«

...random.
In all cases, while i am waiting for something to happen on a
webpage, the number of packets sent and the number of packets received
do NOT change until afterwards.
This does not always happen; seems ranDUMB.

Case #1, trying to register my new credit card online: most of the
time after i fill out the form on a particular page and click on "next"
or "continue", there is that "freeze" which can be as "short" as a few
seconds to minutes (causing timeout to show as a page).

Case #2, in g-mail account, moving from incoming mail to spam (or
trash), same "freeze" (also happens in my localnet webmail account).

Case #3, after typing in my username and password, clicking on sign
in, same freeze.

Case #4, NNTP, click on a conversation title and NADA, so go to a
different one (previous or next one listed),see whole conversation, go
back to the "unseen" one, and all is OK. VERY sporadic, delay time not
noticed and packet counts not observed either as the "dead time" is too
short.

Ref: cases #1-3:
Sometimes, the black progress bar, bottom right, will lengthen some,
slow or stop, lengthen some more, then stop "forever".

"Fixes" tried:
A) click on stop and then re-load (works sometimes);
B) when i see the "timeout" message i try clicking on its "try again"
button (works sometimes);
C) if webpage has a "continue" or "next" or equivalent button, i try
that (usually works).

Have done full Avast! AV scan; clean.
Looked at what was running with ProcessExplorer; no joy.
I have even gone to the extreme of killing 5 of the shields, leaving
only the web shield, the network shield and the script shield running;
no joy.

Is there anything i can run in the background that will "sniff" out
what is preventing communications to/from the net?

Robert Baer wrote:
> ..random.
> In all cases, while i am waiting for something to happen on a webpage,
> the number of packets sent and the number of packets received do NOT
> change until afterwards.
> This does not always happen; seems ranDUMB.
>
> Case #1, trying to register my new credit card online: most of the
> time after i fill out the form on a particular page and click on "next"
> or "continue", there is that "freeze" which can be as "short" as a few
> seconds to minutes (causing timeout to show as a page).
>
> Case #2, in g-mail account, moving from incoming mail to spam (or
> trash), same "freeze" (also happens in my localnet webmail account).
>
> Case #3, after typing in my username and password, clicking on sign
> in, same freeze.
>
> Case #4, NNTP, click on a conversation title and NADA, so go to a
> different one (previous or next one listed),see whole conversation, go
> back to the "unseen" one, and all is OK. VERY sporadic, delay time not
> noticed and packet counts not observed either as the "dead time" is too
> short.
>
> Ref: cases #1-3:
> Sometimes, the black progress bar, bottom right, will lengthen some,
> slow or stop, lengthen some more, then stop "forever".
>
> "Fixes" tried:
> A) click on stop and then re-load (works sometimes);
> B) when i see the "timeout" message i try clicking on its "try again"
> button (works sometimes);
> C) if webpage has a "continue" or "next" or equivalent button, i try
> that (usually works).
>
> Have done full Avast! AV scan; clean.
> Looked at what was running with ProcessExplorer; no joy.
> I have even gone to the extreme of killing 5 of the shields, leaving
> only the web shield, the network shield and the script shield running;
> no joy.
>
> Is there anything i can run in the background that will "sniff" out
> what is preventing communications to/from the net?

Cases 1,2,3 probably use HTTPS protocol. That will use specific outgoing
port numbers (like 443 instead of 80). The conversation with the server
is encrypted.

However, that observation isn't helping matters at all, because
I haven't been able to find any issues with https web pages yet.

I don't think attempting to debug the web browser, is going to lead
to answers fast. I've tried that before, and it sucks.

(Firefox, printing debug output in a Command Prompt window...
I compiled my own copy of Firefox, so I could debug it.
Waste of time. Takes about 2GB of downloads or so, to get the
tools to do it. At least the tools are free.)

