>Yes. As a note to any non-Mac developers who may be observing, this latest
>release involved cleaning up the code so that it would compile with GCC 4.0
>on Mac OS X. Hopefully that will also make it easier for Linux developers
>to begin working with the code. (Of course, the code is still full of Mac
>system calls and data structures, but it's a step in the right direction ;)
I am going to install the compiler on a dedicated machine in my office to get back to C programming, starting with Quartz since it is one of the emergencies and I am still familiar with graphic procedures in BP2.
> > One detail is missing: with this version graphics are not displayed on
> > MacOS 10.6 or higher.
I must correct : the limit is MacOS 10.5, not 10.6. Currently this version does not display graphics on an Intel machine running 10.5.8.
>I put a note about the graphics problem in the file "Known Issues.rtf" that
>accompanies the release. I still have not tested the application myself on
>an Intel Mac or on Snow Leopard at all. On an Intel Mac with OS X 10.5, you
>_should_ be able to get graphics with BP 2.9.7 beta by running it with
>Rosetta. The following Apple support article explains how to do this:
>
> Intel-based Macs: Forcing a Universal application to run with Rosetta
> http://support.apple.com/kb/HT1730
GOOD NEWS: It works! You get graphics when using Rosetta.
However there is a trick for activating Rosetta: the first time you open get-info on Bol Proccessor icon, the "Run with Rosetta" checkbox might appear dimmed. No panic, close the window and open it again! ;-)
Still, this should not refrain my motivation to rewrite graphic code...
> > I suggest that once graphics have been fixed we jump to version "2.10". ;-)
>
>I've been wondering whether we would run out of version numbers before we
>were ready for "3.0.0" ;-)
It would be easy to remember that version 2.9 is for MacOS 9 and 2.10 for MacOS 10...
>Of course, fixing the graphics problem will be a priority along with any
>work on porting to Linux. I would be pleased if someone else wanted to pick
>up the task of learning Quartz, etc. but I could do it eventually if needed.
There is also some interesting work that I would love to assist if anybody is interested : interfacing BP2 with the Mbrola diphone speech synthesizer: http://tcts.fpms.ac.be/synthesis/mbrola.html
Not only BP2 would start singing, but this would also be a fascinating tool for checking models of speech rhythm. This idea is supported by my colleague Daniel Hirst, a prominent specialist of speech prosody who designed a model similar tio polymetric expressions.
>Just an idea: Sourceforge now allows projects to run a long list of
>supplied web applications including MediaWiki. So, if desired, we could
>host the documentation wiki on http://bolprocessor.sourceforge.net/.
Great. Altogether we need to make sure that exports of the wiki are possible and can be archived in a long-term preservation system. Unfortunately, since MediaWiki is using a database this procedure is not certain. This is one of the reasons why I adopted PmWiki which only manipulates text files (in Unicode). Among the advantages is that you can move the entire site (data + program) to another web server in a couple of mouse-clicks!
In addition, crdo.fr is a long-term web location. We booked the domain for 10,000 years. ;-)
>I see that there have been 7 downloads of the 2.9.7 OS X binary already. I
>hope everyone is enjoying a slightly improved BP2!
The old interface looks so funny on a 24-inch monitor. ;-)
Bernard
|