Expanded Psionics Handbook: no limit
The original definition of astral construct, given in Expanded Psionics Handbook, has no limit on the number of astral constructs you could create, other than the limit on the number of power points you can spend in a day/the number of actions you can use before older constructs begin to expire.
Expanded Psionics Handbook is one of the most well-designed books for 3.5, widely acclaimed for solid and cohesive design.1 3.5 Psionics is an elegant system that provides a lot of important limitations that traditional spellcasting lacks. This makes it easier to use and more consistently balanced. It is also very popular, literally spawning an entire third-party publisher (Dreamscarred Press) to carry on its legacy, as well as a massive homebrew effort to translate traditional spellcasting to psionic mechanics.
Complete Psionics: harsh limits
Complete Psionics rewrites a number of powers from Expanded Psionics Handbook, including astral construct, and includes this line:
You can only have one astral construct shaped at a time.
No justification for this change is given, but several feats and a prestige class are offered to partially undo this nerf.
Complete Psionics is one of the most widely-panned books for 3.5. In contrast to Expanded Psionics Handbook, Complete Psionics is full bad ideas,2 filler,3 and complete disregard for the existing psionics system that had been in place.4 And then there is the issue of the “errata” themselves: instead of properly publishing errata for Expanded Psionics Handbook for free on the Internet, Wizards of the Coast put “errata” into Complete Psionics. This has neither the support of the official rules, nor is it an appropriate way to errata things: if there were issues with Expanded Psionics Handbook, that should be in an errata for that book.
So who wins? Expanded Psionics Handbook
The official errata documents available on Wizards website explain the process of adjudicating between conflicting rule sources: the “primary” source wins. In the case of astral construct, the primary source is Expanded Psionics Handbook.
Note that this contradicts the FAQ, which claims the most recent version wins. However, the FAQ itself is now contradicting the errata documents – and the errata documents are the primary source on errata. Thus, the errata documents’ rule wins, and thus Expanded Psionics Handbook wins. The FAQ is never the primary source on any issue, and any contradiction between the FAQ and the published rules is, under the errata rules, officially to be decided in favor of the published rules. Since the FAQ frequently makes such contradictions, the FAQ is basically worthless, even for its stated goal of clarifying the rules (since they get the rules wrong so often that one can put little faith in those rulings).
This is a good thing: there were no problems with shapers under Expanded Psionics Handbook
The nerf to astral construct was utterly inappropriate. Shaper psions were an interesting option for players, but not one that itself caused problems. Far more powerful options are available – both to psions, and especially to other classes like clerics, druids, or wizards. But shapers were interesting and competent, and many found them a great deal of fun to play.
The Complete Psionics nerf kills that character concept. A solitary astral construct is extremely limited, and the feats and prestige classes used to partially undo that nerf are far too expensive. I could perhaps see playing a shaper-nerfed-by-Complete Psionics in an intentionally low-power game, but such a game should include numerous houserules limiting powers across many classes. Complete Psionics’s rule is not appropriate to a typical game of D&D.
So even if you don’t buy the errata argument – and I’ll freely admit, it’s pretty weak to claim that a published book is wrong when it says it is changing things – using Complete Psionics on this point is a bad idea for most games. I cannot more strongly recommend that astral construct be left alone, free to make as many constructs as you have time/power points for.
Excepting the soulknife; even the best books always have something. The soulknife is atrocious, even weaker than the monk. Still, an overly-weak class is arguably superior to an overly-powerful one.
The divine mind is a psionic character – that is, a character who derives their magic from their inner power, from their own mind and soul – who can fall as a paladin can. The class is a complete travesty.
A similar case is the anarchic initiate, a class that is all about infusing powers with chaos. You can enter with Wild Surge, and indeed, it improves and expands on the concept of Wild Surge. But it can alternatively be entered with the Overchannel feat, which psions can choose as a bonus feat, and either way it requires 8 ranks in Knowledge (the planes) – which psions get in class, but wilders don’t. Thus, the class that is dedicated to expanding on the wilder’s key class feature, is easier to enter and more effective for psions than it is for wilders.
There are a ton of feats in Complete Psionic that are literally
____ Mind Blade
When you reshape your mind blade, you can change it into a _____.
Prerequisite: Ability to generate a mind blade, shape mind blade class feature.
Benefit: Any time you wish to reshape your mind blade using your shape mind blade class feature, you can add the ______ to your shape repertoire. You are proficient with your ______ mind blade—you are treated as if you possess the feat ______ Weapon Proficiency (_____ mind blade). The weapon is sized appropriately for you and deals damage as a _____.
It’s a disgusting waste of space and customer’s time and money.
Refers both to unnecessary nerfs like astral construct, as well as to completely inappropriate powers like synchronicity, and things like the divine mind. Complete Psionics is the source of most broken psionic tricks, and its material is just very poor in general.5
Just to be fair to Complete Psionics, the ardent, soulbow, and Practiced Manifester are pretty solid. Those are literally the only good things I can think of from the book, though.