BPBible on WebConnect beta and release plans
1 view
Skip to first unread message
Jonathan Morgan
Dec 17, 2010, 10:23:19 AM12/17/10
to bpbibl...@googlegroups.com
Hi all,
I have just merged all of the changes on the webconnect branch into trunk. [hopefully correctly - I hate large merges].
This means at least two things:
1. Anyone wanting to work from trunk will need to have a fairly recent copy of my wxWebConnect with PyWebConnect bindings (BPBible relies on the latest change there from around a week ago).
2. We think the WebConnect version is fairly stable and forms the base for future development.
We are planning on releasing the first beta of 0.5 fairly soon (in fact, one driver for merging with trunk was to follow the recently invented policy of only releasing betas and official releases from trunk). However, it still crashes in ways we feel a beta shouldn't do (like not starting up if there are no books installed and not working at all on Linux without hardcoding the XULRunner path).
The original goals of switching to wxWebConnect were to revive a stagnating project with a rendering engine that could do what we needed without requiring a complete rewrite of the application (like the XUL branch did), while still reusing as much of the work put into the XUL branch as we could. Barring any unexpected show-stoppers, I think we have met those goals. We also aimed to make a release of 0.5 by the end of the year, or at least to have a beta out by the end of the year. Given the amount of change in both user interface and code, I think it would be not just unrealistic but actively harmful to try and force out a full release by the end of the year, but it is very nearly beta-ready and it is likely that we will be able to start the beta process by the end of the year. After that the final release schedule will largely be determined by how long it takes to complete known outstanding tasks and by comments we receive from users. Mid-January would be ideal, but I expect mid-February to be more realistic.
I'm hoping that after the 0.5 release is out and the WebConnect approach has been validated in the Real World (TM) we can get back to delivering functionality incrementally in point releases every two - six months without having the endless problem of deciding how making a wxHTML based change will affect our future plans.
Jon
I have just merged all of the changes on the webconnect branch into trunk. [hopefully correctly - I hate large merges].
This means at least two things:
1. Anyone wanting to work from trunk will need to have a fairly recent copy of my wxWebConnect with PyWebConnect bindings (BPBible relies on the latest change there from around a week ago).
2. We think the WebConnect version is fairly stable and forms the base for future development.
We are planning on releasing the first beta of 0.5 fairly soon (in fact, one driver for merging with trunk was to follow the recently invented policy of only releasing betas and official releases from trunk). However, it still crashes in ways we feel a beta shouldn't do (like not starting up if there are no books installed and not working at all on Linux without hardcoding the XULRunner path).
The original goals of switching to wxWebConnect were to revive a stagnating project with a rendering engine that could do what we needed without requiring a complete rewrite of the application (like the XUL branch did), while still reusing as much of the work put into the XUL branch as we could. Barring any unexpected show-stoppers, I think we have met those goals. We also aimed to make a release of 0.5 by the end of the year, or at least to have a beta out by the end of the year. Given the amount of change in both user interface and code, I think it would be not just unrealistic but actively harmful to try and force out a full release by the end of the year, but it is very nearly beta-ready and it is likely that we will be able to start the beta process by the end of the year. After that the final release schedule will largely be determined by how long it takes to complete known outstanding tasks and by comments we receive from users. Mid-January would be ideal, but I expect mid-February to be more realistic.
I'm hoping that after the 0.5 release is out and the WebConnect approach has been validated in the Real World (TM) we can get back to delivering functionality incrementally in point releases every two - six months without having the endless problem of deciding how making a wxHTML based change will affect our future plans.
Jon
Reply all
Reply to author
Forward
0 new messages