http://img16.imageshack.us/img16/1456/v3628running.gif

Paul

Paul wrote:
> Robert Baer wrote:
>> ..random.
>> In all cases, while i am waiting for something to happen on a webpage,
>> the number of packets sent and the number of packets received do NOT
>> change until afterwards.
>> This does not always happen; seems ranDUMB.
>>
>> Case #1, trying to register my new credit card online: most of the
>> time after i fill out the form on a particular page and click on
>> "next" or "continue", there is that "freeze" which can be as "short"
>> as a few seconds to minutes (causing timeout to show as a page).
>>
>> Case #2, in g-mail account, moving from incoming mail to spam (or
>> trash), same "freeze" (also happens in my localnet webmail account).
>>
>> Case #3, after typing in my username and password, clicking on sign
>> in, same freeze.
>>
>> Case #4, NNTP, click on a conversation title and NADA, so go to a
>> different one (previous or next one listed),see whole conversation, go
>> back to the "unseen" one, and all is OK. VERY sporadic, delay time not
>> noticed and packet counts not observed either as the "dead time" is
>> too short.
>>
>> Ref: cases #1-3:
>> Sometimes, the black progress bar, bottom right, will lengthen some,
>> slow or stop, lengthen some more, then stop "forever".
>>
>> "Fixes" tried:
>> A) click on stop and then re-load (works sometimes);
>> B) when i see the "timeout" message i try clicking on its "try again"
>> button (works sometimes);
>> C) if webpage has a "continue" or "next" or equivalent button, i try
>> that (usually works).
>>
>> Have done full Avast! AV scan; clean.
>> Looked at what was running with ProcessExplorer; no joy.
>> I have even gone to the extreme of killing 5 of the shields, leaving
>> only the web shield, the network shield and the script shield running;
>> no joy.
>>
>> Is there anything i can run in the background that will "sniff" out
>> what is preventing communications to/from the net?
>
> Cases 1,2,3 probably use HTTPS protocol. That will use specific outgoing
> port numbers (like 443 instead of 80). The conversation with the server
> is encrypted.
>
> However, that observation isn't helping matters at all, because
> I haven't been able to find any issues with https web pages yet.
>
> I don't think attempting to debug the web browser, is going to lead
> to answers fast. I've tried that before, and it sucks.
>
> (Firefox, printing debug output in a Command Prompt window...
> I compiled my own copy of Firefox, so I could debug it.
> Waste of time. Takes about 2GB of downloads or so, to get the
> tools to do it. At least the tools are free.)
>
> http://img16.imageshack.us/img16/1456/v3628running.gif
>
> Paul
Well, this is horsesh*t..i formatted a drive and put Win2K on it (ie:
working from scratch).
Running "baer" - NO AV software, did not even allow the ASUS NVIDIA
video driver to load the NVIDIA and ForceWare Network Access Manager.

Cannot say anything about cases 1,2 or 3, but case 4 NNTP still see
the same BS and one time when clicked on sci.electronic.design or some
such, i got a message "document is empty" (i think that was the
wording), so clicked on another ?feed? and then back and all was OK.

Am starting to get ****ed.

Robert Baer Inscribed thus:

> Paul wrote:
>> Robert Baer wrote:
>>> ..random.
>>> In all cases, while i am waiting for something to happen on a
>>> webpage, the number of packets sent and the number of packets
>>> received do NOT change until afterwards.
>>> This does not always happen; seems ranDUMB.
>>>
>>> Case #1, trying to register my new credit card online: most of the
>>> time after i fill out the form on a particular page and click on
>>> "next" or "continue", there is that "freeze" which can be as "short"
>>> as a few seconds to minutes (causing timeout to show as a page).
>>>
>>> Case #2, in g-mail account, moving from incoming mail to spam (or
>>> trash), same "freeze" (also happens in my localnet webmail account).
>>>
>>> Case #3, after typing in my username and password, clicking on sign
>>> in, same freeze.
>>>
>>> Case #4, NNTP, click on a conversation title and NADA, so go to a
>>> different one (previous or next one listed),see whole conversation,
>>> go back to the "unseen" one, and all is OK. VERY sporadic, delay
>>> time not noticed and packet counts not observed either as the "dead
>>> time" is too short.
>>>
>>> Ref: cases #1-3:
>>> Sometimes, the black progress bar, bottom right, will lengthen some,
>>> slow or stop, lengthen some more, then stop "forever".
>>>
>>> "Fixes" tried:
>>> A) click on stop and then re-load (works sometimes);
>>> B) when i see the "timeout" message i try clicking on its "try
>>> again" button (works sometimes);
>>> C) if webpage has a "continue" or "next" or equivalent button, i try
>>> that (usually works).
>>>
>>> Have done full Avast! AV scan; clean.
>>> Looked at what was running with ProcessExplorer; no joy.
>>> I have even gone to the extreme of killing 5 of the shields, leaving
>>> only the web shield, the network shield and the script shield
>>> running; no joy.
>>>
>>> Is there anything i can run in the background that will "sniff" out
>>> what is preventing communications to/from the net?
>>
>> Cases 1,2,3 probably use HTTPS protocol. That will use specific
>> outgoing port numbers (like 443 instead of 80). The conversation with
>> the server is encrypted.
>>
>> However, that observation isn't helping matters at all, because
>> I haven't been able to find any issues with https web pages yet.
>>
>> I don't think attempting to debug the web browser, is going to lead
>> to answers fast. I've tried that before, and it sucks.
>>
>> (Firefox, printing debug output in a Command Prompt window...
>> I compiled my own copy of Firefox, so I could debug it.
>> Waste of time. Takes about 2GB of downloads or so, to get the
>> tools to do it. At least the tools are free.)
>>
>> http://img16.imageshack.us/img16/1456/v3628running.gif
>>
>> Paul
> Well, this is horsesh*t..i formatted a drive and put Win2K on it
> (ie:
> working from scratch).
> Running "baer" - NO AV software, did not even allow the ASUS NVIDIA
> video driver to load the NVIDIA and ForceWare Network Access Manager.
>
> Cannot say anything about cases 1,2 or 3, but case 4 NNTP still see
> the same BS and one time when clicked on sci.electronic.design or some
> such, i got a message "document is empty" (i think that was the
> wording), so clicked on another ?feed? and then back and all was OK.
>
> Am starting to get ****ed.

