|
|
<Voltar para Apostila e dicas> Don't Blame the Modem This section is probably of most interest to owners of V.90, K56Flex
& X2 modems who are tearing their hair out because they are the proud owners of 56k modems that simply will not
connect at 56,000 bit/s. Once again, there is much below that will be of interest to all modem users. Contents A Little
History
Brief explanation
The situation now
V.90 The one-eyed man's view What Can
You Do to get better connection speeds?
DUN shows 57,600 or 115,200 bit/s
DUN takes a long time to connect
DUN Server Types [+ screen shot]
Check it out practical measures to improve connection speed
How long is yours? Speed impact of the modem line
And how many of them? Speed impact of other telephones
And are they unusual? Speed impact of other devices
And is it corroded? Speed impact of noise
Or too thin? Speed impact of telephone gain / AGC
Or out of your control? Speed impact of SNR NETCOM
x2 Line Tester A Little History In 1998 - if you believe what each company says - Rockwell Semiconductor
Systems and US Robotics both won the commercial war that they had been waging against each other in the
previous year for the new standard in modem telecommunications. The reality, of course, is that modem users lost
& are still suffering now, a year later. Brief
explanation: In February 1988 the ITU introduced V.90, a new
internationally-agreed standard for modem-to-modem
communications that promised 56,000 bits-per-second [bit/s] downstream-rates and
33,600 bit/s upstream
rates; this standard was ratified in September of the same year. Now, what you
have just read is already untrue and part
of the hype that always surrounds these things. 56k bit/s and 33.6k bit/s are
theoretical maximum downstream/upstream
rates and almost never achieved in practice - indeed there are people that will
never ever achieve even the
maximum upstream rate in either direction, but more of this later. Before V.90
was introduced Rockwell [now Conexant,
the largest manufacturer of modem chipsets in the world] and US Robotics [now
taken over by 3Com] had attempted to use their
commercial muscle to enforce mutually-incompatible and proprietary standards
developed by themselves - K56Flex™
from Rockwell and X2™ with US Robotics. This latter situation had led to ISP's
having to have multiple dial-up
lines - one for K56Flex™, one for X2™, one for ISDN and another for previous
standards. V.90 now promises to do
away with all of this nonsense, but of course we all know about promises. The
situation now: Modem chipset manufacturers and software writers are desperately
trying to produce silicon and software
drivers that conform to the standard, and succeeding, more or less. The
situation is compounded by many things:
At the ISP's server Headend is the digital equivalent of a modem This
communicates at 56k upstream and
33k downstream - the reverse of the V.90 modem in your system.
These are made by different manufacturers Freeserve's Digital Modems are
provided, as an example, by
Cisco, and declared upon the Freeserve support pages to be a Beta version
of the V.90 standard. Digital Modem
manufacturers - just as with personal modem manufacturers - have managed
a better-or-worse conformance to
the standard.
V.90 was ratified in Sep 98 and your modem chipset - let alone the modem
- had probably already left the
manufacturing plant by then. Most of the chipset manufacturers do not
produce the modem, only the chipset.
Software takes twice as long to produce as the person writing it thinks...
...and is then full of bugs [the polite word for errors]. Firmware is
software in silicon.
Windows98 came out before V.90 This means that even a Windows98 standard
modem driver will not connect
at V.90 speeds.
Any of the above would make life difficult; put them together, & it
is a miracle that any modem connects at all. As
one example, your modem - which is more-or-less compliant with a standard
- has to connect with the Headend
Digital Modem - which is more-or-less compliant with the same standard.
Get the picture? V.90 As
long as you do not mind being led by a one-eyed man, here is as untechnical an
explanation of V.90 as I can manage.
As a non-telephone engineer I'm grateful to Alan Fowler for his input on this. The
telecommunication system, which our mothers use to ask why we haven't phoned
them recently, & we use to connect
to the internet, was originally designed to allow one human to talk to another
human. This technically means turning
sound waves into electrical energy at one end, pushing that electrical energy
down a copper wire connecting the two ends,
and then turning electrical energy back into complaints/laughter/love or
whatever at the other end. Now, since all
things including conversation are a two-way exchange two copper wires are needed
- the famous twisted pair. This is simple
enough once a microphone and a speaker has been invented & each party has
one of each and there are only two people -
you possibly know the story of when Alexander Graham Bell invented a third
telephone and joined it to the first pair of
telephones. He tried to dial one of the first two phones & found the line
was engaged. Another
issue comes with distance, and distance is always going to be an issue, as why
use a phone if you are close together?
