about every desktop for Linux looks and works like a Windows or Mac system.  I've tried over ten distros of Linux and about 25 OSes overall, and I've never used any one of them for more than a week.

Ironicly, the only OS I've used in the last 5 years that I found really interesting was an old copy of OS/2 (version 2.0).  It was a lot of fun to play with, but certainly not very stable or flexible.  I learned a lot of BAD design tips to avoid by using OS/2.  

I agree with the "appliance for the masses."  Part of the reason why everything looks like Windows/Mac is because the alternatives are trying to appeal to the mass market.  Only power users really seek out alternatives.

Irishmike:  What is a keen example of what I am talking about is a new feature of Windows (Vista) which looks keenly like an idea that Sun Microsystems had out on their open-source project called Looking Glass. The Windows implementation of "flip" and "flip 3D" looks exactly like Looking Glass.

My feeling is that we should be looking for ways to consolidate information to reduce clutter, rather than find new, flashy ways to flip between windows.  Compare Windows Explorer to Total Commander.

Irishmike:  What I don't get is why Amiga, Inc or Hyperion don't simply write their own GUI on top of a *NIX base (just the way Apple did) and make it therefore able to run on any platform that the kernel would support?

I think it's because Hyperion seriously underestimated the improvement in graphics on hand-held systems, and were trying to make a really fast proprietary 2D solution with 3D just tacked on, like the original Amiga.  Any embedded system powerful enough to have a PPC processor needs more than that, and anything less is going to run on ARM.

Amiga Inc. was on the right track with AmigaDE.  Too bad they just didn't release anything that looked like it really was based on a next-gen platform.  Slot machines?  Come on.

I don't like Java that much, but after seeing a demo of SavaJe, it's really changed my impression of where we need to go as far as OS design is concerned.  I've seen quite a bit of that platform and what it can do.  Plus, Java programmers are plentiful.  That means a lot to companies that want to build "toy" applications for their products, like games, messaging services, etc.

GreggBZ:  We've hit a sort of plateau on so many consumer technologies.

When it comes to tools, yes.  I'd prefer only a modernization of UNIX, but a brand new programming language.  Programming, especially GUI development, needs to be made much easier.

A few years ago, I read The Mytical Man Month, and I was very surprised to hear the author say (in the 80's), that computers were basicly fast enough to do anything... it was the system architecture and team management that was the problem.  Granted, the author began his programming career in the 60's.  The longer you work on computers, the more you realize speed is not the solution.  I don't think a lot of people really understand what responsiveness means.

Dovegrace:  As for the original topic, the only thing I can think of that would impress ME would be a true 3D GUI manipulated by, for example, something similar to Nintendo's old Power Glove...

Given that a sheet of paper is still one of the most powerful and frequently used tools used by man, I'm not sure a virtual 3D environment is the solution.  2D is still the best for data representation.  There's a reason why the virtual reality craze died out in the early 90's.

I see most "VR" controllers as short-lived trends.  Tablet screens, like on TabletPCs and the NintendoDS, have a huge amount of potential.  If early PDA's hadn't been so proprietary and sluggish, they would have survived the onslaught of other trendy gadgets, like cell phones, and evolved into a much more mature platform.  I'm disappointed that laptop computers are still so popular.

BTW, the Power Glove was developed by Abrams/Gentile Entertainment.

[Waccoon]

People who complain about outrageous system requirements are massively impatient, as they want cutting-edge software with ancient hardware.  Most OSes, including Windows, work just fine on older hardware.  Most games made by GOOD developers adapt to the hardware available, rather than force you to buy a $4,000 "gaming rig."

Focus on function, please, not speed.

PS - I spent the same amount of money on my PC two years ago as I did for my A1200.  I see no reason to spend $1,000+ for my next computer, which will be a Conroe.

[Waccoon]

ALewis:   Actually, there is an very very real use for a 3D interface, even on a 2D screen. Vista, OSX, et al are *NOT* 3D interfaces. They merely present a 3D look and feel, much like games such as Wolfenstein.

What you described in that article rarely works.  As usual, it looks good on paper, but is not useful in practice.

Remember those "fake" 3D games like Welltris, and those various (and hideously annoying) 3D breakout games where you shot the ball at a wall in the distance?  I shudder to think how that would work with icons!

Here's what I wrote as a comment for your article:

It makes more sense to group windows or rip interface elements from one window and move them to another, or use the zooming paradigm, where windows can be made much smaller and bigger at will (solving a lot of problems for visually impared people to boot).  There aren't many people who have solid relational skills to handle interfaces with depth, especially "fake" depth.  You should always avoid "stacking" stuff, if possible.

