Email Robert |
Daynotes
Journal
Week of 12/14/98
Friday, July 05, 2002
A (mostly) daily
journal of the trials, tribulations, and random observations of Robert
Bruce Thompson, a writer of computer books. |
TTG Home
Robert Home
Daynotes
Home
Last Week
Next Week
Monday
Tuesday
Wednesday
Thursday
Friday
Saturday
Sunday |
Monday,
December 14, 1998
A minor format change starting this week. Rather than using one big
two-cell table, I decided to create a two-column table with cells for each
day of the week. The advantage is that I can put the set of navigation
links in the left hand column with each day's text. I can't see any
disadvantages, but if you do, please let me know. Thanks to Bo Leuf for
the idea. He created a similar form for Pournelle's web site, and I
shamelessly stole his idea.
* * * * *
The book of the week this week is The
Thieves' Opera: The Mesmerizing Story of Two Notorious Criminals in
Eighteenth-Century London by Lucy Moore, and the title says it
all. This book chronicles the rise and fall of two criminals in
18th-century London. Jonathan Wild was the ringleader of a ubiquitous
theft ring that operated on both sides of the law. By day, as Thief Taker
General, Wild operated with official government approval as an agent who
helped theft victims recover their property. By night, as London's
Criminal Mastermind, he managed the theft ring, which he ruled with an
iron hand. Then there was Jack Sheppard, an individualistic and rather
likable thief who refused to become a part of Wild's ring, insisting on
operating independently. Set against a richly detailed backdrop of London
as it really was 250 years ago, the real-life characters in this book were
the inspiration for Dickens' Fagin and the Artful Dodger, and later for
Doyle's Professor Moriarty. This is Moore's first book, and she's done a
wonderful job. At age 28, she has a big future. Recommended.
* * * * *
And this message from Roger G. Smith, which scared me until I realized
he was talking about yesterday's Daynotes (with the old format) rather
than today's in the new format (which I hadn't published yet):
Don't know what you did today, but I have to
scroll right about three times to read a line with IE 3. I had not
changed browser settings, and could not find anything that would
correct. Comes Opera to the rescue with its "toggle document
settings," button beside the progress bar. One click and the table
now wraps the text within the browser window. Just another useful
feature in Opera.
Hmm. I don't know what could have caused that. Okay, I
found it. It was Robert Morgan's letter about MP3 rippers and players. I
used IE4, did a View-Source, and scrolled down until I found the long
line. For some reason, the text in Robert's letter hadn't wrapped, leaving
all of one paragraph as one long line. When I viewed the original page in
my browser, it was wrapping fine. I'm not sure why IE4 wrapped the display
properly and IE3 doesn't. At any rate, it's fixed now. Thanks.
* * * * *
I have to take my mother to the dentist today, and tomorrow it's my
turn to visit the dentist. With the holidays and all the other stuff going
on, it's going to be tough to get much done this week, but I have to. |
TTG Home
Robert Home
Daynotes
Home
Last Week
Next Week
Monday
Tuesday
Wednesday
Thursday
Friday
Saturday
Sunday |
Tuesday,
December 15, 1998
FedEx showed up yesterday afternoon with the Red Hat Linux Power Tools,
which are now sitting on my desk with the copy of Red Hat Linux 5.2 that
arrived last last week. I'll probably do an upgrade 5.2 installation to
the 5.1 copy of Red Hat Linux running on freya, but I want to do
more than that. I think what I'll do is retire sherlock as a
Windows NT Server box and turn it into a multi-boot client system running
Windows 98, Windows NT Workstation 4, and Red Hat Linux configured as a
client.
The only problem with doing that is all the dependencies around here. Sherlock
is currently my WinGate server, which provides Internet access to the
whole network. So before I convert sherlock to a multi-boot
client, I need to relocate WinGate to a stable server. The obvious
candidate is bastet, the Windows NT Server box that I'm building
as my network's "resource server," but that'd mean relocating
the modem and phone line as well. Not to mention reconfiguring all of the
proxy applications on all of the clients to point to bastet
instead of sherlock.
Then there's the fact that I'm running WinGate Pro 2.0 right now on sherlock.
I have WinGate 2.1, but I've never bothered to install it. WinGate Pro 3.0
is due out this month, so installing what I now have seems stupid if I'll
only have to turn around and install 3.0 when the upgrade shows up. And,
come to that, the Red Hat Linux 6.0 update is due out soon, and is a major
upgrade, so it may not make sense to spend a lot of time messing with 5.2.
Anything I change around here affects other things.
* * * * *
And this from Roger G. Smith:
Went back and looked at 1207xxxx, still
over-runs the screen. Checking in Opera... ah! Here's another one:
"sneakily and without your permission retrieve from another site
content that you did not ask for and do not want", <--- IE302a
does not wrap this line. Insert a space after the "e" in site,
maybe. Or not. Toggle document setting restored the width of all but
this line, making it easy to find. My respect for Opera grows almost
daily.
