About the bounty: Up today the original question is not fully answered. I'm asking for the actual reasons of the merge. I've also deleted my own comments so question and answers can stand one in front of the others instead of... feelings.

was a tag used by the community but now it's gone. I don't see any discussion about this... deprecation. SEDE query (Good 'til the next data dump)

Edit: The same holds for . SEDE query (Good 'til the next data dump)

Edit 2: From @ChrisF's comment and Rebecca Chernoff' brief answer, it's confirmed that this was an action taken by a moderator. I don't agree with it, but I might be wrong. I can't decide without an explanation. I think the reason is part of what happen. But again, I can't make an informed without an explanation (Do note that this affects 230 questions).

Edit 3: Because anyone letting himself by comments instead of answers might wander why it's important the actual reason for this merge (as been part of the question) against general answers about merging or opinions not analyzing these questions in particular and how they meet some establish criteria ("citation needed") resulting in damage for both askers and answerers, here are some numbers ("Do measure by the same yardstick!"):

There are at least five tags related to Javascript. These are the numbers of the questions tagged with those and without :

: 620 https://stackoverflow.com/questions/tagged/javascript-events+-javascript

: 75 https://stackoverflow.com/questions/tagged/javascript-framework+-javascript

: 99 https://stackoverflow.com/questions/tagged/unobtrusive-javascript+-javascript

: 4 https://stackoverflow.com/questions/tagged/javascript-performance+-javascript

: 32 https://stackoverflow.com/questions/tagged/serverside-javascript+-javascript

closed as off-topic by TRiG is Timothy Richard Green, Glorfindel, ale, rene, Nathan Tuggy Nov 8 '16 at 18:44

This question appears to be off-topic. The users who voted to close gave this specific reason:

  • "This question pertains only to a specific site in the Stack Exchange Network. Questions on Meta Stack Exchange should pertain to our network or software that drives it as a whole, within the guidelines defined in the help center. You should ask this question on the meta site where your concern originated." – TRiG is Timothy Richard Green, Glorfindel, ale, rene, Nathan Tuggy
If this question can be reworded to fit the rules in the help center, please edit the question.

  • 1
    Do you have any examples of questions that were tagged with that tag so we can look through the revision history? – Daniel DiPaolo Apr 27 '11 at 16:19
  • 1
    Google cache: The xpathengines tag page. The revision histories of the posts don't contain any hint on the removal of the tag, but it seems to have been renamed to xpath. – sth Apr 27 '11 at 16:44
  • 2
    Looks like it was renamed in the database by a dev. I don't know why a synonym wasn't set up (which would have been the way I would have done it). – ChrisF Apr 29 '11 at 20:04
  • 1
    I could have sworn there was another MSO question about this, but it's not coming up in search, or even the deleted questions list. – Pops May 2 '11 at 16:32
  • Sorry, @Chris seems to know more than I do. – Pops May 2 '11 at 16:46
  • There is by the way already a synonym on xpath tag stackoverflow.com/tags/xpath/synonyms, but it has 0 votes. I can't upvote it and I assume a lot of people also not. Anyway, the tag synonym approach has to be revised as it gets too little attention. – BalusC May 2 '11 at 19:10
  • @Alejandro: I have expressed my position on this important problem in a separate answer. – Dimitre Novatchev May 4 '11 at 3:46
  • @Alejandro Sorry for the confusion, I appear to be misinterpreting some of your comments, and confusing them with comments Dimitre has made. – Adam Davis May 4 '11 at 21:51
  • @mmyers: Exactly. Or, as I said in another comment, they behave like an elephant in a china shop. – Dimitre Novatchev May 5 '11 at 2:43
  • 1
    @Adam: That's actually the scene I was thinking of. (I'm glad to see my character has moved up to A-list actors since the Stack Overflow movie last fall.) – mmyers May 5 '11 at 17:37
  • 1
    @mmyers You were always on my a-list. Am I being too intence? – Adam Davis May 5 '11 at 18:59
  • 4
    possible duplicate of Question title that doesn't describe the problem – Adam Davis May 6 '11 at 1:20
  • 2
    @Alejandro, except that it doesn't precisely describe the content of those questions as a group. The questions have nothing to do with each other in the slightest, and that makes it a horrible tag. – Charles May 9 '11 at 16:13
  • 1
    @shog9, @Rebecca Chernoff and all SO Managment: "Congratulations"! @Alejandro has cancelled his account at SO. He was one of the best specialists in XSLT, XPath and XML. You have successfully eliminated one of your leading specialists ! – Dimitre Novatchev May 13 '11 at 3:42
  • 2
    @Dimitre: every time I get dragged back to this question, I have to page past five of your deleted "answers" and dozens of inflammatory comments. If you actually care about this site, its users, or at least the little bit of respect you've managed to accumulate within your area of expertise, you'd do well to take a long, hard look at the way you've behaved here and how it reflects on them. I know you don't buy it, but... I don't have a dog in this fight. See ya in the funny pages... – Shog9 May 13 '11 at 4:28