I've had similar behavior with bad net card drivers in M$win and bad
cards under Linux.

--
Best Regards:
Baron.

Baron wrote:
> Robert Baer Inscribed thus:
>
>> Paul wrote:
>>> Robert Baer wrote:
>>>> ..random.
>>>> In all cases, while i am waiting for something to happen on a
>>>> webpage, the number of packets sent and the number of packets
>>>> received do NOT change until afterwards.
>>>> This does not always happen; seems ranDUMB.
>>>>
>>>> Case #1, trying to register my new credit card online: most of the
>>>> time after i fill out the form on a particular page and click on
>>>> "next" or "continue", there is that "freeze" which can be as "short"
>>>> as a few seconds to minutes (causing timeout to show as a page).
>>>>
>>>> Case #2, in g-mail account, moving from incoming mail to spam (or
>>>> trash), same "freeze" (also happens in my localnet webmail account).
>>>>
>>>> Case #3, after typing in my username and password, clicking on sign
>>>> in, same freeze.
>>>>
>>>> Case #4, NNTP, click on a conversation title and NADA, so go to a
>>>> different one (previous or next one listed),see whole conversation,
>>>> go back to the "unseen" one, and all is OK. VERY sporadic, delay
>>>> time not noticed and packet counts not observed either as the "dead
>>>> time" is too short.
>>>>
>>>> Ref: cases #1-3:
>>>> Sometimes, the black progress bar, bottom right, will lengthen some,
>>>> slow or stop, lengthen some more, then stop "forever".
>>>>
>>>> "Fixes" tried:
>>>> A) click on stop and then re-load (works sometimes);
>>>> B) when i see the "timeout" message i try clicking on its "try
>>>> again" button (works sometimes);
>>>> C) if webpage has a "continue" or "next" or equivalent button, i try
>>>> that (usually works).
>>>>
>>>> Have done full Avast! AV scan; clean.
>>>> Looked at what was running with ProcessExplorer; no joy.
>>>> I have even gone to the extreme of killing 5 of the shields, leaving
>>>> only the web shield, the network shield and the script shield
>>>> running; no joy.
>>>>
>>>> Is there anything i can run in the background that will "sniff" out
>>>> what is preventing communications to/from the net?
>>>
>>> Cases 1,2,3 probably use HTTPS protocol. That will use specific
>>> outgoing port numbers (like 443 instead of 80). The conversation with
>>> the server is encrypted.
>>>
>>> However, that observation isn't helping matters at all, because
>>> I haven't been able to find any issues with https web pages yet.
>>>
>>> I don't think attempting to debug the web browser, is going to lead
>>> to answers fast. I've tried that before, and it sucks.
>>>
>>> (Firefox, printing debug output in a Command Prompt window...
>>> I compiled my own copy of Firefox, so I could debug it.
>>> Waste of time. Takes about 2GB of downloads or so, to get the
>>> tools to do it. At least the tools are free.)
>>>
>>> http://img16.imageshack.us/img16/1456/v3628running.gif
>>>
>>> Paul
>> Well, this is horsesh*t..i formatted a drive and put Win2K on it
>> (ie:
>> working from scratch).
>> Running "baer" - NO AV software, did not even allow the ASUS NVIDIA
>> video driver to load the NVIDIA and ForceWare Network Access Manager.
>>
>> Cannot say anything about cases 1,2 or 3, but case 4 NNTP still see
>> the same BS and one time when clicked on sci.electronic.design or some
>> such, i got a message "document is empty" (i think that was the
>> wording), so clicked on another ?feed? and then back and all was OK.
>>
>> Am starting to get ****ed.
>
> I've had similar behavior with bad net card drivers in M$win and bad
> cards under Linux.
>
Thanks for the idea; you might be correct, but i will try everything
else first.
The "net card" is part of an ASIC on the ASUS motherboard, and the
driver is from their own CD, so it is a little hard to suspect either one.
Not impossible - eg: the ASIC might literally be cracked.

