- Inkscape Devel - Inkscape Mailman

RE: [Inkscape-devel] RE: offset width modifying
by bulia byak 31 Dec '03

31 Dec '03

I've tried the new offset object - really impressive! For those who haven't tried it yet, you can now interactively adjust the inset/outset value by dragging a control point (similar to e.g. rounding a rect) and it will be redrawn in real time. You can also vary the radius attribute of the object in XML editor to achieve the same result. The quality and precision of the outline is very good (at least compared to the previous attempts to do multiple in/outsets). A few comments: - The new attributes that you add must be in the inkscape: namespace, not sodipodi:. - The inset/outset commads still require that a path has a stroke. Please change them so that they decrement/increment the radius attribute, by say 1 pt with each command, regardless of stroke (converting a path into offset if necesary). Ideally we'll need to have two pairs of commands, say one incrementing radius by 1 pt and the other (with alt) by the distance corresponding to 1 visible pixel at the current zoom (similar to alt-arrows). - The statusline description of the object can be more useful if instead of simply "Offset" it would say e.g. "Path inset by 1.5pt" or "Path outset by 3pt". - Even though the path quality is now very good, I still noticed occasional "splotches" that would be nice to fix. I'm now uploading one example to the bug tracker so you can see what I'm talking about. _________________________________________________________________ MSN 8 with e-mail virus protection service: 2 months FREE* http://join.msn.com/?page=features/virus&pgmarket=en-ca&RU=http%3a%2f%2fjoi…

3 2

gtkmm2 rollback
by MenTaLguY 31 Dec '03

31 Dec '03

Well, I've made an executive decision... The packaging situation with gtkmm2 is really pretty dire at this point on several platforms (even RedHat), so I'm going to roll back our gtkmm2 dependency for now. I hadn't expected it would trail Gtk2 packaging by THIS MUCH, but it seems to be... Hopefully the situation will resolve itself in the next month or two, and we can go back to using it. -mental

2 1

Re: [Inkscape-devel] Re: dnd [was: Re: export patches]
by bulia byak 31 Dec '03

31 Dec '03

>Okay, had to change > >rnewdoc = sp_repr_read_mem (svgdata, data->length, NULL); > >to > >rnewdoc = sp_repr_read_mem (svgdata, data->length, SP_SVG_NS_URI); > >now it should behave exactly like the import function. Updated patch >attached and posted to the tracker. Works for me ;) Thanks! Indeed it works well now. I think you can commit it. Of course it'll need more testing when/if Gimp supports svg dnd too. And by the way, why can't I drag an svg document from my KDE desktop into Inkscape? Is this version of dnd Gnome-only? _________________________________________________________________ Protect your PC - get McAfee.com VirusScan Online http://clinic.mcafee.com/clinic/ibuy/campaign.asp?cid=3963

5 8

Nuke the fake config?
by Jon A. Cruz 31 Dec '03

31 Dec '03

Can we get rid of gen_nr_config.c and nr_config.h? We 'should' just be using standard config approaches and config.h. This is even in gen_nr_config.c /** * A little utility function to generate header info. * * Yes, it would be possible to do this using more "native" autoconf * features, but I personally find this approach to be cleaner. * * The output of this program is generally written to art_config.h, * which is installed in libart's include dir. **/ To me, that translates as "I don't understand the standard ways, so I'll subset and re-invent a wheel myself". Given that we're trying to keep cross-platform including Win32 and OS X support, it probably is time to kill it. Are there any issues I'm missing? Or anything from those building on Win32 that would complicate things?

3 2

can't compile
by bulia byak 31 Dec '03

31 Dec '03

There's a mismatch in declarations: desktop-snap.h has the last argument of _free_snap as const, but *.cpp has it without const. Please fix one way or the other. _________________________________________________________________ The new MSN 8: advanced junk mail protection and 2 months FREE* http://join.msn.com/?page=dept/bcomm&pgmarket=en-ca&RU=http%3a%2f%2fjoin.ms…

6 12

SPOffset
by MenTaLguY 30 Dec '03

30 Dec '03

I've been playing with SPOffset today -- great stuff! It has some awesome potential as an artistic tool. That said, there are a few issues that bug me: 1. something's going a bit wonky with the associated NRArenaItem -- an assertion keeps triggering about the object's bounding box. Unfortunately, until I figure out what NRArenaItem::state actually means, I'm not sure of the best way to fix this. 2. I think perhaps Inset/Outset should modify the original path, rather than making a duplicate. Its behavior of duplicating the path and leaving them both selected surprised and frustrated me. 3. I'm not sure the use of xlink:href is really SVG-compliant in this case ... also, the behavior of following the original path was also a little surprising and frustrating, once I managed to get the original path selected without selecting the outset. -mental