My understanding is as follows:

xpath and xquery are essentially languages for dealing with XML data, such as HTML.

The generic terms "xpath engines" and "xquery engines" refer to the fact that there are many, many implementations of xpath and xquery.

It would be akin to marking C and C++ questions and . If the question is about

  • The compiler itself (ie, NOT a question about the language)
  • Multiple compilers (ie, NOT a question about how one compiler implements it)

then it would be reasonable to use those tags.

There are precious few questions that need to be marked as such - those questions which really concern a number of engines, perhaps, but most of those questions really only need to discuss the one engine that the OP is using, and so having a tag for that specific engine might be useful.

Looking at the google cache, though, I can plainly see that most of those questions are not about the multitude of xpathengines, and that the tag is superfluous for them.

There may be questions that would fit that tag, however the practical reality is that the tag is being egregiously abused by those who don't know that their question has no need for it. It's apparent that a lot of users are adding it because they don't know what else to add.

Tags are meant to make it easy to search for specific types of questions. It's quite obvious that these tags were useless - they didn't make it easier to find true engine questions. When people search Google for "xpathengines" the search results would be littered with BAD matches from stackoverflow, which lowers the page ranking of the site as a whole, and the questions individually.

Since people are using them incorrectly, and they have a negative impact on on site searches ans well as off site searches, then it's reasonable to ban them, even if there are some valid uses for them. The bad, unfortunately, outweighs the good.

  • 5
    @Dimitre Then please explain it to me. Your answer doesn't give us enough information to understand why it's important that the question "How to find child elements of a specific node with tinypath" requires the xpathengines tag. You are the expert, you should be able to help us understand why those tags are necessary, and when they should be used. So far your answer lacks this basic information, and so you're forcing us to simply accept your word, or worse - search google and try to make sense of it ourselves. Educate us. – Adam Davis May 4 '11 at 14:43
  • @Alejandro "What did happen with “xpathengines” and “xqueryengines” tags, and why?" suggests that you are interested in what happened to them and why. Perhaps if you reviewed your question and edited the title and the body, then you might get the answers you are looking for. – Adam Davis May 4 '11 at 14:49
  • @Alejandro So If I have a question about what arguments are valid for a particular compiler's implementation of a given method, then I should be including the c-compiler tag along with, say, the gcc tag? I don't understand - what value is added to the question by adding the generic term for all xpath engines when the question is clearly about one particular xpath engine? Either I don't understand the situation clearly enough (so educate me!) or you and Dimitre don't understand the purpose of tags. – Adam Davis May 4 '11 at 15:01
  • @Alejando They were removed by a moderator. It is because they were believed to be redundant. – Adam Davis May 4 '11 at 15:02
  • 10
    @Dimitre @Alejandro Why are you two refusing to educate us? Are you claiming that we are of such low intelligence that attempting to do so would be a waste of your precious time? You've already implied that people who don't understand should keep their nose out of your business. Well, my nose is in your business. You can either educate it, or suffer the consequences of an uneducated audience moderating your contributions. It's your choice, but I don't understand why you insist on keeping us in the dark. – Adam Davis May 4 '11 at 18:04
  • @Adam Davis: There are good books on both subjects. Hope this link may help (at least for XPAth): stackoverflow.com/questions/339930/… – Dimitre Novatchev May 4 '11 at 18:40
  • 4
    @Dimitre Thanks for the link. I've added two of them to my reading list and will reassess whether to change my behavior as soon as I understand the difference. Until then, or until someone explains it in a paragraph or two, I'm afraid I'm going to have to concur with the moderator who first made the change. – Adam Davis May 4 '11 at 19:12
  • 4
    @Dimitre It would be very nice if we could sort this out right now, but making a decision without understanding the situation would be worse than waiting until we understood it. However, for your general question about how we should approach tag changes in the future, check out Should experts be consulted prior to tag changes? and add your thoughts. – Adam Davis May 5 '11 at 2:32
  • 2
    @Dimitre I am not part of SO administration in any capacity, and am a user of these sites just like you. Please don't let my poor behavior or choices reflect badly on SO or its administration. – Adam Davis May 5 '11 at 3:10
  • 5
    @DimitreNovatchev, Shog9 had nothing to do with these tags outside of responding in this discussion after the merge had been done. – Rebecca Chernoff May 5 '11 at 4:02
  • 9
    @DimitreNovatchev, the who is not important. The fact that neither you nor Alejandro are answering the important questions, such as Charles' comment on this answer is very telling. – Rebecca Chernoff May 5 '11 at 4:25