A simple twisted-pair wire will work up to a distance of a number of miles, but
eventually the sound at the other end
will get fainter & fainter until inaudible. This is due to the resistance of
the copper wire to the electrical signal [if an
alternating signal - such as sound waves translated into electrical energy - is
used, the term impedance is used instead; an alternating current
suffers more loss in a copper wire than does a direct current.] This translates
into the 'gain' on our telephone line. A telephone
which is close to the exchange needs very much less gain than one which is far
from the exchange. Telephone gain is
one of the aspects which will be talked of below in terms of getting a better
communication speed from the modem. Modems
are designed to allow a computer - which talks in 'bits' of information - to
communicate along a system which was not
designed with them in mind. The modem does this by mod-ulating [convert computer
bits into noise] and dem-odulating
[convert the noise back into computer-speak] - hence mod-dem. The problem comes
when the other person sounds
very tinny. This would happen a lot with early telephone exchanges or
international calls, and still does now with cellular
phones - so much for progress. This is due to both the restricted bandwidth of
the telephone system [meaning that the
telephone attenuates both the bass & treble of the human voice into
extinction] and also frequency shift or distortion. Humans are much
cleverer than computers and, normally, much less pedantic. If mother sounds like
she has her head in a tin can, or is
accompanied by shoosh-ing noises, or is very faint, or is insisting on playing
her Val Doonican records at full volume in the
background again, we can still make out the words. A computer would process this
ever so much more quickly than us
and then throw up a dialog saying 'fatal exception' and shut down. No contest. A
computer, however, can communicate
along a twisted-pair copper wire at high speed if allowed to do so in a way that
suits its own, pedantic self - a
network card, as an example, will do this at 10,000,000 bit/s, no problem - but
even with a clear line the practical maximum
speed of voice-based communication is close to the V34bis standard of 33,600
bit/s. Enter now the digital telephone
network. Telco's [Tel-ecommunication
Co-mpanies - technology produces acronyms like frogs produce spawn] across the
world have been replacing
their old electro-mechanical exchanges with brand spanking-new digital exchanges
for the last two decades now. This
white-heat of technological innovation has led Britain to the point where even
parts of Scotland & Wales are telecommunatically
digitalised. Of course, other parts are not, and folks served by these older
exchanges can neither release
their caller-id info nor communicate faster than V34bis with a modem. Such
advances have also been avoided throughout
Russia, which means that if you decide to visit our Slavic cousins you had
better take an old 14,400 bit/s modem
with you should you desire to keep in touch with the 'civilised' world.
Elsewhere these changes now mean that instead
of having a pretty switchboard girl put us on hold it can be done by a piece of
silicon instead, and we can listen to some
music whilst we wait. If you are at all unsure as to which type of exchange your
telephone is serviced by, here is a simple
way of finding out: the next time you make a telephone call [try the speaking
clock, it doesn't mind] press any button on the
telephone. If you hear a tone the exchange is digital. If it is a series of
clicks it may not be, & I suggest you ask your Telco to
find out. The
problem for a digital telephone network is that it still had to connect &
work with the old, analogue [voice-band] exchanges.
As an example, I've still got a BT speaker-telephone that I bought in the
late-70's for £100 - equivalent to approx.
