Hmm, well for me:

• Better memory management within the editor itself. I have to preview projects in a window or I risk having to sit there for several minutes before the layout is workable again (this is on an 8GB/10,000RPM development PC).
• Multi-core support. Perhaps a drop box for "Single/Multi/Auto" etc.
• Path movement directly implemented.

Not sure what else off the top of my head. Other than just pure optimization.

Nothing is wrong with the site. It's just their name makes people question it.

Legit or not, the much better idea would be to go with just about any other company that doesn't just come and advertise randomly on the forum.

You need to establish yourself first and this is not the way to do it. There are plenty of other proven indie publishers to contact out there.

Only request is to have an option to have the smoke disperse more over time until it's back to no smoke at all.

Don't know if that was already planned/available/whatever but the plugin would be useless for me without such a feature, so thought I'd check!

From what I can tell, it seems to be "skipping" frames because the attack animation isn't properly resetting to the first frame. If I add a condition to do so, it works like it should even when accelerating.

Hard to tell if it's not resetting to the first frame properly because of a conflict or because there is actually a bug there.

EDIT: Added fix. Does it work for you? Needs version 99.92

Yes, certainly look at the Wiki first when you are stuck with something

While not complete, it's a great starting point!

Well, without even opening the cap file I can tell you it was made with a version of Construct over a year old.

It is very common for old cap files to not work correctly with the latest version (thus why you needed to open and re-save those events). Best option would be to try and take a look at newer cap files and learn from those.

Most A.I. in games today is based on a combination of two systems. The first being a waypoint system. These would be key points in the map that the A.I. should be traveling to at some point, for whatever reason.

The second is a hierarchy of importance. The A.I. might have a routine set for exploring, then one set for defending itself, and then another for chasing after the player. The routine chosen would be based on factors like Player health, enemy health, ammo remaining or are they near a power up etc. These different routines are ordered by importance and the best one is chosen based on the situation the A.I. is in.

So, the combination of the two then (for example) might be this: A.I. is wandering from point to point and spots the player. A.I. has full health and ammo so begins a chase. A.I. health drops too low so it looks for a nearby point to run to and get health, etc etc.

Sorry if that was obvious stuff, but all of this can be set up in Construct relatively easily with groups of events, private variables and containers.

I think nVidia made a dumb mistake there. Why would anyone make hardware accelerated physics an important part of their game, if it was going to be exclusive to either ati, or nvidia architectures? They should have licensed it to ATI, or made the implementation free, and earned their money through licensing it to game developers. They doomed it to being only a minor decorative thing, instead of something that could have drastically enhanced gameplay. And did any non-casual games ever use fluid? Now I'm sure someone will make an opencl or directcompute physics engine, or implementation of havoc.

Oh well yes I agree with you, I meant more that the positive is the fact we dodged a bullet where they could have become successful on their own, making it so we all need physics cards.

I also wish they had gone for a more open approach. Only working through nVidia hardware is unfortunate, despite being better than an actual card.

Carmack has a much better approach to this style of rendering from what he has said both in your link and elsewhere.

That being said, his last engine wasn't used nearly as much as the Quake engines before it. If that trend continues with Id Tech 5, I wonder if he can make the impact that would be needed to adopt.

Well that would be because the on key pressed happens very quickly, then the event goes back to checking if the player is idle or moving.

You could try adding the condition to the idle and run events to check if the shoot animation is playing, and if it isn't, then play them. Just make sure your animation for shooting is not going to loop forever, and that it is timed how you want it with repeats.