- Inkscape Devel - Inkscape Mailman

mockup: styles editor / xml editor docked toggle
by Andy Fitzsimon 27 Jul '06

27 Jul '06

okay so the title is horrible but take a look at this mockup. for the day we support external styles this sort of functionality would be pure gold

2 4

WEBFORM Info on License
by Gianni 27 Jul '06

27 Jul '06

Message from Inkscape submission form Please note, that the submitter may not read this list. CC your answer to the submitter name: Gianni email: gianni@...1407... submitted the following: Hi, my name is Gianni and I\'m writing from Italy. I\'d like to know if it\'s possible to develope InkScape and personalize it: to put my logo and credits etc...I mean a personal license to deliver it with a sw that we already develop. Thanks.

3 2

WEBFORM size of download + preset shapes
by J Vaux 27 Jul '06

27 Jul '06

Message from Inkscape submission form Please note, that the submitter may not read this list. CC your answer to the submitter name: J Vaux email: scholara@...36... submitted the following: hi I\'m an artist thinking of downloading inkscape - i have windows but no net access at home will the files fit on an 128 mb flash drive? also the faq doesnt state the size of the files or whether you have any preset shapes? i already have the gimp

3 2

Re: [Inkscape-devel] inkscape teeshirt redesign!
by MenTaLguY 27 Jul '06

27 Jul '06

On Thu, 2006-07-27 at 00:01 +0100, Alan Horkan wrote: > So in short, are other colours possible? I think you might get a choice of colors with the "black" teeshirt option (I'll have to play with it), but otherwise CafePress hasn't been set up for printing on non-white colors really. It's one reason we might want to look into a different service actually. -mental

1 0

wiki upgrade
by Aaron Spike 26 Jul '06

26 Jul '06

The wiki moved this last weekend. I've given a few days to shake out any remaining bugs. Things seems to be working equally as well or better than at sf.net. This can only mean one thing: time to upgrade. Shall we move to the latest and greatest that mediawiki has to offer? Aaron Spike

2 1

Fedora Core 3 and Inkscape problems
by donn 26 Jul '06

26 Jul '06

Hello, new to list. I tried to get help via the IRC (don't have gaim) and one chap there tried to help me, but in the end suggested I try this list. SYSTEM: Fedora Core 3 : 2.6.12-1.1381_FC3 Machine : AMD i386, 1gig RAM, loads of drive space. PROBLEM: Cannot run Inkscape 0.44 Yum repo only has 0.43 - usually DAG. It appears he tried to compile 0.44 and had some problems. So, no yum update for me (or any FC3 users) Tried the Autopackage from the inkscape download page. It works in that it installs without overt error. Problem comes when trying to run inkscape: inkscape: /usr/bin/../lib/libpng12.so.0: no version information available (required by inkscape) inkscape: /usr/bin/../lib/libstdc++.so.6: version `GLIBCXX_3.4.4' not found (required by inkscape) My system has those two exact filenames (libpng12.so.0 and libstdc++.so.6) so a fake link technique is not going to work. I then tried to compile it. Harsh dependency errors. I don't have the time to go through the spin-cycle of sub-sub-sub compiles and yums to do that! So, I am in a fix. No Inkscape and not much hope! Any ideas? EXTRA QUESTION: Is there some way to put a complete, runnable Inkscape into a single tar.gz file like they do with Thunderbird and Firefox? I have never struggled to install the latest versions of those two apps. Sure, they take forever to reach a yum distro, but at least I can go fetch the tar file and get on with my life! Inkscape 0.44 even works on my Windows, but it does not work on Linux! There's a terrible irony here somewhere... Thanks, Donn. -- "If there is anyone in the audience who believes in telekinesis, please raise my hand!" --James Randi --

