In other words, do you think it would take time away from other bugs that still need fixing?
I think it was your idea in the first place to resist the temptation to add many features until the bugs are fixed.
That's why that "thought bubble" isn't in the bullet point list. It belongs in V3.x along with OCS/ECS/AGA capabilities. And we have all speculated on what we want when we get around to working on that version
.
To make it clear, V2.10 is purely a bug-fix version.
A few of the reported bugs refer to the AMOS Help system - missing and incorrect links, topics cut off in the middle, etc. Plus, quite a few of the AMOS Help topics are blatantly incorrect! To me, that is a much of a problem as something that just doesn't work. If the Help is wrong, the user loses confidence. So part of my bug-fix process is to get a better quality and quantity of help available to the user. As I'd already got a swag of new documentation available in the AMOS Reference Manual project, that content is being incorporated into the bug-fixed help system. I've taken a little time out from the AMOS source files to ensure that I can generate both AMOS Help files and the Reference Manual from the same database. Otherwise, I'd be rewriting all the stuff I'd already got in the database
. The same will eventually apply to a possible HTML version of the Reference Manual.
The database
contents are by no means complete. That's another on-going project and will take some time yet before everything is in the new format and accurate.
It's also why I kept bleating about some feedback on the new Help File format. It's a major change to the look-and-feel of the help files.
As the database contents grow (something I get back to every now and then) I can then just issue Help File updates at the push of a button. The programming for that is now nearly complete.
It also gives me a break from wading through the sources checking the implications of bug fixes and ensuring the integrity of the code. Which can be a real pain at times. The latest fix (I'm in the middle of it) is a good example. It's the one about long (out-of-range) numbers being accepted by the editor, but truncated or otherwise mashed around. There's no error message for that and no mechanism to handle the message. I'm re-using some of the existing code routines but have to double-check for any unexpected repercussions from the changes. At that point in the code, I'm up to my neck in stack frames that are maintained over many hundreds of lines of code. I think I can see the way to do it...
For that bug, there's also the problem of how to release the fixes. The editor configuration file needs to be changed without upsetting the user's personal settings. So I'll have to build an updater! And so it goes on.
I released that alpha version to give people a chance to okay or veto the help file formats and to see that something is happening. And, as the changes were minor, I could get away with a simple "just backup and replace these files". All future fixes will need a formal updater package for release as many more files will be affected and configuration files will need to be changed in-situ. So it may be a little while before the next batch of fixes appears. At which stage I should be able to raise the level to a beta release.
So don't worry too much. I'm still on track and fully committed
(or should that be committable?
)