Robert Baer Inscribed thus:

> Baron wrote:
>> Robert Baer Inscribed thus:

Snipped

>> I've had similar behavior with bad net card drivers in M$win and bad
>> cards under Linux.
>>
> Thanks for the idea; you might be correct, but i will try
> everything else first.
> The "net card" is part of an ASIC on the ASUS motherboard, and the
> driver is from their own CD, so it is a little hard to suspect either
> one.
> Not impossible - eg: the ASIC might literally be cracked.

I would stick a network card into one of the slots and disable the on
board one. You may need to install the driver but thats a minor issue.
That will at least rule out an on board card problem !

Some of the mainboards used a socket which had the complete network
circuitry built into it. I don't recall any problems with those. The
other was as you say a chip on the mainboard. Asus had a period where
there were severe quality control problems and also piracy issues with
fake mainboards being sold as genuine.

HTH.

--
Best Regards:
Baron.

Baron wrote:
> Robert Baer Inscribed thus:
>
>> Baron wrote:
>>> Robert Baer Inscribed thus:
>
> Snipped
>
>>> I've had similar behavior with bad net card drivers in M$win and bad
>>> cards under Linux.
>>>
>> Thanks for the idea; you might be correct, but i will try
>> everything else first.
>> The "net card" is part of an ASIC on the ASUS motherboard, and the
>> driver is from their own CD, so it is a little hard to suspect either
>> one.
>> Not impossible - eg: the ASIC might literally be cracked.
>
> I would stick a network card into one of the slots and disable the on
> board one. You may need to install the driver but thats a minor issue.
> That will at least rule out an on board card problem !
>
> Some of the mainboards used a socket which had the complete network
> circuitry built into it. I don't recall any problems with those. The
> other was as you say a chip on the mainboard. Asus had a period where
> there were severe quality control problems and also piracy issues with
> fake mainboards being sold as genuine.
>
> HTH.
>
Have had and used this MB for about 2 years or so without seeing this
problem.
Will see if there is a way to disable the on-board ethernet and try a
3rd-party board.
Any reasonably-priced suggestions?

Robert Baer Inscribed thus:

> Baron wrote:
>> Robert Baer Inscribed thus:
>>
>>> Baron wrote:
>>>> Robert Baer Inscribed thus:
>>
>> Snipped
>>
>>>> I've had similar behavior with bad net card drivers in M$win and
>>>> bad cards under Linux.
>>>>
>>> Thanks for the idea; you might be correct, but i will try
>>> everything else first.
>>> The "net card" is part of an ASIC on the ASUS motherboard, and
>>> the driver is from their own CD, so it is a little hard to
>>> suspect either one.

Just a thought... Is the driver on the CD correct for the Windows
version you are using ?