2 1

Broken links on the website
by Bryce Harrington 26 Jul '06

26 Jul '06

Anyone have a few spare cycles to fix these busted links? Bryce ----- Forwarded message from Victor Westmann <victor.westmann@...400...> ----- Date: Tue, 25 Jul 2006 23:33:38 -0300 From: Victor Westmann <victor.westmann@...400...> To: Bryce Harrington <bryce@...961...> Subject: Re: [Inkscape-translator] How can I start translating the pt_BR version of Inkscape? Bryce, here is a short list of the links that points to pages with no text information: General: download.php doc/index.php Community discussion.php mailing_lists.php report_bugs.php (Feature requests lead to a '500 - Internal Server Error') Developers status/index.php doc/devdocs.php svn.php This is it! Hope to have helped you all. I like and use Inkscape a lot. If theres any other task you want me to do, and Im able to, just ask! -- []s Victor ----- End forwarded message -----

1 0

Re: [Inkscape-devel] inkscape teeshirt redesign!
by MenTaLguY 25 Jul '06

25 Jul '06

[ NOTE: I am not a lawyer (just someone who reads TOSes and EULAs obsessively), and the following might not apply in your jurisdiction anyway, so it should definitely not be construed as legal advice -- if you need a legal opinion, please consult a lawyer, preferably one specializing in IP law. ] On Tue, 2006-07-25 at 14:08 +0930, Janet Hawtin wrote: > cafe press now claims rights to your artwork. I think that's a bit misleading -- the two rights claims in their TOS are ss. 4.3 and 6. First: 4.3 Licensing Your Content to CafePress.com. You will retain ownership of the Content that you upload to the Web Site. You hereby grant to CafePress.com a royalty-free, worldwide, transferable, nonexclusive, right and license to use such Content, in all media existing now or created in the future, as CafePress.com deems necessary to enable you to use the Create & Buy Service to create, produce and purchase Products for so long as your Content remains uploaded to the Web Site. CafePress.com may sublicense the rights that you grant it in this Section to a third party subcontractor only for purposes of producing your Products and providing the CafePress.com Service. This is a non-exclusive license limited to what CafePress "deems necessary to enable you to use the Create & Buy Service to create, produce and purchase Products". It's intended to protect them from situations like where someone tries to demand royalties or damages from CafePress on account of their use of the artwork to make or sell the products on the person's behalf. Can this be abused? Yes, because the interpretation of that condition is left to CafePress' discretion (although there may be legal limits to that discretion). On the other hand, the license grant is not perpetual and can be revoked by withdrawing your content. The only other IP claim is section 6: 6. Submissions. When you submit questions, comments, suggestions, ideas, message board postings, material submitted via web forms, contest entries, communications or any other information ("Submissions"), you grant CafePress.com permission to use such Submissions for marketing and other promotional purposes, including the right to sublicense. You agree that CafePress.com will have no obligation to keep any Submissions confidential. You will not bring a claim against CafePress.com based on "moral rights" or the likes arising from CafePress.com's use of a Submission. This Section does not apply to your Content that you use in connection with the Create & Buy Service or Shopkeeper Service. This explicitly doesn't apply to your artwork or to any store/product content (but it is the reason I don't touch the CafePress community forums). > printmojo is an alternative that doesnt. Actually, Printmojo's TOS includes the same sort of license grant as CafePress' section 4.3 (for exactly the same reasons): INTELLECTUAL PROPERTY RIGHTS PrintMojo does not own anything you upload to our servers. You retain all rights to your artwork, designs and intellecutal property. To legally operate, we need to have the ability to use the images you upload. This is called a license. Your PrintMojo account allows you to give us a free license to use the images for the following purposes: Printing onto items your buy as inventory to sell in your shop. Creating product shots and ads to help you promote your store. We will always say the images we use of yours are yours. You are free to use your images with any other service. Admittedly, there are three differences to CafePress' section 4.3: 1. The language seems less precise overall (hopefully they had a lawyer check it out). 2. The scope of the license is much narrower than the one CafePress requires (very good). 3. They don't specify how the license may be terminated or whether it is transferable (which is bad). How big a problem is #3? My guess is not much, given the limited scope of the license, but that's REALLY a question for a lawyer. I'd tend to think those things are normally spelled out for a reason. -mental