£1,000 now. This is a pulse-dial phone rather than tone-dial, yet I can still
use it successfully to dial a number. If the
network had gone completely digital then everyone would have had to buy digital
telephones - which is the situation with ISDN
now - and not talk to their neighbours who were still connected to an old
voice-band exchange. Instead the network
compromised by being digital internally and voice-band externally. Voice calls
are carried internally across a 64,000
bit/s channel and, since the trunk-lines connecting exchanges are far wider than
this, many voice-calls can be multiplexed
across the same line. For most domestic subscribers only the final part of the
line - the copper wires from exchange
to house - are voice-band, and some commercial concerns install a digital link
direct to their premises. This means
that a typical domestic voice call gets at least two conversions - first is
analogue-to-digital, then digital-to-analogue -
although curiously there can be other conversions along the way. These
conversions inevitably introduce errors and give a ceiling
to the maximum speed at which a data-call could be made of 34,822 bit/s [Shannon
1948] in a channel capable of 64,000
bit/s. V.90
performs it's magic by assuming that the network is digital on the downstream
side and voice-band on the upstream side. Now,
I know, I know - why not assume that it is digital both ways? but that's how it
is, so try to live with it or get an ISDN
link. The ISP has a direct digital link to the main exchange and thus one of the
conversions is absent, reducing potential
errors and increasing the maximum potential speed at which communication can
take place. Notice, however, that word
'potential' - other factors on the line can reduce the actual
practically-achievable speed, & this leads nicely into the next
section: What can you do to get better connection speeds?. First, a quick bucket
of cold water [called reality] for owners of 56k modems - you will never achieve connection speeds of 56,000 bit/s,
not ever, and any speed over 33,600 is a 56k connection. Another initial question, of course, is 'What speed is my system
connected at?' The easiest way to find the starting connection speed with a dial-up line is to hover the cursor over the little icon sat
in the system tray after connection [the system tray is the part of the Task Bar that contains the time, and the Task Bar is the
grey-strip normally at the bottom of the screen that contains the Windows 'Start' button]. This icon looks like two computer screens connected by a line
and belongs to Dial-Up Networking [DUN]. As with all things in life DUN can lie. DUN shows 57,600 or 115,200 bit/s then the modem's .inf file
is either the wrong one, or inaccurate, & DUN is reporting
the DTE speed rather than the DCE speed [go here for more information]. If the
.inf file is the correct one then go to Extra
Settings and enter:-
Accepted Wisdom for Showing DTE Speed
Chipset
Extra Settings
Motorola
atq=1\v=4
Rockwell PCI
at+mr=2
Other Rockwell
w2
USR
&a3
All
at&f1 The last one sets the modem to factory-defaults.
Unfortunately even this desperate measure will not work for some modems,
& one reason may be that the Responses section of the .inf file is missing
such as the following [this comes from my own
modem's .inf file]:-
[MfgAddReg]
...
; Add the following rates to support V.90
...
HKR, Responses, "CONNECT 45333", 1, 02, 00, 15,b1,00,00,
00,00,00,00
HKR, Responses, "CONNECT 46666", 1, 02, 00, 4a,b6,00,00,
00,00,00,00 Now,
whilst K56Flex is in steps of 2,000 starting at 28,000 & finishing at 60,000
bit/s, V.90 is in steps of 1,333, leading to some
very strange connect-speeds indeed. If the .inf file is missing any of these
speeds then the computer may default to show
the DTE speed. If you are determined to correct this terrible error then it
means either hacking the .inf file & re-installing
the modem and/or hacking the Registry direct. With my modem this is HKEY_LOCAL_MACHINE\System\CurrentControlSet\Services\Class\Modem\0000\Responses
but you can find your own
by using mdu.exe. This was developed by & is copyright BVRP Software, &
can also be sourced at www.modemhelp.com.
It principally sucks info from the Registry but will also print out the ati
responses if the com port is
selected. Beware, as wrong entries in the registry can bring your computer
grinding permanently to a halt, & in this you are on
your own. DUN takes
a long time to connect Whilst this is the speed of connection rather than the
connection speed it seems to be
very common. What follows is specific to Freeserve but may have relevance for
other ISP's, particularly if you also have AOL.
The typical picture is of dialling out with long, long periods of hand-shaking
but no connection with an eventual
time-out resulting in the whole process being repeated over again, perhaps many
times, until connection is eventually
made [or not]. Whilst there is excellent help on the Freeserve support pages
here is a quick troubleshooter for this one:-
Check the Network & DUN settings A screen shot of the minimum Network
entries needed is here. Unless
your setup requires other network-facilities any others should be removed.
Do not change settings in entries here
unless you know exactly what you are doing - changes here over-ride those
in DUN with no warning. Indeed it
may be worthwhile to follow instructions on the same page as the
screen-shot to remove & re-install DUN and
thus the dial-up adaptor & retrieve your Freeserve account as to
obtain those defaults.
The DUN settings that are important here are on the Server Types page of
DUN Properties. It is reached My Computer->Dial-Up
Networking->highlight the Freeserve connectoid & go File |
Properties->Server Types page.
If yours looks any different than the screen shot above then a probable
answer has been found for DUN taking a long time to connect. For the curious,
NetBEUI is a Windows-for-Workgroups peer-to-peer protocol and
IPX/SPX is a Netware protocol.