These tags were merged by a moderator.

Basic details about tag merges: tag merges are done when the tags have similar meanings, are splitting questions that can be grouped, make it more difficult for users to find questions, etc. Tag merging also increases visibility of questions since users set up filters. (these have been mentioned several times in the comments, but adding them here since the comments are a disaster).

  • 5
    Thanks. But this is a little bit... brief answer. Might we know the reason? I know that this was retagged as support but I'd tagged first as discussion, also. Should I also tag this as request for undo? – user150068 May 2 '11 at 19:15
  • @Alejandro, I don't know anything other than what I posted. – Rebecca Chernoff May 2 '11 at 19:19
  • @Alejandro, um, what? – Rebecca Chernoff May 4 '11 at 0:02
  • @Alejandro: Do you mean the questions disappeared? – Dimitre Novatchev May 4 '11 at 16:34
  • 1
    @Alejandro, that might have been me. The tag showed up in the "list of new tags" on the 10k review list. I couldn't tell how the question was about xpath engines in general, so I untagged the question. This was before I saw the drama here on Meta and realized what I'd be stepping in, of course. If I'd known that there was controversy about the tag, I'd have left them alone until there was consensus. This is the question I edited. I have not reverted my edit -- for now. I suggest that the change remains until there's a clear forward path. – Charles May 4 '11 at 20:34
  • 5
    @Alejandro, if you want anyone to be able to get behind your crusade here, perhaps explaining why that question is tagged incorrectly would help. – Rebecca Chernoff May 4 '11 at 20:50
  • 4
    @Alejandro, looking at the merge can be handled after someone explains why it was bad. These rants by you and Dimitre are not helping. I suggest creating another question, do not make that question a rant, and clearly explain why these tags are not the same. You can use the question Charles pointed to for explaining. – Rebecca Chernoff May 4 '11 at 23:32
  • 3
    @Alejandro, the tags were merged. Done deal. You can choose to harp over that, or you can be productive and answer the questions asked in this discussion about why [xpath] and [xquery] are not suitable, particularly Charles' comment here. "Against what can I argue?" is the wrong approach. Don't argue. Calmly answer and explain. – Rebecca Chernoff May 5 '11 at 4:01
  • 3
    @DimitreNovatchev, the tags were merged because they are similar enough to go under the same grouping. This has been said many, many, many times throughout this thread. – Rebecca Chernoff May 5 '11 at 22:32
  • @Rebecca Chernoff: Nobody ever answered the question why the users of the tags were completely ignored and their opinion was never asked and taken into account. Due to actions like this I don't trust SO's administration about anything they say -- just look what they are doing. The users have only theoretical rights to create new tags and to tag questions into these. Regardless how much effort was put by the tagging users (badges they had been "rewarded" for this) SO abruptly destroys this enormous work. Go tagging after that. The tag structure will soon be flat as the brains who did this. – Dimitre Novatchev May 5 '11 at 22:41
  • @Rebecca Chernoff: SO's administration are not a specialists in this area to decide what is similar and what degree of similarity is needed in order to justify a decision to merge such tags. This has been said, to quote you, "many, many, many times throughout this thread". SO administration's decision reflects and promotes ignorance -- go on like this and mess up every single tag to show to users who and what SO administration is. Go on like this to show every user how much SO "cares" about them. – Dimitre Novatchev May 5 '11 at 22:48
  • 7
    @Dimitre: the idea that you can show up and claim "I am a specialist, you must listen to me" or "I have rep, you must listen to me" is antithetical to the fundamental philosophy of SO: Any given user can answer well or poorly on any given question, and you don't start out with +10 on your answer just because you answered well somewhere else. Allowing one user to abuse others simply because he feels he "owns" a given tag and can therefore dictate who uses it is beyond absurd, and your raging assertion that it should be is comical. Might I recommend a nice cup of tea? – Shog9 May 5 '11 at 23:02
  • 4
    @DimitreNovatchev, it has been answered, over and over. Tags exist things to group questions so users can find questions they are interested in easily. Thus, merges are done when tags are separated out making it difficult for users to find questions. We don't collude against our users. – Rebecca Chernoff May 5 '11 at 23:58
  • 1
    @DimitreNovatchev, meta.stackexchange.com/questions/89877/… - you've been asked to contribute to that constructively, but you haven't. If you feel strongly about this issue, which you seem to, this is a chance to help change things. – Rebecca Chernoff May 6 '11 at 1:37
  • 6
    @DimitreNovatchev, I agree with everything in your last comment. Both users and team members have tried to work with you and Alejandro. I'm sorry we are unable to resolve things. – Rebecca Chernoff May 6 '11 at 1:52