2 2

SoC students (and others): refactoring ahead!
by MenTaLguY 25 Jul '06

25 Jul '06

There are some upcoming changes in the codebase you need to know about. Namely, I'm nearing the end of the first stage of C++ification of SPObject: conversion from GObject to sigc++ signals. This means that instead of this kind of thing to connect to "release" or "modified" on an SPObject: g_signal_disconnect_...(G_OBJECT(object), ..., widget); ... g_signal_connect(G_OBJECT(object), "release", G_CALLBACK(&foo), widget); ...you'll need to do something like this: widget->release_connection.disconnect(); ... widget->release_connection = object->connectRelease(sigc::bind<1>(sigc::ptr_fun(&foo), widget)); I'll be glad to give you any assistance you need to convert your code to use the new API, which I've made available beforehand to ease the transition. The old API hasn't gone away yet, but it will on Saturday. -mental

1 0

Architectural fix for defs (for marker bug 1298718)
by Bryce Harrington 25 Jul '06

25 Jul '06

Over the past couple evenings after the DDC talks, mental and I have been hacking on one of the marker bugs, and last night managed to find the fix to it. It seems to work okay but I haven't had a chance to really exercise it so there may be some side effects... Patch is attached. Please test out gradients and other things that use defs, undo/redo, etc. Here's the bug: http://sourceforge.net/tracker/index.php?func=detail&aid=1298718&group_id=9… seems like a fairly obscure issue, that not many users would run into, but it is a key underlying bug for being able to alter markers (and anything else that uses defs). The actually fix was a one-liner (as most things are), but in finding it we've also done: * Added clearer, more complete comments/documentation for the modify/update paths in SPObject and Document. * Factored out common code in SPObject subclasses for getting a child list; now this is provided by SPObject's new childList() member. * Refactored updateDisplay() and requestDisplayUpdate() to highlight their similarities; maybe this could suggest a better refactoring in the future. Now they're consistent at least. * Added new pseudo-private member function _updateDocument() to the Document class, which factors out the flag-based update code to a single place. sp_document_ensure_up_to_date and sp_document_idle_handler have been refactored to move their shared logic into _updateDocument(). * An Action enumeration thingee originally implemented in sp-item-group has been promoted up to SPObject. Honestly I don't quite understand the purpose of it, but it doesn't appear to hurt anything, so hopefully whomever added this knows what it's for and can put it to more use. Anyway, now that this is fixed, the way is cleared for further enhancements to the marker code. Bryce P.S., Mental, can you commit this? I'm still having trouble committing to svn. Index: src/sp-object.h =================================================================== --- src/sp-object.h (revision 12719) +++ src/sp-object.h (working copy) @@ -206,6 +206,10 @@ SPObject *lastChild() { return _last_child; } SPObject const *lastChild() const { return _last_child; } + enum Action { ActionGeneral, ActionBBox, ActionUpdate, ActionShow }; + /** @brief Retrieves children as a GSList */ + GSList *childList(bool add_ref, Action action = ActionGeneral); + SPObject *appendChildRepr(Inkscape::XML::Node *repr); /** @brief Gets the