I agree... there's no real reason for a stand-alone css3 tag.
There are only 13 questions with css3 that are not already css. See this search result. In those cases it is clear that each one would be satisfied with a simple css tag.
Here is the synonymization suggestion.
I find error-handling encompasses all others, in a more language-agnostic way. For example VB6/VBA doesn't have exceptions, but that doesn't stop anyone from handling errors.
Therefore, I'd make error-handling the "master" and exception-handling the synonym.
The tag wiki for exception sadly doesn't say what questions it should be used for, but I'd argue ...
Synonymizing these tags would be an absolutely wrong and harmful decision.
Let's look across the Stack Exchange network to see what sites like "Unix & Linux" or "Super User" say about Linux, Unix, and other operating systems:
Is Linux a Unix? Which is answered with "it depends", basically.
Is Mac OS X UNIX? Which is answered mostly with "yes".
Templates and Generics, though similar enough in some ways, are different enough in other ways to the point where a person expert in one will have little or no valuable insight on the other. A person who follows templates will not likely follow generics (and vice versa). This makes it different from other language-agnostic tags like linked-list, etc...
From a technical standpoint, if we look into the way C++ templates are implemented vs how Java generics are implemented, there is a big difference. C# generics are implemented in a way more similar to C++ templates, but there are a few differences as well. Certainly there's more, but these are the languages I know about.
From a semantic standpoint, my ...
I think that is very reasonable, for 2 reasons:
As you stated, we don't have individual tags for Python 3.5, 3.6, etc. and having python-2.x would be consistent.
While Python 2.6 has some significant differences compared to 2.5, namely new packages such as multiprocessing and json, as well as some compatibility module future_builtins incorporating ...
The adventure-game was probably created in reference to the original text based computer game, Adventure. It was written by Will Crowther and Don Woods in the late 1970's. You can play the game at the link, or take my word for it that it really is just a primitive RPG game. I don't see a useful distinction between these two tags, assuming that was the intent ...
Lately, the term ECMAScript has been becoming relevant again, as ECMAScript 6 is the name for a proposed major overhaul of the language. That said, the term is only meaningful if the tag has a version number, such as ecmascript-6.
Let's retag each ecmascript ...
I say do it. Wikipedia considers them the same thing and as far as I know, they are.
However, there might be a bit of confusion for any Java developers and that might be a drawback. Apparently both exists as separate entities, so we'll need to be prepared to explain that tags such as these are not language specific.
I would caution against making a sweeping decision that applies to all language-version tags.
Some languages, like java, are conservative, and introduce new features with very few incompatible changes between major versions. It makes sense to abolish java-8.
Other languages are more innovative. For example, python has introduced breaking changes from ...
If you mean the original meanings from when these genres were created (in the 1970s/80s):
Adventure game meant a single-player text-based game (maybe with illustrations) based on exploring and interacting with unknown locations/people/objects/circumstances. Not necessarily anything to do with role-playing, which came later.
Role-playing game originally ...
These tags should not be synonymised.
I will double check the objective-c-runtime-tagged questions and be certain that they are tagged correctly, but synonymising these tags would actually be quite harmful.
From the Objective-C runtime wiki:
The Objective-C runtime is a runtime support library provided with an implementation of the Objective-C language. ...
Go ahead with the rename, but don't synonymize until all 2.6 and lower questions have explicit mentions in the question (regardless of the tag) it's targeting 2.6.
For reasons half the Python community pretends not to understand, a version is very important in Python. Sometimes you can't/won't/shouldn't upgrade and this should be respected.
I don't think this solves anything.
Maybe I'm different to most as I normally pick well know categories and groups. Think [math] rather than [binomial expansion].
There's not many easing questions. (I had to add a second search term because it comes back with "ease" otherwise which isn't what we're searching for.)
Most of the questions seem to be ...
There are a lot of sorts around, and their spelling varies. The rationale for the irregularity is likely that their names are irregular. Radix sort has a space between them, quicksort has not. It is like this on Stack Overflow as well.
I disagree with the question's proposed solution: a synonym from sql-server to t-sql.
I am also going to disagree with the quoted text. This is wrong:
Transact-SQL is central to using Microsoft SQL Server. All
applications that communicate with an instance of SQL Server do so by
sending Transact-SQL statements to the server, regardless of the user
I'm not quite sure about this one. After comparing the questions, I can see some differences:
Questions tagged as asking-questions mostly involve asking about whether or not a certain question can be asked here.
Questions tagged as questions involve things such as feature requests or anything else about existing questions (which are also not about specific-...
I think that tags are useful even if one tag is a subset of another. For example, on Stack Overflow, there is the tag c#, but there are also its subsets: c#-2.0, c#-3.0, c#-4.0, c#-5.0. But this does not make such tags useless (for example, I am subscribed to c#-5.0, but not to the other subtags or the supertag).
So, I don't see any reason why the two tags ...
In my experience error-handling is a process that is common to all languages, when there's an error, you handle it (if you're a decent programmer).
As has been pointed out by NHGrif, many languages have fine-grained control of what error conditions exist, and often differentiates these error subtypes with names that include Error, Exception, and possibly ...
TL;DR: angular and angular.js are two different things, therefore not the logical synonyms, therefore de-synonymization is right in my opinion.
Looks like there is more than one thing here to think about.
Firstly, in my very humble opinion, CR.SE should stick to names that are meaningful/accepted in the communities around the specific technologies. ...
I agree that error-handling is the most generic and others should be synonyms of it. I actually tried to propose that synonym this morning, but exception-handling has more questions and the system wouldn't let me. (That's what brought me to meta actually.) For simplicity's sake, I see no issue with it being the other way around. I would like to see someone ...
I propose synonymizing master interval ← synonym range.
"Interval" is a less ambiguous term than "range", since it invokes the idea of interval arithmetic. "Range", on the other hand, has multiple meanings:
the set of allowable values for a variable
the set of possible output values of a function
the difference between the lowest and highest ...
There is a difference between the two genres, though RPG elements are now commonly added to just about every other genre - @smci's answer covers the difference well, I think.
The important thing is whether the difference is important in the context of Code Review. If we already understand the goal of the code, we shouldn't need to know that this is an ...