I logged into my old account and found an old question I asked:

https://stackoverflow.com/questions/32711321/setting-element...

culi 4 days ago | parent | next [-]

The linked answer seems like a valid guess for a relevant dupe. Like I said in my comment, "I understand a few eggs got cracked along the way to making this omelette" but I really don't think this was as widespread of a problem as people are making it out to be.

They also have Meta Stack Overflow to appeal if you think your question was unfairly marked as a dupe. From what I read, it seems that most mods back off readily

	▲	zahlman 4 days ago \| parent [-]
		> From what I read, it seems that most mods back off readily If a reasonable, policy-aware argument is presented, yes. In my experience, though, the large majority of requests are based in irrelevant differences, and OP often comes across and fundamentally opposed to the idea of marking duplicates at all.

▲

zahlman 4 days ago | parent | prev [-]

That was not closed by a moderator. In fact, it was closed automatically by the system, when you agreed that the question was a duplicate. Because of my privilege level I can see that information in the close dialog:

> A community member has associated this post with a similar question. If you believe that the duplicate closure is incorrect, submit an edit to the question to clarify the difference and recommend the question be reopened.

> Closed 10 years ago by paradite, CommunityBot.

> (List of close voters is only viewable by users with the close/reopen votes privilege)

... Actually, your reputation should be sufficient to show you that, too.

Anyway, it seems to me that the linked duplicate does answer the question. You asked why the unit-less value "stopped working", which presumably means that it was interpreted by newer browsers as having a different unit from what you intended; the linked duplicate is asking for the rules that determine the implicit unit when none is specified.