author-visible label for this object. */ Index: src/document.h =================================================================== --- src/document.h (revision 12719) +++ src/document.h (working copy) @@ -115,6 +115,8 @@ Inkscape::EventLog& getEventLog() const; + bool _updateDocument(); + private: SPDocument(SPDocument const &); // no copy void operator=(SPDocument const &); // no assign Index: src/sp-item-group.cpp =================================================================== --- src/sp-item-group.cpp (revision 12719) +++ src/sp-item-group.cpp (working copy) @@ -592,13 +592,15 @@ } void CGroup::onUpdate(SPCtx *ctx, unsigned int flags) { - SPObject *child; SPItemCtx *ictx, cctx; ictx = (SPItemCtx *) ctx; cctx = *ictx; - if (flags & SP_OBJECT_MODIFIED_FLAG) flags |= SP_OBJECT_PARENT_MODIFIED_FLAG; + if (flags & SP_OBJECT_MODIFIED_FLAG) { + flags |= SP_OBJECT_PARENT_MODIFIED_FLAG; + } + flags &= SP_OBJECT_MODIFIED_CASCADE; if (flags & SP_OBJECT_STYLE_MODIFIED_FLAG) { @@ -608,9 +610,9 @@ } } - GSList *l = g_slist_reverse(_childList(true, ActionUpdate)); + GSList *l = g_slist_reverse(_group->childList(true, SPObject::ActionUpdate)); while (l) { - child = SP_OBJECT (l->data); + SPObject *child = SP_OBJECT (l->data); l = g_slist_remove (l, child); if (flags || (child->uflags & (SP_OBJECT_MODIFIED_FLAG | SP_OBJECT_CHILD_MODIFIED_FLAG))) { if (SP_IS_ITEM (child)) { @@ -626,17 +628,6 @@ } } -GSList *CGroup::_childList(bool add_ref, Action) { - GSList *l = NULL; - for (SPObject *child = sp_object_first_child(_group) ; child != NULL ; child = SP_OBJECT_NEXT(child) ) { - if (add_ref) - g_object_ref (G_OBJECT (child)); - - l = g_slist_prepend (l, child); - } - return l; -} - void CGroup::onModified(guint flags) { SPObject *child; @@ -650,7 +641,7 @@ } } - GSList *l = g_slist_reverse(_childList(true)); + GSList *l = g_slist_reverse(_group->childList(true)); while (l) { child = SP_OBJECT (l->data); l = g_slist_remove (l, child); @@ -662,7 +653,7 @@ } void CGroup::calculateBBox(NRRect *bbox, NR::Matrix const &transform, unsigned const flags) { - GSList *l = _childList(false, ActionBBox); + GSList *l = _group->childList(false, SPObject::ActionBBox); while (l) { SPObject *o = SP_OBJECT (l->data); if (SP_IS_ITEM(o)) { @@ -675,7 +666,7 @@ } void CGroup::onPrint(SPPrintContext *ctx) { - GSList *l = g_slist_reverse(_childList(false)); + GSList *l = g_slist_reverse(_group->childList(false)); while (l) { SPObject *o = SP_OBJECT (l->data); if (SP_IS_ITEM(o)) { @@ -723,7 +714,7 @@ NRArenaItem *ac = NULL; NRArenaItem *ar = NULL; SPItem * child = NULL; - GSList *l = g_slist_reverse(_childList(false, ActionShow)); + GSList *l = g_slist_reverse(_group->childList(false, SPObject::ActionShow)); while (l) { SPObject *o = SP_OBJECT (l->data); if (SP_IS_ITEM (o)) { @@ -742,7 +733,7 @@ void CGroup::hide (unsigned int key) { SPItem * child; - GSList *l = g_slist_reverse(_childList(false, ActionShow)); + GSList *l = g_slist_reverse(_group->childList(false, SPObject::ActionShow)); while (l) { SPObject *o = SP_OBJECT (l->data); if (SP_IS_ITEM (o)) { Index: src/sp-switch.cpp =================================================================== --- src/sp-switch.cpp (revision 12719) +++ src/sp-switch.cpp (working copy) @@ -75,9 +75,9 @@ return NULL; } -GSList *CSwitch::_childList(bool add_ref, Action action) { - if ( ActionGeneral != action ) { - return CGroup::_childList(add_ref, action); +GSList *CSwitch::_childList(bool