|
Wonderland of information superhighway
Information superhigh-way is a now-obsolete term that was popularised in the early 1990s by the Clinton-Gore administration to describe ways of expanding the internet beyond its then-current state. It is often associated with the U.S. politician and former Vice President of the United States, Al Gore, though the exact origins of the term are unknown.
Nam June Paik, a 20th century South Korean born American video artist, claims to have coined the term in 1974. "I used the term (information superhighway) in a study I wrote for the Rockefeller Foundation in 1974. I thought: if you create a highway, then people are going to invent cars. That's dialectics. If you create electronic highways, something has to happen."
The Superhighway Summit was held at UCLA's Royce Hall on 11 January 1994. It was the "first public conference bringing together all of the major industry, government and academic leaders in the field [and] also began the national dialogue about the Information Superhighway and its implications." The conference was organized by Richard Frank of the Academy of Television Arts and Sciences and Jeffrey Cole and Geoffrey Cowan, the former co-directors of UCLA's Center for Communication Policy. It was introduced by former UCLA Chancellor, Andrea L. Rich.
'We have a dream for...an information superhighway that can save lives, create jobs and give every American, young and old, the chance for the best education available to anyone, anywhere,' Gore said."
Geoffrey Cowan, the former co-director of UCLA's Center for Communication Policy, indicated that the key concept of the Information Superhighway was Interactivity or "the ability for the consumer to control it, to decide what they want to receive, and the ability of the technology to respond to highly sophisticated consumer demands."
The participants underscored the point that the major challenge of the Information Highway would lie in access or the "gap between those who will have access to it because they can afford to equip themselves with the latest electronic devices and those who can't."
The transition toward an Internet
The term "Internet" was adopted in the first RFC published on the TCP protocol (RFC 675: Internet Transmission Control Protocol, December 1974). It was around the time when ARPANET was interlinked with NSFNet, that the term Internet came into more general use, with "an internet" meaning any network using TCP/IP. "The Internet" came to mean a global and large network using TCP/IP. Previously "internet" and "internetwork" had been used interchangeably, and "internet protocol" had been used to refer to other networking systems such as Xerox Network Services.
As interest in wide spread networking grew and new applications for it arrived, the Internet's technologies spread throughout the rest of the world. TCP/IP's network-agnostic approach meant that it was easy to use any existing network infrastructure, such as the IPSS X.25 network, to carry Internet traffic. In 1984, University College London replaced its transatlantic satellite links with TCP/IP over IPSS.
Many sites unable to link directly to the Internet started to create simple gateways to allow transfer of e-mail, at that time the most important application. Sites which only had intermittent connections used UUCP or FidoNet and relied on the gateways between these networks and the Internet. Some gateway services went beyond simple e-mail peering, such as allowing access to FTP sites via UUCP or e-mail.
- Internet
 ^ TOP OF THIS PAGE ^ MAIN PAGE
Stable version of kernel comes
The current development kernel release is 2.3.35. The patch includes the usual large set of fixes (271 files changed), a driver for Quicknet Internet PhoneJACK and Internet LineJACK cards, more Sparc fixes, a reorganization of the Arcnet driver, lots of Adaptec SCSI driver changes, USB updates (including a new HID (human input device) driver), and a number of other changes.
The current stable kernel version is (finally!) 2.2.14. This release has been long in coming, and should be well received - it contains a lot of important and useful fixes.
The first 2.3.x "things to fix" list has been posted by Alan Cox. The list covers a lot of ground - one wonders how all of that stuff is going to get done anytime soon. But, of course, the posting of a list like this causes an immediate flood of additions... The most popular items which did not appear on Alan's list would appear to be:
32-bit UID support. Linus has said in the past that he wants to incorporate 32-bit user ID's, so this one may yet happen.
A journaling filesystem. None of the journaling filesystem alternatives are currently in a state that is ready for 2.4. Ext3 needs a fair amount of work on how it handles buffer memory before it can go in - too much for this release. Reiserfs, on the other hand, might just happen: Hans Reiser says a 2.3 port is close, and Linus has said that inclusion in 2.4 is a possibility. Meanwhile, XFS from SGI has still not been released (though they have begun to make small pieces available). Thus, the only journaling filesystem alternative for 2.4 is reiserfs, and that remains uncertain.
