|In the next few days (barring unforeseen problems) I'll be posting the first test release of Scribe v1.87. The main features being less crashes, improved internationalization (include new translations) and more SPEED!
There are lots of optimizations in the new version, starting with:
This is mostly due to choosing better algorithms in key loops for the aforementioned parts of Scribe. Most of this you don't notice until you've got a lot of data in Scribe. So don't be too concerned if it's not obvious when you upgrade. All it means is that Scribe is growing with you and it'll scale better as you keep collecting email.
- Compacting folders is now much quicker (at least the non IO bound parts).
- The user interface caches a lot of the string mappings now. As a side effect lots of base controls now support glyph sub in Win32.
- Loading and sorting folders is faster.
- Bayesian filtering's word database build runs significantly faster due to better string management.
There is more to be done on the optimization front but to some degree a lot of the incrementation approaches have been milked for their benifits. Increasingly the next option for optimization requires a total rewrite of part of the code. Which I won't do this side of v2 for the sake of stability.
v2 will bring a swath of new technologies to Scribe. A proper filesystem/DB backend for lightning fast queries and hard core corruption resistance. More flexible user interface and class leading features. New powerful plugin API. Deeper Scripting support. Open config file formats (xml). And more stability.
There is even the possibility of open source. But I'm not promising anything.
All in all I think this year will be pretty exciting, and I'll looking forward to creating some great software. Who knows? we may even see the return of BeOS(zeta) versions of various programs ;)