A tick in the box 'Log on to network' indicates
a missing or corrupt Dial-up adaptor in Network.
an account with AOL
If this does not solve the problem then the best course of action is to
start the modem troubleshooter here. If that
doesn't find the problem & your system is OK then the likelihood is
that the ISP is having software and/or
hardware problems of their own.
Something to experiment with is to put a tick into Enable Software
Compression; this has speeded up
downloads for some people with some files. True hardware [i.e. not soft]
modems implement compression in silicon
- this is the LAP-M, MNP & V42Bis standards - so ticking this option
should not make a difference. However, it
has helped some & you may wish to experiment. Check your house Having
sorted out the computer & the modem as potential culprits attention now
moves along the Line of Communication to the
house. One first check - is the power supply to the computer earthed?
Non-earthed power supplies have been reported
to halve modem connection speed.
How long is yours?
Shorten the modem line-cable. Many people have a line extension from the
phone socket to the modem which
is many metres long, and this modem line is responsible for far more
speed loss than most people realise. Indeed,
someone reported a 10,000 bit/s increase in speed just by replacing the
modem cable for one from another
manufacturer, yet of the same length. The ideal is to have as short a
length of phone cable between the telephone
exchange and the modem as possible and, whilst this does lead to visions
of plugging the modem into a socket
positioned on the outside of the exchange, there are many practical steps
which everyone can take inside of their
own house to shorten their internal high-loss lines. If you are
contemplating replacing your own modem line-cable
do remember that there is no such thing as a standard modem cable, and
that you will need to obtain the correct
replacement - take the old one along for comparison. Indeed, one computer
manufacturer recently supplied cables
which caused every telephone in the house to 'screech' on connection.
And how many of them?
Detach every other device. Whilst a telephone in every room may seem
convenient it will eventually cause
problems. This is usually considered in terms of Ring Equivalence Number
[REN] & means that if you have too
many devices on the telephone line one or more of them will not ring if
someone phones you, but it also impacts
the connection speed that the modem can establish. The issue here is one
of the 'gain' on the line, & it is mentioned
above and below. Each device will suck some of the juice from the line
& may not leave enough for the modem.
Testing this out is easy enough - detach every device [be it telephone,
fax, answering machine, or whatever] - and see if it
makes a difference. If it does then re-attach them one-by-one to see
which one(s) make a big difference.
BT Relate phones These have been reported to reduce internet connection
speeds. Try unplugging before net
connections.
And are they unusual?
Sky Digiboxes. These are reported to reduce connection speed. Please be
aware of any contract details on
minimum length of connection to the line.
DACS boxes If you or a neighbour have had a second or new line installed
by BT & it has been achieved with a
DACS box then you will never achieve connection speeds above 33,600
bit/s. This is particularly aggravating for
people that installed the second line specifically for internet
connection! DACS is a means of multiplexing two
voice-band telephone numbers across one physical telephone line &
will be betrayed by the presence of a box on
the outside of the house. It is obviously cheaper for the Telco to do
this than to install another physical twisted-pair
line, but the trade-off is in reduced bandwidth for at least one
connection. BT's Terms of Trading state that it is
required to provide a line capable of carrying faxes [i.e. 14,400 bit/s]
and therefore installing a DACS line is well
within this contract. It is possible to stipulate before installation
that any additional lines are achieved with a Direct
Exchange Line [DEL] but too late afterwards. One ray of hope may be that,
usually, only one of the DACS lines is
of highly-reduced bandwidth & a call to 151 may be able to get the
modem put onto the high-bandwidth line.
Get a cable line... It is an odd situation. As mentioned above, the same
twisted-pair wire on an internal network
will allow 10 million bit/s, American experience with digital subscriber
line technology has shown the existing
infrastructure to be capable of 1.5 million bit/s or more, and even BT
will drive it at 64,000 bit/s, yet when 2 lines
are multiplexed it becomes capable of no more than 14,400 bit/s. It has
to be said that those with the best
connection speeds usually seem to be on cable lines. 'Nuff said.
..for data connection. A [British] Telewest customer discovered that the
box in the street had the line to each
customer designated as either a voice or data customer. Persuading the
engineer to switch his line to data doubled
his connection speeds.