>>> Not impossible - eg: the ASIC might literally be cracked.
>>
>> I would stick a network card into one of the slots and disable the on
>> board one. You may need to install the driver but thats a minor
>> issue. That will at least rule out an on board card problem !
>>
>> Some of the mainboards used a socket which had the complete network
>> circuitry built into it. I don't recall any problems with those.
>> The other was as you say a chip on the mainboard. Asus had a period
>> where there were severe quality control problems and also piracy
>> issues with fake mainboards being sold as genuine.
>>
>> HTH.
>>
> Have had and used this MB for about 2 years or so without seeing
> this problem.
> Will see if there is a way to disable the on-board ethernet and try
> a 3rd-party board.
> Any reasonably-priced suggestions?

You should be able to disable the on board network card in the BIOS.
Press DEL or F1 when booting. Sometimes the key presses can be missed
so I stick a blank floppy in the drive to stop the machine. Then
CTRL+ALT+DEL will re-start the boot sequence.

A cheap 10/100 card can be obtained for £3 - £5 in the UK. I would
think similar in the USA.

--
Best Regards:
Baron.

Robert Baer wrote:
> Baron wrote:
>> Robert Baer Inscribed thus:
>>
>>> Baron wrote:
>>>> Robert Baer Inscribed thus:
>>>>
>>>>> Baron wrote:
>>>>>> Robert Baer Inscribed thus:
>>>>
>>>> Snipped
>>>>
>>>>>> I've had similar behavior with bad net card drivers in M$win and
>>>>>> bad cards under Linux.
>>>>>>
>>>>> Thanks for the idea; you might be correct, but i will try
>>>>> everything else first.
>>>>> The "net card" is part of an ASIC on the ASUS motherboard, and
>>>>> the driver is from their own CD, so it is a little hard to
>>>>> suspect either one.
>>
>> Just a thought... Is the driver on the CD correct for the Windows
>> version you are using ?
>>
>>>>> Not impossible - eg: the ASIC might literally be cracked.
>>>>
>>>> I would stick a network card into one of the slots and disable the on
>>>> board one. You may need to install the driver but thats a minor
>>>> issue. That will at least rule out an on board card problem !
>>>>
>>>> Some of the mainboards used a socket which had the complete network
>>>> circuitry built into it. I don't recall any problems with those.
>>>> The other was as you say a chip on the mainboard. Asus had a period
>>>> where there were severe quality control problems and also piracy
>>>> issues with fake mainboards being sold as genuine.
>>>>
>>>> HTH.
>>>>
>>> Have had and used this MB for about 2 years or so without seeing
>>> this problem.
>>> Will see if there is a way to disable the on-board ethernet and try
>>> a 3rd-party board.
>>> Any reasonably-priced suggestions?
>>
>> You should be able to disable the on board network card in the BIOS.
>> Press DEL or F1 when booting. Sometimes the key presses can be missed
>> so I stick a blank floppy in the drive to stop the machine. Then
>> CTRL+ALT+DEL will re-start the boot sequence.
>>
>> A cheap 10/100 card can be obtained for £3 - £5 in the UK. I would
>> think similar in the USA.
>>
> Well, i paid about $2.50 (free shipping) to get what the included
> "user guide" says it is a Realtek 8169 10/100/1000Mbps LAN card.
> That is the only clue as to maker.
> A small 3" CD came with it that supposedly includes software drivers
> for all Win OSes from Win95 to Win7, for DOS, Linux, Netware, and maybe
> the kitchen sink as well.
> Hell, it could even have a driver for Win9 and 10 also...
> BUT.
> Absolutely NO way to tell, as it is Z-rated (worse than X-rated or
> Y-rated)...it is completely unreadable.."Not High Sierra or ISO-9660
> format" or some such.
>
> So, i snoop on the web and find what seems to be drivers...Realtek has
> various drivers but they cover a lot but NOT Win2K; other places say
> they cover Win2K and a lot of other OSes, BUT,,, BUT,,, motorBUTT.
> The unZIPping gives "not a Win32 program" or some such.
>
> This is real bullshit.
>
>

I bought a couple of those. TPLink brand. With RTL8169SC chips
on them. Turned out to not be a particularly good purchase.

The problem with 8169, is it uses a lot of interrupts per packet
processed. By my math, around 5 interrupts per packet. Whereas
the motherboard NIC on my system, uses 1 interrupt per packet.

With the processor on this computer, that card will do around 70MB/sec.
(I get 117MB/sec using the motherboard Marvell NIC chip instead.)
By extrapolation, you'd need a Core2 processor running at 4GHz, to make
that 8169 run at full gigabit rates. It's a bit of a CPU hog.

