As of today, AP V2.0 is largely feature complete. It has plenty of bugs and things that don’t quite work right, but I have not been concentrating on fixing them until I have most of the development stuff finished. Here’s a list of things still to do:
- Fix up the Script system. There have been several changes to the application that require a fair number of mods to the scripting interface (for example, local/global observations have gone away).
- Printing. The printing of objects/observations is currently broken, since changes were made to the app. Also the current printing format and mechanism gets lots of complaints, so I need to rethink it. Also, also, the current code is arcane and rats-nest-like. It really needs major refactoring (think: remodelling).
- Removing the Windows MDI interface. The MDI (Multiple Document Interface) paradigm is “old-fashioned”, according to Microsoft, so I’m hoping to lose it. It should also remove a lot of problems with maximising windows and so on, but does add some other quirks that I need to deal with.
- Adding a Windows installer. Mac folks won’t understand this, but Windows users expect an installer with every application. Most of my support problems are a side-effect of not having this. It has become even more critical with the advent of Vista, which is so paranoid I’m surprised it doesn’t command the CPU to explode if an unauthorised disk, device, or application is inserted.
- Changing the application so that it stores files in “approved” places, rather than all in one folder. This is no now the way to do things properly, both on the Mac and on Windows. Especially true in the case of Vista.
- Rewriting the user manual. This is a pain in the butt, but does bypass a lot of support problems (“RTFM”). If only Adobe Framemaker was available for Mac OS X.
- Revise the Plan Wizard to make it more sensible. This might not make it into V2.0, especially as I can’t think of an easy way to revise it without a lot of work. I’m currently thinking that a community effort might work (i.e. you tell me what you favourite objects are, and I create a database that everyone can access).
- Revise web site. V2 will use astroplanner.net as its base of operations. Currently there’s no matching website.
Some of the above can happen during the alpha period. However, V2.0a1 won’t be as friendly as previous alpha versions. There will be a lot of bugs. V2 is designed to run concurrently with V1. Although there are many changes in file format, most V1 files are converted/copied to V2 format for you, including plan documents. Currently, catalogue file format remains unchanged, so you don’t have to panic about big downloads.
So when is that V2.0a1 version appearing you may well ask? Well there are various problems that are stopping that from happening (some real, others in my mind):
- The compiler I am using produces code that has major memory leaks on Mac OS X Leopard. This is, apparently, an Apple bug and difficult for the compiler vendor to work around. However, I believe the Mac OS X 10.5.2 release might fix this. I’d like to see that happen before I get folks to test the software on Macs.
- The compiler has a major bug that prevents the scripting system from working properly on Windows. I’m hoping this will be fixed in January/February 2008.
- I hate releasing software that has known bugs, even as an alpha, but I might just bite the bullet and do so anyway.
- To do the Windows and Mac Intel testing/debugging, I really need a new computer. I have made the decision to delay this until Apple announces its new models (mid January 2008), in the hope that it will prevent me from buying a computer that becomes obsolete in a few weeks (like this one did…). My current computer/s is/are just too darn slow for development work.
- The longer I postpone releasing V2.0, the better experience it will be.
- I’m somewhat burned out and not firing on all 12 cylinders right now, which is slowing things down a bit.