Progress of the BPBible 0.5 release
1 view
Skip to first unread message
Jonathan Morgan
May 16, 2011, 11:13:38 AM5/16/11
to bpbibl...@googlegroups.com, BPBible Translators
Dear all,
Since BPBible 0.5 has been taking longer than even I expected (after an end of December initial beta, I thought February optimistic but March achievable), I thought saying something about progress would be a good idea. Basically, there have been many, many changes made (see for example https://www.ohloh.net/p/bpbible/contributors/76858439787638), but there were far more quirks in the switch of rendering engine and the continuous scrolling mental model behind it than I expected. The task list keeps on growing, then shrinking, then growing as I and others find new problems and regressions, etc., etc. Exposing it to real world usage has led to a different set of problems being found from those found when implementing and testing in a "systematic" manner.
So far, we have had three betas. Every beta so far I have had significant problems reported to me within 24 hours. Some of these have been due to new people using the beta and finding different problems, more than I like have been regressions introduced by the changes in the beta in an area that I haven't really tested in ordinary usage.
At present I'm thinking of trying to knock off all the remaining major problems in the next couple of weeks, issuing another beta, then perhaps finalising a few things, putting out a new beta that is hopefully closer to a release candidate. Whichever one the intended final beta is would then be put out
Two platforms that still need more work (which I think should be done by the next beta - no guarantees though) are Linux and PortableApps. For PortableApps the existing Windows app should largely be PortableApp compliant once the build system has been updated, but none of the betas have yet been made into a preview and that larger audience may have other useful feedback. For Linux, the issue is principally building the wxWebConnect binaries. I have not done any work on this since December or January. At that point it was possible to build (at least on Ubuntu and Debian), but very fiddly. I know what needs to be done to make the process easier, but haven't done it. Even then it will rely on using XULRunner 1.9.2 only, which I gather might not be so easy on "security conscious" distros like Ubuntu. I'm hoping adding support for XR 2.0 shouldn't be that hard, and I have a plan to do it, but I would prefer to get 0.5 out since the last release was 9 months ago and the new one is a lot better (I would actually typically recommend using the betas instead, and they are what I have been using since beta 2). I have actually been working on this iteration of 0.5 for over a year, and work has been reclaimed from earlier iterations that have been being worked on pretty much since 0.4 was released 2.5 years ago.
Anyway, more patience is needed than I expected or wanted, but I hope the end result is worth it.
Further ahead the prognosis becomes more uncertain, but I would hope to have another release within two months containing bug fixes and enhancements that we would have put into 0.5 if we could justify delaying it. At some point alternate versification, install manager, and better user annotations support will become priority items, but their priority will be determined in part by how hard they appear to be, what development resources and time we have available, and what people feel like doing at the time. [I also have a host of other crazy ideas, but it's unlikely that many of them will see the light of day this year].
Any questions?
Jon
Since BPBible 0.5 has been taking longer than even I expected (after an end of December initial beta, I thought February optimistic but March achievable), I thought saying something about progress would be a good idea. Basically, there have been many, many changes made (see for example https://www.ohloh.net/p/bpbible/contributors/76858439787638), but there were far more quirks in the switch of rendering engine and the continuous scrolling mental model behind it than I expected. The task list keeps on growing, then shrinking, then growing as I and others find new problems and regressions, etc., etc. Exposing it to real world usage has led to a different set of problems being found from those found when implementing and testing in a "systematic" manner.
So far, we have had three betas. Every beta so far I have had significant problems reported to me within 24 hours. Some of these have been due to new people using the beta and finding different problems, more than I like have been regressions introduced by the changes in the beta in an area that I haven't really tested in ordinary usage.
At present I'm thinking of trying to knock off all the remaining major problems in the next couple of weeks, issuing another beta, then perhaps finalising a few things, putting out a new beta that is hopefully closer to a release candidate. Whichever one the intended final beta is would then be put out
Two platforms that still need more work (which I think should be done by the next beta - no guarantees though) are Linux and PortableApps. For PortableApps the existing Windows app should largely be PortableApp compliant once the build system has been updated, but none of the betas have yet been made into a preview and that larger audience may have other useful feedback. For Linux, the issue is principally building the wxWebConnect binaries. I have not done any work on this since December or January. At that point it was possible to build (at least on Ubuntu and Debian), but very fiddly. I know what needs to be done to make the process easier, but haven't done it. Even then it will rely on using XULRunner 1.9.2 only, which I gather might not be so easy on "security conscious" distros like Ubuntu. I'm hoping adding support for XR 2.0 shouldn't be that hard, and I have a plan to do it, but I would prefer to get 0.5 out since the last release was 9 months ago and the new one is a lot better (I would actually typically recommend using the betas instead, and they are what I have been using since beta 2). I have actually been working on this iteration of 0.5 for over a year, and work has been reclaimed from earlier iterations that have been being worked on pretty much since 0.4 was released 2.5 years ago.
Anyway, more patience is needed than I expected or wanted, but I hope the end result is worth it.
Further ahead the prognosis becomes more uncertain, but I would hope to have another release within two months containing bug fixes and enhancements that we would have put into 0.5 if we could justify delaying it. At some point alternate versification, install manager, and better user annotations support will become priority items, but their priority will be determined in part by how hard they appear to be, what development resources and time we have available, and what people feel like doing at the time. [I also have a host of other crazy ideas, but it's unlikely that many of them will see the light of day this year].
Any questions?
Jon
Jonathan Morgan
Jun 28, 2011, 10:05:56 AM6/28/11
to bpbibl...@googlegroups.com, BPBible Translators
Following up from this email, I have had a lot of other things to do, but have finally released beta 4 today. There are still a few known issues, the biggest one of which is a bug with focusing and list controls which affects at least Manage Topics, Search, and the dictionary window. I tried to have it fixed for beta 4, but nothing I have thrown at it yet has worked. If the worst comes to the worst 0.5 can be released with it not working, but it's something that would really affect usability if left in.
Jon
Jon
Reply all
Reply to author
Forward
0 new messages