Nubies' Corner
The internet is a puzzling place, but a
few bits of information can make it a whole lot easier to navigate
successfully.
How the internet works:
User end: At the computer
terminal, three things are necessary for an internet connection.
First is an operating system geared for net. Second is a modem for
data transfer in serial form to the phone line (or high speed access
conduit). Third are clients for interacting with data on the net. The
most common is the browser (Netscape, IE, et al), but there are many
other clients for various internet and intranet interactions.
System: On the Macintosh, Open
Transport (OT) is the system protocol for transfer of data. OT is
responsible for serializing the data and seeing that it gets to the
correct port (usually the modem port). There is also modem firmware,
drivers, and scripts (for various applications) to make the modem
function in synch with the processor. And last but not least, there
is the connection protocol. Apple Remote Access (ARA) is the system
native means of connecting, but there are others that can be used if
there are compatibility issues. Personally, I'm still using FreePPP.
In System 9.x, configuration of these components is done with the
Internet Control Panel, usually on first launch of the system,
auto-launching Internet Setup Assistant. But these components can be
altered at any time. Familiarity with these items is highly
recommended to keep the net connection as smooth as possible. Contact
with the internet provider for specific settings is a must, and this
information must be correctly entered. Never be afraid to call a
provider for information.
Modem: I won't bother talking about
every modem out there. There are just too many out there, and more
all the time. The three most common types out there are 56k Flex,
Cable, and DSL. 56k Flex is the protocol that most providers offer,
as it encompasses most of the previous protocols into a single
package, accepting protocols from 14.4k to 56k. It uses audio signals
to verify data transfers and speed transfers, but it is still
primarily digital. The speed of a 56k modem is usually less than 56k,
the telephone connection being the most limiting factor for speed.
The noisier the line, the slower and less stable the connection.
Weather also play havoc with line conditions, with heat, humidity,
lightning static, etc. DSL (digital subscription line) is a high
grade phone line designed for high speed digital transfer. Speeds of
150k - 500k are common, and in some conditions, even faster speeds
are possible. Likewise, Cable is a faster medium, but cable has a
distance limitation and will slow down when shared throughout the
network. The more people using the cable system, the slower the
overall transfers.
Clients: When first connecting to the
internet, most people assume that the browser is the only software
for transversing the internet. Nothing could be further from the
truth. FTP, e-mail, instant messaging, IRC, Hotline, intranet
clients, MUSH, time server clients, weather clients, application help
clients, Java clients, direct user to user transfer clients, tracers,
and remote control clients are but a very few of the available
clients for the internet. The term client here is not someone buying
goods and services, but rather the user-end application meant for
connecting to a server of the same type (server/client). The mark of
maturity of an internet user is the number of clients the user has
mastered. There is no short cut for getting on the net and using the
clients for gaining practical experience. Some are simple, some quite
complex, but all provide an incentive for using while allowing for
two way flow of data. There are always new clients to try out, no one
ever masters all of the available clients. Many never get past the
browser, which in itself is a configuration monster to tackle.
Provider: The provider is the
user portal onto the net. Depending on the service, this portal can
be full time (dedicated) or part-time (dial up). The dedicated line
is naturally more expensive, but many users need full time
connections, especially the server owners. (More on servers in the
next section.) With most providers, the dial up system starts with a
handshake protocol. With a 56k modem, the most common protocol is
called Point to Point Protocol (PPP). PPP is a three stage
process.
Initiating connection: Upon dialing
into the proper number, the user modem and the provider modem must
synchronize and establish transfer and feedback channels. This is
noted by the all too familiar sound we hear when using the 56k modem
is commanded to connect. This procedure sets the transfer rates to
the same speed (relatively, since it is actually asynchronous, but
getting in the same ballpark is necessary), establishes a
verification channel (parts of the stream is constantly being
compared by echoing back a portion of the received data to be sure
the sequences match), and then sends the connection to the next
stage.
Log/pass: Providers have a separate
processor that keeps track of all who use the connection service.
