Um, there is no built-in BIOS support. There is a software loader (Bootcamp) which creates a bridge from the old BIOS entry points to the new EFI firmware...
Um, no, interface bandwidth is a serious issue on many printers, and will almost certainly be an issue for this printer as well - consider that a full-bleed A4 page has about 139 million addressable dots at 1200 DPI (which sounds like what this printer will provide). For 1-bit printer data with the typical packbits (RLE) compression, you'll need to push 8.2MB of data per page per color on average for photos. For a 60ppm printer, that translates into a required bandwidth of 8.2MB per second for grayscale output and 24-32MB/sec for color output!
Now, while the USB spec does support devices up to 480Mbits/sec, I've never seen a printer come close to supporting that speed - this printer would need to support approximately half the maximum USB bandwidth to provide 60ppm printing. Even the high-end Epson LFP printers can't pull that many bits over USB, which is why you need to use Firewire or the latest network cards to keep them printing...
Anyways, I'm more inclined to take that 60ppm figure with a large dose of salt, and figure that, like most other manufacturers, they mean 60ppm printing plain black text with no graphics/images. The engine may be capable of high speed, but the interface will be the bottleneck for getting data to the printer.
Not only do people point that out every day, but they fail to point out that Apple has a lot less to worry about when it comes to upgrades and their existing customer base. Didn't they basically break tons of apps with the 10.2 or 10.3 upgrade by switching GCC versions?
No, actually. Apple includes all of the older GCC libraries so you can run apps all the way back to 10.0. The GCC switch only affects developers, and they can choose which SDK to build against (10.2.x, 10.3.x, or 10.4.x in the current Xcode release, get an older Xcode to build for older versions of OS X).
The 6809 had an indexed + offset addressing mode (i.e. index register + offset to access memory, not just index register), plus it had two index registers instead of the 1 provided by the 6502.
I still have my old 6809 reference manual by Lance Leventhal - 8-bit multiply (producing a 16-bit result) was a mere 5 clock cycles, and a full 16-bit multiple could be implemented in about 25 clock cycles. Motorola's 68xx 8-bit CPUs were definitely the best ever made... It's too bad that they could never get the 680x0 series to run as fast as other 32-bit CPUs, as it was really easy to program, too...:(
So the RIAA can't get Apple to change their pricing, now they try to get laws enacted to invalidate their existing contracts with Apple and require them by law to change their pricing...
> I read this and I imagine Data saying "Captain. I have an idea. If we reverse > the tachyon coefficient to the digital anomoly drives, we can invert the neutrino > wave probe." And Patrick Stewart says, "Do it."
Actually, I believe the correct phrase is "make it so"...:)
In many cases, hardware vendors simply do not produce the kind of documentation that you would need. The hardware and software developers work together, so there is a lot of undocumented internal knowledge that goes into those drivers.
Also, in many cases there is technology or code that is licensed and *can't* be released as open source because the people that own it don't want it released.
Finally, even if the developers want to do open source, their management may not (yet) be on board with open source.
Given time and good experiences, I think most hardware vendors will provide open source drivers. We just need to encourage them to do this rather than beating them over the head with it!:)
Actually, you *can* configure things statically with CUPS - all local printers work this way, and you can add remote printers (lpadmin -p remotename -E -v ipp://remoteserver/printers/remotename) the same way. It is just that *normally* you can let CUPS figure out what printers and servers are on your network for you...
Since this would be a civil case, the standard is "the preponderence of the evidence". In any case, click-through EULAs are not always enforceable and, as you say, you can access the binaries *without* agreeing to the EULA or running them...
Both printer detection and printer driver lookup are getting a big boost in CUPS 1.2. That said, you won't get a popup saying "you just connected a HP DeskJet 3850, would you like to use it now?" on your desktop unless the Linux distributors do it...
Even with CUPS 1.1.x, the Linux distributors are not taking advantage of the full capabilities of CUPS... Maybe someday...
It isn't quite that simple. Drivers for non-PostScript printers require extra filter programs (at a minimum a raster filter) to print, and those filters are pretty much always binary programs without source code. Those drivers won't work on Linux, even on PPC.