Okay. I went back in and got rid of all the hyphens in that
phrase, so it should break properly now. That's not one that showed up as
a long line when I did View-Source. That's unfortunate, because I'd
thought that View-Source was a quick and easy way to find those long
lines. As far as Opera, perhaps I'll have to download it and take a look.
And I almost posted this message without first removing the hyphens
that caused the problem in the first place.
And I'd better get some work done before I have to go to the dentist. I
never get any work done after I've visited the dentist, so the rest of
this day will be burned.
|
TTG Home
Robert Home
Daynotes
Home
Last Week
Next Week
Monday
Tuesday
Wednesday
Thursday
Friday
Saturday
Sunday |
Wednesday,
December 16, 1998
And I didn't get anything done after getting home from the dentist. He
used one of those ultrasonic cleaner things, which leaves my entire head
throbbing. To make matters worse, I seem to be coming down with some sort
of bug. And Barbara rightly insists that I check out the health insurance
proposals she's gotten so we can pick one before she officially leaves the
library system and we have no health coverage.
I happened to see the TV news last night, which I watch once a year or
so. It told me that President Clinton was about to be impeached, that
Burlington Coat Factory was unknowingly selling coats trimmed with
domestic dog and cat fur, and that a woman murdered a pregnant teenage
girl in order to steal her unborn baby. Now I remember why I don't usually
watch the news. On the other hand, they mentioned that if Clinton is
impeached, the Senate could be fully occupied with the trial for a year or
so, which I suppose is a bright spot. At least, per Will Rodgers, they
won't be able to do much damage to the rest of us while they're busy with
that.
I also got back another chapter from my editor yesterday. He says, as
usual, that it needs pruned. He's right, also as usual. And Pournelle and
I are still awaiting the contract copies for our joint book, but it seems
that that's just a matter of time. Things always move slowly this time of
year.
* * * * *
And this mail from Bo Leuf regarding my borrowing the new page layout
he did for Pournelle:
And you're welcome to it, too. People really
liked the recurring daylinks along the left margin to judge by the
positive response Jerry and I both got.
Actually, Pournelle's version is table ..
/table for header and each day, as distinct from your tr .. /tr within a
single table. Same overall effect visually though. The reason for
multiple tables was to try and have each section render independently
without having to wait for the bottom of the page to load. (This, it
turns out does not always work in all browsers... oh well.)
I didn't realize that until I just went back and looked at
the source for Pournelle's page and mine. Given what you say about all
browsers not rendering the tables independently, I don't think I'll go
back and change mine again.
* * * * *
And I'd better get back to work on the book. Deadlines are looming. I'm
doing the best I can, but I'm not going to meet them. And miles to go
before I sleep. |
TTG Home
Robert Home
Daynotes
Home
Last Week
Next Week
Monday
Tuesday
Wednesday
Thursday
Friday
Saturday
Sunday |
Thursday,
December 17, 1998
FedEx showed up yesterday with the Promise Technology FastTrak eval
unit. Now all I need is some disk drives to connect to it. The docs
strongly recommend using identical drives. That means I need two identical
drives to test RAID 0 striping or RAID 1 mirroring, and four identical
drives to test RAID 0+1 striping/mirroring. I can come up with two Seagate
4.3 GB Ultra-ATA drives. Four is more of a problem. But I supposedly have
at least two and maybe four eval Maxtor Ultra-ATA drives on the way, so
we'll see what happens. At this point, I can say that the card appears to
be well constructed, and the docs look decent. I'm looking forward to
putting this thing through its paces.
* * * * *
Just finished filling out the application for Blue
Cross/Blue Shield medical insurance and put it in the mail. We need to get
something in place before Barbara's group coverage expires at the end of
January. After reading all the horror stories about the price of medical
insurance, I was pretty pleased with this policy. It covers routine stuff
fully with a small co-pay, $20 office visits, $10 prescriptions, etc.
There were a variety of deductibles available. We chose the highest, at
$2,500. It pays 80% on hospital stays and so on, with a $2,000 per person
stop-loss/maximum out of pocket. For Barbara at age 44 and me at age 45,
the total premium is $187/month. That's assuming they approve us, which I
expect they will. I've never had any medical problems, and Barbara's have
been pretty minor, allergies and so forth. The only potential problem I
see is that I smoke a pipe, which is a no-no in these politically correct
times.
* * * * *
I notice that we started bombing Iraq again. I believe
that the timing is a complete coincidence, but then I also believe in
Santa Claus, the Tooth Fairy, and the Easter Bunny. It must be a complete
coincidence, because the last time this happened it was a complete
coincidence too. And the time before that. I feel really bad for Mr.
Clinton. It just so happens that every time he really, honestly needs to
bomb someone for very good reasons, he just happens to be having a
personal crisis that small-minded people can point to incorrectly as the
cause. Geez.