This is most often called the Name Server. It's usually a very small
processor whose sole function if to verify the transmitted log in
name and password. Most user system protocols allow for this
information to be stored for automatic transmission, but some prefer
to keep this step manual to prevent unauthorized internet access.
When the user name and password match, the connection is then passed
on to the router.
Router: The internet interface truly
begins at the first router. Until you reach a router port, you aren't
on the internet. The internet is essentially a massive set of
interconnecting routers. Routers are dumb computers, meaning that
user interface is of secondary importance. It is machine to machine
gating mechanism. To it, there are commands, addresses, and data, and
its purpose is to move data as quickly and efficiently as possible.
If you want to see routers in action, I suggest grabbing a good trace
route client (such as
WhatRoute for Mac) and seeing
where your connections go and how long they take.
DNS name servers: The internet operates
on a four 8 bit set address, with the numbers in the address being
from 0 to 255 (8 bit max). This gives a 32 bit address, and it's
usually expressed as 217.92.121.6 (xxx.xxx.xxx.xxx) with all leading
zeros implied but never written. These addresses have assigned call
names (me.com, you.com, us.net, them.org, otherthem.nl,
englishmen.co.uk, evenotherthem.dk, ruskies.ru) which are held in a
router used name server called DNS registry. Also adding the prefix
to a server name (other.me.com, that.you.com, all.us.net) is a means
of assigning a port number (255.40.119.6:80). Using ports allows the
router to assign which entry into the server that the connection is
made.
Servers: Data has to come from
somewhere, namely another computer that speaks a language that can be
understood by the user. Servers are normal computers that are
specialized for the purpose of data distribution.
A server can be any computer,
literally. With the proper software, any user computer can be reached
by another computer. If it is distributing data on command, it is
acting as a server. What kind of server it is depends on the software
being run to power the server. Server software can be as simple as
SnapperHead
that allows a user to show their desktops to another user's browser,
or it can as complex as OS X Server and WebObjects for Mac G4 servers
for fully customizable e-solutions (such a buzz word!). Commercial
servers are often a group of processors, each performing separate
functions, like page feed, ftp, mail, shopping cart, chat, streaming,
database, customer info, etc. In between lies a world of computer
variations, no two exactly the same. For each client available, there
is a server software to feed it data. Many can be set up on dedicated
DSL or cable modems by individuals, while others pipe directly into
the trunk with T1 and T3 modem connections for high volume traffic
where many hits are being aimed on a daily basis.
The biggest boon to the net are the
search engine servers. These are massive databases that accumulate
contact information about the net and present the data in a
searchable interface. There are two basic types, those that require
submissions in order to be listed, and those that search the web on
their own (spiders, crawlers, etc). The crawlers seem to be the more
popular these day, since the debut of
Google, as the submission based
engines have a tendency to get clogged with advertising, favoritism,
clutter, and misdirection that try to steer you their ($) way instead
of your own. The advantage to the submission engines is that they
tend to be censoring mechanisms for use with those who don't want to
encounter all possibilities on the net. Hence they often take on the
guise of web portals and try to appeal to certain social segments.
The crawlers tend to be more user direct and efficient, getting one
more quickly to the information sought, but they are also wide open
to the entire web contents. Crawlers don't take you into areas unless
you ask for it, but they can take you anywhere. Here, the issue is
primarily one of access control, and those who regulate themselves
have the advantage of a more efficient search with less commercial
clutter and distraction.
Nothing can be said about all servers
except that they dispense data on demand with the proper protocol.
Most can be understood with the proper experience, none are alien to
the human race. But they are fraught with transfer controls,
permissions for access, scripting, and esoteric languages. The art of
spanning the platforms with net based applications is a tricky one.
One seldom reaches everyone with a server. The art of operating a
server is reaching as many users as possible, though some get wrapped
up in specific protocols and offer data that only a few can use.
