It appears that the drawview has a "cachebitmap" attribute already, which sets the cacheAsBitmap flag on the Flash movieclip which it is using. . So it should already be being set on drawview by default.
See the code in lps/components/extensions/drawview.lzx to see how it is being initialized. On Sat, Apr 19, 2008 at 4:50 AM, Gilad Parann-Nissany < [EMAIL PROTECTED]> wrote: > Thank you for the input. Could you give us concrete code examples of: > > 1. calling ActionScript directly in our OL code > > 2. using drawview (or any other technique) to create a bitmap cache of our > SWFs. Note: we can limit ourselves to "simple views' (views that do not nest > other views) as a first step if necessary. > > Gilad > > > > Gilad Parann-Nissany > > > ----- Original Message ----- > From: "Max Carlson" <[EMAIL PROTECTED]> > To: "P T Withington" <[EMAIL PROTECTED]> > Cc: "Gilad Parann-Nissany" <[EMAIL PROTECTED]>, "Max Carlson" < > [EMAIL PROTECTED]>, "laszlo-user" <[email protected]>, "Ammar > Tamazi" <[EMAIL PROTECTED]>, "Elias Khalil" < > [EMAIL PROTECTED]>, "Lou Iorio" <[EMAIL PROTECTED]> > Sent: Thursday, April 17, 2008 7:18:24 PM (GMT+0200) Auto-Detected > Subject: Re: Understanding Flash memory usage when using Laszlo > > > > P T Withington wrote: > > On 2008-04-17, at 09:12 EDT, Gilad Parann-Nissany wrote: > >> Hi > >> > >> With the help from this forum we have been making progress > >> understanding our memory usage. See the separate thread: [Laszlo-user] > >> Understanding memory analysis when using _LzDebug.whyAlive() > >> > >> We're now at the point were we see that the "smoots" taken up by > >> Laszlo objects are not all of our memory usage. It seems a large part > >> (large percentage of total RAM usage, maybe even more than 50%) may be > >> used elsewhere, perhaps in the Flash objects that are "underneath" the > >> Laszlo objects. > >> > >> Questions: > >> > >> 1. Is there any tool or technique that allows us to analyze exactly > >> how this memory is being used by Flash? ("underneath" the laszlo > >> objects, I mean) > > > > I am not aware of such a tool. There may be a tool provided by Adobe. > > Actually, this will be a nice benefit from the move to swf9. Flex > builder (and presumably fdb) have profiling tools that can show this > kind of memory usage. > > >> 2. Are there any guidelines to help us optimize this Flash part of > >> our memory usage? > >> 3. More generally is there any way to use Flash-specific > >> optimization techniques (going to the ActionScript where necessary). > >> We know that in pure Flash there are several optimization techniques; > >> the following points are examples of things we wish to try > >> 4. Can we in general call pieces of ActionScript from Laszlo script? > > > > Yes. The OL compiler does not prevent you from doing that, but neither > > is it supported. > > > >> 5. Bitmap caching: > >> > >> > >> • SWF resources can be translated once into bitmaps (on the > >> client side) and cached > >> • you probably know this is known as bitmap caching. > >> • Then the bitmaps are used repeatedly instead of rendering the > >> SWF objects repeatedly. > >> • Can we do Bitmap Caching for all those vector graphics which > >> do not change frequently (from Laszlo, but possibly with specific > >> usage of Actionscript if necessary) > > > > I believe Max (cc-ed) has experimented with this. There may even be > > support in LZX for it. > > We do have support for this in drawview. We can experiment with adding > support for vector resources as well. There seemed to be side effects > for things with nested views, for example. > > >> 6. > >> Garbage collection: we would like to trigger the Flash garbage > >> collection just to test what is going on (as a test not permanent > >> change). Any way for us to do this? > > > > I am not aware of a programmatic way of doing that. In most browsers, > > if you minimize the window the app is running in the Flash GC will run. > > > > Again, there are probably tools provided by Adobe for these purposes, > > but I have not used them. There are probably more tools in the open > > source Flex compiler, so the swf9 runtime should be easier to measure > > and tune in this respect. > > -- > Regards, > Max Carlson > OpenLaszlo.org > > -- Henry Minsky Software Architect [EMAIL PROTECTED]