The rationale for the merge

The [xpath-engines] tag was merged into [xpath] because it was creating a distinction that non-expert users could neither understand nor properly use.

There are other problems with such tags as well; these have been discussed to the point of exhaustion in other answers here. In particular, see Adam's:

Tags are meant to make it easy to search for specific types of questions. It's quite obvious that these tags were useless - they didn't make it easier to find true engine questions. When people search Google for "xpathengines" the search results would be littered with BAD matches from stackoverflow, which lowers the page ranking of the site as a whole, and the questions individually.

Further references (read all answers and discussion):

Appealing to tradition w/ JavaScript

Regarding your last edit:

is... annoying. Strictly-speaking, it should be , which in turn should be (that is, two tags) - but because there's this persistent (and... totally understandable) belief that some form of browser DOM is actually part of the JavaScript language, it's not. This tag originated very early on in the life of the site, and has proved itself extremely popular... Indeed, it is perhaps the single best example of why this sort of hyphenated subgrouping is a bad idea: many, if not most of those 600+ questions you linked to are JavaScript questions, but they won't show up on anyone's feed unless they've thought to use wildcards.

Same goes for the rest, with the exception of , which is the proper name for a specific technique (and not a language or language subset). Actually, might fall into this category as well, but I'm skeptical: those questions would be better off tagged as [javascript] + server-side engine.

So how does this apply to [xpath-*]?

Well, it's a poor analog... The biggest problem with retagging most of these - especially javascript-events - is that they imply two tags: so you can't just merge them and be done; those 600+ questions without the primary tag would still have to be re-tagged manually, a non-trivial effort.

But it does illustrate something very important: this is the kind of mess that can result if namespaced-tagging is left unchecked. Merging the 200-odd [xpath-engines] questions was probably an easy decision, since the collateral damage was effectively non-existent (I've seen all of two questions so far that shouldn't have been tagged [xpath] as a result); merging [javascript-events] would immediately guarantee hundreds of mis-tagged questions. You don't want to wait until there are 2K+ instances of a tag to clean up the mess - it's much more likely to not get done.

But I don't want to see questions which aren't Pure XPath™!

If the idea of dealing with messy platform-specific details fills you with dread, simply exclude the platform: https://stackoverflow.com/questions/tagged/xpath+-java+-c#+-php+-javascript

The system provides numerous tools for highlighting and filtering questions. If the results aren't to your liking, consider that adding tags that describe the platform is preferable to removing information that others depend on.