* * * * *
And a box of twenty comp copies of Windows
NT TCP/IP Network Administration showed up via FedEx yesterday in
the same delivery as the Promise card. A pretty standard contract term is
that the publisher supplies ten or twenty free copies, called "author
copies" or "comp copies". I know some authors who actually
give comp copies as Christmas presents. Now, many authors are cheap, and I
like getting free stuff as much as the next person, but that strikes me as
stepping over the line. Then, too, people always want you to sign the
book, which always makes me feel funny. Not so much any more as it once
did, but it still seems strange. I'd have no hesitation at all about
signing a fiction book I'd written, but signing a computer book just seems
odd for some reason. |
TTG Home
Robert Home
Daynotes
Home
Last Week
Next Week
Monday
Tuesday
Wednesday
Thursday
Friday
Saturday
Sunday |
Friday,
December 18, 1998
The FedEx lady showed up at dinner time yesterday teetering under a
load of eval units from APC. I've bought APC UPSs for years, and have
seldom had anything but the best results with them. But the most recent
APC UPS I had around here was the two-year old BackUPS 600 that's
currently sitting under Barbara's desk. I wanted to get a look at some of
their more recent products, so I requested eval units of several of them.
I unboxed one of them last night, a BackUPS Pro 650, and plugged it in
to charge. After my experience with the TrippLite 675 a couple of months
ago--the original unit shorted out and threw sparks--I now always plug in
the UPS by itself before connecting any equipment to it. One change since
I last bought a UPS. The federal government now mandates that UPSs must be
shipped with the battery disconnected. Something to do with safety
regulations, although how leaving the battery connected can affect
shipping safety is difficult to say.
At any rate, I unboxed the BackUPS Pro 650, and was very pleased with
it. On average, APC UPSs may cost a bit more than their competition, but
the difference is worth it. APC is known for using high-quality inverters,
big batteries, and so on. Stuff that adds a little bit to the cost, but a
lot to the value. APC also gets the small things right.
For example, one of the minor aggravations of using other UPSs is that
all of the power receptacles are arrayed on the back of the UPS panel
itself. That's fine if all you need to connect is regular power cords, but
it makes things difficult when you have one or more power bricks to
connect. No matter which way you do it, using a power brick blocks one or
more of the other receptacles. With competing UPSs, I always end up
connecting a corded surge protector simply to extend the outlets to allow
me to connect the power bricks. Although a minor inconvenience, this does
cost money and add to the clutter under the desk. Rather than putting all
of the receptacles on the back panel, the BackUPS Pro replaces two of them
with short tail circuit cables which resemble the female end of an
extension cord. These are just long enough to allow a power brick to
connect directly to the UPS without blocking access to other receptacles.
Similarly, APC includes a couple of Velcro cable ties to help control
the rats' nest of cables that typically surround the back of a UPS.
Getting small details right tells me that they've also gotten the major
things right. Although I'll be doing detailed testing and evaluation of
these APC products for the book, I can tell you now from long experience
that when you need a UPS, you should look first to APC.
* * * * *
Yesterday was Barbara's last actual day of work for the Forsyth County
Public Library. Although she's technically still an employee through the
end of the year, she decided to take some of her accrued leave through the
holidays. That brought up an interesting issue. At work, she sends and
receives email on an antiquated DEC VAX system. When she got home
yesterday evening, she asked me if she could connect to the VAX from home
to retrieve her final email before the end of the year. At first thought,
that seemed like a no-brainer. Of course she could. But then I got to
thinking.
Although I believe that the Forsyth County MIS department attempted at
one point to make VAX mail retrievable by POP, they never did succeed in
getting the POP server working. That means that Barbara can't retrieve her
mail across the Internet. Instead, she has to dial in to a terminal server
using an asynchronous dial-up client that supports VTxxx emulation. I
checked around, and I don't have even one system here that has what is
needed to do asych dialup. That made me sit up and take notice. It hasn't
been all that long ago that bulletin board systems and terminal servers
were the main means of connecting to remote systems. No more. Winsock,
PPP, and the Internet have so completely replaced asych dialup that it's a
more or less dead technology.
I told Barbara that we had two possibilities. First, I could dig out an
old copy of ProComm Plus for Windows or QModem, install it, configure the
dial up account, and let her dial in to get her mail.. That seemed a lot
of work for little return, so I suggested as an alternative that she visit
the branch library near here, log on with her own account, and read her
mail there. Asynch is truly dead. Do any bulletin board systems even still
exist?
* * * * *
And this from Roger G. Smith:
Yes, you should have one or more systems
around capable of doing terminal emulation and async' dial-up. The
programs are Windows 95, and windows 98 -- specifically, Hyperterminal
in Win98. Whiile not exactly a feature-fest, it does suppore several
emulations, and will initiate connections on an installed modem. WinNT
3.51 also came with a program called "Terminal" that had at
least one "VT" emulation, so I'm gonna go way out on a limb
and assume that NT 4x workstation and server do also. These programs are
probably more than "good enough" to read and reply to
messages, and have the advantage that Barbara won't have to leave the
house just to check e-mail. She can capture/save any message she wants
to keep, and import a text file into Outlook for later reply -- just not
at the same time you're dialed up if only one machine has a modem
installed.
