Re: LogOn Scripts? - POP3 timeout?

From: Robert Aldwinckle (robald_at_techemail.com)
Date: 01/10/05 

Date: Mon, 10 Jan 2005 11:56:20 -0500

"Neil Gould" <neil@myplaceofwork.com> wrote in message
news:t6mEd.13228$by5.11055@newssvr19.news.prodigy.com...
> Recently, Robert Aldwinckle <robald@techemail.com> posted:
>
>>>>> "Neil Gould" <neil@myplaceofwork.com> wrote in message
>>>>> So... I have a question related to the issue I'm trying to deal
>>>>> with. Where are the "0x800CCC0D" error messages generated, at the
>>>>> ISP server or from within OE? BTW -- I've followed threads
>>>>> related to this error message for over a year, and read the MS
>>>>> info on it, but find nothing relevant to my issue, nor has my ISP
>>>>> been able to address this matter.
>>>>>
>>>> It's generated by OE. What's the whole error message?
>>>>
>>> Depends. Sometimes I get a "long version" about ports POP3 25 & 110
>>> being unavailable, other times it's suggests that perhaps the
>>> account isn't configured correctly. Both are bogus, because as I've
>>> mentioned, doing a refresh (F5) connects immediately. This also
>>> doesn't happen all the time, but it does more often than not. It's
>>> mainly an annoyance that the ability to do logon scripting would
>>> have gotten around.
>>
>> Finally we are getting a description of the problem
>> and not discussing a self-prescribed solution to
>> (the previously unstated) symptoms. ;)
>>
> [...]
>> Does this occur often enough that you could endure the discipline
>> of continuing diagnosis?
>>
> As I've mentioned earlier, I've been "diagnosing" this problem for over a
> year, now. With all due respect, I don't believe I'm in the dark about the
> nature of the problem, only about viable solutions. ;-)

I think your above question "where are the error messages generated"
demonstrates that you have been musing about the symptom but I don't
see any clear signs of your diagnosis. I asked you about frequency which
can make a big difference in how successful a complete diagnosis would be
but for some reason you seem reluctant to share details.

>
> The problem is caused by having to wait for the server to open the POP3
> ports 25 & 110 when I'm going on-line. If it takes more than a few
> seconds, then OE times out and I get one of the error messages.

Are you saying you have a problem with both your SMTP and the POP3
connections? I only have a problem with my POP3 connection (port 110).
The POP3 connection is always started before the SMTP one by my OE.

> A while back I asked here about being able to reset this timeout value,
> but no joy with that "solution" either.

Which was?

> Logon scripting would be a quick and easy
> fix, but unfortunately, that isn't supported either.
>
> [...]
>
>> I have reproduced the symptom while using telnet and think I
>> understand
>> what is probably happening but never thought to use Refresh as
>> a workaround. Great idea! But when do you do this?
>>
> I can hit F5 immediately after receiving the error message and the
> connection is then made instantly 100% of the time.
>
>> While a timeout value is still in effect? Won't it affect the other
>> operations?
>>
> There is no relation between the logon timeout problem and any of those
> settings, and none of my other usual activities are affected.

How long do you have to wait for the error message?
Isn't that related to your timeout value?
E.g. I mentioned that I now use 30 seconds as my timeout value
and that this means that when the Read is going to fail I can reply Stop to it,
thus generating a consistent error condition. The most important point
about waiting until I can reply Stop is that it allows other operations to continue.

>
> [...]
>> The problem has never been consistent enough for me to bother
>> trying to trace it to get an even better understanding of what I guess
>> will turn out to be a deficiency in OE's timeout handling.
>>

To elaborate, I think there is something which isn't working quite right
with the Wait function since (statistically) the number of times that I have
tried it and it has failed compared to the number of times that I have had to
wait that long when using telnet to simulate (what I assume to be the conditions
that OE has at the time of error) or had a failure there just doesn't match.
However, as I mentioned I haven't done any tracing to understand this
any better than that guess.

A long time ago I experimented with maximum timeout value.
What I can recollect about that test is that then there were two symptoms
to deal with, an early error symptom (which perhaps is what you are seeing?)
or an intolerably long wait for the Wait/Stop dialog. My POP3 server
now seems too reliable to try simulating that possibility with telnet but
I can easily imagine that it should be capturable there too if it occurred
often enough.

BTW I tried pressing F5 after a reasonable wait but before the Wait/Stop
dialog was presented and got a message about downloading newsgroups!
Then the Read worked without an error. This was with a Synchronize All
so it makes sense that multiple connections might be being started simultaneously.
That idea had not occurred to me before either. I will start using netstat
to check that out. Hopefully the discipline of doing that will come more easily
than activating tracing--I forgot to do that and so don't have any diagnostics
yet for how an intervening F5 works in terms of connections and requests.
I have captured a WAIT and a STOP but haven't analysed them yet.
I would still like to capture a "normal case". I didn't realize how often I was
resorting to using STOP and just letting the read occur whenever it could later.

;)

Robert
---