If we open source our frontend codebase, how many meaningful contributions will we get in the first 30 days?
Basic
27
Ṁ1116resolved Aug 22
Resolved as
20%1D
1W
1M
ALL
"Meaningful contribution" will be fairly subjective, but it could look like "pull request that clearly took more than 1 hour of effort" or "catching and reporting serious bug". We'll use the PROB resolution mechanism, with each 5% = 1 meaningful contribution.
Note that this question closes in 2 weeks, but we won't resolve this market until the end of the year (with up to a 30-day extension into Jan 30 2023 if we open source our codebase on New Year's Eve.)
Feb 13, 5:00pm: the question actually closes in 4 weeks
Feb 18, 2:10pm:
- If open sourced before Jan 1 2023: resolution will be 30d after open source date
- If not: resolve to N/A on Jan 1 2023
Close date updated to 2022-04-13 11:59 pm
Close date updated to 2022-04-30 11:59 pm
This question is managed and resolved by Manifold.
Get
1,000and
3.00
1,0003.00Sort by:
[Austin here] This is what I'm counting:
https://github.com/manifoldmarkets/manifold/commit/c9f36449884c887d6dd5443e3f44ccdd9f5e38e8
https://github.com/manifoldmarkets/manifold/commit/2a5b68977b5176d52c76d0f0df52cebe83773ad0
https://github.com/manifoldmarkets/manifold/commit/c5763e6ec3b3974ba4423e348a7d7491e2f01a3f
https://github.com/manifoldmarkets/manifold/commit/1e0845f4b99af6c071c0b5312894b01bb15d22bc
And what I'm excluding:
https://github.com/manifoldmarkets/manifold/commit/a373f08b4ff7c90b4e046af4a04059299b1d217e
https://github.com/manifoldmarkets/manifold/commit/92aa56ba2050188f2bf459521906c76111c8e38c
https://github.com/manifoldmarkets/manifold/commit/0e64e0f9f99962ec2526d6d640ca63ea8300be57
Thanks for everyone's contributions!
I think it's quite likely that you would get more than 10 contributions (= 50% PROB), so bumping this up a bit. My main reasoning is that many of Manifold Market's users are really tech-savvy, and there is quite a lot of engagement in the "Feature Request" tags on the market.
Personally, I have a few ideas for contributions already. I would also imagine that existing outside contributions like https://outsidetheasylum.blog/manifold-markets-faq/ would quickly be merged into the project, if it were open-source.
1) are you worrying about legal fork ? Then you can use license more restrictive than typical open source and still allowing contributions. This does not solve someone using code illegally.
2) in general, sadly, open source makes little sense for most commercial projects
3) that should be freeform answer version, with numbers as only valid answers, not probability (that is a complete failure in this role, and not useful for most roles)
4) releasing only frontend would require provide some mockup of backend that would need to match behaviour exactly (may range from "done already" to "obnoxious")
Copied from a DM: I'd say that [fear of a fork is] maybe 20% of our reluctance [to open source]; then there's around 30% weight that "the value isn't clear yet" and 50% "we think it would slow down our velocity"
I'll caveat this all by saying I'm the biggest proponent of open-sourcing our code among us devs; previously I worked at Streamlit which had a heavy open-source component. And I do think it's a matter of time; I'd put 60% for us to be significantly open-source within 6 months. Especially as we move more from being a "product" to a "platform"; crypto would be one
But all that said, being open-source introduces a significant operational and maintenance burden. As a project, we're still very very young, and want to be able to maintain high shipping velocity and the ability to drastically pivot without worrying about breaking external dependencies. https://panelbear.com/blog/why-not-open-source/ is a post that resonated with my experience at Streamlit.
Then there's the more EMH-type argument: if you look at all the successful SaaS startups, almost none of them are open source today. Reddit was famously open at one point and then closed off access; I'm struggling to think of a marketplace or social startup that is succeeding off the open source route. And you would think there are big reasons to publicize such wins!
(The projects that I'm most excited about that look like this, incidentally, are Supabase and the LessWrong platform ForumMagnum)
So yeah -- hope that helps explain why we aren't open right this moment. Again, I'm very hopeful that we will open source down the line, and would appreciate if you think there's anything we're overlooking (since I can then go and use those to convince my cofounders!)