Anarchists may have public access to vast libraries of blueprints, but just because access to these libraries is "free" does not mean that they are without cost. Modeling off of torrent "seeds/leeches", having someone come in and schlurp down a full copy of a library of blueprints without giving anything back is looked down upon in anarchist circles. In game terms, yes, PCs still need to use rep and networking to get access to the blueprints, which are not on some generic server, but being hosted by an actual person (of one variety or another) in exchange for the rep boost. Taking more than they give back results in a ding to the rep scores of the offenders. (Of course, attempting to hack the server and make a copy without compensation or thanks is still a possibility).

Another related point that occurred to me involves the realization that, while many blueprints are developed by anarchists, for fellow anarchists, not all of them are. Many are still developed by the hypercorps, who are not above also hacking anarchist archives and either corrupting blueprints or inserting nefarious backdoors into them. Therefore, there is a constant hacking war going on, as the hypercorps try to corrupt the anarchist designs and the anarchists try to keep them clean and usable. This takes time, that most precious of commodities. So, in line with the Seed/Leech Solution, just coming in and taking full copies of blueprints that some transhuman has spent considerable time making sure were clean and usable is still going to cost rep, while making sure that unverified blueprints are clean will take the PCs time as well, meaning that, either way, it's not "free", meaning that there is still a cost involved, which makes me happy.

A more fully fleshed out thought to my original solution is, fine, the PCs have their mesh implants and maybe a couple of ectos packed full of every blueprint they could lay hands on. However, there's considerable difference between using a fabber or CM on a hab and using nanofabrication in the field, and one of the biggest issues is lack of rare elements. Having a design that can make railguns/cornucopia machines/giant rubber swords/etc in the Martian Outback, surrounded by all of that handy oxidized iron, silica and other useful trace elements, is very different from designs that can make those same things on a small iceball, where the nearest heavy elements are already in use in one's gear. Simply put, if the blueprint/recipe call for 211 g of thorium or 73 mg of uranium, then that blueprint isn't going to do anyone much use when those materials are unavailable. Now, the better blueprints will have possible substitutions, but there's only so much substitution one can do and retain functionality--engineering for these sorts of things is always about acceptable tradeoffs.

Whenever something is being fabricated, raw materials are needed. These materials can be gained in the following ways: Purchased or acquired with rep; raw materials for an item cost one item category lower than the finished item, subject to the GM's approval (rare elements, such as radioactive elements or antimatter, will typically cost at the lower bounds of the same cost category in credits). Materials for trivial items can be acquired with "half" of a trivial favor--i.e. two items worth of trivial items can be acquired with a single favor. Salvaged from existing items. Feeding an identical or related item into the fabber to be disassembled and reconstituted will usually suffice (as usual, subject to GM approval) Scrounged from junk or the environment; this uses the Scounging skill (and adds a new specialty of "Raw Materials" to the skill). General rule of thumb is that this is a Task action, with a situational penalty based on the cost category of the item (Trivial -0; Low -1; Moderate -5; High -10; Scarce -20). Chemical sensors/spectroscopes will generally add bonuses. Failure on these tests to acquire appropriate raw materials for the fabbed item means that either the item cannot be made, or requires substitutions for elements that are available locally. Modifying or selecting a blueprint for substitutions can be as easy as gaining a blueprint that already has a variety of acceptable substitutions and selecting one (recommend that these blueprints be one cost category higher than normal, due to their versatility) or programming a substitution into an existing blueprint. This second option is a Programming (Nanofabrication) Task action, with penalties depending on the intricacy of being fabricated being highly recommended. Failure means that the programmer is simply not up to the task. Severe/Critical Failure indicates that the element(s) in short supply are too integral to the original design and cannot be substituted. Success indicates a functional item, although due to the realities of engineering tradeoffs, a penalty for using that item is suggested. Excellent/Critical Success indicates that the programmer managed to push off the design tradeoff, making an item that is fully functional, and perhaps even superior to the original design. Side Note: While in the process of writing up this potential house rule, one thought did occur to me. I've seen considerable discussion on this forum regarding how the exponential growth made possible by the existence of freely available cornucopia machines "breaks" the game's economy. I happen to agree with this point, but wish to point to this house rule as a possible rules patch to at least ameliorate the issue--yes, exponential growth is still an issue, given unlimited resources, but, let's face it--the society in Eclipse Phase claims to be "post-scarcity", but actually isn't and there aren't any places with "unlimited resources". Metals, rare elements, volatiles... all of these and more are all scarce and not freely available. Therefore, using this rule to emulate if the raw materials are even available or usable could help solve or at least diminish that particular issue.