Reflections on the Variety of Presentation of the Unit 6 Material

Two aspects of the variety of formats of presentation of the Unit 6 material which struck me as interesting had to do with the podcasts.

The first aspect had to do with my expectations about the accessibility of content with respect to a delivery format. Prior to viewing the Harvard Computer Science Podcast, my expectation of the podcast format was that of exclusively audio content delivery. Given this expectation, I was skeptical that the information content could be effectively conveyed to me in audio-only format. I was relieved to discover my notions of podcasting to be dated, and thought the actual audio-video presentation of the Harvard Computer Science Podcasts to be highly effective. However, I began to wonder why this should be the case. Because overall, the video component of the podcast did not substantively contribute additional quantitative information absent from the audio. Much of the video content was an image of a professor delivering a lecture. The same amount quantifiable content would have been delivered if the format were solely audio, but I'm sure I would not have judged it to be an effective presentation. My guess is the distinction speaks to a certain vague subjective factor in effective content delivery. I think much of this has to do with content user expectations. I expect to see a lecture if I am listening to a lecture. Watching the video of the lecture added something to the audio information which I think was mainly of a subjective quality. But the addition, at least for me, was very important. But on the other hand, maybe I'm making too much of this. I do listen to radio broadcasts of baseball games.

The second aspect had to do with the RSS component of podcast format, and the effect of this technology on my understanding of the models of networking architecture. Specifically, what is the designation of an architectural component which, to me, is neither precisely a client nor a server? The content of the Harvard Computer Science Podcasts is hosted by the content creators, presumably by servers of Harvard. But users request this content not from Harvard servers, but from iTunes. So my question is, what is the name for functional role of iTunes in this instance? Maybe I am making too much of this, but I think what it adds to the simple client-server model is interesting. This is the second time I have had the impression that RSS creates interesting effects with networking architecture. At the Computers in Libraries Conference in 2007, I listened to a presentation by a librarian at the Rand Corporation. He was using RSS in an innovative way, providing a greater degree of freedom for Rand staff to edit web pages on the Rand LAN. According to him, his utilization of RSS allowed Rand staff to edit web pages without messing up the web page architecture.

I would say the Unit 6 material is relatively "deep," and the variety of presentations of the same repeated content was helpful. For instance, I found the Warriors of the Net animation to be very helpful for me to conceptualize TCP/IP. The same material, for me, was effectively reinforced in detail in Nemeth's Linux Administration Handbook. I found the Wikipedia articles on local area networks to be very helpful gaining a conceptual understanding of the applied assignment.

Installation Routine Experiences

Installation Routine Experiences

I completed the assigned installation routines without significant issues. I built a server GUI by installing the X window system and the IceWM window manager, and verified installation of the server GUI via http://localhost/.

I installed Ubuntu Desktop on a virtual machine (VM) free of any major issues. I think I am going to move the .iso files of my Ubuntu operating systems from my machine's C Drive to its D Drive, because the C Drive is getting full. While on the subject of drives, I need to read about installing VMware virtual machines on a portable flash drive. If the result would mean being able to carry around a portable Linux OS, then such a thing would be useful.

I utilized wget and dpkg to install the Webmin admin GUI. Notable are the large variety of available modules (http://www.webmin.com/standard.html). I ran ifconfig to find the virtual server's IP address on the second line of eth0. I might be making a mistake, but according to my notes the server's IP address changed from a previous day. Hmm. I connected to the Webmin GUI on the IceWM window manager on the X window system through Port 10000 from my host machine browser via https://localhost:10000.

Generally all these routines completed without major issues. My big issue for this week was the installation of vim, which kept me busy Monday and Tuesday. The issue took many hours to resolve, and is detailed in the Unit 5 Activities section.

The New York Times: Charging by the Byte to Curb Internet Traffic

The New York Times
Printer Friendly Format Sponsored By

June 15, 2008
Charging by the Byte to Curb Internet Traffic
By BRIAN STELTER

Some people use the Internet simply to check e-mail and look up phone numbers. Others are online all day, downloading big video and music files.

For years, both kinds of Web surfers have paid the same price for access. But now three of the country’s largest Internet service providers are threatening to clamp down on their most active subscribers by placing monthly limits on their online activity.