The _last_ BBS that I know of, not that I'm
paying any attention to the BBS world now, is one I built all the
systems for and has been up 24/7 ever since without a hardware related
incident or downtime. Hmmm, OK, one floppy drive died after three years.
Anyway, he's finally gotten tired of paying the phone bills to serve
just a few users -- not to mention the bill for the satellite downfeed
-- and is pulling the plug 31 December. Bye, bye, CyberGold.
Agree that APC's are the best in their
class.
You're right. I should have, but I don't. Years ago, I
started clearing the communications options when I installed Windows,
assuming that I'd have no need for HyperTerminal, Chat, and so on. So at
this point, I literally don't have a machine in the house that has any
dialer/terminal emulator installed on it. I suppose I could install them,
but it doesn't seem worth the trouble. Barbara stopped at our local branch
library and logged on to check her remaining VAX mail.
You're right about BBSs. As recently as three years ago, I
was running a two-line MajorBBS system here. That makes me wonder what
ever happened to Galacticomm, the folks who created MajorBBS. I guess
they've probably recast themselves as a web software provider, but it must
be disconcerting to have the entire technological foundation of your
business collapse almost overnight. I guess that's the way buggywhip
manufacturers felt when automobiles started to get popular.
And another from Ken Scott:
Windows NT and 95/98 come with an acceptable
terminal client, called HyperTerminal. It will do VT100 emulation, and
works nicely. There are free updates for HyperTerminal on
www.hilgraeve.com. I use HyperTerminal Person Edition 3.0 for all my
telnetting and the little bit of direct dialing that I need to do. I
haven't found any reason to pay the big bucks for something like
ProComm, etc. Hoping this is helpful.
I knew that Windows bundled HyperTerminal, but I never
install it. I guess with disk space rapidly dropping near a penny a
megabyte, that no longer makes any sense. I wasn't aware that there were
updates available from Hilgraeve, though. Thanks for letting me know. I
may just download the updates and go ahead and install HyperTerminal so
that I'll be ready the next time I need an asynch dialer/terminal
emulator.
And mail from Robert Caruso about how I format reader mail for display:
Instead of using the fixed face/indented font for
reader mail and the bold/italic for your replies, why don't you just use
different colors to set off the different elements? The fixed font is hard
to read and the indents just waste space.
Okay, I'm always willing to try something new.
I'd been using those font conventions to make it immediately evident
visually which portions of the text were the reader mail, which were my
response, and which just part of the regular text. I'll give it a try your
way, although I'm not sure which color scheme will work best. FrontPage 98
has a limited number of standard colors to choose from, and some of the
clash with the background or are very difficult to see against it.
I'd be interested in hearing anyone's opinion who cares to give it. Is
the old style indent/monospace and bold/italic method easier to read, or
do you prefer that I use the same type face with different colors? Any
suggestions for color schemes that don't suck would also be appreciated. |
TTG Home
Robert Home
Daynotes
Home
Last Week
Next Week
Monday
Tuesday
Wednesday
Thursday
Friday
Saturday
Sunday |
Saturday,
December 19, 1998
I ran my regular weekly network backup yesterday afternoon, and
immediately ran up against the problem we'd foreseen. Until now, I always
stuck the latest backup tape in Barbara's purse to serve as a
field-expedient off-site backup. Now that she's going to be working at
home, that method will no longer serve. We've considered several
alternatives. Making a weekly run to the bank to deposit the tape in a
safe deposit box seems a bit much, not to mention the fact that I couldn't
get to the backup tape outside of bank hours. We used to spend nearly
every weekend at our friends, the Tuckers, and I considered asking Steve
and Suzy to keep a tape for me. The problem is, we're both very busy now,
and don't see each other every weekend any more. Barbara offered to take
the tape to her parent's house, but that's on the other side of town, and
puts the tape a 40 minute round trip away. I just called Steve, and he
said I was welcome to store a tape at their place. One problem down.
* * * * *
Last night I unboxed the APC Smart-UPS SU-1000/Net, connected the
battery, and set it to charging. I was surprised when I opened the case to
connect the battery just how large that battery was. Actually, there are
two of them, each a 12 volt, 11 amp-hour unit. In total, they look to be
nearly the size of the battery in my first Volkswagen, and larger than the
one in the Honda 750F motorcycle I rode years ago. It's no wonder the
thing weighs 20 kg/44 lbs.
As a rough means of calculating run time, I multiplied out the numbers.