  • @Shog9: You wrote: "Well, it's a poor analog" Sorry, my bad. But this clearly shows two standards. And that is the reason for asking about this specific merge of xpath-engine with xpath. After 12 days there is no answer about this: you are talking about your opinion of javascript-* and @Rebecca Chernoff wrote that she doesn't know anything other than what she posted: it seems that no moderator could neither ask to the responsible nor answer on his behalf. – user150068 May 9 '11 at 13:05
  • 3
    @Alejandro: there are waaay more than two standards... That's why this Meta site has been host to thousands of discussions on the various aspects of this, and why there are plenty of tags that violate current standards (btw: look up "appeal to tradition"). But yes, your question has been answered. Repeatedly. According to the person involved, Adam nailed it in his answer right off the bat, so you're probably not going to get yet another answer, since frankly we've all spent a little bit too much time dealing with this and still have to go deal with Dimitre throwing a tantrum on SO. – Shog9 May 9 '11 at 14:50
  • @Shog9: @Adam Davis' answers says that "it would be reasonable to use those tags" but "the bad, unfortunately, outweighs the good". I've showed you that this is not the case for xpathengines with concrete examples. About my question of the reason of the person involved: if you are not such person (as you have denied this consistently) but you know through confidence that any of the shared opinions was his reason for this merge, you could add that to your answer and I would consider this question answered. – user150068 May 9 '11 at 15:52
  • @Alejandro: done. – Shog9 May 9 '11 at 16:12
  • @Shog9: Ok, I will award the bounty to you, but where is the link to the "non-expert users could neither understand or use" policy? – user150068 May 9 '11 at 19:48
  • 1
    @Alejandro: not sure there's any specific policy statement there, but... You can consider this question a definition of policy if you want. For future reference, if you plan to start re-tagging questions on an existing subject with a brand-new tag, post a question here on Meta and try to get at least some consensus. Doubly so if you plan to remove a broad, well-known tag. See also: meta.stackexchange.com/questions/18878/… – Shog9 May 9 '11 at 20:03

Suddenly I got a bump on the head. I said to me: "Apparently I did something wrong". Then I asked those around me: "What did I do wrong?". And they answered: "Someone gave you a bump on the head. Explain yourself!"

Let's see the first five questions in the query that @Shog9 kindly made for us:

From https://stackoverflow.com/tags

A tag is a keyword or label that categorizes your question with other, similar questions. Using the right tags makes it easier for others to find and answer your question.

From https://stackoverflow.com/tags/xpath/info

enter image description here

Do note:

  • None of the questions ask for an XPath expression.
  • None of the answers contains an XPath expression (other than a "place holder")
  • All the questions have referenced either the host language or the environment where an XPath engine implementation must be aviable.
  • Most of the answers are code of the host language.
  • None of the answers reference themself with the XPath specs.
  • Most of the links point to specific XPath engines implementations' documentation


From @Charles' comment:

Is there ever a case where a question would be tagged with x.+engine where it wouldn't be tagged with xpath/xquery?

Let's see the next five question on the SEDE query:

  • Performing complicated XPath queries in Scala:

    What's the simplest API to use in scala to perform the following XPath queries on a document? [...some legal XPath expression...] (s is defined as a nickname for a particular namespace)

    Most upvoted answer:

    Well, I don't understand the s: notation, and couldn't find it on XPath spec either. However, ignoring that this would look like this: [... Scala code workaround...] I can't answer about namespaces, though. No clue how to work with them on searches, if it's even possible. The docs mention @{uri}attribute for prefixed attributes, not does not mention anything about prefixed elements.

    Do note: We went from a perfectly legal XPath expression to a Scala code workaround.

  • Glassfish v3 and XPath.compile error:

    We are migrating fairly new web application from tomcat to glassfish and everything is working well except for some xpath code that is used to parse some xml returned from a third party service. [...] Here is a snippet that throws an null pointer exception: [...one method from an XPath engine API...] If this change is made in the code: [...other method from another XPath engine API...]

    Falls in the same category than first five

  • XPath loop with Selenium and perl:

    I'm trying to extract all name values in input fields using selenium and perl. [...] This xpath works in finding all relevant matches: [...valid XPath expression...] However, how to extract the value of the @name attribute for each individual matches? I understand the principle is to construct a loop: [...code...]

    Most upvoted answer: a code workaround...

    Replace 1 with your counter. [...]

    Do note: @Dimitre's answer (a correct XPath expression) is downvoted. We might think that something XPath language can do, in fact it can't.

  • XPath must be a literal?

    I have a stored proc with a parameter which I want to have contain an XPath expression which can then be used to retrieve a value in a select, but the r.value(...) exception message says that I can only use literals.

    Falls in the same category than first five

  • Sitecore query does not support count() function?

    Sitecore query is based on XPATH, but looks like it does not support count() function. My query is this: [...something that is not standard XPath...]

    Accepted answer:

    Try this: [...other thing that is not standard XPath...]