add_ref, SPObject::Action action) { + if ( action != SPObject::ActionGeneral ) { + return _group->childList(add_ref, action); } SPObject *child = _evaluateFirst(); @@ -120,7 +120,7 @@ _releaseLastItem(_cached_item); SPItem * child; - for ( GSList *l = _childList(false, ActionShow); + for ( GSList *l = _childList(false, SPObject::ActionShow); NULL != l ; l = g_slist_remove (l, l->data)) { SPObject *o = SP_OBJECT (l->data); @@ -162,7 +162,7 @@ NRArenaItem *ac = NULL; NRArenaItem *ar = NULL; SPItem * child; - GSList *l = _childList(false, ActionShow); + GSList *l = _childList(false, SPObject::ActionShow); while (l) { SPObject *o = SP_OBJECT (l->data); if (SP_IS_ITEM (o)) { Index: src/sp-object.cpp =================================================================== --- src/sp-object.cpp (revision 12719) +++ src/sp-object.cpp (working copy) @@ -458,6 +458,22 @@ } } +/** + * Retrieves the children as a GSList object, optionally ref'ing the children + * in the process, if add_ref is specified. + */ +GSList *SPObject::childList(bool add_ref, Action) { + GSList *l = NULL; + for (SPObject *child = sp_object_first_child(this) ; child != NULL; child = SP_OBJECT_NEXT(child) ) { + if (add_ref) + g_object_ref (G_OBJECT (child)); + + l = g_slist_prepend (l, child); + } + return l; + +} + /** Gets the label property for the object or a default if no label * is defined. */ @@ -1168,26 +1184,31 @@ void SPObject::requestDisplayUpdate(unsigned int flags) { + g_return_if_fail( this->document != NULL ); + if (update_in_progress) { g_print("WARNING: Requested update while update in progress, counter = %d\n", update_in_progress); } + /* requestModified must be used only to set one of SP_OBJECT_MODIFIED_FLAG or + * SP_OBJECT_CHILD_MODIFIED_FLAG */ g_return_if_fail(!(flags & SP_OBJECT_PARENT_MODIFIED_FLAG)); g_return_if_fail((flags & SP_OBJECT_MODIFIED_FLAG) || (flags & SP_OBJECT_CHILD_MODIFIED_FLAG)); g_return_if_fail(!((flags & SP_OBJECT_MODIFIED_FLAG) && (flags & SP_OBJECT_CHILD_MODIFIED_FLAG))); - /* Check for propagate before we set any flags */ - /* Propagate means, that this is not passed through by modification request cascade yet */ - unsigned int propagate = (!(this->uflags & (SP_OBJECT_MODIFIED_FLAG | SP_OBJECT_CHILD_MODIFIED_FLAG))); + bool already_propagated = (!(this->uflags & (SP_OBJECT_MODIFIED_FLAG | SP_OBJECT_CHILD_MODIFIED_FLAG))); - /* Just set this flags safe even if some have been set before */ this->uflags |= flags; - if (propagate) { - if (this->parent) { - this->parent->requestDisplayUpdate(SP_OBJECT_CHILD_MODIFIED_FLAG); + /* If requestModified has already been called on this object or one of its children, then we + * don't need to set CHILD_MODIFIED on our ancestors because it's already been done. + */ + if (already_propagated) { + SPObject *parent = SP_OBJECT_PARENT(this); + if (parent) { + parent->requestDisplayUpdate(SP_OBJECT_CHILD_MODIFIED_FLAG); } else { - sp_document_request_modified(this->document); + sp_document_request_modified(SP_OBJECT_DOCUMENT(this)); } } } @@ -1228,29 +1249,30 @@ update_in_progress --; } +/** + * Request modified always bubbles *up* the tree, as opposed to + * request display update, which trickles down and relies on the + * flags set during this pass... + */ void SPObject::requestModified(unsigned int flags) { g_return_if_fail( this->document != NULL ); - /* PARENT_MODIFIED is computed later on and is not intended to be - * "manually" queued */ + /* requestModified must be used only to set one of SP_OBJECT_MODIFIED_FLAG or + * SP_OBJECT_CHILD_MODIFIED_FLAG */ g_return_if_fail(!