|
Currently, on the article editor at the following URL...
http://www.codeproject.com/script/Articles/Edit.aspx?aid=xxxxx
One can spend a long time editing an article, and then an accidental "navigate" off the page causes all that work to be LOST!
This is especially problematic because the editor often loses key-focus, and browsers often map the backspace key to "back", so when the editor loses key focus while you are backspacing, often it causes a "back navigate" which then loses the article content.
However, there is an EASY fix for this and other problems related to navigating away. Simply set up a javascript handler for "onbeforeunload". If the user accidentally tries to navigate off the page, the dialog will give them the chance to cancel the navigate before it destroys all their work. All that's needed is the following javascript added to the page:
window.onbeforeunload = function() {
return 'You have unsaved changes!';
}
Then, a small javascript handler should be added to the buttons which "validly" leave the page which turn off this handler.
|
|
|
|
|
We already have this in place. If it's not working for you, which browser are you using?
cheers,
Chris Maunder
The Code Project | Co-founder
Microsoft C++ MVP
|
|
|
|
|
Hmm. I was using google chrome when it ate a bunch of my work by navigating away. This was on first time article creation if that is handles any differently than editing.
I'm not sure what caused it to navigate away. I think I was hitting backspace a bunch and somehow the editor dropped my cursor, causing the browser to treat the backspace as 'back'. However, I work on think pad keyboards, so I sometimes accidentally hit the navigation buttons which they locate near the arrow keys.
I'll see if I can reproduce it later.
|
|
|
|
|
I'll double check with the latest build of Chrome and see if it's a Chrome specific issue.
cheers,
Chris Maunder
The Code Project | Co-founder
Microsoft C++ MVP
|
|
|
|
|
I just tested this myself.
In Firefox (3.6.10) I correctly receive the onbeforeunload warning dialog.
In Chrome (8.0.552.224), I do not receive an onbeforeunload warning dialog when clicking "back" from the editor.
Apparently there is an "incompatible" way to do onbeforeunload, and a "compatible" way. It looks like in Chrome/Safari/Webkit, you need to do:
function stopUnload() {
return "warning";
}
window.onbeforeunload = stopUnload;
Instead of attachEvent().
http://www.webreference.com/dhtml/diner/beforeunload/bunload4.html
http://code.google.com/p/chromium/issues/detail?id=4422
|
|
|
|
|
Yes, I was actually going through my code and in one place I have the traditional method, and in another the "window.onbeforeunload = stopUnload" method with a note "must be done this way".
Needless to say I'm going back and making the code consistent 
cheers,
Chris Maunder
The Code Project | Co-founder
Microsoft C++ MVP
|
|
|
|
|
Thanks for fixing this.... Codeproject is an excellent site!
|
|
|
|
|
Chris,
My guess is that it doesn't work when the page never finishes loading (usually happens when one of the ad banners is hung). So when this happens, and you have typed into the text box and then accidentally hit the Back button or did something else that causes the page to navigate away, you risk losing all the text you typed in.
|
|
|
|
|
When forum notifications from the code project arrive to my mailbox at hotmail.com, I get a nasty-looking warning from their SmartScreen:
"This message looks suspicious to our SmartScreen filters."
The warning has a pink background and a little red shield icon, hinting that the message may contain a security threat of some sort. When I click the [dismiss] button of the warning, the logo (logo225x40.gif) appears, but the warning stays.
It would be nice if you could figure out what part of the e-mail "trips" the SmartScreen, and re-arrange the auto-generated content in a different way, so that the warning does not appear.
|
|
|
|
|
You probably already know this but just add the address(sender) to your safe/white list. Problem solved...that's what I did and now I don't get those messages anymore.
|
|
|
|
|
Thanks for the suggestion. Unfortunately, this did not fix my problem: although "@codeproject.com" filter is on my "white list", the warning from SmartScreen still comes up.
|
|
|
|
|
Sorry to hear. I just checked my Hotmail account and everythings is fine. My notification of your reply did not trigger the SmartScreen super-duper interceptor radar thingy. 
|
|
|
|
|
We'll look into it here to see if there's anything obvious (or even unobvious) we canb do our end
cheers,
Chris Maunder
The Code Project | Co-founder
Microsoft C++ MVP
|
|
|
|
|
The voting stars are not showing properly...at least with this article: See Here[^]
When you hove over the stars the chart goes behind everything else.
I am using Windows XP SP3 w/IE 7.
|
|
|
|
|
That looks very similar to a problem I reported the day the new article header got launched: a random tooltip may be showing, as if the mouse was over a widget (e.g. the bookmark widget). Never had it happen on the voting bar though. Chris has looked into it already and called it a timing issue. Anyway, reloading the page seems to suffice.
|
|
|
|
|
Luc Pattyn wrote: Anyway, reloading the page seems to suffice.
Sorry, no cigar...still doesn't work.
|
|
|
|
|
Interesting.
I can't get it to fail (FF3.6, Chr7, IE7; all on Vista).
Try CTRL/F5!
|
|
|
|
|
Luc Pattyn wrote: Try CTRL/F5!
I did...twice...no luck.
Luc Pattyn wrote: all on Vista
This might be the problem. Vista and Windows 7 might be working fine but XP may not. I wonder if this site is tested on XP SP3 or just Windows 7?
|
|
|
|
|
Slacker007 wrote: all on Vista...This might be the problem
I doubt it.
Slacker007 wrote: I wonder if
I don't know. My guess is they use Win7+Chr7 most of the time. Not sure though.
OTOH, if it really is a timing problem, then CP staff themselves are worst positioned to solve it as they are bound to get best server response.
|
|
|
|
|
Looks fine to me (IE8/Windows 7).
|
|
|
|
|
Although the right of the article looks strange, as some of the widgest are overlapping.
|
|
|
|
|
When I type the correct email id and password, I log in in a jiffy, but when I make a typo, it takes noticeable amount of time before it reports an error and shows that page that it shows. Why does this happen? Is is just me?
PS: This is not quite a bug I suppose but it would be helpful if people could respond as I could learn some bit about web programming.
|
|
|
|
|
it is customary to include a deliberate delay before reporting a safety related attempt has failed; that makes it harder for bots to crack the system in a brute force attack. So I guess that is what is going on.
|
|
|
|
|
Thanks
|
|
|
|
|
Every page I open loads almost immediately but the page indicator still shows the loading symbol.
The loading bar says: 'waiting for a.lakequincy.com'
And the advertisements don't show up.
Just thought you might want to know
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
