clearing up the browser memory problems (although Jordi already mentioned it
I think)
if you have javascript objects/variables referencing DOM elements (html
divs, images, mouseevents etc) you need to make sure those are deleted
(de-attached from the html elements) or the html elements will keep on
existing in memory for the rest of your browser session.
so every DynLayer has a reference to atleast an HTMLDIVElement.. that has to
be removed by setting the variables/objects to null
so:
DynLayer.elm=null
would remove the link between the DynLayer.elm object and the actual
HTMLDivElement, making the DIVElement get removed from memory as well.
Note that it might stay in memory even after a refresh, but I think that's
because IE only does memory-freeing at a certain interval, so it will be
gone in your next refresh or after a few minutes/seconds (my perception of
it)
...to much freetime today
Pascal Bestebroer ([EMAIL PROTECTED])
Software ontwikkelaar
Oberon Informatiesystemen b.v.
http://www.oibv.com
> -----Oorspronkelijk bericht-----
> Van: [EMAIL PROTECTED]
> [mailto:[EMAIL PROTECTED]]Namens Stephan
> Tolksdorf
> Verzonden: dinsdag 19 juni 2001 15:40
> Aan: Doug Melvin
> Onderwerp: Re[13]: [Dynapi-Dev] Attention...
>
>
> > Then test it.
> > I'm sure we don't need to hold your hand.
>
> I asked on this list to learn the facts the core developers are
> basing their opinions on. Sure I can do everything on my own, but - as
> I see it - we all do profit if we exchange our experiences and try to
> improve our understanding of the issue together.
> Your answer wasn't productive at all.
>
> Ok, I now tested the different methods for two hours, at least I tried
> to do so (JavaScript and Browsers are a cruel programming
> environment esp. for memory debugging!).
> I couldn't notice any difference in memory handling for the three
> different routines (IE5.5 on Win2k, IE4 on Win98 and Mozilla on
> Win2k): c = null, delete c and doing nothing.
>
> I based my conclusion on the memory the browser process
> allocated while
> playing with the attached html page.
> Does someone know a better testing method?
>
> Stephan
>
_______________________________________________
Dynapi-Dev mailing list
[EMAIL PROTECTED]
http://lists.sourceforge.net/lists/listinfo/dynapi-dev