Can't handle thousand of nods since it is not math base but object base which is way heavier

The basic usage worked great out of the box. After getting familiar with the code (the code is well documented with comments) I was able to extend the classes fairly easy and modify it so that a grid is automatically created based on GameObject Tiles already existing within my scene (instead of populating the grid from scratch with a FOR loop). I had to play around a bit, but overall this asset saved lots of hours of development.

While I can't use this in my project due to the nature of my project, their support was absolutely fabulous. I did play with their demo and it worked like a charm. It was very fast and responsive.

I own both the popular plugins: A* pathfinder project pro and this one, there are great assets targeting on different purpose, for lightweight grid base solution, this one is the most properly fit solution.

simple pathfinding is ok. but... pathfinding without agent size considered. no support. mailed the problem but no reply. however, this can only be used in tile game, tried integrating with isometric, got huge work to do when turning collider2d dynamic obstacle logic to isometric way.

Played around with this for the past several months, I'll list pros and then constructive criticism. Ultimately though after several months I dropped this asset from my project altogether. There are free options which do a better job than this asset until the dev can fix the wasted memory and huge GC this thing uses when you have more than a handful of Agents pathfinding. Pros: - Fairly well documented (but still lots of room for improvement). - A few tools to monitor your implementation/path requests/grid/thread usage. - Fairly easy to implement (there are easier implementations in other free solutions). - Code is fairly well commented Constructive criticism: - As stated earlier, it poorly manages memory and wastes it by creating heavy variables and arrays within loops that are discarded immediately after that iteration. Multiply this by tens of thousands of iterations within a few frames and it creates unacceaptable amount of GC. For this reason calling a pathfinding request is extremely heavy on memory if you're using a grid larger than several hundred nodes. You definitely feel the slowdown if you have more than a handful of Pathfinding Agents - although you can try to mask this issue with the multithreading included and also limit how many nodes out a search can be performed. - No adaquate way to alter a pathfinding request midway for moving targets without making extremely costly new pathfinding requests - which then recalculate the entire path rather than from where the target is to a closest node on the path - which would save a heck of a lot of CPU time. If you have more than a handful of mobs requiring to follow a moving player, you're going to run into efficiency problems. - Documentation says tile weighting is available in the PDF but it actually isn't implemented. I had to manually program it into the pathfinding cost algorithm. - Rather trying to deal with the inadequacies of poor memory management, instead this tool tries to bury them with multithreading option. You can offload the pathfinding to multipe threads, which would be great, but in this case it is just masking memory and non-optimized pathfinding requests. There is no path pooling or path recycling. - I did a comparison between this tool and the free A*Pathfinding Project (free version) using the same 2d grid. For 50 pathfinding AI going across a map of 100x100, A*Pathfinding Project was 200fps faster, fewer GC spikes, and was more optimized to prevent recycled variables from being wastefully dumped and cleared out of memory. A* Pathfinding Project also ran on the main thread and still managed to be faster compared to this 2D AStar asset. - The "stress test" included is poorly coded with 128x128 tilemap default made of gameobjects each calling "lateupdate", in other words there are 16384 tiles all making a useless call every frame aka a waste of CPU time - but this is just a minor inconvenience easily fixed, nonetheless it is this sort of poor planning that frustrates me with this tool. - The included tools to show visual grid use a crap load of wasted memory because again, the dev didn't think about optizing memory usage and instead has arrays being created and discarded for every tile it iterates over - aka GC spikes in the profiler. Finally, if you want a pathfinding system that is basic enough to use but optimized for more than a handful of pathfinding agents, this asset is not yet going to work for you - but has potential if the creator redesigns some fundamental flaws in how he organized his project. It has the potential but sadly updates are slow and long in between - no doubt the dev has a full time job and this is a hobby script asset he made in his spare time.

This is what a professional asset looks like! Scalable, editor visualization, well written code, I love it!

this helps me really. easy to use, have built in performence check tool, support implement, clear document.... and more! taste of pro!

I bought this asset, in general it is good, I was very pleased, until I came across the fact that there is no support for the layers. In the end, I threw away the money, but I hope the author will finish the work. Layers are pretty simple to implement a thing, so I'd like to see them in at least some form, at least in the API.

look nice wait try