If you're going to plug books on/. then the least you could do is buy it, read it and post a proper review instead of a link to the advertising summary.
You can't make two phonecalls at the same time to two different people. Also, you're charged for the calls you make (as well as the fixed line charge!)
If you have multiple people sharing an IP address, then they can all use the connection at the same time. They all can't get maximum bandwidth, but email & web rarely require maximum bandwidth anyway unless you're downloading something large.
No one is denying Monty Python has greatly influenced culture, but you can't reasonably claim that they have had a particularly noticeable effect on the development of technology.
Any child who is thought to be at risk of committing a crime by the police, schools or social services, will be put on the database
Later on
street gangs provided a safer and more caring environment than their homes or classrooms
1) Is it necessary to treat all children as potential criminals because some are in gangs?
2) If these children are safer in gangs than at home or in the classroom then - instead of putting their names into a huge database - wouldn't it be better to take them into care where they will be safe?
Using the Rio Receiver (and other unimportant info)
Chris Uriarte, chrisjur@cju.com
Last Updated: November 24, 2001
Warning: I'm a very technical person AND an audiophile. Therefore, is only natural that I'd be very critical on any piece of technology that produces any type of sound. Despite my criticisms, be sure to read my final thoughts and recommendations at the end of this review.
What is it?
The RIO receiver is a small devices that allows you to play your digital MP3 and WMA music through any stereo system. It attaches to an Ethernet network or a HPNA (phone line) network, allowing it to access digital music files stored on a Windows PC. It's manufactured and sold by SonicBlue [http://www.sonicblue.com], the people who bought the Rio [http://www.riohome.com] line of digital music equipment from Diamond. The same device is re-branded and sold by Dell [http://www.dell.com] as the Dell Digital Audio Receiver The street price for the unit is about US$200. You can find out more about it at its website [http://www.riohome.com/products/receiver.htm]. Here's a picture for your viewing enjoyment:
What are the tech specs?
Without going into the nitty gritty details, here's some tech specs on the unit:
Unit measures 9"x9"x3"; contains small LCD screen, a few buttons and a jog dial on the front.
74MHz ARM7 Cirrus processor
Runs a tiny version of Debian Linux
4MB of DRAM
512KB of flash memory
CS8900 based 10MB ethernet interface
Broadcom HPNA interface. HPNA is some ass-backwards networking setup that allows you to run data over your home phone lines - be afraid, be very afraid!
One set of standard RCA audio output jacks
One set of standard speaker connectors
Why buy it?
The advantage of purchasing the unit is that it gives you access to your digital audio collection through any stereo in your house that's close enough to a phone line or ethernet port on your home network.
Why I bought it.
Let me set the stage by saying that I have over 500 CDs and that I am addicted to music. One day, I realized something important: my music collection was taking over my life. It was getting too large to handle and my fickle nature makes me want to listen to many different types of music in a relatively short period of time. CDs tend to get stale very quickly when they are in my CD player. After having, what alcoholics sometimes refer to as, "a moment of clarity", I realized that it was time to get my act together. I discovered that my car contained 70 CDs - it was a $30,000 CD holder that happened to also be able to get me back and forth to work. There were about 100 CDs laying around in my home office and about 30 in my office at work. There were numerous CDs stuck in changers and players and CD holder and travel pouches and attache bags and just about everywhere I could put them. I was out of control. My girlfriend was about to stage an intervention when I told her that I was going to buy a third CaseLogic car visor CD holder - which I somehow thought I could rig into place, despite the fact that my car, like most others, only has two sun visors. It was bad, and I needed help. One day, I had a vision - If I made my music collection digital, I could access it from anywhere - my stereo, my car (through a car MP3 player), on the road (through a portable MP3 player), at the office, etc. I really hated my trusty Sony CD 200-disc changer, which I know has swallowed a number of CDs that I've been looking for the last 5 years. I had a dream about being able to "dial up" my music - without having to know which slot it lived in or figuring whether it was even in my house at all. I thought the Rio Receiver could help me fill these dreams. (I know, It's sad. One day I'll seek professional help.)
Where to buy it.
When I was first searching for it (November 2001), the unit was somewhat difficult to find. There were only about five online retailers carrying it, and most were out of stock. I settled on the dependable, but expensive, Crutchfield [http://www.crutchfied.com], where I purchased it for $179. Note that I have seen street prices as low as $156. Check out CNET's price comparison service [http://shopper.cnet.com/] to try to hunt one down.
What do you get with it?
The box it's packaged in is roughly the size of a toaster over. Open it up and you'll find:
the Digital Audio Receiver itself
Home PNA adapter card - (Note: It appears that some models do no include the PNA adapter. That's good...avoid it at all costs)
Remote Control
2 "AA" batteries
one 6" RCA stereo patch cord
one 6'6" telephone cord (probably only included if you get the HPNA card)
one 12' 6" telephone cord (ditto)
one 6' 6" AC power cable
one CD containing the necessary server software installation and owner's manuals
lots of cardboard, styrofoam, plastic, foam and other things that kill the environment. Crutchfield is also sure to pack the receiver box in a larger box small enough to hold a child - however, the child would never survive after suffocating from the grotesque amount of styrofoam peanuts inside.
It comes with a "quick start" sheet, which is good enough to actually get it up and running pretty quickly. The full documentation is on the CD-ROM, or available on the Rio website in the Support section. However, the full documentation is scarce on juicy details, although worth a quick read.
Setting it Up
Assuming that you already have an Ethernet network in place, all you need to do is the following:
1. Install the Windows software. Reboot (of course). Run the software and specify drives and directories that contain your MP3 and WMA files. Sit back for a minute and let it catalog your files.
2. Plug the unit in. Hook up the RCA jacks to your stereo. Plug Ethernet cable in the back. Turn the unit on.
3. Select your artist/album/playlist from the LCD screen and start playing music.
The whole setup process literally took me less than 10 minutes, most of which was spent chasing runaway styrofoam peanuts across my hardwood floors.
What does the Windows Software Do?
Yes, you must - I repeat - you MUST use the Windows software for this thing to work (see notes on the Linux hack below, however). The software is not exactly what I would call "feature filled". The software has two purposes in life:
1. Catalog the digital music files on your computer system.
2. Communicate with the Digital Audio Player.
That's it. And you don't have options to play with those two either. It is truly as if SonicBlue had a meeting and said, "What is the absolute, minimal amount of work we can do to make this piece of software work?". It then appears that they took that list of minimums and cut it in half. It is barebones, to say the least.
Those who own other Rio products may be use to getting some type of CD ripping or digital file management software with players. This, however, doesn't include anything like that.
The Windows software, dubbed the "Audio Receiver Manager", doesn't actually allow you to manage anything. It does manage to discretely sit in your system tray and is loaded when you startup and log into windows. Since I always keep my main computer on and logged in, it's not a big deal to me, however, the audio manager will not start until you log into windows (i.e. enter your username password, if you get prompted for one). I'd like to see this made into a Windows NT/2000 service for the next release.
Basically, you tell the software where on your system to look for MP3/WMA/M3U files. You have the ability to specify entire drives, which can be very slow, or individual folders. You must update the database every time you want to make new music available to your receiver. During the import process, the Audio Manager must stop playing music - apparently it has difficulty walking and chewing gum at the same time.
Here's some screenshots of the Windows program:
(Click for larger images)
A Note on Image 2 above: The Audio Manager shows that it imported about 1500 tracks and 120 playlists in 16 seconds. This particular import was run on a folder of MP3s that was already imported, so it ran rather quickly. When I first ran the import on this folder, it actually took close to 40 seconds to import all the tracks (not bad at all). However, this was done on a rather speedy Athlon 1.33Ghz with 768MB RAM and 7200 RPM drives. Your mileage may vary. The Rio documentation, for example, shows a screenshot of an import that was smaller, which took a considerably longer amount of time.
Things that particularly suck about the Windows-based Audio Receiver Manager software:
It is pretty much un-configurable. Choose your DHCP range and choose your folders to scan, that's it.
There's no sort of "preferences" available to save your settings, which would be nice. Even though I only search for MP3s under one directory (D:\MP3, for example), it always defaults to search the entire C: and D: drives. I then have to remove and re-select D:\MP3 EVERY TIME. This is maddening.
Somehow, SonicBlue thought it would be helpful to have this thing search your MusicMatch and RealJukebox playlists - I can certainly see the value to this feature. However, you don't have the ability to disable searching these programs and the internal track lists and playlists are imported from these programs every time you update your database. This causes duplicate track and artist names to appear on your Receiver. For example, if you tell the receiver to search D:\MP3 for your of your tracks and you have previously imported the track names into Real Jukebox, all your tracks will be duplicated when you bring them up on the Receiver's LCD. This is fine if you want to "manage" all your music from RealJukebox, but I don't. Even after removing the tracks and playlists from RealJukebox, there were still dupes....I finally just uninstalled RealJukebox (it's a piece of junk, anyway). If you use MusicMatch, it will only import your MusicMatch playlists, which can be handy at times - but it's all or nothing....if you create playlists in MusicMatch, they will ALL show up on the Receiver, as you don't have any choice to exclude specific playlists.
As mentioned before, the software sits in the system tray and is launched from your Startup folder. Therefore, if your machine reboots or you log out of Windows 2000/NT/XP, the Audio Manager software is not running - so no music for you.