    Do note: This is not XPath. I won't help others looking for an XPath answer to their questions.

Because the only answer wich is claiming to be the actual answer to this question is @Shog9's one:

@Shog9: [...]. About my question of the reason of the person involved: if you are not such person (as you have denied this consistently) but you know through confidence that any of the shared opinions was his reason for this merge, you could add that to your answer and I would consider this question answered. – Alejandro

@Alejandro: done. – Shog9

We should discuss it.

The rationale for the merge

The [xpath-engines] tag was merged into [xpath] because it was creating a distinction that non-expert users could neither understand nor properly use.

Saddly, there is no reference for this other than a circular definition:

@Alejandro: not sure there's any specific policy statement there, but... You can consider this question a definition of policy if you want. [...] - Shog9

But let see the most relevant links provided by @Shog9 for "further references":

Congratulations Eric Lippert for finally winning the c#-language badge.

IMHO, [c#-language] doesn't appear to signify anything that [c#] doesn't already indicate.

This is not the case for and because the further begins where the former specification ends (see my epilogue).

I certainly don't see anything common to the questions currently tagged that would appear to require a separate tag

I've already present six common properties of the questions with , all of them about the content of the questions and the prerequisites in order to answer (enviroment, host language against just an input sample and a clear determination of the parts of the XML instance that OP wants to address for and its synonym ). But, wait! ...

indeed, this one would appear to require the consideration of a specific implementation...

That said, there is a [ecma262] tag for questions on the ECMAScript standard itself or questions on its implementation that are potentially orthogonal to JavaScript use or any specific implementation of it.

So, there are tags that deals "with messy platform-specific details". And getting better:

So I could see a similar tag ([c#-specification] or [ecma334] say...) used for questions on the C# standard that aren't directly concerned with the use of language or its common implementations.

Wich is exactly the case for .

Assuming there are actually questions that would benefit from such a thing...

So, we end again with the benefits measure: as I proof there will be some damage for people trying to learn XPath if we expose them to answers dealing with non-standard complain implementations.


Is this distinction between and some invention? From http://www.w3.org/TR/xpath20/#dt-implementation-defined

Certain aspects of language processing are described in this specification as implementation-defined or implementation-dependent.

  • [Definition: Implementation-defined indicates an aspect that may differ between implementations, but must be specified by the implementor for each particular implementation.]

  • [Definition: Implementation-dependent indicates an aspect that may differ between implementations, is not specified by this or any W3C specification, and is not required to be specified by the implementor for any particular implementation.]

A language aspect described in this specification as implementation-defined or implementation dependent may be further constrained by the specifications of a host language in which XPath is embedded.

  • Determining the boundaries of a problem is a key for its resolution. The XPath specifications define the XPath language scope and also is telling us what is out of its scope, what belongs to the XPath implementations scope. As such, the standard API for XPath engines invocation is part of DOM specification in http://www.w3.org/TR/DOM-Level-3-XPath/ instead of the XPath specification.

  • it's not a meta tag: it reference the content of the question.

  • fully categorize those questions together: from the point of view of XPath specification (authority) the only proper answer is "that's implementation defined/dependent". In order to be answered the prerequisites are the XPath implementation reference, the host language and somes time the environment when the former is unknown. Those are not the prerequisites for an question.

  • Does it worth it? For sure it would hurt people trying to learn XPath. And we are all here to learn.