Also, setting up a consistant work area with custom window styling helps.    This work area is saved and restored whenever you log in.  You don't have to fumble with large folder views if you create groups of your favorite folders with special borders with all the useless peripheral buttons removed.  Today's GUI toolkits don't allow much flexibility as far as removing "standard" window elements.

We've had picture within picture on our TVs for a while.  We've had virtual desktops for a while.  Why not combine both ideas?  Isn't that why windows were created?

Application support for clever organization depends on what GUI toolkit is used.  Regardless of what tools the OS offers, developers will usually respect system-wide programming guidelines, but will rarely, if ever, use only system tools to write their software.  Java suffers from this problem considerably (probably given how many times Sun rewrote their toolkit -- and it still sucks).

BTW, WinFS was canned, at least in its exiting form.  I doubt it would have been truly useful, as it depends on metadata to acutally store information, and most content creators do a horrid job of setting this vital information.  We still sort our MP3s by filename!

What would make more sense is a flexible rules system, allowing people to write rules about how information is displayed, based on type, age, quantity, filename syntax, etc., and allowing you to apply rules to certain folders.  Think regular expressions meets folder views.  For example, you might have a natural sorting rule, which will sort things as "1, 2, 03, 4, 40".  XP does this by default, and I find it annoying as my UNIX servers and application tools do not.  Another rule is that any folder that contains a catalog would automatically count the number of records and display it as a property in the folder view.  XP can be configured to do some of these things, but it requires some nasty registry hacking.  A clean, CLI-based way of doing this would be nice, with a simpler GUI-based front-end for people who can't handle scripting.  It's basicly filtering, which is what UNIX people have been doing for years by piping the output of command-line programs to text files.  Except here, we do it in realtime, inside a folder.

I think this is what the technology inside WinFS was meant to do, but nobody thought about using it that way.

Oh yeah, a CLI bar would be nice, so we could select a bunch of files inside a GUI, which would be stored as a default array.  Then, write a single command line, such as:

"pngout $select {$select}*.pn2"

That would take every file selected, make a PNG out of it, and rename it so the file exetension would be ".pn2".  Doing this as a traditional batch is sloppy; usually this is a job for Perl, which is a terrific way to add REAL bloat.

[Waccoon]

ALewis:  stick a floppy in an XP rig an format it.. and watch the system come to a halt

This is a hardware issue.  Get a USB floppy (free of horrible ISA legacy hardware), and your problems will disappear.

Adz:  Funny, I was thinking about this earlier today, computing really doesn't excite me anymore, even in the days of 486 and Pentium, it was quite exciting, not as exciting as the C64/Amiga days, but it was certainly better than now.

I agree 100%.  The problem is that there was more emphasis on creativity back then.  You know, when your Amiga came bundled with DPaint.  These days, it's all about shoveling content you don't want into your face.

That's why I like Oekaki.  Artists (sometimes very talented ones with tablets) get together and draw pictures online in a forum-like atmosphere.  Paintchat is the same, but it's in realtime, like a chat room.  That's a million times more fun than browsing web sites.

Oh, if only Windows came with Oekaki and not Solitaire...

BTW, the name is weird because it's a Japanese thing.

Adz:  However, having read some articles pertaining to Windows Vista RC1, things do look a little bleak. E.g. machines running a GeForce 7900 were experiencing excessive fan noise because the chip was running at 100% just trying to render the bloody GUI. How on earth is that supposed to be exciting?

Unfortunately, GPU manufacturers haven't put any serious thought into throttling like CPU manufacturers have, so as long as 3D is active, processor utilization will always be close to 100%.

As with CPUs, 100% utilization doesn't mean the chips are being stressed to their limit.  Hopefully, the introduction of Vista will force GPU developers to rethink their tactics, which really should have been done five years ago.

Adonay: Hmm breaktrughs or new exiting things in todays computerworld ? Hmmm not that big ones but for me at leased Is the chip market the upcoming PPU "when it becomes usefull" physics processing unit

Blech.  PPUs are not as accurate as CPU calculations, which is one reason they are faster.  This fad is pretty dumb -- a throwback to math coprocessors.  As more standardized tools become available, PPUs will die quickly in favor of more generalized vector units.

Nycran: There is a new platform, it's called the Internet. This platform is not specific to an OS or hardware.. it is something else entirely.

LAMP?  XML?  AJAX?

I think I'm going to be sick. :egad: