3 ms·

Nice idea. I realize at this point you're just trying to feel the waters and measure demand. Good for you. But even for a first iteration landing page, I think

Nice idea. I realize at this point you're just trying to feel the waters and measure demand. Good for you. But even for a first iteration landing page, I think you need to demonstrate more awareness of how hard it is to outsource software development (even in small morsels). Submit > payment > success just doesn't convince me, on the contrary, it makes me wonder if you ever delegated software development tasks to others. Not counting issues such as domain knowledge and integration, that were already mentioned by others, communication is extremelly tricky. For small-sized jobs, producing a full specification of both funcional and non-funcional requirements and going through a few cycles of submit-review-resubmit can take more time than required to just do the job myself. With all those caveats, I would have use for services like these as well and I hope you can do a good job of communicating why someone would go to Codacy instead of Odesk, Freelancer.com or RentACoder.

jaimefjorge 14y ago
Thank you for your encouragement! The idea of the submit > payment > success is that it should be easy for devs to get started sending problems. In the background, however, you will be able to specify a number of requirements and details for each task. For instance, for every problem we try to detail a solution for it. For more complex tasks, some more discussing will be required. However, given the small problems that we are getting and the awesome description capabilities of the people sending them, this is surprisingly cool. However, we still have a long way to go. Just to be able to help people out with their work is already a very cool thing for us. Thank you for your constructive opinion and the references.
- sitkack 14y ago
  I would limit the problem domain so that delegation is easier. 1) submit unit tests in format XYZ 2) tests pass 3) receive code Generate high level reports off of submitted code (kloc, cyclomatic, whathavs). Another one would be Design the urls for a REST interface and the JSON responses, worker implements code to support REST using framework A/B/C. Or, implement algorithm in paper FOO using Python, Julia or Ocaml, should be able to read this dataset... It has to be fairly constrained or else it is hard to specify and complete projects. They should come back within the same order of magnitude that one would complete them in anyway w/o either party being overworked. One could also, instead of money, implement it as a sort of quid-pro-quo, I see a simple task (for me) that I could knock out, this gives me credits to spend on other projects. Virtual currency of code. Also, licensing is an issue. I wouldn't want to work on code I couldn't keep. This just seems wrong to me. The only way I am ok with it, is if the rates are really high and that doesn't seem amenable to a site like this.

jaimefjorge 14y ago

Thank you for your encouragement! The idea of the submit > payment > success is that it should be easy for devs to get started sending problems. In the background, however, you will be able to specify a number of requirements and details for each task. For instance, for every problem we try to detail a solution for it. For more complex tasks, some more discussing will be required. However, given the small problems that we are getting and the awesome description capabilities of the people sending them, this is surprisingly cool. However, we still have a long way to go. Just to be able to help people out with their work is already a very cool thing for us. Thank you for your constructive opinion and the references.

sitkack 14y ago
I would limit the problem domain so that delegation is easier. 1) submit unit tests in format XYZ 2) tests pass 3) receive code Generate high level reports off of submitted code (kloc, cyclomatic, whathavs). Another one would be Design the urls for a REST interface and the JSON responses, worker implements code to support REST using framework A/B/C. Or, implement algorithm in paper FOO using Python, Julia or Ocaml, should be able to read this dataset... It has to be fairly constrained or else it is hard to specify and complete projects. They should come back within the same order of magnitude that one would complete them in anyway w/o either party being overworked. One could also, instead of money, implement it as a sort of quid-pro-quo, I see a simple task (for me) that I could knock out, this gives me credits to spend on other projects. Virtual currency of code. Also, licensing is an issue. I wouldn't want to work on code I couldn't keep. This just seems wrong to me. The only way I am ok with it, is if the rates are really high and that doesn't seem amenable to a site like this.