3 ms·
I do hope he is speaking entirely about software patents rather than all patents in general. There may be issues with non-software patents but under any system
by drucken 15y ago
I do hope he is speaking entirely about software patents rather than all patents in general.
There may be issues with non-software patents but under any system their problems are trivial in scope and solution space compared to software patents.
- rdl 15y agoEven most modern "real" (non-software) patents don't seem to be useful for actually teaching people how to implement technologies, which was one of the major purposes of requiring the disclosure.
- zanny 15y agoI'd argue the 20 year long patent law in place now is way too long for a publicly granted monopoly of production.
- rdl 15y agoThe possible exception being in regulated industries like pharma where it's entirely possible for a drug to take 15 years to develop and test. I think healthcare and pharma probably need to be treated differently, but even there, patents don't seem to be making the world a better place, at least not in the way they were originally intended.
- zanny 15y agoIf you used a patent at launch system, where you could only copyright / patent something going to market, you could avoid the corner case. It is a good practice in general. You can only patent that that you can actively push to market in a reasonable time frame (even if it is in small quantities). Under a policy like that, drugs couldn't be issued patents until after they pass the FDA, but the FDA could internally have rules against letting someone else copy a design and pass it through faster. I guess that would make copyright law easier at the expense of bureaucracy.