And is it corroded? Try the quiet line
test. The Bible tells us that the moth and rust doth corrupt, and this most
certainly is true at the
telephone junction box. Typical signs are a noisy line. BT maintains an
automated quiet line test [amongst other facilities] at 17070. My own
experience with BT from many years back indicates that their twisted-pair line
is
actually a twisted-threesome of which only 2 lines are used. The
engineers can switch this at the exchange without
physical intervention and possibly bypass a poor connection on the third
wire [worked for me]. Other points are
unshielded wires passing close to electrical motors/machinery introducing
interference/hum onto the line. The
biggest culprit here, of course, is probably the computer. Line cards
within the local or main exchange are also
capable of generating noise.
Or too thin?
Pump up the volume. The final aspect in being able to affect the speed of
connection concerns the 'Gain' on the
line [see also here & here]. This means, quite simply, is the line
too faint? To use, once again, BT as the example, it
normally subjects the line to Automatic Gain Control [AGC]. This usually
works well for a voice call but is not
necessarily most effective for modem communication as the modem usually
contains AGC within it's own codec,
& the two can conflict. Many users have reported dramatic speed
improvements after asking for the gain to be
increased/decreased. Just in case you are now confused by the fact that
the gain may need to be put up or down,
it is quite simple. 'Gain' talks of how much voltage the Telco is using
to drive the telephone signal down the line. They do not want to melt the cables so the
voltage is quite low, but a voltage it is nonetheless. If your house is
close to the exchange this voltage can be quite low, whereas if far from
the exchange it will need to be much
higher. Also, across time corrosion at the various junctions will
increase the resistance on the line & the voltage will
need to be a little higher. Dialling 151 & asking for the gain to be
changed will automatically cause AGC to be
switched off. If you explain that it is to allow a better internet
connection the engineers are usually very helpful.
Or out of your control?
Check the SNR. There is one more item that impacts dramatically on
connection speed - but there is nothing that
you can do about it - and this is the signal-to-noise power ratio [SNR]
on the entire line. Now, this is a topic for real
techno-sluts & take what follows under advisement as I'm a little
rusty on these matters. SNR is usually expressed
in positive decibels [dB] and the original measure of logarithmic power
ratio was, of course, the Bell [after Alexander
Graham Bell - who else]. dB is used because you need to shout ten times
harder to be heard twice as loud. Or
something such. The US telephone network has a theoretical noise floor of
39.5 dB and realistically achieves 35 to
36 dB [here more is better - I have no info on equivalent figures for the
UK network]. 56k connection requires 44.4 dB or
better. Modems make a much more intimate examination of the line than you
might expect when they connect, &
line quality is only one of the factors. Some modems also are able to
give a report following connection which
includes line quality. First,
Use HyperTerminal and follow the instructions until you enter +++
following the words login:. At this point the
modem is still connected on the line to the ISP server & can report
on many things to do with the connection. The
command strings vary modem-to-modem:
Checking Signal-to-Noise Power Ratio [SNR] during Connection
Rockwell Chipsets
at&v1
at%l
for line signal
at%q for line quality
All in -dB.
A line quality under 25 is OK.
Lucent
LT WinModem Chipsets
ati11
8 Estimated Noise Level
[No
base reference - more is worse]
USR Chipsets
atl6
ati11
aty11
All
in +dB.
A line quality over 40 is good.
Don't forget to enter ath0 to hang up.
NETCOM x2 Line Tester Netcom [in the UK] run a very nifty line-tester
that will produce multiple results, and
amongst which is the SNR [shown below in dark blue] plus a graphical
display of the line frequency-response. If the
line can support x2 it can also support V.90. Use HyperTerminal &
enter the lines as shown below. The
connection is standard V34bis, so don't get the collywobbles at what
appears to be a slow connection. A sample
from my own modem follows [the colour graphical output is much prettier,
by the way, but shown here in text for speed of
display]. You will also find a Zoltrix Phone Line Test here.
+++ate1
OK
atdt0845 079 8022
CONNECT 31200/LAPM
Connected to ARCHOST v1.05
03-04-99 05:39:00
Caller number: 13380
Do you want ANSI/PC color graphics? (Y/N) >N
Please wait, testing in progress...
+-----------------------------------------------------------------------+
|
This connection will support x2!
|
+-----------------------------------------------------------------------+
Press any key to continue...