12 volts X 11 amp-hours X 2 batteries = 264 volt-amp-hours. Times 60
minutes per hour = 15,840 volt-amp-minutes. Divided by 1,000 VA = 15.84
minutes. So that UPS should theoretically run under full load for about 16
minutes. But that ignores two factors. First, inverters are not 100%
efficient. Second, the amp-hours that a battery delivers depends on the
number of amps it's providing. In other words, a battery that will deliver
1 amp for 1 hour (a one amp-hour battery) will deliver 0.5 amp for more
than two hours, and will deliver 2.0 amps for less than half an hour. The
relationship isn't linear. The heavier the load, the smaller number of
amp-hours the battery will actually deliver.
Taking both of those factors into account, I'd guess that this unit
would deliver about 40% to 50% of theoretical run time under full load, or
about 6 to 8 minutes. I just checked the APC web page, and found that
SU-1000 run time is rated at 7 minutes under a 900 VA load, and 36 minutes
under a 300 VA load, which seems about right. Note that cutting the load
by a factor of three raises the run time by a factor of more than five.
Once I have a moment, I'm going to relocate this UPS under my main
desk, and use it to power the three Windows NT Server boxes--bastet,
sherlock, and kerby--that reside there. APC also sent me
the Share-UPS hardware needed to shutdown multiple servers from one UPS,
so I'll be installing and testing that as well.
* * * * *
And the following from Bo Leuf:
"Asynch is truly dead. Do any bulletin board
systems even still exist?"
Yes and no :) Apparently few technologies really
die all that dead -- there will always be that die-hard group of
enthusiasts. But likely there are few serious places where non-TCP/IP
dial-up is still being used. The programs have all become
"legacy" applications.
I've run BBS systems since the mid-80s, starting
as support line for my then-Atari-business shop. Numerous
"hobby" networks based on BBS and FidoNet protocol still exist,
both regional and international, albeit with a steadily declining user
base of participating systems and user base. I still run a 24/7 mailer
system on my business line which daily pulls in network mail from the
appropriate hubs, but quite honestly I don't know that I would
bother if it wasn't for the fact that I can use a legacy (atari) machine.
It sits in a corner totally silent, with spun-down HD 99.99% of the time,
and demands no attention whatsoever from me. Perhaps more important is
that the mailer can identify and receive faxes. Most of the network area
posting tend these days to be about what to do concernign the lack of
postings in the areas, and how to patch routings whenever a system leaves
the network...
For some years (say 90-96), several BBS systems
served as Gateways to the Internet, and in particular provided newsgroup
access for interested users. This provided a measure of extended
popularity, at least until the Internet-in-every-home explosion came
along. (I have wondered, tongue-in-cheek, based on the fact that a number
of these BBS systems use non-PC platforms with other date-rollover
characteristics, whether I could quickly write a y2k SF novel where these
networks turn out to be the only global message networks functional come
January 2000. Even if the phone lines go down, there are RTTY links...
Then again, not worth it, we will very soon now experience the real
story.)
I _have_ noted now and then that some companies
will still mention BBS-style async-dial-up customer support lines, even MS
until recently, but the acceptance of using web sites is so common now
that I would not be surprised if these lines are no longer there.
As you note, terminal programs have largely
become obsolete. Modems instead come bundled with fax-programs as a
complement to the ubiquitous Internet "packages", where once it
was a "lite" communications package. The continued
popularity of fax, a legacy technology indeed, is interesting in itself,
and probably says something about business conservatism.
Good points, all. Which brings up a question I've
always had: Why is faxing via Internet not a common application by now? I
understand that traditional fax protocols require deterministic,
predictable data delivery while TCP packet delivery is stochastic and
unpredictable. Still, that shouldn't be a problem since fax reception can
be de-linked from fax printing. There's no reason that fax data arriving
via the Internet can't be cached locally and then printed from the local
cache.
In one sense, it's possible now to
"fax" via the Internet, of course. One can simply scan a
document and email it. Still, it seems to me that there would be a large
demand for Internet-enabled fax machines and applications. Any person or
company with a full-time Internet connection could use one for two-way
faxing, and even individuals with dial-up connections could use one for
outbound faxing.
Given the huge number of faxes sent every day via
expensive long-distance telephone calls, I'd think Internet faxing would
have become a monster application by now. Many large companies could save
thousands or even hundreds of thousands of dollars a month. Even small
companies could save quite a bit by avoiding long-distance fax calls.
Unless I'm missing something, all that's needed is a standard set of
protocols. I wonder why IETF hasn't done this.
One reason, of course, may be whose ox is being gored. I remember
reading figures a year or so ago that said that more than 25% of all
long-distance "voice" calls were in fact fax calls. It wouldn't
surprise me a bit if the majority of overseas calls during business hours
were fax calls. If that's the case, the long distance companies have a
strong incentive to do whatever is needed to keep that business. If tens
of millions of faxes are being sent by long distance every day now,
replacing those per-minute revenues with Internet fax calls at nothing per
minute would put a major crimp in long distance company revenues and
profit. One word that telcos hate is "bypass" which is how they
refer to customers communicating by alternative channels. The widespread
adoption of Internet faxing would constitute the biggest case of bypass
the phone companies have ever experienced.
|
TTG Home
Robert Home
Daynotes
Home
Last Week
Next Week
Monday
Tuesday
Wednesday
Thursday
Friday
Saturday
Sunday |
Sunday,
December 20, 1998
I got the APC Smart-UPS SU-1000/Net UPS installed under my desk. I was
going to run all three of the machines on my desk--bastet, sherlock,
and kerby--from the APC and move the TrippLite 675 under my
credenza, but there's really nothing there at the moment that needs a UPS.
