Hi Will,

> I can absolutely see, and believe you on how painful this conversion will be to do, but
> obviously you'll have to do it sooner or later so I'd advocate why not sooner?

Well one factor is just basically fatigue - the comparison to the Mac version that you make is a good one, and the Mac version took a huge concentrated 6 month effort for me to produce. During that time other new feature development was totally suspended which is bad.

I don't think that I have enough energy to go through another similar slog right after finishing one up. And since it would involve no new modeling features, the end result would actually be of no benefit to a really large amount of users anyway...

So weighing in on the negative side there is a huge amount of work and sacrificing a bunch of feature work in order to do it, and on the positive side there is a benefit to only a smaller group of people... The cost/benefit analysis here tells me to hold off for now.

Then on top of that now that I have a Mac version I don't think that it would fly very well to make a 64-bit version for only one platform and not the other, and right now the mechanism that MoI uses for the Mac version does not support 64-bit Windows programs yet. So figuring out what to do about that only adds to the amount of work involved...

So it's definitely not looking like it will happen anytime soon, it is extremely unlikely that a 64-bit version will pop out as a kind of surprise like the Mac version sort of did.

Maybe it's possible sometime in the v5 timeframe. I am absolutely 100% certain that it won't happen for v3, the Mac version took up all the energy for such a large single-feature type effort for now. And I wouldn't mind doing an entire release without anything like that at all, so it might not happen for v4 either.

Some previous discussion here as well:
http://moi3d.com/forum/index.php?webtag=MOI&msg=5172.5

It may be that the best way forward that would include the Mac as well would involve a major architecture change with a kind of internal 64-bit cross-compiled database engine that still had a 32-bit layer for UI. But I'm not entirely sure about that, it will require a lot of planning and design work to accomplish that.


> I can absolutely see, and believe you on how painful this conversion will be to do, but obviously
> you'll have to do it sooner or later so I'd advocate why not sooner?

Well, it's actually not obvious - certainly one valid strategy is to never do a 64-version at all, ever... I don't think that's probably what will happen, it will probably be more like sometime in the future when more of the modeling modeling features that take priority have been done. But in any case it is not coming up anytime very soon anyway.

If 64-bit-ness is a requirement for what you need to do, then it means that MoI is just not the right tool for the job currently, I have never shied away from telling people that MoI is not the right software for them when they are trying to do something outside of what it is good at.


- Michael