Subject:
|
Re: Uh Oh... (was Traffic Error )
|
Newsgroups:
|
lugnet.admin.general
|
Date:
|
Tue, 20 Apr 1999 16:03:27 GMT
|
Viewed:
|
922 times
|
| |
|
|
In lugnet.admin.general, Todd Lehman writes:
> In lugnet.admin.general, edboxer@aol.com (Ed Jones) writes:
> > Just out of curiosity, was Mr. McDonald posting from the webpage instead
> > of newsgroups.
Please, call me Tom. I've been known to call you Ed when you're not looking :)
> Yup! Exactly, that's what we're looking into...
>
>
> > The reason I ask is....
> >
> > When posting/replying from the web page (which I am doing now), upon
> > clicking the Post Message button, you receive a confirmation message
> > screen. But, when you click on the browser <Back> button, it takes you
> > to the reply that you wrote and just sent. Is it possible that Mr.
> > McDonald did this and thought that he had written the reply but not
> > sent the reply?
> >
> > Of course this would be dependant on his keeping his web browser open
> > the entire time. He could have visited other sites and by hitting the
> > <Back> button suddenly came upon the message that he thought he had sent.
>
> That was my thought. It's the only logical explanation, given the data in
> the webserver logs.
I just thought of another logical possibility that eluded me last night: I
probly had opened another browser window when I was typing my message to be
able to write while reading one of Todd's other messages in a different
offshoot of the thread, or perhaps to browse a picture to get my facts
straight, and I inadvertantly sent the same message text from each of the two
windows.
> The bummer of it is: Although it would be easy enough to embed a unique key
> in a hidden form field on the /news/post/ page to ensure that the same form
> couldn't be used twice to post the same message twice, that sort of thing
> would prevent a form from being used twice for legitimate reasons -- i.e.,
> posting two different messages to the same group. If a unique key were
> embedded in the form (generated on-the-fly when the form page was loaded), a
> user would then have to reload the form page for subsequent posts, which is
> not only unintuitive but would require explanation on the page, which
> complicates the system somewhat unnecessarily and makes the human work for
> the computer rather than vice-versa.
Ugh. From an administrative standpoint it does make sense to sometimes not have
things be uniquely keyed. I'm in charge of our music information database
program at work using Paradox 8 :P, and I don't uniquely key unless it's
absolutely necessary.
> Another workaround would be to do some sort of checksum analysis on the
> content to make sure it wasn't identical to previously posted content, but
> it would have to take into account things like adding a missing sentence or
> blank line. And if the content was similar but different, would it
> automatically go and delete the previous article?
I don't like machines automatically deleting like that without permission. They
make efficient servants, but they're low on wisdom. But with regard to the
message body checksum, maybe it should be identical with message subjects being
identical too.
> Probably best all around would be some kind of warning system to the user.
> If the server suspected that the Post Message button was clicked twice, it
> could put up a big shiny red box asking the user whether he/she *really*
> wanted to proceed.
Well then maybe here's your reason to implement that "90% redundant text"
message you've been infrequently mentioning. More than one tired user has
perused these pages at the end of a tedious day. And as you also have pointed
out on occasion, it might also help to keep replies a bit shorter. You could
title the message box "Tombox" ;-)
Todd, without me, you're site shouldn't/wouldn't be idiot-proof! :)
-Tom McD.
"Idiotboy strikes again!" :D
|
|
Message has 1 Reply:
Message is in Reply To:
 | | Re: Uh Oh... (was Traffic Error )
|
| (...) Yup! Exactly, that's what we're looking into... (...) That was my thought. It's the only logical explanation, given the data in the webserver logs. The bummer of it is: Although it would be easy enough to embed a unique key in a hidden form (...) (25 years ago, 20-Apr-99, to lugnet.admin.general)
|
27 Messages in This Thread:
 
  
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
This Message and its Replies on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
