Glauber’s salt is an appealing material for both phase-change thermal energy storage and thermochemical energy storage, but the usual ways to make it might be inconvenient at small scales, particularly in places where the police consider such activities suspicious. So how can you make it from non-suspicious materials?
It occurs to me that gypsum is much more water-soluble than calcite, particularly at pH above 7, so in an equilibrium between solid gypsum, solid calcite, and some aqueous solution with some other cation that is highly soluble with both anions, there will be very little gypsum left. So perhaps you can boil drywall (plaster of Paris) in baking soda for a while (months?) to get Glauber’s salt, or just boil the baking soda to liberate the carbonate ion and then soak the plaster in the resulting liquor (for months?) after cooling.
Aside from its potential use as a source of mirabilite, this remineralization process might be a useful way to strengthen objects initially shaped from plaster of Paris (or composites cemented with it), at least if they survive it intact rather than falling to pieces or swelling out of shape. Plaster of Paris is easy to mold and sets up in a few minutes; it’s easy to abrade, cut, and burnish to precise dimensions once it’s set; it’s a fairly non-corrosive environment; and it adheres well to a wide variety of functional fillers, including traditional fillers like quartz sand, horsehair, and various forms of cellulose fibers, but also I think higher-performance fillers like fiberglass which are never used with it in practice. It would be very valuable to convert the resulting weak plaster or plaster-cemented composite, which will simply crumble with enough exposure to water, to the much stronger calcite or a calcite-cemented composite like lime mortar. (Matweb gives limestone as 5-25 MPa tensile, 14-255 MPa compressive, but that’s presumably full of cracks; I’m having a hard time finding properties of the pure minerals.)
Lime-mortar buildings have stood for thousands of years in rainy areas, and calcite can also withstand much higher temperatures than plaster of Paris for many cycles. (Plaster of Paris won’t melt, but it dehydrates below 200°, shrinking and losing strength, while calcining calcite back to quicklime requires temperatures over 600°, normally over 800°.) Moreover, the use of lime cement in the traditional form imposes very significant restrictions: the pH of the slaked-lime cement mix before it sets is high enough to degrade most glass fibers, limiting the strength of the final result. It also attacks most candidate thickeners and thixotropic additives, and so without a large amount of filler (normally quartz sand), lime cement tends to just turn into a puddle. If calcite objects could instead be created by remineralization of plaster of Paris, it would avoid these problems.
Even if the remineralization process required a large amount of other feedstocks relative to the gypsum, it could be very valuable: small additions of plaster of Paris with ordinary portland cement have of course been tried, and while the result does set up quickly, it never achieves much strength. If this is due to gypsum in the structure that could be converted to the much stronger calcite in such an easy way, many new possibilities appear.