Playing your Music From the Receiver
When it comes down to doing what it is meant to do, the Rio Receiver does a very good job. It's very easy to navigate through your music collection, even if you have a large number of tracks, artists and albums. I have close to 150 albums worth of digital music and I can easily find what I'm looking for in a few seconds.
Music selection is done using a simple jog-dial knob on the front of the unit. You have the ability to search for music by artist, album, song title, genre or m3U playlist. All this information is determined by the ID3 tags [http://www.id3.org] contained within your MP3 files - therefore, it's VERY important that all of your MP3 files are embedded with at least the artist, album and song title. (See my "Tips and Tricks" below for details on this). The Rio DOES NOT CARE about the organization of your directory structure - all it wants to know is the contents of your ID3 tags.
Overall Quality and Performance
Overall, the sound quality is very good. Of course, sound quality greatly depends on the bitrate used to encode your digital music files. I encode all my MP3s at 192 bits using LAME [http://www.mp3dev.org/mp3/] - the quality is excellent. The Receiver just takes a digital stream of the music files over the network and decodes it, so there is no loss in quality due to the network transfer or any weird types of streaming protocols. At 192bits, no one can tell that the music is sourced from am MP3 file.
The unit has a 10-second buffer, just in case there are any network slowdowns. I wanted to figure out just what it would take to make the thing stop or skip or do whatever it may do if the network gets really busy. So I wrote a simple Perl script to transfer a 200MB file between my Windows box and my Linux box over HTTP using 15 simultaneous threads (i.e. 15 downloads occuring at the same time). This is, of course, way more than what I normally see on my home LAN, which is usually just comprised of me and my girlfriend surfing the web (mostly me, of course). I should note that my home network uses a NetGear 10/100 Hub (not a switch) and a Linksys 10MB in the Living room - so the effective throughput will never exceed 10MB (not including protocol overhead). With 2 Skinny J's The Whammy playing in the background, I started the network torture test. Despite the fact that my Linux box's four-year-old hard drive protested, the Rio Receiver didn't skip a beat (no pun intended). So I think you can be pretty sure that this thing will work well, even in a busier network environment.
Did you say it runs Linux?
Yes, it runs Linux. Many people tend to get very aroused about this these days. I am a Linux advocate, as I like to advocate any technology that helps you work or play better, but I'm not a Linux zealot. But for the more curious, the unit has already been hacked, although it appears to be a bit more difficult to do than it should be. Jeff Mock has put together a wonderful collection of information [http://www.mock.com/receiver/] related to hacking the Receiver. Even though I was greatly tempted see what you can do with the thing, I figure I bought it to do what it does best - play music. I'm happy with its feature set and don't feel the need to crack it open. If this unit gets more popular and there are a number of good hacks available out there, you may see more people trying ot hack it.
What Operating Systems Does it Work With? (Do I have to use Windows?)
The Rio Audio Receiver Manager software officially supports Windows98, 98SE, 2000 and ME. Although not listed, I'm sure it also works on XP, as the application nothing more than simple HTTP and UDP server. It will not run on Windows95 or WindowsNT 4.0.
There are a few efforts underway to make the unit work with non-Windows servers. The first is a home-grown project put together by Jeff Mock [http://www.mock.com/receiver]. Jeff has done an excellent job at reverse engineering the Rio's protocols and has put together a number of scripts that allow you to use a Linux machine as the file server. At this time, his software does not support playlists, but after looking at the code and the protocols, it is probably an easy thing to add. Be aware, though, installation of Jeff's code is not a trivial task - you need to have knowledge about Linux, Apache and Perl and there's not a lot of "HOW-TO" documentation available (who has time to document, anyway?...especially for open-source projects that don't pay the bills?). It took me about 40 minutes, but I got Jeff's scripts running. Again, he did a great job with this.
Second, there's an open source project called Digital Audio Server [http://sourceforge.net/projects/das/], which I have not tried out. It looks as if it might be a bit more complete and simpler to use than Jeff's solution.
There's a third project called Jreceiver [http://sourceforge.net/projects/jreceiver/], which is open source servlet-based audio server built to support network-based MP3 players like the Rio. However, it appears to be quite immature at this time. It's completely servlet-based, but does not support Tomcat, which is a major drawback, in my opinion.
Are there competitors?
Yes, in fact there is one generally-available competitor that I am aware of: The Turtle Beach Audiotron [http://www.audiotron.net/audiotron.asp]. I like this particular product a lot. Had it not been for the Rio's attractive $150 street price, I probably would have bought the Audiotron. There are a few major differences in the way that these boxes operates. First, the Audiotron is a full-size (but thin) stereo rack component, so it can blend nicely into your existing system. Second, the Audiotron accesses music over a simple SMB share, so you can simply export a folder from your Windows or UNIX box and make it available to the Audiotron. There's no server-side software required. Third, you can actually READ the LCD display on the Audiotron from a distance greater than 10 inches. Forth, the Audiotron has a TOSLINK optical output, which is nice to have. Last, but not least, the Audiotron comes with a suite of useful software, including a web-based front-end. The street price for the Audiotron, however, is about double the price of the Rio. You can read Rob Malda's review of the Audiotron on Slashdot [http://slashdot.org/article.pl?sid=01/10/12/14232 33].
Tips, Tricks and Things to Be Aware Of
Use ID3 Tags - as mentioned before, the Rio gets all it's track, artist and album info from the ID3 tags embedded in the MP3 file, so make sure you ID3 tag your MP3s. The easiest way to ID3 tag your MP3s is to do it when you rip them from the CD using a good ripping program with CDDB/FreeDB integration (Check out Audiograbber [http://www.audiograbber.com-us.net]). If have existing MP3s that are not tagged, MusicMatch [http://www.musicmatch.com] does a good tagging job for no cost. Musicmatch will also identify all the MP3s on your system that are not tagged.
Use RealJukebox or simple file system management - the Rio Audio Receiver Manager WILL import all tracks and playlists you have setup in RealJukebox and you don't have any say over it. Therefore, you should manage your MP3s through RealJukebox or not use RealJukebox at all. If you search for MP3s on your hard drives AND have a RealJukebox tracklist built, you will get duplicate track entries when you bootup the Rio.
Listening to Full Albums - There's only one way to listen to full albums the way they were intended to be heard on the CD: create an m3u playlist for the album. If you simply choose "Albums" through the Rio controls, it will list out the unique album names on the system (without the artist name attached to it). If you choose an album, the Rio plays through tracks in the album in alphabetical order - not in the order they appear on the CD. The solution is to create an m3u playlist for each album and then access the album through the "Playlists" option on the Receiver. Again, a good CD ripping program should create these playlists for you when the rip is complete.
Album Playlists help if they include Artist Names - Some CD ripping programs will create m3u playlists and name the file after the album you just ripped. The Rio displays playlists according to their file system name. Therefore, if you don't have the artist name in the playlist file name, you may have a tough time identifying and finding the album playlist your are looking for. The solution is to name your playlists in some format like "Artist-Album.m3u". Sadly, my favorite software, Audiograbber, does not let you configure the file name format of the playlists. So I wrote a Perl script that goes through your MP3 directories and renames the playlists to "Artist-Album.m3u" format IF you organize your MP3's in the following fashion:
+/MP3-Root-Directory
+/Artist-Name
+/Album-Name
/Track.mp3
/Track.mp3
/Track.mp3
etc.
I think this is the typical way most people organize their MP3's, so the Perl script may be helpful. Also, since I think Playlists are so important, I wrote a Perl script that hunts out any album directories that DO NOT contain playlists and prints them out for you. (BTW, if you didn't know this already, you need to install Perl [http://www.cpan.org] on your machine for these to work).
The LCD Screen is Small - You need to be pretty close to the unit to read the LCD, so choosing tracks from a distance is impossible. This used to piss me off, but I figured that with the old way, I would have to dig out another CD anyway (and chances were that CD I was looking for was MIA).
Tiny Gaps Between Songs - the Rio inserts an every-so-tiny gap between tracks...so small you don't really notice it. I consider this a good thing. For live albums and albums with songs that continue onto the next track without a gap, everything sounds like it did on the CD...the Rio does an excellent job with this. I tested Jerry Seinfeld's I'm Telling You For the Last Time (a live comedy album), 311's Live album (a live concert album) and Rx Bandit's Progress (a traditional studio album where the first 5 tracks flow into the next track) - all of them sounded great.
Final Thoughts and Impressions
At $150 street price, this thing is a no-brainer. Despite some of my gripes, it does what its supposed to do and does it well. I highly recommend it. I no longer require a CD player in my den. Everything gets ripped to MP3 and then played on the Rio. If you already have a home network in place and a lot of digital music, it's pretty much just plug-and-play. I've already enjoyed using the Rio for the short time that I've had it and I'm sure that I will get much more enjoyment out of it in the future. Good job, SonicBlue.
For the tech weenies:
-----
How does it work? (Warning: very useless technical information follows...you may want to stop here.)
Given my particular background, I was more interested in how the Receiver actually gets music from the central PC. After further analysis, I became convinced that the great Rube Goldberg was the chief architect of Rio. I thought the whole thing could have been made a lot simpler, especially compared with the file-sharing simplicity of the Turtle Beach Audiotron (see "Competitors" above).