Version 3 NFS. Another major kernel release with a sub-standard NFS implementation would be a shame. There's been no word on whether the NFSv3 work that has been done will be merged or not.
Linus had wanted to get a pre-2.4 series going before the end of the year. Not only did that not happen, but it appears that it is still rather distant at this point. Some things can not be rushed; 2.4 will come out when it is ready.
Now that Y2K has wimped out, it's time to worry about the year 2038 problem. 2038, of course, is when the 32-bit time_t value that Unix systems use to represent times overflows. Some people want to try to deal with the problem now; others feel less urgency.
One point of view says that we'll all be using 64-bit systems by then; at some point we just redefine time_t to be a 64-bit value, recompile everything, and the problem goes away. There are, however, a couple of problems with that approach:
Old hardware has an amazing ability to hang around in crucial roles long past when it seems it should have been junked. That 32-bit Pentium doorstop in the corner of the machine room may still be doing something important when the rollover happens. Embedded applications, where size and power consumption are crucial, also tend to use less capable hardware for long periods of time.
Making time_t into a 64-bit quantity and recompiling everything will make a lot of programs work. But it will do nothing for all of the databases and file formats which contain 32-bit quantities. Quite a bit of application fixup work will be required to deal with all of those problems.
A reprieve could be had by simply redefining time_t to be an unsigned quantity - it would then be good until slightly after the beginning of the 22nd century. The only problem there is that some applications actually use negative time values to indicate times before the epoch.
The solution would seem to be to design a migration path now. With almost forty years in which to make things work correctly, one would assume the a reasonably painless transition could be made. In practice, many of us may well find ourselves being called out of retirement in 2037 to deal with the last-minute fixes...
A beta version of RealTime Linux V3.0 has been released. This version is based on the 2.3 kernel series, and does not (yet) contain much that is new at the API level. Note that RTLinux 2.x is still under active development as well...
A programming guide for Linux USB drivers has been released by Detlef Fliegl. It documents the structure of the Linux USB subsystem, and should be a valuable resource for those wanting to write USB drivers.
- Internet
^ TOP OF THIS PAGE ^ MAIN PAGE
Nokia 5700 Xpress adding a new
twist to mobile music
Nokia released the Nokia 5700 XpressMusic, an accessible 3G smartphone augmented by a dedicated audio chip for enhanced music performance. The Nokia 5700 XpressMusic features an iconic twist design that easily switches between four modes at the flick of the wrist - music player, 2 Megapixel camera, video call and smartphone. Combining the benefits of a music-oriented, multipurpose device featuring 3G dataspeeds, the Nokia 5700 XpressMusic is expected to retail for an estimated EUR 350 before taxes and subsidies during the second quarter of 2007. Smartphone 3G can download and play tracks from range of Internet music stores.
"The Nokia 5700 XpressMusic adds a new twist to the mobile music experience with compatibility for tracks purchased from many Internet music stores," says Heikki Norta, Senior Vice President, Mobile Phones, Nokia. "In addition the Nokia 5700 XpressMusic features a dedicated audio chip for improved music performance which consumers can further enhance by selecting from an ever increasing line of compatible audio enhancements."
Using an optional 2GB MicroSD card, the Nokia 5700 XpressMusic supports up to 1500 tracks that can be enjoyed using Bluetooth stereo headphones, with their own favorite 3.5mm plug-equipped headphones and the included headset adaptor or through built-in stereo loudspeakers. Loading songs onto the Nokia 5700 XpressMusic is fast and easy as consumers can choose from a broad range of supported digital formats, including WMA, MP3, AAC, as well as eAAC+ and MP4. When purchasing music online, shoppers again have the luxury of choice as they can select from a range of Internet music stores, also those utilizing Windows Media Player Digital Rights management (WMDRM). The optimized music player in the Nokia 5700 XpressMusic supports album art, playlists, plus a 5-band equalizer and audio visualizations.
3G and video
More than a music-only device, the Nokia 5700 XpressMusic brings videos to life on a vibrant color screen supporting up to 16 million colors for exceptional video playback. With 3G high-speed data connectivity, the Nokia 5700 XpressMusic makes browsing, downloading and streaming multimedia content much more convenient and faster than before. A video call feature and the built-in 2 Megapixel camera make it easy to share moments with friends and family.