  • 3
    Actually our answer was: "You got bumped on the head because you went through a passage we thought no one used, and there are low beams in that path. If you give us a good reason why people should be able to use that path safely without getting bumped, we'll make it safe, otherwise we suggest you use a different path" – Adam Davis May 4 '11 at 23:58
  • 2
    I don't see a compelling reason the [xpath] tag is only for questions asking for an XPath expression. A question about an XPath engine is still a question about XPath isn't it? – Rebecca Chernoff May 5 '11 at 0:02
  • 2
    @Alejandro: As far as examples go, those questions don't look very related (apart from having a connection to XML/XPath). And the rather meta relation of you adding the xpathengines tag to them. I guess I'm saying, if you tagged five questions and 200+ other users took and ran with it, great; if you tagged 230 questions, um, that doesn't help your cause... And if it's somewhere in between, [more data needed]. At this point, I strongly suspect both you and Dimitre are missing the point made on another question, that a tag understood only by a handful of experts in another tag isn't usable – Shog9 May 5 '11 at 0:20
  • 2
    @Alejandro: regarding your original question: did you or Dimitre engage in any discussion before adding that tag to scores of questions? If so, pointing to it could help this discussion considerably. – Shog9 May 5 '11 at 0:24
  • 2
    I read through the chat transcript of the room from the beta chat - I didn't see a discussion on [xpathengines] and [xqueryengines]. – Rebecca Chernoff May 5 '11 at 1:46
  • 2
    Okay, let me find another way to ask the same question. Is there ever a case where a question would be tagged with x.+engine where it wouldn't be tagged with xpath/xquery? If so, why not, and can you give clear examples of questions that would fall under both, either, or just one of the set? They don't have to be links to existing questions, make them up to demonstrate the categorization differences. I think that would help clarify the discussion greatly. – Charles May 5 '11 at 2:04
  • 2
    @DimitreNovatchev, I'm not saying that xpath expressions and xpath engines are the same. I'm saying they both deal with xpath in some way. Dogs and cats are both animals. Speaking of not wanting to waste your time, an answer to Charles is still noticeably absent. – Rebecca Chernoff May 5 '11 at 5:00
  • 2
    Interestingly, @Dimitre, your stance at the time was admirable: you wrote, "Let's avoid all the etimological, philosophical and historical aspects and just be practical." I'm saddened to observe that you've abandoned practicality in this instance. – Shog9 May 5 '11 at 7:17
  • 3
    Regarding your next five examples: first is about using xpath in scala. Next is about an error encountered using glassfish. Next is about using xpath with selenium. Next is about tsql, and final is about Sitecore's xpath-like query language. NONE OF THESE ARE RELATED, except for a (in some instances very tangential) connection to xpath. They aren't about xpath engines. They're about software that is or contains an xpath engine, or something like one. That last one wasn't about xpath and wasn't originally tagged as such. And adding [not-xpath] because your answer got down-voted? Ugh. – Shog9 May 5 '11 at 16:10
  • 3
    @Alejandro: why tag them in the first place? They are at least tangentially related to xpath, which your tag appeared to recognize: your mistake was in trying to hold them at arm's length. Seeing that [not-xpath] tag in there was a revelation to me: this isn't really about categorizing the tags themselves at all, now is it - it's you and Dimitre trying to establish some arbitrarily-defined xpath-purgatory for questions that aren't "pure" enough. That's missing the entire point of the tag system. Either a question has some connection to xpath (tag it) or it doesn't (don't). – Shog9 May 5 '11 at 18:54
  • 4
    The tag adds zero value to question askers searching for help, because it has no concrete meaning! You've just been using it as a place to stash questions where xpath isn't the sole focus of the question. @Shog9 is right on the money here. The engines tags are misuses of the SO categorization system, and should not exist. Questions can belong in a tag if they are even tangentially related to the tag itself. It's perfectly fine for any of those example questions to be tagged xpath (except the last one, probably), because the questions are related to xpath. – Charles May 5 '11 at 19:09
  • 2
    Alejandro: The merge already happened. At this point, I'm more interested in questions like your last one, which aren't really xpath questions at all - those need to just have the tag removed. I'll be going through more of these as I have time, but I'd like to encourage you to do likewise, keeping in mind @Charles's point that users looking for questions on using xpath with scala are probably going to search for "xpath scala" and not "scala something else that's not xpath" – Shog9 May 5 '11 at 19:15
  • 8
    @Dimitre I think you'd see more people willing to work with you on this if you'd stop insulting them at every turn. – Adam Lear May 6 '11 at 19:33
  • 6
    @Dimitre I'm not just talking about @Shog9. In general, you've been pretty abrasive to anyone who didn't see things your way. I don't think anybody reading this question cares about your past interactions with other users, nor should we. We're just here to try and help you resolve this particular problem without getting insulted in the process. – Adam Lear May 6 '11 at 21:03
  • 6
    @Dimitre: again, if you've any legitimate complaint against me, you do yourself and the site a disservice by failing to report it. And if you don't... Well, you haven't exactly helped poor Alejandro by using his question as a platform for your nonsense. – Shog9 May 6 '11 at 21:55

The truth is that many questions are asked about particular implementations of XPath, and these questions have almost nothing significantly related to XPath. Merging them together with the true XPath questions makes finding/answering the true questions difficult and decreases the overal value of SO for any user that has submitted an XPath question. I would even start tagging some questions as "screenscraping".