+-----------------------+-----------------+-----------------------------+
| x2-STATS Version 1.01 | 03-04-99 05:39
| Caller 13380
|
+-----------------------+-----------+-----+-----------------------------+
| Elapsed Time
00:00:12 | Modulation
V.34+ |
| Blocks Received
1 | Speed
31200/31200 |
| Blers
0 | Symbol Rate
3200/3200 |
| Blocks sent
22 | Carrier Frequency
1829/1829 |
| Link Naks 0
| Trellis Code
64S-4D/16S-4D |
| Blocks resent
0 | Nonlinear Encoding
ON/ON |
| Link Timeouts
0 | Precoding
OFF/ON |
| Chars sent 0
| Shaping
ON/ON |
| Octets sent
158 | Preemphasis
6/0 |
| Chars lost
0 | Rx Lev/TX Lev/SNR 20.0/18.2/44.2
|
| Chars Received 0
| Echo Loss Near
Far |
| Octets Received
1 | Roundtrip Delay
18 |
| Protocol
LAPM | Retrains Request/Grant
0/0 |
| Block Size
128 | Fallback
Enabled |
| Window Size
15 | HST Line Reversals
0 |
| Compression
NONE | HST Equalization
Long |
| Dictionary Size
| SV: 07/26/96
DSP: 06/14/96 |
| String Length
| Reason:
Online |
+-----------------------------------+-----------------------------------+
Press any key to continue... +---------------------------------------------------------------+
| -18 | . x x . . . . . . . . . . . . . . . . . . . . . . |
0 |
| -20 | . X X X X x . . . . . . . . . . . . . . . . . . . |
2 |
| -22 | . X X X X X X X x . . . . . . . . . . . . . . . . |
4 |
| -24 | X X X X X X X X X X X x x . . . . . . . . . . . . |
6 |
| -26 | X X X X X X X X X X X X X X X x x x . . . . . . . |
8 |
| -28 | X X X X X X X X X X X X X X X X X X X x . . . . . |
10 |
| -30 | X X X X X X X X X X X X X X X X X X X X X . . . . |
12 |
| -32 | X X X X X X X X X X X X X X X X X X X X X X . . . |
14 |
| -34 | X X X X X X X X X X X X X X X X X X X X X X x . . |
16 |
| -36 | X X X X X X X X X X X X X X X X X X X X X X X . . |
18 |
| -38 | X X X X X X X X X X X X X X X X X X X X X X X x . |
20 |
| -40 | X X X X X X X X X X X X X X X X X X X X X X X X . |
22 |
| -42 | X X X X X X X X X X X X X X X X X X X X X X X X . |
24 |
| -44 |=X=X=X=X=X=X=X=X=X=X=X=X=X=X=X=X=X=X=X=X=X=X=X=X=x=| 26 |
|Level+---------------------------------------------------+Atten|
| 0
0 0 0 0 0 1 1 1 1 1 1 1 2 2 2 2 2 2 3 3 3 3 3 3
|
| 1
3 4 6 7 9 0 2 3 5 6 8 9 1 2 4 5 7 8 0 1 3 4 6 7
|
| 5
0 5 0 5 0 5 0 5 0 5 0 5 0 5 0 5 0 5 0 5 0 5 0 5
|
| 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
|
+---------------------------------------------------------------+
Press any key to continue...
Thank you for using the NETCOM x2 Line Tester. We hope you found what
you were looking for!
The NETCOM Operations Team
Press any key to continue...
Caller number: 13380
Thank you for calling.
Disconnecting now.
NO CARRIER
+++ath0
OK
My Motorola soft modem regularly achieves 49,333 bit/s on this line. Finally, you can go here if you fancy trying to tweak your modem or here
if you want to optimise your internet connections. Both of these sites are mostly used by games fanatics - I have no
personal experience of the help offered. In addition, this site's Downloads & Links page has a whole slew of sites offering various
form's of help.
broken links | mail me | modem experience All of the views expressed on this page are my own and are not intended
to represent those of any other person or commercial entity, particularly those of Freeserve Ltd., The Microsoft Corporation or any
other software/hardware manufacturer. All advice given is drawn from personal experience and/or the experience of others and is genuinely
intended to assist. Please take responsibility for your actions - I cannot and will not be held responsible for any damage caused to your hardware
or software by actions taken which are inspired by writings on this or other pages. All efforts have been taken to ensure the veracity of what
is written - if you know of any errors or omissions then please let me know & I will amend/add it at the earliest opportunity.
Since 20 February 1999
All trademarks respected as the property of their respective owners By Manegher |