The Nokia 5700 XpressMusic is based on the world's leading S60 software running on Symbian OS. S60 enables consumers to personalize their device with a wide choice of compatible applications that can be downloaded to the Nokia 5700 XpressMusic, including games, navigation, entertainment, productivity and creativity.
Nokia is a world leader in mobile communications, driving the growth and sustainability of the broader mobility industry. Nokia connects people to each other and the information that matters to them with easy-to-use and innovative products like mobile phones, devices and solutions for imaging, games, media and businesses. Nokia provides equipment, solutions and services for network operators and corporations.
- Internet
^ TOP OF THIS PAGE ^ MAIN PAGE
Linux: Driver support solves it
A few trends emerged from the comments and I want to address them directly in this and future posts. This post focuses on our strategy for device drivers. All of the practices described here are available today-you can enact them yourselves too.
At least half of the comments effectively said "we want Free Software, GPL-licensed drivers which are maintained in kernel.org, for all hardware in Dell systems." This request is not new to us-it's been our standard operating procedure for the last 8 years on PowerEdge servers, which today have no closed-source drivers necessary. For new Linux desktops and notebooks, we'll use drivers already in the mainline kernel.org kernels for as many components as possible. In these cases, the drivers will be included in your distribution of choice. This includes storage, wired networking, power management, USB, and more.
For device types where a choice exists between a component with a non-Free driver and one with Free driver availability, in our Linux offering we'll opt to bundle the component with the Free driver. Wireless network adapters is one such example; Printers are another. We recommend Linux users buy our printers which have PostScript engines in them, as opposed to those which don't and for which no Linux drivers are yet available. The Tech Specs tab for each printer on dell.com show if it has PostScript or not.
Some components, particularly some video cards, have working 2-D open source drivers, open source 3-D drivers actively being clean-room written by the community, and closed-source 3-D drivers available from the video card manufacturer. In these instances, while we continue to encourage the development (by all parties) towards open source drivers, we will provide the closed-source drivers for people who wish to use them.
The last category is devices for which no open source drivers are available at all, such as software-based modems. In our desktops these are add-in cards, so you can substitute a hardware-based modem available from your local electronics store quite easliy. However, we can't substitute hardware-based modems in our notebooks without redesigning and significantly increasing the price of the system. If it's important to you to have a hardware-based modem, you would add one into your PC Card or ExpressCard slot.
Dell recognizes the importance of open source, GPL-licensed drivers which are maintained upstream in kernel.org. They allow users the widest choice of Linux distributions, effectively taking the specific hardware and distribution out of the decision-making process and let you focus on solving your business problems. We will work with our hardware partners to develop, test, and maintain Free drivers, and continue to make progress towards that goal for all drivers. Most drivers are in good shape now, but there's clearly longer-term work to be done. Work that we're doing now at the driver level will pave the way for more Linux offerings in the future. There's no way to please everyone, but I'll continue to share more details around our strategy as we have new developments. Stay tuned.
Some components, particularly some video cards, have working 2-D open source drivers, open source 3-D drivers actively being clean-room written by the community, and closed-source 3-D drivers available from the video card manufacturer. In these instances, while we continue to encourage the development (by all parties) towards open source drivers, we will provide the closed-source drivers for people who wish to use them.
What message are you sending to graphics vendors who actively refuse to participate in open source? What incentive do they have to make the job of open source support easier than the inevitable reverse-engineering?
What work are you doing to ensure that the vesa fallback driver provides an acceptable experience? Many laptops usually do not provide the ability to set the resolution to that of the built-in panel using the vesa bios.
The last category is devices for which no open source drivers are available at all, such as software-based modems. In our desktops these are add-in cards, so you can substitute a hardware-based modem available from your local electronics store quite easliy. However, we can't substitute hardware-based modems in our notebooks without redesigning and significantly increasing the price of the system.
That's fine, we know how to write the necessary software to drive these devices, if we know where they are and how to talk to them. Do we have documentation for these devices? Will Dell work to facilitate getting this information to developers? Many times simply asking the IHV for the documentation as an independent developet gets you flat refusal or silence in response.
- Internet
^ TOP OF THIS PAGE ^ MAIN PAGE
|