Everyone hits sites they can't use, and the call for universal access
is a strong one. However, not everyone hears the message, and there
will always be sites loaded with bells and whistles that only a very
few users can use. Everyone must expect connection failures when
browsing the net. It is the server owner who loses by alienating
users, but some server owners do get a bit myopic about their server
projects.
Among the servers on the net are Mac
servers that deal with Mac file protocol.
The one area where Macintosh is most
profoundly different from all other platforms is dual file fusion. In
essence, and file on a Mac can contain one or two files, which are
called forks (data and resource). All other platforms contain only
one file. If a file on Mac is to be shared with other platforms, it
must be reduced to data fork only. For web based formats (HTML, JPEG,
GIF, FLASH, et al), any client that transfers the file from user to
server must remove the resource fork from the file before it will be
accepted. This takes place automatically by the Mac based client. For
these net based formats, the resource forks are not necessary for net
use (or Mac use), though some of the function of the file on the Mac
(like previews, custom icons, etc) will be lost. These functions
usually appear in secondary files on other platforms. In non-net
formats (like applications, QuickTime movies, etc), the resource
forks are often necessary for the file to function at all. In these
cases, a file must be encoded to a single fork. The most popular way
of doing this on a Mac is with StuffIt and binhex (.sit.hqx). This
allows the complex files to arrive and then decode into the two part
Mac file protocol.
However, when files are transferred
directly from a Mac to a Mac, the resource fork need not be stripped.
The internet will transfer the dual natured Mac files without damage,
as long as they are not written to a processor of another platform.
They can be transferred as is without encoding from client to server
or server to client. There are a number of direct file sharing
protocols for Mac, including the native FileSharing for TCP. They
need not even require an extension (ie: .gif) for transfer to take
place. As long as it's Mac to Mac with nothing in between except
routers, it's the same as any Mac network.
But unless the protocol is Mac to Mac,
the files should be data fork only and named with the appropriate
extension so that the MIME protocol can identify the type of file it
is (see Helpers section in the browser preferences). Every server
must have a MIME protocol for the file types by extension if it is to
arrive intact. If a file arrives as an unknown, even though the
extension was correct, it is because the server does not recognize
the MIME type for the extension. Correction for this is for the
server own to place the file's MIME type in the server MIME database.
There is also the text protocols to consider, but this more a case of
running files from a different platform through a converter
application. Mac has automatic conversion available from a number of
system sources, depending on the system version.
So, that is the net in a nutshell.
There is a lifetime of details to be found online, no end to
investigations and learning curves, and no end to the potential for
the money that can be poured into it. But at the same time, it can be
a global contact system that greatly reduces the cost of contact. The
art of the internet is finding resonance that reinforces our
lifestyles.
The current digital age is certainly an
art renaissance (I say this despite how ugly the net can be). The
introduction of Photoshop, paint applications, vector art
applications, 3D applications, animation applications, audio/video
applications, music generation and encoding, and more has used the
internet to change the very mediums that we accept as art.
Imagination is breaking new ground daily at a rate never before seen
in human history.
Time will record the net as the
beginning of a new age in human culture. People are finally breaking
the appearance barrier and being taken seriously solely for their
expressions. While the "DotCommers" scramble for jobs trying to
commercialize the net (after all, it's where the money arises to
support the net structure), there is a world of people exchanging
resonance in web pages, chat rooms, software, information, and so
much more. A little common knowledge of the net helps a great deal
when trying to navigate it. A little more is needed when dealing with
the specifics, and that is gained by venturing on through the router
maze.
Hardware
CD-r burners are the current rage for
saving files to backup inexpensively. There is no less costly medium
for storing files off drive, with 700M disks available for as little
as $0.50 each. CD-RW, which can be rewritten sell for under $2.00
each.
The CD-RW drives are many, so I won't
get into tech specifics about which brand to purchase. My
recommendation is to spend a few extra bucks and get a drive capable
of 8 X writing and 24 X reading. The speed of a disk is usually
labeled with figures like 4 X 4 X 8, 12 X 10 X 32, etc. These numbers
are multiple factors applied to a standard base speed for CD play.
