Module Updates XPToolkit August 22 Submitted by Peter Trudelle <trudelle@netscape.com> Highlights Mike Pinkerton implemented drag-start events coming from Gecko that D&D clients can listen for in order to determine if a drag has started. It won't actually begin the drag, just be a hint to anyone listening that the user has made a gesture that could be interpreted as a drag. It will be up to the individual clients to determine if this is actually a drag. (pinkerton) David Hyatt rewrote the broadcaster/observer system from scratch and added some syntactic enhancements.  He also updated the  Broadcasters and Observers documentation to cover this. (hyatt) Lowlights Accomplishments Resolved 35 bugs in the last week, see our  Bug List  for details. Eric Vaughan landed GFX scrollbars, but they'll remain disabled by default until a few remaining build problems are resolved. (evaughan) Stuart Parmenter met with clipboard developers to identify and resolve remaining issues.  As a result, Rod Spears identified and fixed a bug that should now unblock QA, and Akkana Peck is working on conversion in her groups code to fix the rest of the major problems. (akkana, rods, pavlov) Stuart also fixed an awful resizing bug on Unix, and made it much better by removing old  hacks that had been there since day 2. (pavlov) David Hyatt has checked in the initial popup re-architecture, although at the moment popups are disabled because the syntax is changing.  (hyatt) David also created a new event called "oncommand", and taught titledbuttons, menus, and key bindings how to field the event, documented in Buttons and menus and keysets (oh my) (hyatt)   Priorities Help XPApps team with dialog & widget issues. (evaughan) Finish the splitter widget. (evaughan) IDL rewrite/overhaul for D&D. (pinkerton) XPMenus ahoy. (saari) Finish Linux D&D for M10. (mcafee) Get nsIWidget::CaptureMouse working properly on unix. (pavlov) Work with ramiro on unexposed areas under xp menus. (pavlov) Finish removing crusty code from widget/* that has never been used. (pavlov) Dialog functionality.  As in, getting them to display. (danm) Work with Gagan early next week to implement auth dialogs within Necko.  This is especially important because other test cases which demonstrate the problems with dialogs are all pretty much unusable right now.  (danm) Finish support for default buttons in dialogs. (sdagley) Complete the popup re-architecture.  (hyatt) Work on fixing the scrolling problems in the tree widget related to the dynamic creation/destruction of frames.  (hyatt) David Hyatt plans to "rewrite tree widget selection from scratch, now that I know what the many problems with it are. Basically I have to write a custom DOM node for the tree so that I have a place to store the selection in the content model for fast access (again, this relates to the lazy instantiation problem. since frames come and go, I can't rely on doing storage of the selection as frames)". (hyatt) Decisions Issues IDL conversion for D&D is proving more difficult than we have time for, in spite of help from Chris Waterson and John Bandhauer.  We'll need help from the Pork Jockeys on this. (pinkerton, waterson, jband) Too much tree horkage this week, especially during prime daytime checkin hours, making it difficult to work on the tip. Loading dialogs with gfx widgets on, especially with lots of text widgets,  can take 10 times longer than loading the native version. (pavlov, alecf) Dialogs are flaky -- the exact same dialog seems to work properly on some machines, but not on others (under the same OS).  As of yesterday, all of the test cases (bugs 7770, 12043, 12137, and auth dialogs) were unreachable because of other bugs, unexplained irreproducibility, or being unwritten. Clipboard support for images has turned into a snake.  We need to resolve the details of what needs to be in by beta & RTM. People   Internationalization and Localization August 21 Submitted by Tague Griffith <tague@netscape.com> Rod Spears found and fixed a bug so copy and paste now works with Japanese. This was a big blocker for us - thanks Rod. Hooked up Henrik Lynggaard Hansen <Lynggaard@netscape.net> translation work. UI strings can now be swtiched from English to Danish with a single click. Started a gallery of examples using the various scriptable internationalization interfaces in Seamonkey. Continued investigating the performance of Seamonkey when viewing Japanese web sites. XML & Layout August 21 Submitted by Nisheeth Ranjan <nisheeth@netscape.com> XML Currently working on plugging in an XSL processor into Gecko. The processor was recently contributed by Keith Visco (MITRE) to mozilla.org (I have been interacting with Keith for some time and he got the go ahead to MPL his source just a week or so back). Gecko has hooks that let an external XSLT processor apply an XSL transformation to an XML document before it is displayed. I'll try to demo the plugged in processer at XML DevCon '99 in Canada this Friday. If I can't get it plugged in by then, I'll at least talk about the XSL hooks in Gecko. I really want to get the XSL processor hooked up by beta, though. We use James Clark's expat parser for parsing XML documents. Till now, almost all the bugs related to XML have been related to the layers above expat and the few that were expat related were killed by James Clark in no time. He is an ideal module owner. I'll be doing more XML bug fixing till beta. Most of the current bugs are related to the new external parameter entity support that went in recently to support localization of XUL files. The xml, xul, rdf content sinks' code needs to be factored better by beta. Vidur and Peter Linss are owning this and I'll be helping out. The error strings reported by expat need to be localized. - the xml documentation on mozilla.org is badly in need of an update. I wanna get to do that after the upcoming XML conference. It would be really cool to wrap expat with XPCOMified SAX interfaces. That would create an XP alternative to Microsoft's COMified XML parser component. Layout I'm transitioning into learning the block-inline code in core layout. Definitely need to do that by beta. I'll be working with others in the gecko team to implement document state storage for session history by beta. Hopefully, one of us in core layout will also be able to get rid of bug 10818, an RFE to implement batched reflows. Previous Updates