That said, any PostScript printer and any non-PostScript printer with an open source driver for CUPS will work equally well on Linux, MacOS X, or any commercial UNIX (if you choose to install CUPS:)
Most of Eric's comments are NOT about CUPS, but instead about the various GUIs that have been written to run on top of CUPS.
Regarding the CUPS web interface, there is actually a LOT of development happening for the new CUPS 1.2 release to make things work much more smoothly, ask the user less questions when they don't need to be asked, and move the web interface to a more task-oriented UI instead of the current function-oriented UI.
For example, in the new web interface the "add printer" button will list any printers that CUPS discovers automatically ("Epson Stylus RX300 on USB port") - you just click on "add listed printers" to add the printers, or "add printer manually" to add one manually. Similarly, printer sharing, remote administration, etc. are now check boxes on the administration page instead of going through the cupsd.conf file.
Anyways, good changes ARE coming for the native CUPS interfaces, and I only hope that the Linux distributors follow suit with their GUIs...
I didn't see this on the UserLinux package lists, but are you planning on including the LSB stuff so that LSB packages (RPM format, not my favorite by far but makes it easier to package software for multiple distributions) can be easily installed on UserLinux?
I did some quick charting against US population and the year; the peak time (so far) for new patents was from 1999 to 2002 with 166,667 per year (FWIW, we are almost there in 2004)
The rate of population increase since 1900 has averaged about 1.37% per year. Patents have increased at more than double that rate at 2.73% per year.
Furthermore, the population growth appears to be slowing while the patent growth is speeding up.
I would guess that this is the result of a lot more businesses getting patents for trivial inventions (or getting separate patents for different parts of their inventions) as well as the dominance of larger businesses which can afford to have full-time patent filing departments - i.e. patents are one more "product" that a company produces.
Slashdot Mirror
You make it a co-op that is owned by the customers...
Um, there is no built-in BIOS support. There is a software loader (Bootcamp) which creates a bridge from the old BIOS entry points to the new EFI firmware...
FLTK is not OpenGL-based, it uses X11 just like any other toolkit.
Don't confuse support for OpenGL with dependence on OpenGL...
Um, CUPS is GPL2/LGPL2. Apple used CUPS for 5 years before they bought it...
Um, no, interface bandwidth is a serious issue on many printers, and will almost certainly be an issue for this printer as well - consider that a full-bleed A4 page has about 139 million addressable dots at 1200 DPI (which sounds like what this printer will provide). For 1-bit printer data with the typical packbits (RLE) compression, you'll need to push 8.2MB of data per page per color on average for photos. For a 60ppm printer, that translates into a required bandwidth of 8.2MB per second for grayscale output and 24-32MB/sec for color output!
Now, while the USB spec does support devices up to 480Mbits/sec, I've never seen a printer come close to supporting that speed - this printer would need to support approximately half the maximum USB bandwidth to provide 60ppm printing. Even the high-end Epson LFP printers can't pull that many bits over USB, which is why you need to use Firewire or the latest network cards to keep them printing...
Anyways, I'm more inclined to take that 60ppm figure with a large dose of salt, and figure that, like most other manufacturers, they mean 60ppm printing plain black text with no graphics/images. The engine may be capable of high speed, but the interface will be the bottleneck for getting data to the printer.
No, actually. Apple includes all of the older GCC libraries so you can run apps all the way back to 10.0. The GCC switch only affects developers, and they can choose which SDK to build against (10.2.x, 10.3.x, or 10.4.x in the current Xcode release, get an older Xcode to build for older versions of OS X).
The 6809 had an indexed + offset addressing mode (i.e. index register + offset to access memory, not just index register), plus it had two index registers instead of the 1 provided by the 6502.
I still have my old 6809 reference manual by Lance Leventhal - 8-bit multiply (producing a 16-bit result) was a mere 5 clock cycles, and a full 16-bit multiple could be implemented in about 25 clock cycles. Motorola's 68xx 8-bit CPUs were definitely the best ever made... It's too bad that they could never get the 680x0 series to run as fast as other 32-bit CPUs, as it was really easy to program, too... :(
> Now, driving your CAR supports mideast oil barons. Easy solution...drill
a de_and_investment/energyrel050328-en.asp
> off the East and West Coast as well as in Alaska.