One of them, Time Warner Cable, began a trial of “Internet metering” in one Texas city early this month, asking customers to select a monthly plan and pay surcharges when they exceed their bandwidth limit. The idea is that people who use the network more heavily should pay more, the way they do for water, electricity, or, in many cases, cellphone minutes.

That same week, Comcast said that it would expand on a strategy it uses to manage Internet traffic: slowing down the connections of the heaviest users, so-called bandwidth hogs, at peak times.

AT&T also said Thursday that limits on heavy use were inevitable and that it was considering pricing based on data volume. “Based on current trends, total bandwidth in the AT&T network will increase by four times over the next three years,” the company said in a statement.

All three companies say that placing caps on broadband use will ensure fair access for all users.

Internet metering is a throwback to the days of dial-up service, but at a time when video and interactive games are becoming popular, the experiments could have huge implications for the future of the Web.

Millions of people are moving online to watch movies and television shows, play multiplayer video games and talk over videoconference with family and friends. And media companies are trying to get people to spend more time online: the Disneys and NBCs of the world keep adding television shows and movies to their Web sites, giving consumers convenient entertainment that soaks up a lot of bandwidth.

Moreover, companies with physical storefronts, like Blockbuster, are moving toward digital delivery of entertainment. And new distributors of online content — think YouTube — are relying on an open data spigot to make their business plans work.

Critics of the bandwidth limits say that metering and capping network use could hold back the inevitable convergence of television, computers and the Internet.

The Internet “is how we deliver our shows,” said Jim Louderback, chief executive of Revision3, a three-year-old media company that runs what it calls a television network on the Web. “If all of a sudden our viewers are worried about some sort of a broadband cap, they may think twice about downloading or watching our shows.”

Even if the caps are far above the average users’ consumption, their mere existence could cause users to reduce their time online. Just ask people who carefully monitor their monthly allotments of cellphone minutes and text messages.

“As soon as you put serious uncertainty as to cost on the table, people’s feeling of freedom to predict cost dries up and so does innovation and trying new applications,” Vint Cerf, the chief Internet evangelist for Google who is often called the “father of the Internet,” said in an e-mail message.

But the companies imposing the caps say that their actions are only fair. People who use more network capacity should pay more, Time Warner argues. And Comcast says that people who use too much — like those who engage in file-sharing — should be forced to slow down.

Time Warner also frames the issue in financial terms: the broadband infrastructure needs to be improved, it says, and maybe metering could pay for the upgrades. So far its trial is limited to new subscribers in Beaumont, Tex., a city of roughly 110,000.

In that trial, new customers can buy plans with a 5-gigabyte cap, a 20-gigabyte cap or a 40-gigabyte cap. Prices for those plans range from $30 to $50. Above the cap, customers pay $1 a gigabyte. Plans with higher caps come with faster service.

“Average customers are way below the caps,” said Kevin Leddy, executive vice president for advanced technology at Time Warner Cable. “These caps give them years’ worth of growth before they’d ever pay any surcharges.”

Casual Internet users who merely send e-mail messages, check movie times and read the news are not likely to exceed the caps. But people who watch television shows on Hulu.com, rent movies on iTunes or play the multiplayer game Halo on Xbox may start to exceed the limits — and millions of people are already doing those things.

Streaming an hour of video on Hulu, which shows programs like “Saturday Night Live,” “Family Guy” and “The Daily Show With Jon Stewart,” consumes about 200 megabytes, or one-fifth of a gigabyte. A higher-quality hour of the same content bought through Apple’s iTunes store can use about 500 megabytes, or half a gigabyte.

A high-definition episode of “Survivor” on CBS.com can use up to a gigabyte, and a DVD-quality movie through Netflix’s new online service can eat up about five gigabytes. One Netflix download alone, in fact, could bring a user to the limit on the cheapest plan in Time Warner’s trial in Beaumont.

Even services like Skype and Vonage that use the Internet to transmit phone calls could help put users over the monthly limits.

Time Warner would not reveal how many gigabytes an average customer uses, saying only that 95 percent of customers use under 40 gigabytes each in a month.

That means that 5 percent of customers use more than 50 percent of the network’s overall capacity, the company said, and many of those people are assumed to be sharing copyrighted video and music files illegally.