So I put sherlock and kerby on the SU-1000/Net and left bastet
on the TrippLite 675. The next thing I planned to do was install the APC
PowerChute automatic shutdown software on all of my NT boxes. Once I'd
done that, I planned to install the Share-UPS to allow the SU-1000/Net to
shutdown both (or all three) servers connected to it.
I ran into a slight problem which has nothing to do with APC. I'm short
of serial ports. I looked at Barbara's machine, thoth, first.
Although it has two serial ports, only one of them works (I must have
gotten the header cable for the other one misaligned). That one operating
serial port is connected to her PalmPilot cradle at the moment. I plan to
install a fax modem and a fax machine back in her office, which means I
need a second serial port for the modem. I suppose I could use an internal
fax modem, but I greatly prefer external modems. That means I need three
working serial ports--one for the Pilot, one for the fax modem, and one
for the UPS. Similar problems in my office. The resource server bastet
has only one serial port, which I'll need for the modem that WinGate will
use. I'll have to think on this for a while. I hate to install an internal
modem, and I hate to add ISA port cards, but I may have no alternative but
to do one or the other.
* * * * *
This from Bo Leuf about Internet faxing:
> ... a question I've always had: Why is
faxing via Internet not a common application by now?
One could ask similar questions about other
things, both past and present. Remember that it took considerable time
and rethinking before telephony became the killer app -- it was long
considered a no-win compared to the telegraph. Possibly we don't see
Internet-faxing because the fax is already so entrenched with fax
machines, fax modems, and fax apps. Everyone knows about fax by now,
while most are unaware that it is even possible via Internet. Businesses
have long had dedicated fax lines with 24/7 fax machines, and many
people have over the years invested in fax machines or fax apps at home.
In addition, for years we have gotten full send/receive fax capability
for free in our modems, often with the fax software bundled.
Print&click from the application.
To some extent, there has been a slow growth of
Internet fax-points, where the main haul (international) is via
Internet, and with optional "gateways" to POTS S/R faxmodems
on either end. But I would hardly call it a growth industry. Considering
that the typical fax only takes a minute or less, cost is not a big
issue here no matter what the distance is, so you have a small cost with
existing infrastructure competing against something free but not yet
fully deployed. I doubt that very many see any significant advantage,
and perhaps instead only see new obstacles and complications.
All good points, and yet I'm still not sure they explain
why a shift hasn't occurred. Technology is all about better, faster, and
cheaper. A new techology that is only one of these things may or may not
be adopted. A technology that is any two of those things is normally
adopted in relatively short order. A technology that is all three should
be adopted overnight.
> traditional fax protocols require
deterministic, predictable data delivery while TCP packet delivery is
stochastic and unpredictable. Still, that shouldn't be a problem since
fax reception can be de-linked from fax printing. There's no reason that
fax data arriving via the Internet can't be cached locally and then
printed from the local cache.
This is of course perfectly possible; all that
is really needed is a new MIME category and a client software (plug-in)
to collect and display the results. Easier than streaming audio or
IP-phone, that's for sure. On the other hand, when is a fax preferable
to a document "original" when you are already on the computer?
-- assuming that the appropriate MIME and viewing software is available.
The main problem is that much information is today "in the
machine", and fax takes it out of easy machine readability. Most
development is concentrated on making and keeping stuff machine
readable.
Some apps are sort-of aware of the potential
however. Maybe someone will decide it's worthwhile to try. One never
knows.
I have FaxMaker installed, and it goes the
other way by allowing traditional faxes to be received and sent via the
email clients on the network. Hence, I can email the fax server and it
will send the fax via the designated port = fax modem.
A new MIME type would also be useful. Many people have
email addresses, but relatively few have their own personal fax numbers.
Delivering a fax to an email address would solve the troublesome problem
of fax routing, and would also avoid the problem of having a private fax
lying in the out tray of a public shared fax machine. The only real need I
see for a facsimile of an original document is for things like signed
contracts, and even that need goes away as public-key encryption become
ubiquitous and legally accepted.
PS: on reader mail layout...
I think the indent (blockquote) is preferable
no matter what font you use (will also adhere to the markup syntax in
that you are in fact quoting). I will agree that it _is_ a bit more work
for the reader to adjust between monospaced and normal faces, so going
proportional overall is ok, but do keep the blockquote. Making your
responses italic is fine, and note that it can also (should?) be set
inside the blockquote to make it more distinct from the normal text.