3 3

node edit now listens!
by bulia byak 30 Dec '03

30 Dec '03

I finally added repr listener to the node editor, which means it now updates its nodes when an object is modified outside of its control (by undo, in XML editor, in another view etc.) This fixes 857997, 852973, 865627, and re-fixes 848355 properly (previous hack in document_undo is removed). "Works for me" but I'd appreciate if it is tested really thoroughly, as the code is still somewhat flakey. E.g. instead of the closed bug I'm now submitting two other bugs related to the new code which I still need to fix. Any insights are appreciated. Some issues: - When to regenerate the nodepath? Not always when a repr is changed, because the majority of such changes are from the node editor itself. One approach: compare the actual repr with a "fake" repr temporarily created from the current nodepath; if different, recreate nodepath. Another approach: When updating repr from within node editor, set up the local_change flag. The listener checks this flag and if it's set, does not recreate nodepath (but zeroes the flag). Right now my code uses both these approaches, the first for the nodepath and the other for knot_holder (the structure used to edit ellipses, spirals etc). Does anyone have an opinion on which approach is more preferable? - Preserving selection is a hack: before regenerating nodepath, I save the list of the numbers of selected nodes, and then re-select them in the new nodepath. This does not always work (see bug 867525). Can anyone propose a better idea? - Objects within groups: they can be transformed via their group, without their repr being changed at all (see bug 867521). What to do with them? _________________________________________________________________ The new MSN 8: advanced junk mail protection and 2 months FREE* http://join.msn.com/?page=dept/bcomm&pgmarket=en-ca&RU=http%3a%2f%2fjoin.ms…

1 0

RE: [Inkscape-devel] Nudge control
by bulia byak 29 Dec '03

29 Dec '03

>Ok. I know it's the silly season but can someone tell me where the nudge >control distance setting for the arrow keys is located? I want to control >the distance setting as the current default one is too high. There must a >control surely but I cannot find it. It's not in the UI yet. Maybe in the next version. For now you can change it in your ~/.inkscape/preferences.xml, by setting e.g. <group id="nudgedistance" value="10"/> (the value is in points). >While I'm on nudge I notice that operation of nudge seems a bit erratic on >occasion. Sometimes an object does not nudge with the arrows until it has >been moved manually with the mouse. When it happens none of the arrows >work. >I'll try and document more precisely when it occurs. Most likely the arrows don't work when the mouse is not over the canvas. It's documented here: http://inkscape.org/cgi-bin/wiki.pl?EventContextGrab If the misbehavior you experience is different, please file a detailed bug report. _________________________________________________________________ Help STOP SPAM with the new MSN 8 and get 2 months FREE* http://join.msn.com/?page=dept/bcomm&pgmarket=en-ca&RU=http%3a%2f%2fjoin.ms…

2 1

Re: [Inkscape-devel] Ungroup All Command
by bulia byak 29 Dec '03

29 Dec '03

>On Tue, 30 Dec 2003, vellum wrote: > > In the real world of drawing pictures I often group objects and then put > > groups within groups etc sometimes many times over. Commonly I have many > > composite groups floating around. Now, if I need to modify something >buried > > in a composite group it can be a real pain getting back down to it. It >would > > be very useful to have an "Ungroup All" command to complement "Group" >and > > "Ungroup". No, "Ungroup all" is not necessary, Ungroup must simply ungroup _all_ selected groups (now it only works if one group is selected). This way you'll be able to ungroup any nested group(s) down to atoms by pressing ctrl-u sequentially. It's in my todo list already. _________________________________________________________________ The new MSN 8: advanced junk mail protection and 2 months FREE* http://join.msn.com/?page=dept/bcomm&pgmarket=en-ca&RU=http%3a%2f%2fjoin.ms…

1 0

Gtkmm
by Daniel Borgmann 29 Dec '03

29 Dec '03

Hello, I have just changed the save confirmation dialog to use Gtkmm. For this to work, I also had to change gtk_init (&argc, (char ***) &argv); to Gtk::Main kit = Gtk::Main(&argc, (char ***) &argv); in main.cpp, otherwise it would crash with lots of "quark" errors. I don't really understand the reasons to be honest, but it seems somewhat logical and it works well. My question is, should I commit this or should the Gtkmm stuff rather happen in a branch? Daniel

4 4