(flags & SP_OBJECT_PARENT_MODIFIED_FLAG)); - - /* we should be setting either MODIFIED or CHILD_MODIFIED... */ g_return_if_fail((flags & SP_OBJECT_MODIFIED_FLAG) || (flags & SP_OBJECT_CHILD_MODIFIED_FLAG)); - - /* ...but not both */ g_return_if_fail(!((flags & SP_OBJECT_MODIFIED_FLAG) && (flags & SP_OBJECT_CHILD_MODIFIED_FLAG))); - unsigned int old_mflags=this->mflags; + bool already_propagated = (!(this->mflags & (SP_OBJECT_MODIFIED_FLAG | SP_OBJECT_CHILD_MODIFIED_FLAG))); + this->mflags |= flags; - /* If we already had MODIFIED or CHILD_MODIFIED queued, we will - * have already queued CHILD_MODIFIED with our ancestors and - * need not disturb them again. + /* If requestModified has already been called on this object or one of its children, then we + * don't need to set CHILD_MODIFIED on our ancestors because it's already been done. */ - if (!( old_mflags & ( SP_OBJECT_MODIFIED_FLAG | SP_OBJECT_CHILD_MODIFIED_FLAG ) )) { + if (already_propagated) { SPObject *parent=SP_OBJECT_PARENT(this); if (parent) { parent->requestModified(SP_OBJECT_CHILD_MODIFIED_FLAG); @@ -1260,6 +1282,9 @@ } } +/** + * This is what actually delivers the modified signals + */ void SPObject::emitModified(unsigned int flags) { Index: src/document.cpp =================================================================== --- src/document.cpp (revision 12719) +++ src/document.cpp (working copy) @@ -746,69 +746,75 @@ ctx->i2vp = NR::identity(); } +/** + * Tries to update the document state based on the modified and + * "update required" flags, and return true if the document has + * been brought fully up to date. + */ +bool +SPDocument::_updateDocument() +{ + /* Process updates */ + if (this->root->uflags || this->root->mflags) { + if (this->root->uflags) { + SPItemCtx ctx; + sp_document_setup_viewport (this, &ctx); + + bool saved = sp_document_get_undo_sensitive(this); + sp_document_set_undo_sensitive(this, FALSE); + + this->root->updateDisplay((SPCtx *)&ctx, 0); + + sp_document_set_undo_sensitive(this, saved); + } + this->_emitModified(); + } + + return !(this->root->uflags || this->root->mflags); +} + + +/** + * Repeatedly works on getting the document updated, since sometimes + * it takes more than one pass to get the document updated. But it + * usually should not take more than a few loops, and certainly never + * more than 32 iterations. So we bail out if we hit 32 iterations, + * since this typically indicates we're stuck in an update loop. + */ gint sp_document_ensure_up_to_date(SPDocument *doc) { - int lc; - lc = 32; - while (doc->root->uflags || doc->root->mflags) { - lc -= 1; - if (lc < 0) { - g_warning("More than 32 iterations while updating document '%s'", doc->uri); - if (doc->modified_id) { - /* Remove handler */ - gtk_idle_remove(doc->modified_id); - doc->modified_id = 0; - } - return FALSE; + int counter = 32; + while (!doc->_updateDocument()) { + if (counter == 0) { + g_warning("More than 32 iteration while updating document '%s'", doc->uri); + break; } - /* Process updates */ - if (doc->root->uflags) { - SPItemCtx ctx; - sp_document_setup_viewport (doc, &ctx); - doc->root->updateDisplay((SPCtx *)&ctx, 0); - } - doc->_emitModified(); + counter--; } + if (doc->modified_id) { /* Remove handler */ gtk_idle_remove(doc->modified_id); doc->modified_id = 