Interesting you mention devices eight times as large like the Newton or the WinCE-based jokes that drained 2 AAA cells in about a day of light usage (I know - I was issued one). Of course they were more powerful. The Newton was very interesting, but costing more than a desktop computer of the time, nobody really bought it (I have one, still working BTW).
The Palm had to do with battery backed RAM and a very simple 68K-like embedded processor. Its API resembled a lot the original MacOS and it was not nearly as bad as people like to think.
Of course, these days I find anything less sophisticated than my Python/PyGnome stack unbearable. But that's the time. By early 90s I found anything less than a Smalltalk environment was ridiculously primitive (and I still do to some extent, but I am more forgiving)
"Am I going to be able to use the WebOS when there's no wireless data connectivity?"
Yes. It supports local storage.
"Can Palm ensure the security of my data while using WebOS?"
Who can?
"What happened to the rumored port of PalmOS to Linux?"
The "classic" PalmOS now belongs to Access. They sell something Linux based, I don't know exactly what. WebOS is built on top of a Linux kernel, AFAIK.
"Since they are abandoning the platform, is it for sale? Are they going to open source it?"
You have to ask Access for that. They ceased development 4 years ago.
The single-core ARM is a deal-breaker for an x86 replacement, even with the coupled DSP and 3D engines. In the same power envelope of an Intel Atom you can do a lot more and people need that "more" to run their software acceptably.
Also, people will need a real keyboard and a much larger screen (600x1024 is a little bit painful to use - I run my netbook with an external monitor most of the time).
I would prefer a general purpose vector unit over a dedicated GPU, but that's me. If someone makes a chip similar to what I described above and builds a netbook around it, I am buying regardless of how many GPU or vector cores are inside.
How about 8 ARM cores with FPU and vector capabilities (that can be used for 3D acceleration), along with hefty caches, on a single chip? Now, imagine all parts can downclock themselves to a full halt when idling or even power down when the OS asks them. IIRC the hotplug stuff in the Linux kernel can already deal with that kind of stuff.
The tech is all around. All we need is someone who is willing to do it.
About 10 years ago I had an IBM z-50. It was about the size of one of the larger Eee's and sported a MIPS processor. At that time (with late 90s battery and display technology) it offered a decent computer (I made it run NetBSD later in its life) with a 10 hour battery life.
A decent laptop built around one of these low-power processors could last a week on a single charge and a big enough battery.
As for desktops, how many ARM cores, FPUs, vector units and cache can be put on a Core i7 die?
I wonder of what past is he speaking. I remember the time a decent Unix workstation had 4 megs of RAM and 100 of disk.
Heck. I remember the time when a decent personal computer had 16K of RAM. My first computer had 2K! I remember how incredible the upgrade to my first Apple II, with its 48K and 1 MHz 6502 processor (that ran circles around the ZX81-clone's Z80).
I am curious as to what parts of an x86 are patented in a way that you can't circumvent by designing it from the ground up yourself.
Are things like "a device or method to manufacture a processor with a ridiculously small number of registers" patentable?;-)
Because if there are things so fundamental that cannot be circumvented this way, then everybody who makes processors (and Nvidia builds GPUs which are specialized processors) infringes them somehow.
I remember my first Linux install. It was actually MkLinux on a Power Macintosh 9300 (or something close). Installing was _very_ easy - just uncompressing data on an empty filesystem.
It booted via an MacOS extension. I guess that made it so easy.
Still, it was pretty cool. I had to compile and install Afterstep by hand.
A year later, I came back to that company (a major ISP here in Brazil) as a consultant only to see just about all of those Macs converted into MkLinux boxes.
Building a distro from scratch is a huge undertaking. The best would be to adopt one or two successful distros and build packages for them that work great in your hardware.
If giving up revenue is the idea, why just not make it free of charge, or, better yet (at least for them) make people pay for an inexpensive lifetime unlimited upgrade and usage permission? No need for re-licensing of components acquired from other sources, no risk of letting people know they use GPL'ed code hidden deep inside stuff etc.
Slashdot Mirror
Interesting you mention devices eight times as large like the Newton or the WinCE-based jokes that drained 2 AAA cells in about a day of light usage (I know - I was issued one). Of course they were more powerful. The Newton was very interesting, but costing more than a desktop computer of the time, nobody really bought it (I have one, still working BTW).
The Palm had to do with battery backed RAM and a very simple 68K-like embedded processor. Its API resembled a lot the original MacOS and it was not nearly as bad as people like to think.
Of course, these days I find anything less sophisticated than my Python/PyGnome stack unbearable. But that's the time. By early 90s I found anything less than a Smalltalk environment was ridiculously primitive (and I still do to some extent, but I am more forgiving)
"Am I going to be able to use the WebOS when there's no wireless data connectivity?"