The Time Warner plan has the potential to bring Internet use full circle, back to the days when pay-as-you-go pricing held back the Web’s popularity. In the early days of dial-up access, America Online and other providers offered tiered pricing, in part because audio and video were barely viable online. Consumers feared going over their allotted time and bristled at the idea that access to cyberspace was billed by the hour.

In 1996, when AOL started offering unlimited access plans, Internet use took off and the online world started moving to the center of people’s daily lives. Today most Internet packages provide a seemingly unlimited amount of capacity, at least from the consumer’s perspective.

But like water and electricity, even digital resources are finite. Last year Comcast disclosed that it was temporarily turning off the connections of customers who used file-sharing services like BitTorrent, arguing that they were slowing things down for everyone else. The people who got cut off complained and asked how much broadband use was too much; the company did not have a ready answer.

Thus, like Time Warner, Comcast is considering a form of Internet metering that would apply to all online activity.

The goal, says Mitch Bowling, a senior vice president at Comcast, is “ensuring that a small number of users don’t impact the experience for everyone else.”

Last year Comcast was sued when it was disclosed that the company had singled out BitTorrent users.

In February, Comcast departed from that approach and started collaborating with the company that runs BitTorrent. Now it has shifted to what it calls a “platform agnostic” approach to managing its network, meaning that it slows down the connection of any customer who uses too much bandwidth at congested times.

Mr. Bowling said that “typical Internet usage” would not be affected. But on the Internet, “typical” use is constantly being redefined.

“The definitions of low and high usage today are meaningless, because the Internet’s going to grow, and nothing’s going to stop that,” said Eric Klinker, the chief technology officer of BitTorrent.

As the technology company Cisco put it in a recent report, “today’s ‘bandwidth hog’ is tomorrow’s average user.”

One result of these experiments is a tug-of-war between the Internet providers and media companies, which are monitoring the Time Warner experiment with trepidation.

“We hate it,” said a senior executive at a major media company, who requested anonymity because his company, like all broadcasters, must play nice with the same cable operators that are imposing the limits. Now that some television shows are viewed millions of times online, the executive said, any impediment would hurt the advertising model for online video streaming.

Mr. Leddy of Time Warner said that the media companies’ fears were overblown. If the company were to try to stop Web video, “we would not succeed,” he said. “We know how much capacity they’re going to need in the future, and we know what it’s going to cost. And today’s business model doesn’t pay for it very well.”

My Experiences with the Text Editors and the Hands-on Assignments

The appearance of the new VMware Workstation 8.04 "feature," the Recovery Menu, suggested for me an investigation into the differences between shutdown and halt. (M)an shutdown says shutdown arranges for the system to be brought down in a safe way. Users on the system are notified. (S)hutdown options include shutdown -P, which requests that the system be powered off after it has been brought down. (M)an halt says when halt is called without options, it simply invokes shutdown. Apparently some halt options cause the system to stop more abruptly than with shutdown. This is underscored by the following post on a Unix (Solaris) forum, found by googling "difference between shutdown and halt":

The shutdown command runs various shutdown scripts and then finally invokes the halt command. Not running those scripts could be a problem depending on what they do. As one example, if you are running an Oracle database, you could lose transactions or even mangle the database....

------------------------------------

Wanting to more frequently use Vim, I downloaded a MS DOS version at http://www.vim.org/download.php. The download was a .zip file. I used a Windows utility to download and unzip the file. (I did not figure out how to download and unzip in DOS. Sorry, maybe next time.) File downloaded and unzipped, I went to the DOS command line and found it, in directory of downloaded files. I did not want the file in a download directory, but in the program files directory. So I tried to moved it to my program files directory with the DOS command

move c:\Downloads\vim71w32\*.* c:\Program_Files

But I got this message:

The process cannot access the file because it is being used by another process.

0 file(s) moved.

What was the other process that would be using the Vim file, preventing me from moving it? I decided at that point to go the easy route and figure out the issue another day. I opened the download folder window, opened the c: window, and dragged and dropped the Vim folder icon from the former into the latter. The DOS version of Vim resembles that of Linux. The help feature does not work. Maybe I will fix it. Anyway, I now have a version of Vim in Windows, which I have the idea of using for typical word processing tasks, such as producing this blog. We'll see how it goes.

