2 ms·
> This is absolutely par for the course, open source means no warranty. No warranty, sure, but if the original authors have a working program and what they pub
by yjftsjthsd-h 4y ago
> This is absolutely par for the course, open source means no warranty.
No warranty, sure, but if the original authors have a working program and what they publish doesn't work then did they really meaningfully open source it? Like, if you ask an Android OEM for their kernel source and they send you a tarball that when compiled can't actually boot their phone, is that above board?
(None of this is specific to the particular authors here, BTW; I will happily criticize literally anyone claiming to do Science or Open Source without publishing enough that others can actually reproduce their work.)
- mandevil 4y agoThe team I work for recently open-sourced a version of our SaaS code. Thanks to docker, we are pretty sure that it will work for you if you follow the regular install instructions- containerization is awesome for that. The instructions on how to set up a local dev environment, without docker? That's a lot harder, has a ton more moving pieces, and if we don't actively maintain and test it regularly (to catch when libraries rot out from underneath us) I suspect that in 6-12 months those instructions will not work for anyone, because this is an incredibly complex beast doing machine-learning related things with a whole lot of dependencies and moving pieces (DB, API, Web interface, caching layer, etc.) so there are a lot of potential failure points. This was the result of a team of 3-4 SW engineers and a UX engineer working on open-sourcing our existing, working, generating actual revenue SaaS code base for 5-6 months (not as our only task but as a very important one). So it was, say, somewhere around 2 man years of effort to take something that I know exists and works and turn it into a reasonable Open Source project. There is a huge difference in the amount of work required between simple "open sourcing" and actual effective Open Sourcing, and I suspect that few if anyone involved in "Open Science" is doing that work absent large funding sources specifically requiring and supporting with money the good kind. Just a simple fork of the repo? I think the result for us would have been the same as what happened here, and I know that our SaaS project works- I can look at my paycheck and see that.