3 ms·
I feel a bit sorry for Riverbank, the company that makes the existing PyQt bindings, because their business model is being wiped out as Nokia replicates their e
by pavlov 17y ago
I feel a bit sorry for Riverbank, the company that makes the existing PyQt bindings, because their business model is being wiped out as Nokia replicates their efforts 1:1. Riverbank relied on dual licensing (GPL + commercial) to monetize their open source work, but this doesn't work anymore as Nokia has converted the entire Qt ecosystem to LGPL.
But such is life for a toolmaker: when the platform provider decides that your product is good enough to be integrated, you're out of business. The best defense is probably to make tools that are not merely API plumbing, but have a user-facing component that is not easily replicated.
- tome 17y agoCould they just have sold to the platform provider? Maybe not in this case, but in general is a possible and good strategy?
- mbreese 17y agoSince the original GPL and the new LGPL bindings use a different technique, I doubt he could have sold it to Nokia. Nokia likely had a different method in mind, or had some sort of technical reason for wanting to use the existing code. Also, I don't think that having your exit strategy be getting bought by the tool-providing company that your code depends on is a good one. They could just as easily use your work as a template for their own, and obsolete your product in the process. I think that it would be too risky.
- pavlov 17y agoThe PySide FAQ says: Nokia’s initial research into Python bindings for Qt involved speaking with Riverbank Computing, the makers of PyQt. We had several discussions with them to see if it was possible to use PyQt to achieve our goals. Unfortunately, a common agreement could not be found, so in the end we decided to proceed with PySide. It sounds like Nokia wanted to make a deal first. Maybe Riverbank underestimated Nokia's willingness to throw sheer manpower at working around them, or perhaps Riverbank thought that reimplementing the APIs through "cleanroom" reverse engineering would be more difficult than it turned out to be.