3 ms·

Would it not be more sensible for the authors to have a reasonably complete implementation for testing, and then request funding to polish it off with a nice UI

by quasque 13y ago

Would it not be more sensible for the authors to have a reasonably complete implementation for testing, and then request funding to polish it off with a nice UI and a security audit? As it is at the moment, it seems quite risky to fund this.

voltagex_ 13y ago
Couldn't you look at it from the other direction? A complete implementation is more important than a pretty UI at the moment. I really hope more people fund it.
- quasque 13y ago
  Yes but there's no guarantee of a complete implementation, and as far as I can tell from looking at the website, no way of getting one's money back if the work is not done. At least if there is already a working prototype, then that inspires confidence in those who might be interested in donating.
  - voltagex_ 13y ago
    I don't see this as much different from any other donation to an open source project, but I can understand your concern.
  - cottonseed 13y ago
    There is a working prototype. Steven Jewel, the original author, wrote the spec and an initial prototype in Ruby. To support a wide variety of platforms, it was decided to rewrite that prototype in C++. The Ruby prototype, which I believe is no longer being maintained, can be found here: https://github.com/jewel/clearskies-ruby https://github.com/jewel/clearskies-ruby
    - quasque 13y ago
      Thanks, I hadn't seen that. Looks like they've since updated their bountysource page with more information, including this link.
cottonseed 13y ago
I sympathize with the attitude, but if people wait until the bulk of the work has already been done to donate to crowd fund open source projects, that means crowd funding is not a viable way to fund the bulk of open source projects. If true, that's a shame. I'm not sure how to fix it and mitigate the funder risk. Maybe have payment go through only after concrete milestones are delivered?
- quasque 13y ago
  That would be much better, although it does raise the question of who would arbitrate on such milestones being reached.

by quasque 13y ago

voltagex_ 13y ago

Couldn't you look at it from the other direction? A complete implementation is more important than a pretty UI at the moment. I really hope more people fund it.

quasque 13y ago
Yes but there's no guarantee of a complete implementation, and as far as I can tell from looking at the website, no way of getting one's money back if the work is not done. At least if there is already a working prototype, then that inspires confidence in those who might be interested in donating.
- voltagex_ 13y ago
  I don't see this as much different from any other donation to an open source project, but I can understand your concern.
- cottonseed 13y ago
  There is a working prototype. Steven Jewel, the original author, wrote the spec and an initial prototype in Ruby. To support a wide variety of platforms, it was decided to rewrite that prototype in C++. The Ruby prototype, which I believe is no longer being maintained, can be found here: https://github.com/jewel/clearskies-ruby https://github.com/jewel/clearskies-ruby
  - quasque 13y ago
    Thanks, I hadn't seen that. Looks like they've since updated their bountysource page with more information, including this link.

cottonseed 13y ago

I sympathize with the attitude, but if people wait until the bulk of the work has already been done to donate to crowd fund open source projects, that means crowd funding is not a viable way to fund the bulk of open source projects. If true, that's a shame. I'm not sure how to fix it and mitigate the funder risk. Maybe have payment go through only after concrete milestones are delivered?

quasque 13y ago
That would be much better, although it does raise the question of who would arbitrate on such milestones being reached.