So I set out to determine the protocols used between the Receiver and the Windows server. I figured that if I could reverse engineer these protocols, I could write my own server software to circumvent the lack of features in Rio version - including getting the server to run on Linux. So I started running Etherpeek sniffs of my network and began analyzing the traffic flow between the Receiver and my PC. I was about a half hour into it when I first discovered Jeff Mock's great "Hacking the Rio Receiver" page [http://www.mock.com/receiver]. Jeff did a far better job that I did in my first 30 minutes, so I used his notes to compare what I sniffed (btw, Jeff, your 100% right on...awesome work). The result of Jeff's sniffing efforts produced a set of scripts that allow you to run a Linux version of the server software (available for download on his site).
Basically, the protocol is simple, but there's a lot of little steps to actually get to the point where the music starts playing. Here's a quick overview:
When the box boots, it sends out a DHCP request ("please, someone give me an IP address"). The Rio Audio Manager software on the PC responds to DHCP requests only for the RIO. This is based totally on MAC address, as I was able to send a spoofed DHCP request using the Rio's MAC from my Linux box. FYI, on the Windows side, I always limit the DHCP IP scope to exactly 1 address, which essentially gives the Rio a static IP. Jeff relies on the OS to do the DHCP stuff - not his server software. You can, however, configure dhcpd to always dish out the same IP to a single MAC address, if you so desire.
After the Receiver is happy with it's IP address, it sends out an SSDP request on UDP 21075 ("someone please tell me where my server is") with the request string "^upnp:uuid:1D274DB0-F053-11d3-BF72-0050DA689B2F\n {Source MAC Address}\n". The Audio Manager responds with the IP address and port of the Windows server. It tends to use TCP Port 12078 on the server side. It actually sends back an URL to an.xml file (http://serverip:12768/descriptor.xml)...the important part is the port 12768, which will be used as the status port between the receiver and server.
The receiver actually does an NFS mount of/tftpboot on the server and grabs a basic Linux file system - sort of like an XTerm. It then reboots a second time, this time bootstrapping with the downloaded kernel. Jeff points out something interesting that I initially missed: the Receiver does an NFS mount of/tftpboot - it does NOT use TFTP to download the boot image from/tftpboot, as you would normally see with diskless clients.
When you first scroll through a list - such as "Artists", the box issues a HTTP query to the app port specified in the SDDP response that looks something like: "GET/query?artist=". The server responds with an HTTP response with a standard HTTP 200 header, Content-type, Content-length, etc. a a response string that looks like "ArtistA.1=61,0,0:ArtistB.1=32,0,0......etc." You can start to see that the software assigns a unique index number to each item in the list, starting with 1, assigned in alphabetical order. Once you select an artist, for example, it will then make a subsequent query for the tracks associated with that artist ("http://10.0.0.25:12078/results?_extended=1&artis t=Artist" ). The server returns a similar list of tracks, each also having a unique track ID number. Once a track is selected, using its unique ID number, the Receiver asks for the tags associated with the mp3 file ("http://10.0.0.25:12078/tags/240", for example).
From that point on, it's time to actually play the music. This accomplished by streaming the data over HTTP, gracefully shooting a flow of TCP PSH's and ACK's across your network. The receiver requests the unique track id over HTTP a la http://10.0.0.25:12078/content/240.
While alive and active, the Receiver and the server software start to miss each other and send UDP status packets out to each other about every five seconds. Basically, the server says "State? (What's going down?)" and the receiver replies with it's current state, which includes its MAC address, the current song it's playing and an associated timecode.
It should be noted that once you sniff the format of these queries, you can simply just insert them into any old web browser to see the results the server sends back. Type in "http://10.0.0.25:12078/query?artist=", for example, and you'll get a list of all the artists in the database. Doing something like "http://10.0.0.25:12078/query?artist=c" returns all artists starting with C - this might be useful in building some type of music search engine or something. Doing something like "http://10.0.0.25:12078/results?_extended=1&artist =Ash" will return all songs associated with the artist Ash. To see all playlists, it sends "http://10.0.0.25:12078/content/100?_extended=1" and to view a specific playlist, it sends "http://10.0.0.25:12078/content/5e90?_extended=1", where 5e90 represents the unique playlist ID it was sent when it queried all the playlists. You get the drift....this isn't exactly rocket science. Given this simple protocol, you can easily write a browser-based frontend to the Windows-based Audio manager software.
I figured that if I wanted to write a web-based front-end for the Audio Manager software, I would need to figure out which port the server is running on. So I created a trivial perl script that mimics the Receiver, sending out a direct SSDP query to the Windows server, which is happy to return the all-important port number that the server software happens to be running on. Once you have that port, you can issue any of the HTTP queries above via a regular web browser or with any piece of code you may just create. Interesting note: by sending out the SSDP request from my Linux box, the Linux box now shows up as a "New Audio Receiver" in the Audio Receiver Software. This causes the server software to send my Linux box a UDP "status" query every five seconds (too bad my Linux box isn't answering).
Lots more that can be done with this...if you have the time and can think up a good use for it.
As a side not, if you're looking for a front-end to your MP3 collection and you're using Apache, take a look at Lincoln Steins's Apache::MP3 [http://www.modperl.com/Songs/], for use with mod_perl. I use it and love it. It gets me access to my entire music collection anywhere that I have an Internet connection.
THEY WERE SUCH TINY DOTS, YET THEY HELD SUCH immense promise. After months of trying, on October 13, 2001, we came into our laboratory at Advanced Cell Technology to see under the microscope what we'd been striving for--little balls of dividing cells not even visible to the naked eye. Insignificant as they appeared, the specks were precious because they were, to our knowledge, the first human embryos produced using the technique of nuclear transplantation, otherwise known as cloning.
With a little luck, we hoped to coax the early embryos to divide into hollow spheres of 100 or so cells called blastocysts. We intended to isolate human stem cells from the blastocysts to serve as the starter stock for growing replacement nerve, muscle and other tissues that might one day be used to treat patients with a variety of diseases. Unfortunately, only one of the embryos progressed to the six-cell stage, at which point it stopped dividing. In a similar experiment, however, we succeeded in prompting human eggs--on their own, with no sperm to fertilize them--to develop parthenogenetically into blastocysts. We believe that together these achievements, the details of which we reported November 25 in the online journal e-biomed: The Journal of Regenerative Medicine, represent the dawn of a new age in medicine by demonstrating that the goal of therapeutic cloning is within reach.
Therapeutic cloning--which seeks, for example, to use the genetic material from patients' own cells to generate pancreatic islets to treat diabetes or nerve cells to repair damaged spinal cords--is distinct from reproductive cloning, which aims to implant a cloned embryo into a woman's uterus leading to the birth of a cloned baby. We believe that reproductive cloning has potential risks to both mother and fetus that make it unwarranted at this time, and we support a restriction on cloning for reproductive purposes until the safety and ethical issues surrounding it are resolved.
Disturbingly, the proponents of reproductive cloning [see Reproductive Cloning: They Want to Make a Baby] are trying to co-opt the term "therapeutic cloning" by claiming that employing cloning techniques to create a child for a couple who cannot conceive through any other means treats the disorder of infertility. We object to this usage and feel that calling such a procedure "therapeutic" yields only confusion.
What We Did
WE LAUNCHED OUR ATTEMPT to create a cloned human embryo in early 2001. We began by consulting our ethics advisory board, a panel of independent ethicists, lawyers, fertility specialists and counselors that we had assembled in 1999 to guide the company's research efforts on an ongoing basis. Under the chairmanship of Ronald M. Green, director of the Ethics Institute at Dartmouth College, the board considered five key issues [see The Ethical Considerations] before recommending that we go ahead.
THERAPEUTIC CLONING: HOW IT'S DONE
The next step was to recruit women willing to contribute eggs to be used in the cloning procedure and also collect cells from individuals to be cloned (the donors). The cloning process appears simple, but success depends on many small factors, some of which we do not yet understand. In the basic nuclear transfer technique, scientists use an extremely fine needle to suck the genetic material from a mature egg. They then inject the nucleus of the donor cell (or sometimes a whole cell) into the enucleated egg and incubate it under special conditions that prompt it to divide and grow [see Therapeutic Cloning: How It's Done].
We found women willing to contribute eggs on an anonymous basis for use in our research by placing advertisements in publications in the Boston area. We accepted women only between the ages of 24 and 32 who had at least one child. Interestingly, our proposal appealed to a different subset of women than those who might otherwise contribute eggs to infertile couples for use in in vitro fertilization. The women who responded to our ads were motivated to give their eggs for research, but many would not have been interested in having their eggs used to generate a child they would never see. (The donors were recruited and the eggs were collected by a team led by Ann A. Kiessling-Cooper of Duncan Holly Biomedical in Somerville, Mass. Kiessling was also part of the deliberations concerning ethical issues related to the egg contributors.)
We asked potential egg contributors to submit to psychological and physical tests, including screening for infectious diseases, to ensure that the women were healthy and that contributing eggs would not adversely affect them. We ended up with 12 women who were good candidates to contribute eggs. In the meantime, we took skin biopsies from several other anonymous individuals to isolate cells called fibroblasts for use in the cloning procedure. Our group of fibroblast donors includes people of varying ages who are generally healthy or who have a disorder such as diabetes or spinal cord injury--the kinds of people likely to benefit from therapeutic cloning.
Our first cloning attempt occurred last July. The timing of each attempt depended on the menstrual cycles of the women who contributed eggs; the donors had to take hormone injections for several days so that they would ovulate 10 or so eggs at once instead of the normal one or two.