0; } - return TRUE; + return counter>0; } +/** + * An idle handler to update the document. Returns true if + * the document needs further updates. + */ static gint sp_document_idle_handler(gpointer data) { - SPDocument *doc; - int repeat; - - doc = static_cast<SPDocument *>(data); - -#ifdef SP_DOCUMENT_DEBUG_IDLE - g_print("->\n"); -#endif - - /* Process updates */ - if (doc->root->uflags) { - SPItemCtx ctx; - sp_document_setup_viewport (doc, &ctx); - - gboolean saved = sp_document_get_undo_sensitive(doc); - sp_document_set_undo_sensitive(doc, FALSE); - - doc->root->updateDisplay((SPCtx *)&ctx, 0); - - sp_document_set_undo_sensitive(doc, saved); - /* if (doc->root->uflags & SP_OBJECT_MODIFIED_FLAG) return TRUE; */ + SPDocument *doc = static_cast<SPDocument *>(data); + if (doc->_updateDocument()) { + doc->modified_id = 0; + return false; + } else { + return true; } - - doc->_emitModified(); - - repeat = (doc->root->uflags || doc->root->mflags); - if (!repeat) doc->modified_id = 0; - return repeat; } static bool is_within(NR::Rect const &area, NR::Rect const &box) Index: src/sp-switch.h =================================================================== --- src/sp-switch.h (revision 12719) +++ src/sp-switch.h (working copy) @@ -38,7 +38,7 @@ virtual gchar *getDescription(); protected: - virtual GSList *_childList(bool add_ref, Action action); + virtual GSList *_childList(bool add_ref, SPObject::Action action); virtual void _showChildren (NRArena *arena, NRArenaItem *ai, unsigned int key, unsigned int flags); SPObject *_evaluateFirst(); Index: src/sp-defs.cpp =================================================================== --- src/sp-defs.cpp (revision 12719) +++ src/sp-defs.cpp (working copy) @@ -83,21 +83,14 @@ flags &= SP_OBJECT_MODIFIED_CASCADE; - GSList *l = NULL; - for ( SPObject *child = sp_object_first_child(object) ; child != NULL; child = SP_OBJECT_NEXT(child) ) { - g_object_ref(G_OBJECT(child)); - l = g_slist_prepend(l, child); - } - - l = g_slist_reverse(l); - + GSList *l = g_slist_reverse(object->childList(true)); while (l) { SPObject *child = SP_OBJECT(l->data); l = g_slist_remove(l, child); - if (flags || (child->uflags & SP_OBJECT_MODIFIED_FLAG)) { + if (flags || (child->uflags & (SP_OBJECT_MODIFIED_FLAG | SP_OBJECT_CHILD_MODIFIED_FLAG))) { child->updateDisplay(ctx, flags); } - g_object_unref(G_OBJECT(child)); + g_object_unref (G_OBJECT (child)); } } Index: src/sp-item-group.h =================================================================== --- src/sp-item-group.h (revision 12719) +++ src/sp-item-group.h (working copy) @@ -79,8 +79,6 @@ gint getItemCount(); protected: - enum Action { ActionGeneral, ActionBBox, ActionUpdate, ActionShow }; - virtual GSList *_childList(bool add_ref, Action action = ActionGeneral); virtual void _showChildren (NRArena *arena, NRArenaItem *ai, unsigned int key, unsigned int flags); SPGroup *_group; Index: ChangeLog =================================================================== --- ChangeLog (revision 12719) +++ ChangeLog (working copy) @@ -1,3 +1,10 @@ +2006-07-18 Bryce Harrington <bryce@...961...> + * src/document.h, src/document.cpp: Refactoring from mental & + bryce to consolidate document update functionality from + sp_document_ensure_is_up_to_date() and + sp_document_idle_handler(). This is the first step in getting + updates of defs fixed. + 2006-07-18 Tim Dwyer <Tim.Dwyer@...38...> * src/graphlayout/graphlayout.cpp,

4 9