Path: g2news1.google.com!news1.google.com!border1.nntp.dca.giganews.com!nntp.giganews.com!local02.nntp.dca.giganews.com!nntp.mozilla.org!news.mozilla.org.POSTED!not-for-mail
NNTP-Posting-Date: Thu, 05 Jun 2008 12:14:14 -0500
Date: Thu, 05 Jun 2008 12:14:14 -0500
From: Boris Zbarsky <bzbar...@mit.edu>
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.8.1.13) Gecko/20080317 SeaMonkey/1.1.9
MIME-Version: 1.0
Newsgroups: mozilla.dev.tech.gfx
Subject: Re: Large images memory usage
References: <g9Kdnbwhwqxlc6fVnZ2dnUVZ_tTinZ2d@mozilla.org> <8d70cff2-735a-4f0f-94ba-1edc8680f236@m3g2000hsc.googlegroups.com> <dbednTespbJ1mdrVnZ2dnUVZ_u2dnZ2d@mozilla.org> <AvmdnbOA7qddZdrVnZ2dnUVZ_jmdnZ2d@mozilla.org>
In-Reply-To: <AvmdnbOA7qddZdrVnZ2dnUVZ_jmdnZ2d@mozilla.org>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
Message-ID: <PZKdnTQhFet6g9XVnZ2dnUVZ_gqdnZ2d@mozilla.org>
Lines: 24
X-Usenet-Provider: http://www.giganews.com
NNTP-Posting-Host: 68.252.230.73
X-AuthenticatedUsername: NoAuthUser
X-Trace: sv3-pDqtBdPPMzlrDbONA68qrTC74k/2QdAwQ/3ZH4G4k2qQQrsBV34Zm5PguABJqHBOO//6Flvl7MAKI3U!4/BIBLy5qiDg0YTWeSF41v1iqPhE+Z29MZ9oH93qRke70vCb5UOxzpr/7yvTOjRsziqFpdeNLsia!0fxRpl1Ae/n0
X-Abuse-and-DMCA-Info: Please be sure to forward a copy of ALL headers
X-Abuse-and-DMCA-Info: Otherwise we will be unable to process your complaint properly
X-Postfilter: 1.3.39

Francis Robichaud wrote:
> You are right, my next step is to generate new pixmaps when surfaces get 
> dirty (eg. scrolling events).

You do realize that the same image could be being painted in multiple places at 
once, and they could be painting different parts of the image, right?

> I'm still looking for a way to read the "container" size.

libpr0n doesn't know that.  And in any case, it would be a set of container 
sizes, not just one size.

> Also, I'm trying to figure out why libpr0n creates the 
> gfxImageFrame which contains the gfxImageSurface

A gfxImageFrame is an abstraction for a single frame of an animated image, as I 
recall.  Does that help?

> and where is the link between the frame and it's parent.

Via an imgRequst(Proxy)

-Boris