You're right. I was focused on the idea that for software to be taxed as a capital investment there had to be a time when it was considered a finished product. Like building a tractor. I guess the analogy then is how the tractor builder is taxed when fulfilling warranties. I suppose tractor business can expense as R&D work that goes into processes that make it easier to fulfill warranties.
Each version of the software is a finished product just as any particular tractor is a finished product. The difference is that I know plenty of physical assets that companies buy which see no changes other than maintenance over their entire lifetime. I don't know of a single piece of software produced which receives only bug fixes.
Yeah devs spend their time fixing bugs, but a large percentage of those bugs are the result of the software needing to work under new conditions or with new version of dependencies.
That fundamentally different than a tractor breaking down from wear and tear while doing the exact same thing it's always done.
>I don't know of a single piece of software produced which receives only bug fixes.
TeX
https://web.archive.org/web/20190428184722/https://texfaq.or...
Interesting. However for the purpose of this discussion it is both:
A) Irrelevant since it's not being run by a for-profit corporation, the kind that would care about these tax bills.
B) It's a bit of a cheat. TeX itself may be approaching an asymptote, but scientists writing papers in TeX do not use it directly or on its own. White it has created a reference upon which to build it has also externalized all that research and development into the thousands of accompanying packages which do keep receiving new features and need updating just to stand still.