Next version?

[HPG]

I'm just curious: I was wondering when we might be seeing the next version of BrainStorm. I imagine a number of suggestions must have piled up by now (I know I've made some), many of which you've responded to enthusiastically. When might we see them implemented, do you think?

[david]

I wish I knew.

We have been wrestling with a particular bug since June. It's rare, but it crashes BrainStorm when it happens. It's been in the program from birth as far as we can tell. It is to do with opening complex, multi-level text files containing lots of namesakes.

We really don't want to release until we've nailed it. We've had a few false starts where it passes all the tests, but then, as soon as we start soak-testing, it pops its head up somewhere else.

It's driving us both nuts (for different reasons - Marck's the one who's trying to nail it, and I'm the one who keeps making it reappear). So, sorry that's not much of an answer. The other thing is that we occasionally have to go out and earn a living.

In case you didn't know, BrainStorm's a spare-time labour of love.

[HPG]

Thanks, David -- that explanation makes a whole lot of sense. My sympathies on both fronts (an untraceable bug, and the need to earn a living).

Cheers
Hugh

[alx]

I recently came across a bug in the official Brainstorm release; it might be the same. Since it's been a very busy period, I didn't look further into it, especially since I gather that the "new" Brainstorm is being developed on a different platform (or is it?)

Anyway, in my case it always occured with fairly large models but not too many namesakes. However, it only occured when I had 3 or 4 multiple windows of the same model. I overcame it simply by maintaining two windows at most. Brainstorm has never crashed since.

If you believe that it's a different bug from the one you are currently researching I can try to reproduce it and send you the said model. I think that the program crashed when autosaving.

cheers
alx

[david]

These are the four known 'crash' bugs. They have been fixed for the next release:

1. Pasting text outline containing 2 adjacent namesakes with a namesaked parent. (You'd need to be pasting from a complex external text file for this to happen.)

2. Deeply nested models could corrupt on load and crash on close. I'm pretty sure that this is the same as 1. Except when we listed it as a bug, we didn't know that. (Marck's asleep so I can't ask him. He'll see this note though.)

3. Auto-save trigger during paste / drop operations caused severe crash. (I'd guess that this is the culprit, unless you've found a new one.)

4. Retrieve to current entry causing loop crash. (I think this means retrieving an entry to itself, which seems an odd thing to want to do, but it could happen by accident, I guess.)

Marck gets a new machine in a week or so. With luck this will enable him to push to the finish line. He's been horribly held up by an ailing .NET development environment.

[srdiamond]

We really don't want to release until we've nailed it. We've had a few false starts where it passes all the tests, but then, as soon as we start soak-testing, it pops its head up somewhere else.

Any progress on that bug?

[david]

All the bugs have been fixed. We are in late beta now.

We were planning a release on May 1 - the sixth birthday of the Windows version of BrainStorm.

This is looking unlikely. (Ploughing through documentation and last-minute checks.)

[david]

And, on May 1, we discovered that the search drop-down boxes worked on the development computer but not on the testing machines.

That's fixed.

We're in a pre-beta test cycle again.