Ubuntu Live CD Experiences

Not long ago I burned a Debian image file on CD. If I recall correctly, at that time I downloaded the image file in binary mode. So initially I tried downloading the Ubuntu file in binary mode. This did not seem to go well. The file was downloading very slowly, and it occurred to me the generally user-friendly instructions did not prompt me switch to binary mode. So I cancelled that process, followed the instructions downloading the file in text mode and all went well. At first the Ubuntu Desktop CD would not boot. So I went into the BIOS and changed the boot order, and that did the trick.

Initially I went through the tutorials in the order they were presented. The CBT videos were a good introduction to the Linux file systems, especially because I could see demonstrations of the commands. The sections of Learning the Shell which list the contents of the Linux directories was useful as a reference to which I frequently returned. I frequently referred to the 15 commands in An Introduction to the Linux Command Line Interface. But I would include the halt command, since I needed it properly shut down the virtual server. Conveniently, being a virtual server, it was running within Windows, so I was able exit the Linux window to look up the halt command.

Generally the topics covered made sense to me. I expect it will be a process of increasing familiarity with them. Typical directories in a Linux file system and basic commands were introduced. I expect I will be referring back to the Learning the Shell tutorial pages about the various directories until I gain a greater familiarity. There are other details I need to review in the not-to-distant future, such as octal numbers for assigning permissions, details of find and grep This should be just a matter of consulting the man pages. Learning the Vi editor looks like it is going to be somewhat of a project. Vi is on the virtual server, so I can practice using Vi while still running Windows. Vi Help is not on the Ubuntu CDs. But it is on the Web at http://vimdoc.sourceforge.net/htmldoc/help.html#help.txt.

Notes on the Linux File System, Disk Drives and Device Nodes

Below are my notes from Sections 2 and 3 of the Introduction to Linux tutorial. The sections are titled "A Look Around the File System" and "Disk Drives and Device Nodes". Also included are some notes from Section 1, the introductory chapter.

I find these notes useful as identification of concepts covered. I'm posting them in case someone besides myself should find them useful.


www.distrowatch.com
www.ibiblio.org/pub/Linux/distributions
www.linuxhq.com
www.AnchorPointBooks.com/etc/linux.html

contents of linux file systems
cygwin- cygnus? - X terminal emulator
ls- list current directory
bin- executable files
linked files end in @
boot- operating system files
dev- device files
etc- system configuration files
home- login home directories
lib- shared libraries
lost+found- used if the system crashes, hold fragments of files
mnt- mount
opt- optional programs
proc- holds running operating system files
root- home directory of the super user
sbin- commands and utilities used by the super user
swap- swap programs in and out of memory as they are running
temp- temp files
usr- catchall, all kinds of stuff
var- datafiles which change when files are running

etc- system configuration files
ls | more - runs list files in current directory and pipes results to more
more, return- scrolls one line at a time
more, space bar - scroll one page at a time
cat- writes a file to the screen

passwd
user name:password:user ID number:group ID number:full name:home directory:name of the program run upon login
-password is in the shadow file, used to be encyrpted in passwd
-usually do not edit passwd by hand

group and shadow files
in /etc
su -superuser
shadow
user name:encrypted password:date(since 1970):restrictions on changing password:number of days to change password:reminder time:number of grace days:date account became disabled:empty field
group
group name:group password:group ID number:can be used to list users in the group

file permissions
every file has an owner, is a member of a group, a type, a set of permissions (read, write, execute),
type: file, directory, symbolic link, device nodes, others too
t rwx rwx rwx - type of file, owner, group, others
ls -l permissions of files in a directory
number of links to the file

change ownership and permissions
touch -if file exists is bring the modification date stamp up to date on a file, or if the file does not exist it creates an empty file
umask -default permission settings, in the form of three octal digits

chmod -change permission settings for a file
chmod 777 myfile give read, write, excecute perm to everyone
or do it without using octals:
chmod O+r myfile give read permission to others
chmod gu+rw myfile give read, write permissions to group and owner (u is owner, (user))
chmod g-w myfile remove write permission from group
chmod a+rW read, write permissions to all