I agree that it looks odd not to use indents, so I'm going
to play around with using indents and blue for the reader mail portion,
keeping my own responses non-indented, but bolded and italiced as before.
* * * * *
And the following from Tom Syroid:
Greetings. Long time no chat <G>
Christmas festivities, computer reorganizing, and my work schedule have
kept me busy of late.
I did end up having to redo my HDD structure to
get NT to dual boot with W98. I tried every trick in my book, but no
dice -- NT definately wants to be on the first partition if you want to
do anything fancy like making it aware of Win98 sharing space with it.
Part of my stubborness, I believe, came from the notion that it shouldn't
have to be so difficult. Obviously there are some people in Redmond
who like to challenge us technical folk. Every time I get up to my
eyeballs in structure and installation, though, I have to wonder how
people with less technical insight and know-how than me manage to get
things functioning as they should. I suppose they don't play around like
we do -- they get things running and leave well enough alone...
I've also spent considerable time playing
around with Office 2000. Overall, I very much like what I see. The fit
between applications is definately more polished and the usual slew of
productivity enhancements is welcome. I recognize that I've only
scratched the surface, though, and look forward to using the web
publishing tools in earnest when I start constructing a web site.
Hopefully we'll have the beginning structure up before the New Year.
Well, I'm not sure that it's fair to blame Microsoft for
the problems setting up a dual-boot environment when Windows NT is
installed in the first partition. When you think about it, it's far more
common for a Windows NT user to want to boot Windows 9x than the converse.
I'd guess a fair percentage of Windows NT systems dual boot, but probably
about 99.5% of Windows 9x systems don't.
As far as Office 2000, thanks for the observations. I
haven't looked at it yet, and I may not. I've been told that the current
beta has a forced registration requirement that allows it to be started
only fifty times before requiring you to register the product. I have
never registered software, and I don't plan to start now. If this forced
registration carries over to Office 2000, I won't be upgrading to it.
Here are some of my personal views on your web
site. I've had these in mind for some time now, and with the thread
developing on your Daynotes page, it seems like the appropriate time...
1. Overall, I like the structure of your site
very much. It was what drew me to return in the beginning before I
started appreciating the content so much. As a matter of fact, I like
your structure so much I plan to model my site after yours. I trust you
don't feel I'm infringing here.
Thanks. Feel free to copy the structure to whatever extent
you wish.
2. I liked your original colour (<G>
that's Canadian for color) schemes/themes better than I do your current
ones. I found the green background unique and unobtrusive. I recognize
that colour and style are uniquely interpreted by each and every reader,
so I leave this as a simple personal viewpoint. I guess, on reflection,
I find your site too look-alike now to Pournelle's, and as I read both
everyday, this is significant to me. Pournelle's site is undoubtably
chocked full of insight and valuable information. It's just that --
well, it's so chaotic, if you'll pardon the pun. It's takes
patience to find what you're looking for. I suspect this colour thing is
psychological for me -- your site has ten-fold better structure than
Pournelle's but is beginning to become too much of a look-alike to
Jerry's. And I mean no offense by this -- just that I'd like to see more
differentiation between the two.
Hmm... That's odd. As far as I could tell, my original
theme didn't have green pages. They appeared tannish here, a bit darker
than the current parchment background, but not much. As far as lookalike,
I plead guilty to stealing his parchment background because I liked it.
I'd thought about using a different color, but I like the effect of
primarily black letters on an off-white or tan background. I'm more
concerned with readability than anything else. I did originate the layout,
however. In fact, I originally did my current layout for Pournelle when he
was looking to change his own layout. I offered it to him, but he'd
already settled on another layout. I decided that I liked what I'd done
enough that I'd just use it myself. I got rid of my older style, using the
FrontPage Expedition theme, simply because it was too graphics intensive.
To the extent that our designs look similar, it's probably because we're
both striving for the same things--simplicity, fast downloading, minimal
bells & whistles, and readability above all.
3. I think you're colouring of reader's
comments is a big improvement and makes the unfolding dynamics of
dialogue easier to track. Ideally, though, I'd like to see you put
readers comments on a separate page as in Pournelle's site. I fully
recognize the added work involved in this step, though, and understand
why you do things the way you do. KISS. There are also merits to being
able to track a thread all in one place. I don't like it, though, when your
daynotes consist almost totally of other people's comments and/or
insight. I come to your page primarily to read your insights and
comments. I do value your reader's insights, but consider them
ancillary. I recognize the fine line here. Perhaps threads over a short
paragraph belong on a separate mail page???
I'd actually thought about doing a separate mail page, but
I decided to stick with just embedding mail in-line with my day notes.
There are a couple of reasons for this. First, as you point out, doing a
separate mail page is more work, which is the last thing I need. Second,
my site probably generates as much traffic in a month as Pournelle's does
in a day. I usually get anywhere from three or four site-related messages
on a slow day to perhaps 15 or 20 on a heavy day. I try to respond quickly
to all of them, but many are not candidates for being posted for one
reason or another. If I went to a separate mail page, I could see having
some weeks with only a couple of messages and others with 20 or more,
which seems a bit too much variation to make a good candidate for a
separate page. Third, I think one of the reasons my site is easy to follow
is that everything is in one place.