The first is record speed (music), the second is write speed (data),
the third is playback speed (read). An 8 X write speed will burn a
CD-r in less than 10 minutes. A 4 X takes about 20 minutes to burn a
full 700M disk.
Do be warned that CD-r and CD-RW are
sensitive media, and that not all burn attempts will be successful.
Improper burns leave a useless disk commonly called coasters, as they
little other value when they turn out bad. Removal of all dust
particles is essential, so a can of clean compressed air, commonly
used to clean photographic negatives before printing, is a must. A
dirty CD-r disk is a guaranteed failure.
Also, care after the burning is a must.
Finger prints, scratches, and other marrings on the surface can
destroy the disk. These are not regular CD's that can be tossed
about, left lying out of their case on a tabletop, maintain
scratches, withstand exposure to direct sunlight, withstand the
mildest soaps or solvents (ethyl alcohol is the only approved
cleaner), or be cleaned with a standard CD cleaner. Always handle a
CD-r by the very edges or the center hub where the surface doesn't
exist. They need to be placed carefully in the trays as well. It's a
rather simple habit to gain, more a case of continued awareness when
handling than hardship. Keep them out of the hands of kids and slobs
is probably the hardest part of using them.
A normal CD is created with a laser
burning holes through a reflective film, which is then encased in the
plastic medium of the CD. The CD-r contains a dye film already bonded
to the surface of the plastic, and the laser of the writer then
evaporates away the bluish dye. So the surface of the CD-r is much
more sensitive to contact, much like the surface of a floppy disk.
True protection with CD-r's comes with being extremely careful, and
are even more secure when multiple copies are made, one master for
creating more that is never touched except to duplicate more, and
however many copies are needed for normal use. Even with multiple
copies, the CD-r is the least expensive storage medium, so don't
worry about cost. The main difference is a little more time spent
burning.
However, the CD-r is not made of tissue
paper. It will withstand many plays, and with care, it will travel,
play tunes in a car CD player, and do all the things that a CD will
do. They will work in any CD drive for which the system is
compatible.
As with all disks, there is a directory, a
set of desktop files, and ID for the type of system. CD-r's can be
burned to HSF+, HSF, Mac/Windows hybrids, Windows, straight
audio........ The selection is made with the burning software. For
the Mac, the most popular burning software for data is Toast by
Roxio, but there many
others, including Apple's new
iTunes, which is music oriented. The distinction between record and
write is a firm one, and the division of CD-r burning software into
data and music is distinct. Music can stand a few bad data bytes
without dire consequence, data and applications often must have
perfect copy to remain intact. Shopping around for the right
burner(s) is a must to get the system best suited for the user's
needs.
The DVD-RAM market is quite similar to
the CD-r/CD-RW, however burners will seldom handle both types of
disks. DVD is written to two sides in a much more compact storage
system. The top grade Mac production G4's have DVD-RAM, which is used
to write DVD disks which can hold two hours worth of encoded
audio/video. However, a single DVD-RW disk can run $40 and up, hardly
competitive for standard storage as is the CD-r and CD-RW.
For anyone with a disk space problem
(Is there ever enough hard drive for the digital fanatic? We're
talking closet space here.......), the CD-r is an ideal means for
keeping data that is not needed for immediate access. I personally
keep all my internet acquired software installers on CD-r, as well as
a backup copy of my hard drives. Thus far, not counting duplicates,
I'm holding 35 written disks, holding over 20 G of data. The cost in
disks is under $30 (counting the 3 coasters I've made). If the same
number of disks were all purchased today at the best price, the cost
would have been right at $20. This equates to $1 a gigabyte. Not a
bad price to feel secure in regard to the problem of data loss.
Systems
The big news out of Apple in March was
the release of Mac OS X (finally!). However, the immediate
sales for OS X do not appear to be as good as wished by Apple. since
there are far too many industry standard applications that have yet
to make the transition from PPC to Carbon. True, there are quite a
few applications that are Carbon based, but the truly professional
application design houses (Adobe, Macromedia, Corel, Microsoft, just
to name a few at the top) have yet to make the Carbon transition.
