From: dennis@made-it.com To: discuss-gnustep@gnu.org Subject: GNUstep Weekly Editorial 15-03-2002 Date: Fri, 15 Mar 2002 05:19:38 +0100 Editorial 15 March 2002 Your editor believed it was time to do some testing on the different Objective-C window managers (WOOM and Interface). Currently both are for developers only. I was not able to get WOOM compiled with a CVS version of GNUstep, while Interface did compile and work. I will be following both projects closely and as soon as one of them hits the 'user' state, I'll get back on this subject. As for now, keep on using Window Maker, or help one of the projects out by testing, submitting bug-reports, and fixes, or just cheer up the coders. Mailing lists In reaction to the last Objective-C++ notes regarding chimera, Stan Shebs from Apple reports he is trying the Apple Objective-C++ patches against the gcc compiler on Linux, while also Pedro Ivo Andrade Tavares is trying to get the patched gcc up and running. Seems like the incorporation of Objective-C++ in gcc is getting arond. Traditional Chinese support is on it's way with a great big thanks to Yen-Ju Chen Lars Sonchocky-Helldorf wrote a little script called preflight.sh that checks your system for GNUstep compliance. It's in the early stages, but has a lot of promises. A little note from Adam Fedor almost went by unnoticed. So I provide the message, with a little editing here: The following project was recently donated to GNU, and apparently will become the 'official' speech program (and possible speech recognition) program of GNU. I think the source is now on savannah or soon will be. ------- The project is to take an existing text-to-speech package and ancillary tools, apps, etc, as written for the NeXTSTEP 3.xx operating system by myself and colleagues, convert it to the GNU/Linux system, and make it available under a GPL. The package is based on several innovations, including: an acoustic-tube simulation of the vocal tract (instead of a formant resonance model); control of the behaviour of this tool using the DRM model due to Carre; creation and delivery of the necessary articulatory parameters derived from ordinary text; the interactive creation and manipulation of rules to create parameters from ordinary text as required to improve speech quality and create databases for different languages; and the interactive exploration of tube model control for demonstration and parameter creation purposes. There are a number of ancillary components such as a 70,000 word pronouncing dictionary, letter-to-sound rules for words not in the dictionary, a service application to provide speech services to arbitrary applications on a GNU/Linux machine, a Pronunciation Editor to allow users to add words to the dictionary, and various tools and documents as required to support development of the system and applications with embedded speech synthesis capability (such as aids for the visually disabled). The basic tube model has already been ported to GNU/Linux. The remainder has to *be* ported, using GNUstep facilities and other techniques. Some papers relevant to the work can be viewed at http://www.cpsc.ucalgary.ca/~hill/ A CD-R with all the initial material except the ported tube model was provided to Richard Stallman by courier. Code changes Richard Frith-Macdonald went on to get GNUstep compiled on Windows and has added some changes to gnustep-make to get it working through cygwin. In gnustep-base Adam and Richard hacked NSDistributedNotificationCenter, more stuff added to make builds on NeXTstep, OPENSTEP and Darwin function corectly and Adam set the versioning to development version 1.3.0 Adam removed a bug that crashed gnustep-gui on Solaris, while Richard concentrated on NSPasteboard so cutting and pasting between different hosts should be better. Adam removed from gnustep-xgps XGDrawingEngine.m and .h Happy Stepping, Dennis Leeuw