* Craig Ringer <craig at postnewspapers.com.au> [2007-03-20 02:45:58 +0900]:
> *LOTS* of work would be needed on the codebase to get it to the point > where it could support a long-running stable server. The Scribus team > has been making improvements with a view to supporting this for a long > time, so the code is closer than it was. It's a frequent request and not > at all new. > > Commonly associated ideas are are: > - Running scripts from the command line > - Command line driven document export (no GUI) > > ... and work towards those has also bought further-out ideas like a > server version closer. > > A long-running server just takes the same idea and adds some > inter-process communication to make it happen without launching a full > new instance for each job - basically just a resource usage > optimisation. However, it must be MUCH more stable, be more careful > about long term memory handling issues like heap fragmentation, and more. Craig, Would you be able to formulate some ideas that are possible to implement within the 3 months of GSoC 2007 by people new to Scribus development, which would bring us closer to this coveted goal? We don't seem to be having a large influx of potential student participants for GSoC 2007 and I'm wondering if it's because people are just intimidated by our current codebase. There are only 5 days left until the gsoc student application deadline and we probably have two potential applicants so far. Anything to lower the entry barrier would help. Regards, Alex.
