What should we do about dead listings?

Question

There are a few pending flags on posts for applications/scripts/libraries where the advertised program no longer works, sometimes meaning that the link to the source code and/or binary is also dead.

It's not helpful for visitors to Stack Apps to find something that they can't even use, so it seems reasonable to do something about these posts…but I'm undecided as to what. Options include:

• Outright deletion
• Closing with another custom reason
• Locking with the historical significance reason
• Editing the post to include a warning
• Merely posting a comment explaining the status in the hope someone happen to see it
• Shrug and do nothing
• Some other option I didn't consider

I don't have very strong feelings at the moment which one is best, so I'm open to hearing about what everyone feels the approach to these dead posts should be.

Answer 1

Yes, obsolete listings (script / app / library) should at the very least, be made obvious at a glance and ideally obvious in search results.

For reference, there are currently only 754 listings that are not placeholders (more on placeholders, below).

1. Delete sparingly; removing legacy content should be a last resort.
   1. Posts can have historical value.
   2. Even obsolete posts can provide useful ideas and/or code.
   3. Stack Exchange does and should strive to preserve high scoring posts that were once valid. ("There are a large number of views, upvotes and inbound links on the post," etc.)
2. EXCEPTION: Many placeholder questions should probably be deleted. More on this below.
3. Keep listings with, say, 5 score irregardless.
4. Only lock posts if they don't meet today's guidelines for acceptable posts (and have too much score to delete). My SWAG is that no listing will merit locking. Otherwise, all listings should be updateable by whomever has the time, motivation, and skill to do so.
5. Closing would be ideal, because it has that nice banner at the top and has good mechanisms in place for dealing with such posts. Note that this might trigger the Roomba on a few posts ^{(currently 4 candidates that aren't already on hold)} -- which may or may not be a good thing.
6. Tagging with obsolete. This is good and can help with both search and signaling listings that might get closed. Plus: Anyone can do it or suggest it. Minus 1: it's not enough to let the new/casual user realize that the listing is now salvage and/or historical. Minus 2: Easy to remove (erroneously).
7. Leading the title with "OBSOLETE". This has the same advantages and disadvantages as tagging, but sends a much clearer signal to searchers.

---

Proposed Actions:

1. EDIT all such listings with the title prefix "OBSOLETE - ", and the tag obsolete. This can be done immediately and will help with all further actions.
2. Get that custom close reason added and use it.
3. Delete and/or lock if really deserved.
4. Possibly start closing abandoned placeholder apps as previously discussed, as these are an obvious source of (usually) no-value clutter.
5. Some of the placeholders that should be closed should also be deleted and since they have may have a (pity/newbie) score of 1, won't get Roomba'd.

Per this SEDE query, here is a list of placeholders that perhaps should be closed. Update: the original list, from July 2018, have now all been closed.

Updatier, of the 13 original very-obsolete posts, all have now been closed:

1. These have been closed but need to be deleted (but may Roomba):

   • Stacked Up Comments [Placeholder]
   • Stack Exchange v2.2 PHP API Client

2. These have been closed and Roomba'd. Good Job! (Need 10K to see):

   • AnswerMachine (Placeholder)
   • Developing a new parenting app (Placeholder)
   • Browse StackOverflow community from iOS devices (placeholder)
   • Post to have write access (Placeholder)
   • PhoneGap Stack Overflow app (Placeholder)
   • Required Post for Write Access for android (Placeholder)
   • (Placeholder) A test app
   • Stack Exchange Eclipse Plugin (Placeholder)
   • Test Application Post (Placeholder)
   • Question analysis application (placeholder)
   • Placeholder for my application

Answer 2

I would vote for the historical lock option.

Straight deletion is harsh and someone somewhere may still need the information.
Creating a custom close reason could conceivably see posts erroneously voted closed.
A warning would have to be consistently worded for all posts to avoid any possibility of confusion in future readers.
A comment, with the best will in the world, won't get noticed.
Doing nothing doesn't address the problem.