With so many gaps in the lineup, OS X will be a cutting edge novelty
that will appeal primarily to those who can afford an extra machine
just to experience all the goodies that OS X offers. But until the
industry standards make the transition, OS 9.x will continue to be
the main Mac system for publication, regardless of the medium, be it
print, electronic, internet, whatever.
This is not to say that OS X is not a
worthy system. It is truly a masterpiece among operating systems.
Allowing multiple systems to function simultaneously with virtually
everything capable of background operation for superior multitasking
is the backbone of the system, with plenty of new features to make
the new Mac lineup a truly superior processor. Adding to the Mac
format is the power of a native Unix structure with Java 2 that makes
it the supreme internet platform available today for all phases of
content, from static to database to streaming to interactive. Here at
Apple.com is where you can
find out all the features it has to offer. For those wanting a hot
running portal to the net, there is nothing better today, hands
down.
But, there is the lag of software
titles in the conversion to OS X that will prevent it from being the
pick of the top. Until the designers of the top software titles make
the switch, OS X will be met with reluctance by those it was designed
to serve. A Photoshop-less Mac is not something many are willing to
make their flagship. Those running MS Office would have no business
software to run. Surprisingly, shareware is the largest sector of the
designer business that is prepared for OS X. The commercial
application designers stood to lose too much money by continuing
development.
In main part, the fault must lie with
Apple. CarbonLib 1.0.4 was stable. It ran all the Carbon apps with no
problem. But it did not do all that Apple wished of it in their OS X
plans. Then 1.1.x was released to the designers, but not the public,
and it was stable enough unto itself. But it represented the first
major change in the OS X basic system. Carbon 1.0.4 based
applications could not reliably run on Carbon 1.1.x. Likewise, 1.1.X
designed applications would run on 1.0.4. Applications were emerging
that had no Carbon support capable of supporting them in the public
circles. Then Carbon 1.2 was released, an extension of 1.1.x with
even more changes to the basic structure, and the bugs began in
public. Then came 1.2.1, 1.2.2, and then 1.2.5, and each version
seemed to introduce new bugs in the older Carbon applications. Even
AppleWorks was crippled. Try using AppleWorks 6.0.x or another
Carbonized word processor application and perform keyboard shortcuts
(Command-) and see what happens to the Finder's access to the
foreground....... freeze freeze freeze. Everyone was forced to revert
to CarbonLib 1.0.4 to operate Carbon apps. This left out a number of
the more advanced Carbon applications from being on most people's Mac
OS processors. The reason for the bugs: Apple changed some of the
basics of Carbon so that it would better implement their plans for OS
X. This change in the basic structure was responsible for the bugs,
scuttling in mid-design many of the designers plans to Carbonize. The
reason Adobe has ceased development was to see what Carbon would be
for the final release of OS X. This is the case for most of the major
design houses, the design halted until a final OS X SDK was offered,
and why they currently have no finished Carbonized applications for
the new system.
So, OS X, originally designed to
require Mac OS 9.0, now revamped to require MacOS 9.1, was released
with a lot of missing pieces, not from the system, but from the
supporting applications at the industry standard level. Rest assured
that as soon as OS X was released, the design houses got into gear
and are now working furiously to turn out full product lines in
Carbon. But this will take months to accomplish. And add another
month for the final products to be printed on CD-ROM's and pre-sales
to gather income, before these tools become a part of the OS X
system. Within a year, most of the applications for Mac OS should
have seen the conversion to Carbon, and then OS X should be a
smashing success, gathering top line users from all platforms. But
until then, Apple will have to take it on the chin for making
improvements in mid-stream.
So, if you are planning to purchase OS
X, be sure that your Carbonized applications are up to CarbonLib
1.2.5 specs before installing them, unless you're just fond of
bugs.