Re: [exim] Failed to get write lock

Αρχική Σελίδα
Delete this message
Reply to this message
Συντάκτης: W B Hacker
Ημερομηνία:  
Προς: exim users
Αντικείμενο: Re: [exim] Failed to get write lock
Daniel Tiefnig wrote:
> W B Hacker wrote:
> [ short cut-off times ]
>> Folks - I'm not being a hard-a** about this.
>
> I think this message should have been addressed to the exim-users list,
> shouldn't it?


Sorry - thought I had done...

>
>> Just letting the *sender* know 'sooner' that they either need to fix
>> the address and re-send, pick up the phone, send an SMS or fax, buy a
>> postage stamp, call FedEx, or give it up as not important anyway.
>>
>> Their choice. An informed choice.
>
> Well, from my experience with endusers, they are not able to read the
> content of bounce e-mails. There must be some genetic defect causing this.
>


Agree they cannot/will not read (much) of them - but they at least see
the 'Mail delivery failed: returning message to sender' in the subject.

>> How would you feel if you placed a phone call and had to listen to a
>> 'busy' signal for *four days* - then finally get:
>>
>> 'the number you have dialed is not in service'
>>
>> ..... at the *end of the fourth day*?
>
> As usual, the comparison isn't really applicabe. After all, e-mail isn't
> a real-time communication system.


In theory, and by protocol - true. But in actual use, it is often
faster than phone calls that hit divert to voice-mail etc.

More importantly, it automagically leaves a 'trail' w/o chasing a
hundred pink 'While you were out' paper sticky notes - or spending more
time retrieving voice-mail messages and making paper notes for action
than it would have taken to accept the calls and JFDWT in the first place.

> If you're sending e-mails, regularly
> waiting for immediate response, your're doing it wrong. Further, you
> have to distinguish between permanent and transitional failures.
>


DOTS. Depends On The Situation. Think *business* use.

For clients with a network of their own and contractor / customer
offices, many a message is as terse as 'yes/no' , 'now, tommorow will be
OK', 'on the 12th', 'we need [name, address, date]' - whatever.

Seriously - many look like IRC traffic where body-content alone is
concerned.

We typically transit them sub-one second same-server, sub 10 seconds,
branch-branch or branch-HQ, and seldom over a minute off-net.

Where both correspondents are online and in 'chat' mode, the replies are
typed and sent very quikly. Then there may be a lull for an hour.

Of course time-zones and all mean that is not 100% of the time.

>> End-users - be they business or personal - simply do not use smtp the
>> way it was used long ago.
>
> That's definitely true OTOH. People just got used to their e-mails
> popping up on the other side as soon as they have sent them. I got
> complaints of e-mails taking one minute to get delivered to the
> receiver's mailbox. But delivery still isn't the same as response.
>
>> The 'competition' is IRC, SMS, and even intercontinental phone and
>> fax that cost a mere fraction of what they once did.
>
> All of them have different concepts. If people used e-mail to substitute
> them for some years, I don't see why they shouldn't go back to using
> them. As long as the biggest part of all e-mails sent is spam, it might
> be a good idea anyway.
>


Spam clogs the network, loads the server front-end, is in *our* face,
but not the user's inbox.

As with sanitation workers, if we are on form, they don't see what we
have done for them. 'Stuff' sure hits the fan when we fail, though...

>> As to the RFC's? They ordinarily lag reality by ten or more years....
>
> Still we should pay some attention to them.
>


Oh - we do.. but for the most part one has to serve-up what those paying
the invoice need and want.

Email has already changed the way folks go about their work.

Email is, as telex had been, less time-zone sensitive than voice.

>> Google it - SMS is displacing smtp at an increasing rate.
>
> As mobile phones take a bigger part in peoples daily life, this seems
> quite natural to me. With mobile e-mail clients spreading on people's
> phones, the trend will likely go back to e-mail again. At the moment
> it's mere a matter of usability for me not to use mobile e-mail a lot.
> I'm hoping for the iPhone, Android, LiMo, etc. to make some progress here.
>
> br,
> daniel
>


Speech-to-text and text-to-speech is where it needs to go - otherwise,
no matter how clever the finger-pinches, or how neat a tiny screen seems
to be, the I/O is already too small.

A hand*set* can be made smaller. But the human hand itself isn't doing
any detectable downsizing...

;-)


Bill