Funny, the US gets more of its oil from Canada than Saudi Arabia, and the trend is only increasing:
http://www.dfait-maeci.gc.ca/can-am/washington/tr
http://www.eia.doe.gov/emeu/cabs/canada.html
So the RIAA can't get Apple to change their pricing, now they try to get laws enacted to invalidate their existing contracts with Apple and require them by law to change their pricing...
> I read this and I imagine Data saying "Captain. I have an idea. If we reverse
:)
> the tachyon coefficient to the digital anomoly drives, we can invert the neutrino
> wave probe." And Patrick Stewart says, "Do it."
Actually, I believe the correct phrase is "make it so"...
In many cases, hardware vendors simply do not produce the kind of documentation that you would need. The hardware and software developers work together, so there is a lot of undocumented internal knowledge that goes into those drivers.
:)
Also, in many cases there is technology or code that is licensed and *can't* be released as open source because the people that own it don't want it released.
Finally, even if the developers want to do open source, their management may not (yet) be on board with open source.
Given time and good experiences, I think most hardware vendors will provide open source drivers. We just need to encourage them to do this rather than beating them over the head with it!
Actually, you *can* configure things statically with CUPS - all local printers work this way, and you can add remote printers (lpadmin -p remotename -E -v ipp://remoteserver/printers/remotename) the same way. It is just that *normally* you can let CUPS figure out what printers and servers are on your network for you...
MySQL as SCO! :)
Since this would be a civil case, the standard is "the preponderence of the evidence". In any case, click-through EULAs are not always enforceable and, as you say, you can access the binaries *without* agreeing to the EULA or running them...
Even try zooming in on the moon at google? I *knew* the moon was made of cheese! :)
That's why the new CUPS 1.2 stuff won't bother asking you for that info... :)
Both printer detection and printer driver lookup are getting a big boost in CUPS 1.2. That said, you won't get a popup saying "you just connected a HP DeskJet 3850, would you like to use it now?" on your desktop unless the Linux distributors do it...
Even with CUPS 1.1.x, the Linux distributors are not taking advantage of the full capabilities of CUPS... Maybe someday...
It isn't quite that simple. Drivers for non-PostScript printers require extra filter programs (at a minimum a raster filter) to print, and those filters are pretty much always binary programs without source code. Those drivers won't work on Linux, even on PPC.
:)
That said, any PostScript printer and any non-PostScript printer with an open source driver for CUPS will work equally well on Linux, MacOS X, or any commercial UNIX (if you choose to install CUPS
Most of Eric's comments are NOT about CUPS, but instead about the various GUIs that have been written to run on top of CUPS.
Regarding the CUPS web interface, there is actually a LOT of development happening for the new CUPS 1.2 release to make things work much more smoothly, ask the user less questions when they don't need to be asked, and move the web interface to a more task-oriented UI instead of the current function-oriented UI.
For example, in the new web interface the "add printer" button will list any printers that CUPS discovers automatically ("Epson Stylus RX300 on USB port") - you just click on "add listed printers" to add the printers, or "add printer manually" to add one manually. Similarly, printer sharing, remote administration, etc. are now check boxes on the administration page instead of going through the cupsd.conf file.
Anyways, good changes ARE coming for the native CUPS interfaces, and I only hope that the Linux distributors follow suit with their GUIs...
Indiana Jones and the Search for the Lost Social Security
My flPhoto application does all but the last (it does support local printing, of course), available at:
http://www.easysw.com/~mike/flPhoto/That's what we used to have in Maryland, but now we have Diebold machines... :(
I didn't see this on the UserLinux package lists, but are you planning on including the LSB stuff so that LSB packages (RPM format, not my favorite by far but makes it easier to package software for multiple distributions) can be easily installed on UserLinux?
I did some quick charting against US population and the year; the peak time (so far) for new patents was from 1999 to 2002 with 166,667 per year (FWIW, we are almost there in 2004)
The rate of population increase since 1900 has averaged about 1.37% per year. Patents have increased at more than double that rate at 2.73% per year.
Furthermore, the population growth appears to be slowing while the patent growth is speeding up.
I would guess that this is the result of a lot more businesses getting patents for trivial inventions (or getting separate patents for different parts of their inventions) as well as the dominance of larger businesses which can afford to have full-time patent filing departments - i.e. patents are one more "product" that a company produces.