We had a glimmer of success in the third cycle of attempts when the nucleus of an injected fibroblast appeared to divide, but it never cleaved to form two distinct cells. So in the next cycle we decided to take the tack used by Teruhiko Wakayama and his colleagues, the scientists who created the first cloned mice in 1998. (Wakayama was then at the University of Hawaii and is now at Advanced Cell Technology.) Although we injected some of the eggs with nuclei from skin fibroblasts as usual, we injected others with ovarian cells called cumulus cells that usually nurture developing eggs in the ovary and that can be found still clinging to eggs after ovulation. Cumulus cells are so small they can be injected whole. In the end, it took a total of 71 eggs from seven volunteers before we could generate our first cloned early embryo. Of the eight eggs we injected with cumulus cells, two divided to form early embryos of four cells--and one progressed to at least six cells--before growth stopped.
Parthogenesis
WE ALSO SOUGHT TO DETERMINE whether we could induce human eggs to divide into early embryos without being fertilized by a sperm or being enucleated and injected with a donor cell. Although mature eggs and sperm normally have only half the genetic material of a typical body cell, to prevent an embryo from having a double set of genes following conception, eggs halve their genetic complement relatively late in their maturation cycle. If activated before that stage, they still retain a full set of genes.
THE ETHICAL CONSIDERATIONS
Stem cells derived from such parthenogenetically activated cells would be unlikely to be rejected after transplantation because they would be very similar to a patient's own cells and would not produce many molecules that would be unfamiliar to the person's immune system. (They would not be identical to the individual's cells because of the gene shuffling that always occurs during the formation of eggs and sperm.) Such cells might also raise fewer moral dilemmas for some people than would stem cells derived from cloned early embryos.
Under one scenario, a woman with heart disease might have her own eggs collected and activated in the laboratory to yield blastocysts. Scientists could then use combinations of growth factors to coax stem cells isolated from the blastocysts to become cardiac muscle cells growing in laboratory dishes that could be implanted back into the woman to patch a diseased area of the heart. Using a similar technique, called androgenesis, to create stem cells to treat a man would be trickier. But it might involve transferring two nuclei from the man's sperm into a contributed egg that had been stripped of its nucleus.
Researchers have previously reported prompting eggs from mice and rabbits to divide into embryos by exposing them to different chemicals or physical stimuli such as an electrical shock. As early as 1983, Elizabeth J. Robertson, who is now at Harvard University, demonstrated that stem cells isolated from parthenogenetic mouse embryos could form a variety of tissues, including nerve and muscle.
In our parthenogenesis experiments, we exposed 22 eggs to chemicals that changed the concentration of charged atoms called ions inside the cells. After five days of growing in culture dishes, six eggs had developed into what appeared to be blastocysts, but none clearly contained the so-called inner cell mass that yields stem cells.
Why We Did It
WE ARE EAGER FOR THE DAY when we will be able to offer therapeutic cloning or cell therapy arising from parthenogenesis to sick patients. Currently our efforts are focused on diseases of the nervous and cardiovascular systems and on diabetes, autoimmune disorders, and diseases involving the blood and bone marrow.
Once we are able to derive nerve cells from cloned embryos, we hope not only to heal damaged spinal cords but to treat brain disorders such as Parkinson's disease, in which the death of brain cells that make a substance called dopamine leads to uncontrollable tremors and paralysis. Alzheimer's disease, stroke and epilepsy might also yield to such an approach.
Besides insulin-producing pancreatic islet cells for treating diabetes, stem cells from cloned embryos could also be nudged to become heart muscle cells as therapies for congestive heart failure, arrhythmias and cardiac tissue scarred by heart attacks.
CLONING AND THE LAW
A potentially even more interesting application could involve prompting cloned stem cells to differentiate into cells of the blood and bone marrow. Autoimmune disorders such as multiple sclerosis and rheumatoid arthritis arise when white blood cells of the immune system, which arise from the bone marrow, attack the body's own tissues. Preliminary studies have shown that cancer patients who also had autoimmune diseases gained relief from autoimmune symptoms after they received bone marrow transplants to replace their own marrow that had been killed by high-dose chemotherapy to treat the cancer. Infusions of blood-forming, or hematopoietic, cloned stem cells might "reboot" the immune systems of people with autoimmune diseases.
But are cloned cells--or those generated through parthenogenesis--normal? Only clinical tests of the cells will show ultimately whether such cells are safe enough for routine use in patients, but our studies of cloned animals have shown that clones are healthy. In the November 30, 2001, issue of Science, we reported on our success to date with cloning cattle. Of 30 cloned cattle, six died shortly after birth, but the rest have had normal results on physical exams, and tests of their immune systems show they do not differ from regular cattle. Two of the cows have even given birth to healthy calves.
The cloning process also appears to reset the "aging clock" in cloned cells, so that the cells appear younger in some ways than the cells from which they were cloned. In 2000 we reported that telomeres--the caps at the ends of chromosomes--from cloned calves are just as long as those from control calves. Telomeres normally shorten or are damaged as an organism ages. Therapeutic cloning may provide "young" cells for an aging population.
A report last July by Rudolf Jaenisch of the Whitehead Institute for Biomedical Research in Cambridge, Mass., and his colleagues gained much attention because it found so-called imprinting defects in cloned mice. Imprinting is a type of stamp placed on many genes in mammals that changes how the genes are turned on or off depending on whether the genes are inherited from the mother or the father. The imprinting program is generally "reset" during embryonic development.
Although imprinting appears to play an important role in mice, no one yet knows how significant the phenomenon is for humans. In addition, Jaenisch and his co-workers did not study mice cloned from cells taken from the bodies of adults, such as fibroblasts or cumulus cells. Instead they examined mice cloned from embryonic cells, which might be expected to be more variable. Studies showing that imprinting is normal in mice cloned from adult cells are currently in press and should be published in the scientific literature within several months.
Meanwhile we are continuing our therapeutic cloning experiments to generate cloned or parthenogenetically produced human embryos that will yield stem cells. Scientists have only begun to tap this important resource.
-----
THE AUTHORS:
JOSE B. CIBELLI, ROBERT P. LANZA and MICHAEL D. WEST are vice president of research, vice president of medical and scientific development, and president and CEO, respectively, of Advanced Cell Technology, a privately held biotechnology company in Worcester, Mass. Cibelli received his D.V.M. from the University of La Plata in Argentina and his Ph.D. from the University of Massachusetts at Amherst. His research led to the creation of the first cloned genetically modified calves in 1998. Lanza has an M.D. from the University of Pennsylvania. He is a former Fulbright scholar and is the author or editor of numerous popular and scientific books, including the text Principles of Tissue Engineering. West holds a Ph.D. from Baylor College of Medicine and is particularly interested in aging and stem cells. From 1990 until 1998 he was founder, director and vice president of Geron Corporation in Menlo Park, Calif., where he initiated and managed research programs in the biology of telomeres (the ends of chromosomes, which shrink during aging) and the effort to derive human embryonic stem cells. Carol Ezzell is a staff writer and editor.
The movement also runs counter to U.S. laws that permit publicly funded schools to enter into exclusive licensing agreements with private companies
Just because something is legal, doesn't mean that not doing that something is bad.
The law permits me to do lots of things that I choose not - I am not behaving counter to the law.
This is brought up again later,
``I think the Bayh-Dole Act is one of the great economic success stories in the nation,'' said Terry Young, executive director of the Texas A&M Technology Licensing Office. He says the law should remain untouched.
Again, irrelevent. The law lets universities do things. If some people don't want to enter into exclusive agreements, then the status of the law is unchanged.
The real problem is the gigantic risk of birth defects
At the moment the real problems are ethical, not scientific. Researchers are not cloning humans to make babies, they're doing it to use the embryonic cells.
One side sees this as a way of advancing medical knowledge. The cells are just cells.
The other side sees it as killing. The cells are already a person.
They didn't use stem cells. That's something else but related.
Cloning is one way to make stem cells for other research.
The Scientific American story says what they did
The next step [after getting permission from the ethics committee] was to recruit women willing to contribute eggs to be used in the cloning procedure and also collect cells from individuals to be cloned (the donors)... In the basic nuclear transfer technique, scientists use an extremely fine needle to suck the genetic material from a mature egg. They then inject the nucleus of the donor cell (or sometimes a whole cell) into the enucleated egg and incubate it under special conditions that prompt it to divide and grow
This is the same technique used to make Dolly the sheep
Slashdot Mirror
If you're going to plug books on
Rant over.
It would be interesting to see an update from them tomorrow with the same graphs as on the Servers in Practice page with today's data.
/. load.
Their site is slowing down under the
You can't make two phonecalls at the same time to two different people. Also, you're charged for the calls you make (as well as the fixed line charge!)
If you have multiple people sharing an IP address, then they can all use the connection at the same time. They all can't get maximum bandwidth, but email & web rarely require maximum bandwidth anyway unless you're downloading something large.
(Nb. my DSL modem talks to a NAT device...)
A gateway is more complicated than a NAT box.
There will always be ways around this but the cable companies want to go after the simple stuff first.
The cable/DSL routers cost $50 or so and are exceptionally easy to use. And that's what worries them.
I'd estimate a lot less than a third.
90% are phone numbers with 4128 in. Usually 800) 754-4128 which is the phone number for reporting lost visa cards.
In fact, I didn't spot a single number that could be considered revealed in the terms this story suggests.
the Cassini space craft was able to prove that Earth has life
Didn't we know that the Earth had life already? I could've sworn I saw a squirrel the other day.
It's all an act, she's done a few things before the weakest link on British TV where she wasn't a bitch.