I'm in this for the long haul, and the traffic levels at my
site are growing constantly, so at some point I may have to go to a
separate mail page just to keep the Daynotes page size reasonable. Until
then, however, I think I'll keep on as I've been doing.
As far as the balance between what I write and what readers
write, I try not to cheat by depending too heavily on reader mail.
However, it's only me, and I'm very busy. As a matter of fact, my editor
called me yesterday to say that they wanted to accelerate production on
the book I'm working on right now. That means that I'll be working 6.5
days a week for the next three months plus, leaving me very little time to
update the site. I'll take some time to do it, of course, simply because I
don't want to lose momentum. But that time will basically come out of my
sleep time.
* * * * *
And this from Chuck Waggoner:
How about a compromise?
The Verdana face looks very much like the
type that IBM used on its 'Secretary' electric typewriter of the '70s.
And, just like on that very expensive electric machine, Verdana IS a
proportional face. What about using that Verdana typeface (I'm presuming
it's available in Front Page) and reducing the indents slightly?
I have always liked the fact that on your
site, the typewriter looking face immediately identifies and reinforces
that the content is incoming mail. Pournelle's site is confusing on this
score, even though he uses different typeface for his own vs. reader
content. Frequently, I find myself having to think about who is writing.
Never had to do that with your site. With the typewriter face, it was
perfectly obvious. Intelligibility seems more important to me than
saving space; my scrollbar works just fine.
Thanks for reminding me why I went to the monospaced face
originally. And you know what? I'm going to go back to doing it the way I
did originally. No one has taken really strong exception to doing it that
way, and you're right--it does make it immediately obvious who wrote what.
FrontPage does indeed provide Verdana, but I don't think I'll use it,
simply because someone who should know once told me I was better off
sticking to standard fonts--Arial, Times Roman, and Courier New.
* * * * *
And more from Bo Leuf about Internet faxing:
> ... A technology that is all three
should be adopted overnight.
That's assuming the ideal free-market
situation and perfectly rational responses from the dominant actors.
Again, I don't think most potential users are even aware of the option.
I was never clear on the MS stand here, and exactly how their fax
service via MSN was supposed to work (i.e. before they totally reworked
MSN and made it sound like CNN-for-the-Net). I got the impression it was
sort of a "gateway" thingy, but required that both ends used
Wintel. I agree that transparent fax management via Office/Outlook would
pretty much make the option ubiquitous, though. I'm sure that MS would
also manage to mangle the standards side of it, as they did with Outlook
email.
Well, the option doesn't really exist right now in usable
form, which was my point. Granted, there are a few IP-enabled fax machines
available, but no one is pushing them. As far as enabling fax
transmissions via email, if Microsoft created such a product using
proprietary standards, I think they'd become de facto standards overnight.
Better still would be if they implemented it using de jure standards, such
as a new MIME type.
I've never had much problem with the way Microsoft
implemented Internet mail standards. Granted, their first release of
Outlook 97 was missing quite a bit, but the Internet Mail Enhancement
Patch (IMEP) fixed most of those problems. I've been using Outlook 98
happily with no problems since it was first released. It's not the most
powerful mailer in the world--I think Pegasus probably still holds that
honor--but it does everything I need to do.
> but relatively few have their own
personal fax numbers.
Oh? I know that surprisingly many people in
Europe have fax machines at home these days, albeit most do tend to use
them on their normal voice numbers in parallel with the answering
machine. Actually it's a real nuisance, because most fax machines are
rarely programmed by the owners to respond with the current
ID&number. Sending a fax is therfore very much a trusting blind
transmission into the great void, hoping you did in fact dial the right
number and remember the area code, because a great many wrong numbers
will still receive faxes and the sender will never be the wiser from the
often cryptic or non-programmed recipient IDs.
I can't begin to count the number of times
there have been local scandals where hospitals or authorities have send
sensitive files via fax that have ended up somewhere completely
different, usually someone's home fax. Several times a year some such
case is recounted in the papers. One common problem is that someone
dials a date from the document header, rather than the number. (Swedish
dates are still commonly written e.g. 981220 on official papers, which
number is often also a perfectly valid local telephone number.)
Certainly many people have fax machines, but I was talking
about relative numbers. For example, in a typical office one might find 50
employees, each of whom has his own email address. But all of those
employees may share one or a few fax machines. A person is far more likely
to have his own email address than his own fax number. Being able to enter
an email address as the recipient for a fax would eliminate most of the
problems you mentioned. Around these parts, the types of errors you
describe would doubtlessly result in some major lawsuits.
* * * * *
And I still need to do something about Christmas presents for Barbara.
Well, it's only the 20th, so I haven't waited until the last minute this
year...
|
|
Coming Soon (I hope)
|
|