Yes. It supports local storage.
"Can Palm ensure the security of my data while using WebOS?"
Who can?
"What happened to the rumored port of PalmOS to Linux?"
The "classic" PalmOS now belongs to Access. They sell something Linux based, I don't know exactly what. WebOS is built on top of a Linux kernel, AFAIK.
"Since they are abandoning the platform, is it for sale? Are they going to open source it?"
You have to ask Access for that. They ceased development 4 years ago.
Actually, the monopoly was killed, from one end, by a company leveraging another monopoly to extend into the PDA segment.
From the other direction came Symbian with Nokia and Sony Ericsson and pretty much every telco on the planet.
Really, they had little chance.
"Is that perception incorrect?"
Pretty much yes.
It all depends on you deciding to trust the JSON you get and eval-ing it.
If you can't trust the source, you should parse it using a safe parser.
"Be a few years ago to build their new OS versions around BeOS?"
THAT was a major blunder. Hope the patents were worth it at least ;-)
Just remember that API was designed to run on very simple devices built in 1998.
Don't they think of the children?
The single-core ARM is a deal-breaker for an x86 replacement, even with the coupled DSP and 3D engines. In the same power envelope of an Intel Atom you can do a lot more and people need that "more" to run their software acceptably.
Also, people will need a real keyboard and a much larger screen (600x1024 is a little bit painful to use - I run my netbook with an external monitor most of the time).
I would prefer a general purpose vector unit over a dedicated GPU, but that's me. If someone makes a chip similar to what I described above and builds a netbook around it, I am buying regardless of how many GPU or vector cores are inside.
How about 8 ARM cores with FPU and vector capabilities (that can be used for 3D acceleration), along with hefty caches, on a single chip? Now, imagine all parts can downclock themselves to a full halt when idling or even power down when the OS asks them. IIRC the hotplug stuff in the Linux kernel can already deal with that kind of stuff.
The tech is all around. All we need is someone who is willing to do it.
About 10 years ago I had an IBM z-50. It was about the size of one of the larger Eee's and sported a MIPS processor. At that time (with late 90s battery and display technology) it offered a decent computer (I made it run NetBSD later in its life) with a 10 hour battery life.
A decent laptop built around one of these low-power processors could last a week on a single charge and a big enough battery.
As for desktops, how many ARM cores, FPUs, vector units and cache can be put on a Core i7 die?
It's actually great. An ARM-based PC would not only be Windows-free, it would be Windows-proof.
Not really. There is still Windows CE, but one can dream...
vt-100
X Terminals
"massive operating systems of the past"
I wonder of what past is he speaking. I remember the time a decent Unix workstation had 4 megs of RAM and 100 of disk.
Heck. I remember the time when a decent personal computer had 16K of RAM. My first computer had 2K! I remember how incredible the upgrade to my first Apple II, with its 48K and 1 MHz 6502 processor (that ran circles around the ZX81-clone's Z80).
I am curious as to what parts of an x86 are patented in a way that you can't circumvent by designing it from the ground up yourself.
Are things like "a device or method to manufacture a processor with a ridiculously small number of registers" patentable? ;-)
Because if there are things so fundamental that cannot be circumvented this way, then everybody who makes processors (and Nvidia builds GPUs which are specialized processors) infringes them somehow.
I remember my first Linux install. It was actually MkLinux on a Power Macintosh 9300 (or something close). Installing was _very_ easy - just uncompressing data on an empty filesystem.
It booted via an MacOS extension. I guess that made it so easy.
Still, it was pretty cool. I had to compile and install Afterstep by hand.
A year later, I came back to that company (a major ISP here in Brazil) as a consultant only to see just about all of those Macs converted into MkLinux boxes.
Well... I have set up the latest Ubuntu on 512 megs and it runs Gnome, Compiz, Firefox and OpenOffice well enough for my mother.
Try that with Vista.
Building a distro from scratch is a huge undertaking. The best would be to adopt one or two successful distros and build packages for them that work great in your hardware.
How much would I need to pay to work there?
It must be like to have Mozart hanging around with your band...
And why people who still can't tell the difference between free as in speech and free as in beer insist on writing for news sites?
Is the ad revenue so low sites can't hire knowledgeable folks?
If giving up revenue is the idea, why just not make it free of charge, or, better yet (at least for them) make people pay for an inexpensive lifetime unlimited upgrade and usage permission? No need for re-licensing of components acquired from other sources, no risk of letting people know they use GPL'ed code hidden deep inside stuff etc.
Quick, easy and it solves their problem.
I am quite sure the BSD code in Microsoft FTP.EXE program is not exactly free anymore.
"BSD licensed code is, in fact, free."
It's just that it doesn't remain so for long ;-)
Isn'z zlib LGPL?
"Sane" is not the opposite of "evil".