She was on 'The Daily Show' as well last week (?), making fun of the show and herself.
Photo of First Extra-Solar Planet?
No. It's a series of photos showing a star getting 2% dimmer as a planet passes between it and us.
Nothing to get excited about unless you're an astro-physicist.
http://www.mousetrap.net/~mouse/cdrw/cheap.html
Doesn't look very thorough though
It is 1000th the linear size.
Therefore it is one-millionth the surface area.
Therefore it has the same power output per square meter.
The fuel is hydrogen, so the waste product is water (in the form of steam, obviously).
You might scald yourself (or drown!) but you want have smog.
It's just the name though.
No one is denying Monty Python has greatly influenced culture, but you can't reasonably claim that they have had a particularly noticeable effect on the development of technology.
Google has just added the ability to index PDFs, word docs etc. So, yes, the information was there before, but now it is much easier to find.
Just search for your credit card number.
By the way, does google have that realtime display of what people are searching for?
From the beginning of the report
Any child who is thought to be at risk of committing a crime by the police, schools or social services, will be put on the database
Later on
street gangs provided a safer and more caring environment than their homes or classrooms
1) Is it necessary to treat all children as potential criminals because some are in gangs?
2) If these children are safer in gangs than at home or in the classroom then - instead of putting their names into a huge database - wouldn't it be better to take them into care where they will be safe?
Using the Rio Receiver (and other unimportant info)
2 33].
/Track.mp3
/Track.mp3
/Track.mp3
n {Source MAC Address}\n". The Audio Manager responds with the IP address and port of the Windows server. It tends to use TCP Port 12078 on the server side. It actually sends back an URL to an .xml file (http://serverip:12768/descriptor.xml)...the important part is the port 12768, which will be used as the status port between the receiver and server.
/tftpboot on the server and grabs a basic Linux file system - sort of like an XTerm. It then reboots a second time, this time bootstrapping with the downloaded kernel. Jeff points out something interesting that I initially missed: the Receiver does an NFS mount of /tftpboot - it does NOT use TFTP to download the boot image from /tftpboot, as you would normally see with diskless clients.
/query?artist=". The server responds with an HTTP response with a standard HTTP 200 header, Content-type, Content-length, etc. a a response string that looks like "ArtistA.1=61,0,0:ArtistB.1=32,0,0......etc." You can start to see that the software assigns a unique index number to each item in the list, starting with 1, assigned in alphabetical order. Once you select an artist, for example, it will then make a subsequent query for the tracks associated with that artist ("http://10.0.0.25:12078/results?_extended=1&artis t=Artist" ). The server returns a similar list of tracks, each also having a unique track ID number. Once a track is selected, using its unique ID number, the Receiver asks for the tags associated with the mp3 file ("http://10.0.0.25:12078/tags/240", for example).
t =Ash" will return all songs associated with the artist Ash. To see all playlists, it sends "http://10.0.0.25:12078/content/100?_extended=1" and to view a specific playlist, it sends "http://10.0.0.25:12078/content/5e90?_extended=1", where 5e90 represents the unique playlist ID it was sent when it queried all the playlists. You get the drift....this isn't exactly rocket science. Given this simple protocol, you can easily write a browser-based frontend to the Windows-based Audio manager software.
Chris Uriarte, chrisjur@cju.com
Last Updated: November 24, 2001
Warning: I'm a very technical person AND an audiophile. Therefore, is only natural that I'd be very critical on any piece of technology that produces any type of sound. Despite my criticisms, be sure to read my final thoughts and recommendations at the end of this review.
What is it?
The RIO receiver is a small devices that allows you to play your digital MP3 and WMA music through any stereo system. It attaches to an Ethernet network or a HPNA (phone line) network, allowing it to access digital music files stored on a Windows PC. It's manufactured and sold by SonicBlue [http://www.sonicblue.com], the people who bought the Rio [http://www.riohome.com] line of digital music equipment from Diamond. The same device is re-branded and sold by Dell [http://www.dell.com] as the Dell Digital Audio Receiver The street price for the unit is about US$200. You can find out more about it at its website [http://www.riohome.com/products/receiver.htm]. Here's a picture for your viewing enjoyment:
What are the tech specs?
Without going into the nitty gritty details, here's some tech specs on the unit:
Unit measures 9"x9"x3"; contains small LCD screen, a few buttons and a jog dial on the front.
74MHz ARM7 Cirrus processor
Runs a tiny version of Debian Linux
4MB of DRAM
512KB of flash memory
CS8900 based 10MB ethernet interface
Broadcom HPNA interface. HPNA is some ass-backwards networking setup that allows you to run data over your home phone lines - be afraid, be very afraid!
One set of standard RCA audio output jacks
One set of standard speaker connectors
Why buy it?
The advantage of purchasing the unit is that it gives you access to your digital audio collection through any stereo in your house that's close enough to a phone line or ethernet port on your home network.
Why I bought it.
Let me set the stage by saying that I have over 500 CDs and that I am addicted to music. One day, I realized something important: my music collection was taking over my life. It was getting too large to handle and my fickle nature makes me want to listen to many different types of music in a relatively short period of time. CDs tend to get stale very quickly when they are in my CD player. After having, what alcoholics sometimes refer to as, "a moment of clarity", I realized that it was time to get my act together. I discovered that my car contained 70 CDs - it was a $30,000 CD holder that happened to also be able to get me back and forth to work. There were about 100 CDs laying around in my home office and about 30 in my office at work. There were numerous CDs stuck in changers and players and CD holder and travel pouches and attache bags and just about everywhere I could put them. I was out of control. My girlfriend was about to stage an intervention when I told her that I was going to buy a third CaseLogic car visor CD holder - which I somehow thought I could rig into place, despite the fact that my car, like most others, only has two sun visors. It was bad, and I needed help. One day, I had a vision - If I made my music collection digital, I could access it from anywhere - my stereo, my car (through a car MP3 player), on the road (through a portable MP3 player), at the office, etc. I really hated my trusty Sony CD 200-disc changer, which I know has swallowed a number of CDs that I've been looking for the last 5 years. I had a dream about being able to "dial up" my music - without having to know which slot it lived in or figuring whether it was even in my house at all. I thought the Rio Receiver could help me fill these dreams. (I know, It's sad. One day I'll seek professional help.)
Where to buy it.
When I was first searching for it (November 2001), the unit was somewhat difficult to find. There were only about five online retailers carrying it, and most were out of stock. I settled on the dependable, but expensive, Crutchfield [http://www.crutchfied.com], where I purchased it for $179. Note that I have seen street prices as low as $156. Check out CNET's price comparison service [http://shopper.cnet.com/] to try to hunt one down.
What do you get with it?
The box it's packaged in is roughly the size of a toaster over. Open it up and you'll find:
the Digital Audio Receiver itself
Home PNA adapter card - (Note: It appears that some models do no include the PNA adapter. That's good...avoid it at all costs)
Remote Control
2 "AA" batteries
one 6" RCA stereo patch cord
one 6'6" telephone cord (probably only included if you get the HPNA card)
one 12' 6" telephone cord (ditto)
one 6' 6" AC power cable
one CD containing the necessary server software installation and owner's manuals
lots of cardboard, styrofoam, plastic, foam and other things that kill the environment. Crutchfield is also sure to pack the receiver box in a larger box small enough to hold a child - however, the child would never survive after suffocating from the grotesque amount of styrofoam peanuts inside.
It comes with a "quick start" sheet, which is good enough to actually get it up and running pretty quickly. The full documentation is on the CD-ROM, or available on the Rio website in the Support section. However, the full documentation is scarce on juicy details, although worth a quick read.
Setting it Up
Assuming that you already have an Ethernet network in place, all you need to do is the following:
1. Install the Windows software. Reboot (of course). Run the software and specify drives and directories that contain your MP3 and WMA files. Sit back for a minute and let it catalog your files.
2. Plug the unit in. Hook up the RCA jacks to your stereo. Plug Ethernet cable in the back. Turn the unit on.
3. Select your artist/album/playlist from the LCD screen and start playing music.
The whole setup process literally took me less than 10 minutes, most of which was spent chasing runaway styrofoam peanuts across my hardwood floors.
What does the Windows Software Do?
Yes, you must - I repeat - you MUST use the Windows software for this thing to work (see notes on the Linux hack below, however). The software is not exactly what I would call "feature filled". The software has two purposes in life:
1. Catalog the digital music files on your computer system.
2. Communicate with the Digital Audio Player.
That's it. And you don't have options to play with those two either. It is truly as if SonicBlue had a meeting and said, "What is the absolute, minimal amount of work we can do to make this piece of software work?". It then appears that they took that list of minimums and cut it in half. It is barebones, to say the least.
Those who own other Rio products may be use to getting some type of CD ripping or digital file management software with players. This, however, doesn't include anything like that.
The Windows software, dubbed the "Audio Receiver Manager", doesn't actually allow you to manage anything. It does manage to discretely sit in your system tray and is loaded when you startup and log into windows. Since I always keep my main computer on and logged in, it's not a big deal to me, however, the audio manager will not start until you log into windows (i.e. enter your username password, if you get prompted for one). I'd like to see this made into a Windows NT/2000 service for the next release.
Basically, you tell the software where on your system to look for MP3/WMA/M3U files. You have the ability to specify entire drives, which can be very slow, or individual folders. You must update the database every time you want to make new music available to your receiver. During the import process, the Audio Manager must stop playing music - apparently it has difficulty walking and chewing gum at the same time.
