-13

Given the enormous backlog of over 20 000 feature requests and over 10 000 open bug reports it is clear that SE can not keep up with the needs/wants of the community at large.

SE could open source the code, at least part of it, in order to facilitate faster development. SE is in a rather unique positions of having a userbase that is highly technically knowledgeable and willing to contribute. For example, the community made the Charcoal SmokeDetector.

However, as pointed out, due to SE Teams (among other things) SE will have an interest in not open sourcing everything.

A discussion about this topic has been had in the past (2009), but this discussion dealt with a different situation than today. How are thoughts/feelings about that now?

Improve this question
13
  • 12
    Open source their code and then… what? Let anyone modify it? Let people suggest modifications? I fail to understand what solely open sourcing SE's code will achieve. Commented Mar 4 at 9:59
  • 5
    I believe the first and biggest hurdle is to separate the code from anything sensitive, similar to Make the mobile apps open source! Commented Mar 4 at 10:00
  • 1
    @security_paranoid as of know we can only suggest ideas. If the code is open source one could make suggest an idea alongside with complete implementation. You would lessen the workload on SEs side massively from implementing an entire bug-fix/feature to reviewing the code. Meaning that in theory more features/ bug fixes can be implemented. Commented Mar 4 at 10:17
  • 8
    Let's say this question gets lots of upvotes. Stack Exchange's response (if they give one at all) is, no. We're no further forward than the previous question are we? Commented Mar 4 at 10:44
  • @RobertLongson Maybe, but we can say that we offered a solution. Commented Mar 4 at 11:53
  • A lot of what's in that "enormous backlog" is stuff that the community has indicated it doesn't want (or is a duplicate). Adding score:3.. is:q closed:no to the searches cuts both of them down by more than half. Commented Mar 4 at 14:16
  • a backlog with the magnitude of thousands of feature request is still way too big Commented Mar 4 at 15:28
  • 4
    @A-Tech I do support SE code being open source in theory, but practically...there is a reason SE has hundreds of developers and "some random dude on the internet" modifying code is likely to cause more harm than good (in terms of bugs and time wasted) Commented Mar 4 at 16:42
  • 4
    I don't think SE has 'hundreds' of developers. Maybe dozens at best. Commented Mar 4 at 17:18
  • 1
    Or you could just use an alternative (for example) that already is open source and also shows much greater respect for its community, etc. Commented Mar 4 at 18:37
  • 2
    Why was this downvoted? This is not a feature request, but rather a discussion, so I do not see how 'downvotes indicate disagreement' applies here. Commented Mar 4 at 19:13
  • 3
    @CPlus It doesn't but people vote like this almost forever and everywhere. Just look at all the negatively received company posts. They aren't feature requests either, and aren't written so badly as to warrant hundreds of downvotes. Commented Mar 5 at 8:28
  • @CPlus Downvoting is the default action on meta no matter the actual content. Commented Mar 5 at 8:30

1 Answer 1

Reset to default
13

I think its a deeply complicated set of problems - and open sourcing wouldn't be a magic wand for a lot of it.

I suspect quite a lot of the 'backlog' is things that might be fixed or obsolete, but no one really kept track or things that are wontfix/not really a important thing as far as the company is concerned. I don't think the raw numbers tell the whole story outside, there's a lot of tech debt over time, and there's not enough caffeine to go through all of it.

Charcoal exists cause existing official tools weren't good enough, and the loose coupling helped a bit. Its a small team, focused on a specific problem with broad impact on the network. To me, it feels like an anomaly. Contrast that, with say replacement mobile apps - which are mostly possible but don't exist.

I'd say there's a handful of issues, that are mirrored both on the company, and community side - that resources aren’t always community focused (and if parts of the code are open sourced, we'd still need company support to review and merge code contributions, and figure out an appropriate licence for these contributions.

That also means that potentially a good community contribution might not work out either cause its one person's passion project, or cause its just not something the company wants to touch on right now.

I'd note there's small parts of SO that're open sourced now and a very quick look suggests most PRs are either by former or current staff, and compared to MSE, the issues are sparse.

Critically, the company's likely to still be calling the shots on what's worked on, so it would be essentially extracting work from the community towards the company's goals as much as ours.

As much as open source could be a solution, we'd likely get a lot more headway from appropriate resourcing, and better alignment between the community's needs/desires and what the company works on.

Improve this answer
0

You must log in to answer this question.

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.