I got the same "miniCD" with drivers on it. At first I was worried
it wouldn't fit my drive, but my drive tray has an indentation
the same size as the miniCD, so it wasn't a problem.

I think you can also get an 8169 driver here. The chip is made
by RealTek, so why not ?

http://www.realtek.com.tw/downloads/...&GetDown=false

If you look on this page, you can see the 57.19 driver I got.

http://www.realtek.com.tw/downloads/...Downloads=true

WinXP 32/64 and *Win2K* Auto Installation Program (SID:1492533)
5.719 2012/7/19 5038k

The link from that page, while I can paste it here, you can't
expect this to work for very long. The link is to an FTP site,
with the username and password embedded in the link.

ftp://WebUser2:YbwD5pB2@207.232.93.2...9_04062012.zip

Now, the file I got a couple years ago was "PCI_Install_XP_2K_5719_11202009".
It's strange that the release number can remain the same, and the
date field end up changed. In any case, that claims to be a Win2K
driver.

In that ZIP is a "WIN2000" folder, and an INF like "Netrtlx.inf".

; DisplayName Section DeviceID
; ----------- ------- --------
%RTL8169.DeviceDesc% = RTL8169.ndi, PCI\VEN_10EC&DEV_8169

So you can give that driver package a try and see if it works. You
don't have to use the setup.exe in the package if you don't want to.
Navigate to the WIN2000 folder, right-click on Netrtlx.inf and select
"install" from the right-click context menu.

Paul

Baron wrote:
> Robert Baer Inscribed thus:
>
>> Baron wrote:
>>> Robert Baer Inscribed thus:
>>>
>>>> Baron wrote:
>>>>> Robert Baer Inscribed thus:
>>>
>>> Snipped
>>>
>>>>> I've had similar behavior with bad net card drivers in M$win and
>>>>> bad cards under Linux.
>>>>>
>>>> Thanks for the idea; you might be correct, but i will try
>>>> everything else first.
>>>> The "net card" is part of an ASIC on the ASUS motherboard, and
>>>> the driver is from their own CD, so it is a little hard to
>>>> suspect either one.
>
> Just a thought... Is the driver on the CD correct for the Windows
> version you are using ?
>
>>>> Not impossible - eg: the ASIC might literally be cracked.
>>>
>>> I would stick a network card into one of the slots and disable the on
>>> board one. You may need to install the driver but thats a minor
>>> issue. That will at least rule out an on board card problem !
>>>
>>> Some of the mainboards used a socket which had the complete network
>>> circuitry built into it. I don't recall any problems with those.
>>> The other was as you say a chip on the mainboard. Asus had a period
>>> where there were severe quality control problems and also piracy
>>> issues with fake mainboards being sold as genuine.
>>>
>>> HTH.
>>>
>> Have had and used this MB for about 2 years or so without seeing
>> this problem.
>> Will see if there is a way to disable the on-board ethernet and try
>> a 3rd-party board.
>> Any reasonably-priced suggestions?
>
> You should be able to disable the on board network card in the BIOS.
> Press DEL or F1 when booting. Sometimes the key presses can be missed
> so I stick a blank floppy in the drive to stop the machine. Then
> CTRL+ALT+DEL will re-start the boot sequence.
>
> A cheap 10/100 card can be obtained for £3 - £5 in the UK. I would
> think similar in the USA.
>
Well, i paid about $2.50 (free shipping) to get what the included
"user guide" says it is a Realtek 8169 10/100/1000Mbps LAN card.
That is the only clue as to maker.
A small 3" CD came with it that supposedly includes software drivers
for all Win OSes from Win95 to Win7, for DOS, Linux, Netware, and maybe
the kitchen sink as well.
Hell, it could even have a driver for Win9 and 10 also...
BUT.
Absolutely NO way to tell, as it is Z-rated (worse than X-rated or
Y-rated)...it is completely unreadable.."Not High Sierra or ISO-9660
format" or some such.

So, i snoop on the web and find what seems to be drivers...Realtek
has various drivers but they cover a lot but NOT Win2K; other places say
they cover Win2K and a lot of other OSes, BUT,,, BUT,,, motorBUTT.
The unZIPping gives "not a Win32 program" or some such.

This is real bullshit.