Here's some screenshots of the Windows program:
(Click for larger images)
A Note on Image 2 above: The Audio Manager shows that it imported about 1500 tracks and 120 playlists in 16 seconds. This particular import was run on a folder of MP3s that was already imported, so it ran rather quickly. When I first ran the import on this folder, it actually took close to 40 seconds to import all the tracks (not bad at all). However, this was done on a rather speedy Athlon 1.33Ghz with 768MB RAM and 7200 RPM drives. Your mileage may vary. The Rio documentation, for example, shows a screenshot of an import that was smaller, which took a considerably longer amount of time.
Things that particularly suck about the Windows-based Audio Receiver Manager software:
It is pretty much un-configurable. Choose your DHCP range and choose your folders to scan, that's it.
There's no sort of "preferences" available to save your settings, which would be nice. Even though I only search for MP3s under one directory (D:\MP3, for example), it always defaults to search the entire C: and D: drives. I then have to remove and re-select D:\MP3 EVERY TIME. This is maddening.
Somehow, SonicBlue thought it would be helpful to have this thing search your MusicMatch and RealJukebox playlists - I can certainly see the value to this feature. However, you don't have the ability to disable searching these programs and the internal track lists and playlists are imported from these programs every time you update your database. This causes duplicate track and artist names to appear on your Receiver. For example, if you tell the receiver to search D:\MP3 for your of your tracks and you have previously imported the track names into Real Jukebox, all your tracks will be duplicated when you bring them up on the Receiver's LCD. This is fine if you want to "manage" all your music from RealJukebox, but I don't. Even after removing the tracks and playlists from RealJukebox, there were still dupes....I finally just uninstalled RealJukebox (it's a piece of junk, anyway). If you use MusicMatch, it will only import your MusicMatch playlists, which can be handy at times - but it's all or nothing....if you create playlists in MusicMatch, they will ALL show up on the Receiver, as you don't have any choice to exclude specific playlists.
As mentioned before, the software sits in the system tray and is launched from your Startup folder. Therefore, if your machine reboots or you log out of Windows 2000/NT/XP, the Audio Manager software is not running - so no music for you.
Playing your Music From the Receiver
When it comes down to doing what it is meant to do, the Rio Receiver does a very good job. It's very easy to navigate through your music collection, even if you have a large number of tracks, artists and albums. I have close to 150 albums worth of digital music and I can easily find what I'm looking for in a few seconds.
Music selection is done using a simple jog-dial knob on the front of the unit. You have the ability to search for music by artist, album, song title, genre or m3U playlist. All this information is determined by the ID3 tags [http://www.id3.org] contained within your MP3 files - therefore, it's VERY important that all of your MP3 files are embedded with at least the artist, album and song title. (See my "Tips and Tricks" below for details on this). The Rio DOES NOT CARE about the organization of your directory structure - all it wants to know is the contents of your ID3 tags.
Overall Quality and Performance
Overall, the sound quality is very good. Of course, sound quality greatly depends on the bitrate used to encode your digital music files. I encode all my MP3s at 192 bits using LAME [http://www.mp3dev.org/mp3/] - the quality is excellent. The Receiver just takes a digital stream of the music files over the network and decodes it, so there is no loss in quality due to the network transfer or any weird types of streaming protocols. At 192bits, no one can tell that the music is sourced from am MP3 file.
The unit has a 10-second buffer, just in case there are any network slowdowns. I wanted to figure out just what it would take to make the thing stop or skip or do whatever it may do if the network gets really busy. So I wrote a simple Perl script to transfer a 200MB file between my Windows box and my Linux box over HTTP using 15 simultaneous threads (i.e. 15 downloads occuring at the same time). This is, of course, way more than what I normally see on my home LAN, which is usually just comprised of me and my girlfriend surfing the web (mostly me, of course). I should note that my home network uses a NetGear 10/100 Hub (not a switch) and a Linksys 10MB in the Living room - so the effective throughput will never exceed 10MB (not including protocol overhead). With 2 Skinny J's The Whammy playing in the background, I started the network torture test. Despite the fact that my Linux box's four-year-old hard drive protested, the Rio Receiver didn't skip a beat (no pun intended). So I think you can be pretty sure that this thing will work well, even in a busier network environment.
Did you say it runs Linux?
Yes, it runs Linux. Many people tend to get very aroused about this these days. I am a Linux advocate, as I like to advocate any technology that helps you work or play better, but I'm not a Linux zealot. But for the more curious, the unit has already been hacked, although it appears to be a bit more difficult to do than it should be. Jeff Mock has put together a wonderful collection of information [http://www.mock.com/receiver/] related to hacking the Receiver. Even though I was greatly tempted see what you can do with the thing, I figure I bought it to do what it does best - play music. I'm happy with its feature set and don't feel the need to crack it open. If this unit gets more popular and there are a number of good hacks available out there, you may see more people trying ot hack it.
What Operating Systems Does it Work With? (Do I have to use Windows?)
The Rio Audio Receiver Manager software officially supports Windows98, 98SE, 2000 and ME. Although not listed, I'm sure it also works on XP, as the application nothing more than simple HTTP and UDP server. It will not run on Windows95 or WindowsNT 4.0.
There are a few efforts underway to make the unit work with non-Windows servers. The first is a home-grown project put together by Jeff Mock [http://www.mock.com/receiver]. Jeff has done an excellent job at reverse engineering the Rio's protocols and has put together a number of scripts that allow you to use a Linux machine as the file server. At this time, his software does not support playlists, but after looking at the code and the protocols, it is probably an easy thing to add. Be aware, though, installation of Jeff's code is not a trivial task - you need to have knowledge about Linux, Apache and Perl and there's not a lot of "HOW-TO" documentation available (who has time to document, anyway?...especially for open-source projects that don't pay the bills?). It took me about 40 minutes, but I got Jeff's scripts running. Again, he did a great job with this.
Second, there's an open source project called Digital Audio Server [http://sourceforge.net/projects/das/], which I have not tried out. It looks as if it might be a bit more complete and simpler to use than Jeff's solution.
There's a third project called Jreceiver [http://sourceforge.net/projects/jreceiver/], which is open source servlet-based audio server built to support network-based MP3 players like the Rio. However, it appears to be quite immature at this time. It's completely servlet-based, but does not support Tomcat, which is a major drawback, in my opinion.
Are there competitors?
Yes, in fact there is one generally-available competitor that I am aware of: The Turtle Beach Audiotron [http://www.audiotron.net/audiotron.asp]. I like this particular product a lot. Had it not been for the Rio's attractive $150 street price, I probably would have bought the Audiotron. There are a few major differences in the way that these boxes operates. First, the Audiotron is a full-size (but thin) stereo rack component, so it can blend nicely into your existing system. Second, the Audiotron accesses music over a simple SMB share, so you can simply export a folder from your Windows or UNIX box and make it available to the Audiotron. There's no server-side software required. Third, you can actually READ the LCD display on the Audiotron from a distance greater than 10 inches. Forth, the Audiotron has a TOSLINK optical output, which is nice to have. Last, but not least, the Audiotron comes with a suite of useful software, including a web-based front-end. The street price for the Audiotron, however, is about double the price of the Rio. You can read Rob Malda's review of the Audiotron on Slashdot [http://slashdot.org/article.pl?sid=01/10/12/1423
Tips, Tricks and Things to Be Aware Of
Use ID3 Tags - as mentioned before, the Rio gets all it's track, artist and album info from the ID3 tags embedded in the MP3 file, so make sure you ID3 tag your MP3s. The easiest way to ID3 tag your MP3s is to do it when you rip them from the CD using a good ripping program with CDDB/FreeDB integration (Check out Audiograbber [http://www.audiograbber.com-us.net]). If have existing MP3s that are not tagged, MusicMatch [http://www.musicmatch.com] does a good tagging job for no cost. Musicmatch will also identify all the MP3s on your system that are not tagged.
Use RealJukebox or simple file system management - the Rio Audio Receiver Manager WILL import all tracks and playlists you have setup in RealJukebox and you don't have any say over it. Therefore, you should manage your MP3s through RealJukebox or not use RealJukebox at all. If you search for MP3s on your hard drives AND have a RealJukebox tracklist built, you will get duplicate track entries when you bootup the Rio.
Listening to Full Albums - There's only one way to listen to full albums the way they were intended to be heard on the CD: create an m3u playlist for the album. If you simply choose "Albums" through the Rio controls, it will list out the unique album names on the system (without the artist name attached to it). If you choose an album, the Rio plays through tracks in the album in alphabetical order - not in the order they appear on the CD. The solution is to create an m3u playlist for each album and then access the album through the "Playlists" option on the Receiver. Again, a good CD ripping program should create these playlists for you when the rip is complete.
Album Playlists help if they include Artist Names - Some CD ripping programs will create m3u playlists and name the file after the album you just ripped. The Rio displays playlists according to their file system name. Therefore, if you don't have the artist name in the playlist file name, you may have a tough time identifying and finding the album playlist your are looking for. The solution is to name your playlists in some format like "Artist-Album.m3u". Sadly, my favorite software, Audiograbber, does not let you configure the file name format of the playlists. So I wrote a Perl script that goes through your MP3 directories and renames the playlists to "Artist-Album.m3u" format IF you organize your MP3's in the following fashion:
+/MP3-Root-Directory
+/Artist-Name
+/Album-Name
etc.