change group
done as superuser
chgrp rpc myfile -change myfile to the group rpc
change owner
chown root myfile -change owner of myfile to root

partition the disk, create filesystems, mount the filesystems

Linux file system
each Linux partition:
boot block (creates a temporary device node so the system can start running)
super block (filesystem information)
inode list (list of disk locations with file locations and sizes)
directory is a file with a list of entries with filenames and index into the inode table
data blocks

device nodes in /dev
device node is also called a device special file
drives are addressed through their device nodes
IDE are hda, hdb, hdc, etc.
hda with 3 partitions - hda1, hda2, hda3
SCSI are sda, etc.
flobby is fd0
ls -l /dev/hda -examine the device node
b - block device like a disk drive, c - a characher device like a terminal (tty device)

partitioning/fdisk
fdisk /dev/fd0 -to partition a floppy
m - menu, follow the menu instructions
ID number -a hex value indicating the file system for the partition (fdisk only makes partitions, not filesystems. The operating systems make the filesystems.)
w -saves the changes

extended and swap partitions
more than 4 partitions, need an extended partition
boot partition -first parition
extended part - second partion
then use logical partitions within the extended partition
a -flags the bootable partition

creating filesystems
operating systems can use each other's file systems but each must create it own filesystem
mkfs /dev/fdo -create a linux filesystem on a floppy disk
creates a new and empty superblock and inode table
linux filesystem is ext (ext ention of the Minux filesystem)
ext 3 is the latest version (with journaling, which helps in system restoration after crashes)
mkfs -t ext3 /dev/hdb1 if ext2 is the default, this overrides it, selecting ext3 for the first partition of the second IDE drive
mkfs -c -t ext3 /dev/hdb1 check for bad sectors
mkfs -t msdos /dev/fdo create a dos floppy
after creating the filesystems, attach them to Linux (mount the filesystem)

mounting filesystems
attaching the filesystem to Linux
create a directory and use the directory as a mount point
for a floppy disk:
mkfs /dev/fdo
mkdir floppy
ls -ld floppy see information for the directory "floppy"
unmounted, the directory is empty
the mount command attaches the floppy disk to the directory
mount /dev/fd0 floppy
now there is a file, lost+found, in the directory "floppy"
umount unattaches the directory "floppy" from the floppy disk, now the directory "floppy" is empty again
/mnt directory with mount points
you can set relationships between the directories and the device nodes in a filesystem table

/ect contains tables of filesystems to be mounted
fstab filesystems table
the filesystems table
device node, directory onto which the disk is to be mounted, filesystem type
iso9660 cd roms
/dev/pts psuedo terminal, a new window
can mount on machine's filesystem onto another, so another machine can be address as if it were local
last two columns of single digits
first digit backup utility (dump)
second digit order of check for integrity (fsck)
eject unmounts and ejects cdrom
mount -a mounts everything in the the fs table

manuals
man more options for the more command
q exits the man page

www.tldp.org Linux documentation project- good place for help

Notes on Software Patents

Microsoft claims the Linux kernel and Open Office each violate over 40 MS patents, and Linux violates over 200 patents.

Open source software developers are prohibited from paying patent royalties by the terms of open source software licensing agreements. Some dispute whether algorithms are patentable, but the compression algorithms for GIF and MP3 are patented, and the patents have been enforced.

According to Technocrat.net (http://technocrat.net/d/2006/6/30/5032), litigating a single patent suit typically costs 2-5 million dollars, and companies can be sued for using software which violates patents.

On its patent policy page (http://www.novell.com/company/policies/patent/), Novell states "In reality, open source software poses no greater risk of patent infringement than does closed source software."

Trivial patents have been granted. Clearly, some of them are not inventions. For instance Amazon's patent of the one-click buy feature: (http://www.gnu.org/philosophy/amazon.html#whyBoycott).

The patent system was created to foster innovation. Much patented software was originally created as open source.

Open source companies are themselves acquiring patents to defend themselves with the threat of counter litigation against those who would sue them.

On April 2008 Red Hat filed a brief in federal court arguing for the limitation of software patents (http://www.press.redhat.com/2008/04/07/red-hat-asks-federal-court-to-limit-patents-on-software/).