User login

Navigation

Contribute

Interested in contributing to FootNotes? Click here to submit news stories to us or visit this site for more information about available opportunities.

Art.GNOME.org

Error: It's not possible to reach RSS file...

GNOME Support

Friends of GNOME

Now's a good time to say thanks for the best desktop environment on the planet by becoming a Friend of GNOME or by upping your current Friends level. Contribute by joining the Friends of GNOME program. All contributions are tax-deductible in the U.S. Thanks to our generous Friends

Links of Interest

Syndicate

Inkscape project formed

Submitted by stro on Thu, 2003-11-06 11:17. Inkscape

Bryce Harrington wrote: Nathan, mental, Ted and myself have decided to embark on our own
direction with the Sodipodi codebase. We have attempted to do this as
part of the Sodipodi project, but we believe we need to try out a new
project structure to have the freedom to be able to explore some
approaches radically different from Sodipodi.

We have recently reworked the Sodipodi codebase to build with a C++
compiler and renamed it 'Inkscape'. We have also set up a project on
Sourceforge and established a new website at http://www.inkscape.org.

Our goal is the creation of a fully SVG-compliant 2D graphics editor:
* Full SVG (plus XML, CSS2) compliance
* Core written in C++
* Gtk-based user interface following the standards set out in the
GNOME HIG
* Emphasis on a small core plus modular extensions for features (a la
Mozilla Firebird)
* Open, community-oriented development processes that welcomes
experimentation
* Baseline is the Sodipodi-Hydra codebase
* Removal of dead code

We are aiming specifically for a general purpose XML/SVG graphics
editor, and as a consequence will emphasize features that tend to fall
outside the Sodipodi target userbase. We would like to welcome anyone
who'd like to do some experimentation to participate with us. Subscribe
at http://lists.sourceforge.net/lists/listinfo/inkscape-devel

We also intend to seek use of third-party libraries in preference to
custom code (where reasonable); specifically, we are exploring Cairo
(formerly Xr), libcroco, Pango, the core Gtk widgets, but also other
existing and widespread facilities. The extension architecture is
partly intended to minimize library dependencies in the core.

Re: c -> c++ is a waste of time

Submitted by Anonymous George on Thu, 2003-11-06 21:58.

Yeah, I think we understand that. That is the reason to really get the core into a nice, clean object system. Then we can extend on that with scripting languages. This way the core functionality is fast, and everything else seems fast because it's using the fast core.

Nice in theory, except that I'll bet most of the time in an svg app is spent in rendering (gnome canvas, xrender, already C) and xml parsing (C), so theres no need to do any more in C or C++.

IMHO most apps would be much better written in python with C/C++ extensions for really time critical things.

1) a programmer can be much more productive in the same amount of time with a highlevel language than with c/c++. This is incredibly important for the free software community when programmer time is often spare time.

2) more robust - it's very difficult to get a hard crash like in c/c++ apps. In my experience often even outright errors do not result in user data loss, but may require the program to be restarted.

Hmm, perhaps I'll put together an advocacy page ;)

» parent