I think this is the typical way most people organize their MP3's, so the Perl script may be helpful. Also, since I think Playlists are so important, I wrote a Perl script that hunts out any album directories that DO NOT contain playlists and prints them out for you. (BTW, if you didn't know this already, you need to install Perl [http://www.cpan.org] on your machine for these to work).
The LCD Screen is Small - You need to be pretty close to the unit to read the LCD, so choosing tracks from a distance is impossible. This used to piss me off, but I figured that with the old way, I would have to dig out another CD anyway (and chances were that CD I was looking for was MIA).
Tiny Gaps Between Songs - the Rio inserts an every-so-tiny gap between tracks...so small you don't really notice it. I consider this a good thing. For live albums and albums with songs that continue onto the next track without a gap, everything sounds like it did on the CD...the Rio does an excellent job with this. I tested Jerry Seinfeld's I'm Telling You For the Last Time (a live comedy album), 311's Live album (a live concert album) and Rx Bandit's Progress (a traditional studio album where the first 5 tracks flow into the next track) - all of them sounded great.
Final Thoughts and Impressions
At $150 street price, this thing is a no-brainer. Despite some of my gripes, it does what its supposed to do and does it well. I highly recommend it. I no longer require a CD player in my den. Everything gets ripped to MP3 and then played on the Rio. If you already have a home network in place and a lot of digital music, it's pretty much just plug-and-play. I've already enjoyed using the Rio for the short time that I've had it and I'm sure that I will get much more enjoyment out of it in the future. Good job, SonicBlue.
For the tech weenies:
-----
How does it work? (Warning: very useless technical information follows...you may want to stop here.)
Given my particular background, I was more interested in how the Receiver actually gets music from the central PC. After further analysis, I became convinced that the great Rube Goldberg was the chief architect of Rio. I thought the whole thing could have been made a lot simpler, especially compared with the file-sharing simplicity of the Turtle Beach Audiotron (see "Competitors" above).
So I set out to determine the protocols used between the Receiver and the Windows server. I figured that if I could reverse engineer these protocols, I could write my own server software to circumvent the lack of features in Rio version - including getting the server to run on Linux. So I started running Etherpeek sniffs of my network and began analyzing the traffic flow between the Receiver and my PC. I was about a half hour into it when I first discovered Jeff Mock's great "Hacking the Rio Receiver" page [http://www.mock.com/receiver]. Jeff did a far better job that I did in my first 30 minutes, so I used his notes to compare what I sniffed (btw, Jeff, your 100% right on...awesome work). The result of Jeff's sniffing efforts produced a set of scripts that allow you to run a Linux version of the server software (available for download on his site).
Basically, the protocol is simple, but there's a lot of little steps to actually get to the point where the music starts playing. Here's a quick overview:
When the box boots, it sends out a DHCP request ("please, someone give me an IP address"). The Rio Audio Manager software on the PC responds to DHCP requests only for the RIO. This is based totally on MAC address, as I was able to send a spoofed DHCP request using the Rio's MAC from my Linux box. FYI, on the Windows side, I always limit the DHCP IP scope to exactly 1 address, which essentially gives the Rio a static IP. Jeff relies on the OS to do the DHCP stuff - not his server software. You can, however, configure dhcpd to always dish out the same IP to a single MAC address, if you so desire.
After the Receiver is happy with it's IP address, it sends out an SSDP request on UDP 21075 ("someone please tell me where my server is") with the request string "^upnp:uuid:1D274DB0-F053-11d3-BF72-0050DA689B2F\
The receiver actually does an NFS mount of
When you first scroll through a list - such as "Artists", the box issues a HTTP query to the app port specified in the SDDP response that looks something like: "GET
From that point on, it's time to actually play the music. This accomplished by streaming the data over HTTP, gracefully shooting a flow of TCP PSH's and ACK's across your network. The receiver requests the unique track id over HTTP a la http://10.0.0.25:12078/content/240.
While alive and active, the Receiver and the server software start to miss each other and send UDP status packets out to each other about every five seconds. Basically, the server says "State? (What's going down?)" and the receiver replies with it's current state, which includes its MAC address, the current song it's playing and an associated timecode.
It should be noted that once you sniff the format of these queries, you can simply just insert them into any old web browser to see the results the server sends back. Type in "http://10.0.0.25:12078/query?artist=", for example, and you'll get a list of all the artists in the database. Doing something like "http://10.0.0.25:12078/query?artist=c" returns all artists starting with C - this might be useful in building some type of music search engine or something. Doing something like "http://10.0.0.25:12078/results?_extended=1&artis
I figured that if I wanted to write a web-based front-end for the Audio Manager software, I would need to figure out which port the server is running on. So I created a trivial perl script that mimics the Receiver, sending out a direct SSDP query to the Windows server, which is happy to return the all-important port number that the server software happens to be running on. Once you have that port, you can issue any of the HTTP queries above via a regular web browser or with any piece of code you may just create. Interesting note: by sending out the SSDP request from my Linux box, the Linux box now shows up as a "New Audio Receiver" in the Audio Receiver Software. This causes the server software to send my Linux box a UDP "status" query every five seconds (too bad my Linux box isn't answering).
Lots more that can be done with this...if you have the time and can think up a good use for it.
As a side not, if you're looking for a front-end to your MP3 collection and you're using Apache, take a look at Lincoln Steins's Apache::MP3 [http://www.modperl.com/Songs/], for use with mod_perl. I use it and love it. It gets me access to my entire music collection anywhere that I have an Internet connection.
The software has two purposes in life:
1. Catalog the digital music files on your computer system.
2. Communicate with the Digital Audio Player.
AND
Rio gets all it's track, artist and album info from the ID3 tags embedded in the MP3 file
So why is a proprietary protocol necessary? Why can't it drag the files down with SMB and/or NFS?
Easier for them. Easier for us.
THEY WERE SUCH TINY DOTS, YET THEY HELD SUCH immense promise. After months of trying, on October 13, 2001, we came into our laboratory at Advanced Cell Technology to see under the microscope what we'd been striving for--little balls of dividing cells not even visible to the naked eye. Insignificant as they appeared, the specks were precious because they were, to our knowledge, the first human embryos produced using the technique of nuclear transplantation, otherwise known as cloning.
With a little luck, we hoped to coax the early embryos to divide into hollow spheres of 100 or so cells called blastocysts. We intended to isolate human stem cells from the blastocysts to serve as the starter stock for growing replacement nerve, muscle and other tissues that might one day be used to treat patients with a variety of diseases. Unfortunately, only one of the embryos progressed to the six-cell stage, at which point it stopped dividing. In a similar experiment, however, we succeeded in prompting human eggs--on their own, with no sperm to fertilize them--to develop parthenogenetically into blastocysts. We believe that together these achievements, the details of which we reported November 25 in the online journal e-biomed: The Journal of Regenerative Medicine, represent the dawn of a new age in medicine by demonstrating that the goal of therapeutic cloning is within reach.
Therapeutic cloning--which seeks, for example, to use the genetic material from patients' own cells to generate pancreatic islets to treat diabetes or nerve cells to repair damaged spinal cords--is distinct from reproductive cloning, which aims to implant a cloned embryo into a woman's uterus leading to the birth of a cloned baby. We believe that reproductive cloning has potential risks to both mother and fetus that make it unwarranted at this time, and we support a restriction on cloning for reproductive purposes until the safety and ethical issues surrounding it are resolved.
Disturbingly, the proponents of reproductive cloning [see Reproductive Cloning: They Want to Make a Baby] are trying to co-opt the term "therapeutic cloning" by claiming that employing cloning techniques to create a child for a couple who cannot conceive through any other means treats the disorder of infertility. We object to this usage and feel that calling such a procedure "therapeutic" yields only confusion.
What We Did
WE LAUNCHED OUR ATTEMPT to create a cloned human embryo in early 2001. We began by consulting our ethics advisory board, a panel of independent ethicists, lawyers, fertility specialists and counselors that we had assembled in 1999 to guide the company's research efforts on an ongoing basis. Under the chairmanship of Ronald M. Green, director of the Ethics Institute at Dartmouth College, the board considered five key issues [see The Ethical Considerations] before recommending that we go ahead.
THERAPEUTIC CLONING: HOW IT'S DONE
The next step was to recruit women willing to contribute eggs to be used in the cloning procedure and also collect cells from individuals to be cloned (the donors). The cloning process appears simple, but success depends on many small factors, some of which we do not yet understand. In the basic nuclear transfer technique, scientists use an extremely fine needle to suck the genetic material from a mature egg. They then inject the nucleus of the donor cell (or sometimes a whole cell) into the enucleated egg and incubate it under special conditions that prompt it to divide and grow [see Therapeutic Cloning: How It's Done].
We found women willing to contribute eggs on an anonymous basis for use in our research by placing advertisements in publications in the Boston area. We accepted women only between the ages of 24 and 32 who had at least one child. Interestingly, our proposal appealed to a different subset of women than those who might otherwise contribute eggs to infertile couples for use in in vitro fertilization. The women who responded to our ads were motivated to give their eggs for research, but many would not have been interested in having their eggs used to generate a child they would never see. (The donors were recruited and the eggs were collected by a team led by Ann A. Kiessling-Cooper of Duncan Holly Biomedical in Somerville, Mass. Kiessling was also part of the deliberations concerning ethical issues related to the egg contributors.)
We asked potential egg contributors to submit to psychological and physical tests, including screening for infectious diseases, to ensure that the women were healthy and that contributing eggs would not adversely affect them. We ended up with 12 women who were good candidates to contribute eggs. In the meantime, we took skin biopsies from several other anonymous individuals to isolate cells called fibroblasts for use in the cloning procedure. Our group of fibroblast donors includes people of varying ages who are generally healthy or who have a disorder such as diabetes or spinal cord injury--the kinds of people likely to benefit from therapeutic cloning.
Our first cloning attempt occurred last July. The timing of each attempt depended on the menstrual cycles of the women who contributed eggs; the donors had to take hormone injections for several days so that they would ovulate 10 or so eggs at once instead of the normal one or two.
We had a glimmer of success in the third cycle of attempts when the nucleus of an injected fibroblast appeared to divide, but it never cleaved to form two distinct cells. So in the next cycle we decided to take the tack used by Teruhiko Wakayama and his colleagues, the scientists who created the first cloned mice in 1998. (Wakayama was then at the University of Hawaii and is now at Advanced Cell Technology.) Although we injected some of the eggs with nuclei from skin fibroblasts as usual, we injected others with ovarian cells called cumulus cells that usually nurture developing eggs in the ovary and that can be found still clinging to eggs after ovulation. Cumulus cells are so small they can be injected whole. In the end, it took a total of 71 eggs from seven volunteers before we could generate our first cloned early embryo. Of the eight eggs we injected with cumulus cells, two divided to form early embryos of four cells--and one progressed to at least six cells--before growth stopped.
Parthogenesis
WE ALSO SOUGHT TO DETERMINE whether we could induce human eggs to divide into early embryos without being fertilized by a sperm or being enucleated and injected with a donor cell. Although mature eggs and sperm normally have only half the genetic material of a typical body cell, to prevent an embryo from having a double set of genes following conception, eggs halve their genetic complement relatively late in their maturation cycle. If activated before that stage, they still retain a full set of genes.
THE ETHICAL CONSIDERATIONS
Stem cells derived from such parthenogenetically activated cells would be unlikely to be rejected after transplantation because they would be very similar to a patient's own cells and would not produce many molecules that would be unfamiliar to the person's immune system. (They would not be identical to the individual's cells because of the gene shuffling that always occurs during the formation of eggs and sperm.) Such cells might also raise fewer moral dilemmas for some people than would stem cells derived from cloned early embryos.
Under one scenario, a woman with heart disease might have her own eggs collected and activated in the laboratory to yield blastocysts. Scientists could then use combinations of growth factors to coax stem cells isolated from the blastocysts to become cardiac muscle cells growing in laboratory dishes that could be implanted back into the woman to patch a diseased area of the heart. Using a similar technique, called androgenesis, to create stem cells to treat a man would be trickier. But it might involve transferring two nuclei from the man's sperm into a contributed egg that had been stripped of its nucleus.
Researchers have previously reported prompting eggs from mice and rabbits to divide into embryos by exposing them to different chemicals or physical stimuli such as an electrical shock. As early as 1983, Elizabeth J. Robertson, who is now at Harvard University, demonstrated that stem cells isolated from parthenogenetic mouse embryos could form a variety of tissues, including nerve and muscle.
In our parthenogenesis experiments, we exposed 22 eggs to chemicals that changed the concentration of charged atoms called ions inside the cells. After five days of growing in culture dishes, six eggs had developed into what appeared to be blastocysts, but none clearly contained the so-called inner cell mass that yields stem cells.
Why We Did It
WE ARE EAGER FOR THE DAY when we will be able to offer therapeutic cloning or cell therapy arising from parthenogenesis to sick patients. Currently our efforts are focused on diseases of the nervous and cardiovascular systems and on diabetes, autoimmune disorders, and diseases involving the blood and bone marrow.
Once we are able to derive nerve cells from cloned embryos, we hope not only to heal damaged spinal cords but to treat brain disorders such as Parkinson's disease, in which the death of brain cells that make a substance called dopamine leads to uncontrollable tremors and paralysis. Alzheimer's disease, stroke and epilepsy might also yield to such an approach.
Besides insulin-producing pancreatic islet cells for treating diabetes, stem cells from cloned embryos could also be nudged to become heart muscle cells as therapies for congestive heart failure, arrhythmias and cardiac tissue scarred by heart attacks.
CLONING AND THE LAW
A potentially even more interesting application could involve prompting cloned stem cells to differentiate into cells of the blood and bone marrow. Autoimmune disorders such as multiple sclerosis and rheumatoid arthritis arise when white blood cells of the immune system, which arise from the bone marrow, attack the body's own tissues. Preliminary studies have shown that cancer patients who also had autoimmune diseases gained relief from autoimmune symptoms after they received bone marrow transplants to replace their own marrow that had been killed by high-dose chemotherapy to treat the cancer. Infusions of blood-forming, or hematopoietic, cloned stem cells might "reboot" the immune systems of people with autoimmune diseases.
But are cloned cells--or those generated through parthenogenesis--normal? Only clinical tests of the cells will show ultimately whether such cells are safe enough for routine use in patients, but our studies of cloned animals have shown that clones are healthy. In the November 30, 2001, issue of Science, we reported on our success to date with cloning cattle. Of 30 cloned cattle, six died shortly after birth, but the rest have had normal results on physical exams, and tests of their immune systems show they do not differ from regular cattle. Two of the cows have even given birth to healthy calves.
The cloning process also appears to reset the "aging clock" in cloned cells, so that the cells appear younger in some ways than the cells from which they were cloned. In 2000 we reported that telomeres--the caps at the ends of chromosomes--from cloned calves are just as long as those from control calves. Telomeres normally shorten or are damaged as an organism ages. Therapeutic cloning may provide "young" cells for an aging population.
A report last July by Rudolf Jaenisch of the Whitehead Institute for Biomedical Research in Cambridge, Mass., and his colleagues gained much attention because it found so-called imprinting defects in cloned mice. Imprinting is a type of stamp placed on many genes in mammals that changes how the genes are turned on or off depending on whether the genes are inherited from the mother or the father. The imprinting program is generally "reset" during embryonic development.
Although imprinting appears to play an important role in mice, no one yet knows how significant the phenomenon is for humans. In addition, Jaenisch and his co-workers did not study mice cloned from cells taken from the bodies of adults, such as fibroblasts or cumulus cells. Instead they examined mice cloned from embryonic cells, which might be expected to be more variable. Studies showing that imprinting is normal in mice cloned from adult cells are currently in press and should be published in the scientific literature within several months.
Meanwhile we are continuing our therapeutic cloning experiments to generate cloned or parthenogenetically produced human embryos that will yield stem cells. Scientists have only begun to tap this important resource.
-----
THE AUTHORS:
JOSE B. CIBELLI, ROBERT P. LANZA and MICHAEL D. WEST are vice president of research, vice president of medical and scientific development, and president and CEO, respectively, of Advanced Cell Technology, a privately held biotechnology company in Worcester, Mass. Cibelli received his D.V.M. from the University of La Plata in Argentina and his Ph.D. from the University of Massachusetts at Amherst. His research led to the creation of the first cloned genetically modified calves in 1998. Lanza has an M.D. from the University of Pennsylvania. He is a former Fulbright scholar and is the author or editor of numerous popular and scientific books, including the text Principles of Tissue Engineering. West holds a Ph.D. from Baylor College of Medicine and is particularly interested in aging and stem cells. From 1990 until 1998 he was founder, director and vice president of Geron Corporation in Menlo Park, Calif., where he initiated and managed research programs in the biology of telomeres (the ends of chromosomes, which shrink during aging) and the effort to derive human embryonic stem cells. Carol Ezzell is a staff writer and editor.
I look at pictures off the internet just before getting into bed. It helps me sleep
This article was first published on newVALUEnews April 2001
The movement also runs counter to U.S. laws that permit publicly funded schools to enter into exclusive licensing agreements with private companies
Just because something is legal, doesn't mean that not doing that something is bad.
The law permits me to do lots of things that I choose not - I am not behaving counter to the law.
This is brought up again later,
``I think the Bayh-Dole Act is one of the great economic success stories in the nation,'' said Terry Young, executive director of the Texas A&M Technology Licensing Office. He says the law should remain untouched.
Again, irrelevent. The law lets universities do things. If some people don't want to enter into exclusive agreements, then the status of the law is unchanged.
The real problem is the gigantic risk of birth defects
At the moment the real problems are ethical, not scientific. Researchers are not cloning humans to make babies, they're doing it to use the embryonic cells.
One side sees this as a way of advancing medical knowledge. The cells are just cells.
The other side sees it as killing. The cells are already a person.
They're not going to grow a living person from this. The aim is to use the cells after they've multiplied a bit and use them medically.
Whether that is better or worse than producing a person is up to you.
They didn't use stem cells. That's something else but related.
... In the basic nuclear transfer technique, scientists use an extremely fine needle to suck the genetic material from a mature egg. They then inject the nucleus of the donor cell (or sometimes a whole cell) into the enucleated egg and incubate it under special conditions that prompt it to divide and grow
Cloning is one way to make stem cells for other research.
The Scientific American story says what they did
The next step [after getting permission from the ethics committee] was to recruit women willing to contribute eggs to be used in the cloning procedure and also collect cells from individuals to be cloned (the donors)
This is the same technique used to make Dolly the sheep
You've looked on Pricewatch right?
The price ranges for SCSI and ATAPI seem to overlap.