The situation with "xquery" and "xqueryengines" is the same.

I happen to be the only SO user so far with gold badge in XPath (and also have gold badges in XSLT and XML). I also happen to be the leader of the XQuery tag. My competency in these areas as contrasted to that of the anonymous moderator should be well established and beyond doubt.

I am officially asking SO to revert this unfortunate and ignorant action and to seriously re-consider giving the right to moderaters from now on to merge tags. The practice of doing this based on agreement with the tag leaders is the only reasonable solution that will prevent such problems form occuring repeatedly.

At least these two things have to be done:

  1. Reverse the stupid act by an anonymous moderator that destroyed useful tags without ever asking the tag users.
  2. The SO administration should only allow tag merging after investigating the opinion of the respective tag users (and especially the tag leaders) and if the feedback received is significantly positive. Moderators should not generally be given the right to merge tags without following this process.

Yes @Alejandro. The vandalism continues.

People, who may not have even a vague notion of XPath and XQuery take decisions to merge tags to "xpath and "xquery", without ever asking the users of these tags.

To me this is similar to burning books in the past dark ages. Yes, burning, because this act is probably not fully reversible.

For your information, there are users who hold the gold badge in XPath, XSLT and XML. Michael Kay. who is the editor of the W3C Working Group on XSLT, developer of the Saxon XSLT and XQuery processors and author of the best books on XSLT and XPath is a SO user. Why don't you at least ask (or better leave to the) competent users in taking such crucial decisions? Because you are so wise? Let your acts speak for themselves.

locked by Shog9 May 4 '11 at 23:39
  • 2
    Can you explain to us plebes exactly what the functional difference is between the xpath tag and the xpathengines tag? And then the difference between the xquery tag and the xqueryengines tag? The second tags in those pairs simply seem like unnecessary specificity -- shouldn't the engine be entirely inconsequential, given that they're all (likely to be properly) implementing a specification? Someone recreated one of the two engine tags today, and I found myself editing it out because the question was about an actual query and didn't mention the engine being used in a meaningful way. – Charles May 4 '11 at 5:20
  • @Charles: Yes, if your practice with XPath and XQuery isn't significant, then you may think so. The same as the people who were allowed to do things they don't understand. Please, read my answer again -- I put there some more information -- for people that don't understand. – Dimitre Novatchev May 4 '11 at 12:38
  • @Dimitre: I appreciate your opinion with which I agree. – user150068 May 4 '11 at 13:36
  • @Alejandro: Yes, I am sure that you and any person that understands the subjectmatter of XPath and XQuery would agree. The problem here is that ignorance and power combined always lead to destruction. Nothing new on Earth. – Dimitre Novatchev May 4 '11 at 13:46
  • 3
    @Dimitre You can also edit your posts, so deleting and reposting to add new information isn't needed. – Adam Lear May 4 '11 at 20:01
  • 4
    As @Anna notes, you can and should edit your answer to add additional information. Re-posting to remove comments or votes you don't like is abusive. Stop it. I've restored your original answer, added your edit, and deleted your re-post: please stick to making a good argument and leave the silly games to users who haven't been using this system long enough to garner gold tag badges... – Shog9 May 4 '11 at 20:13
  • 4
    @Dimitre, could you explain how the catch-all x.+engines tags are a superior solution than marking questions with the appropriate xpath/xquery tag and a tag for the specific engine implementation(s) being asked about? Or, put another way, is there enough question volume about engine implementations in general, excluding questions either comparing 2+ engines or about the behavior of a specific engine? I easily see the use case for engine-specific tags, but I'm just having a really hard time understanding the need for another catch-all tag. I believe specificity wins here. – Charles May 4 '11 at 20:19
  • @Charles: There were 230 questions tagged in such way. I don't think is good to ask @Dimitre for justification now after what it seems to be a policy change. – user150068 May 4 '11 at 21:01
  • 2
    Agree completely with @Charles' latest comment, +1 on that. -1 on this answer for the tone (and also it not being an actual answer). The mods aren't out to get you and your precious tag, they were trying to clean up what looked like an unnecessary tag. I recommend you outline what you believe is appropriate material for those two tags that would not be covered better by a different set of tags. – Daniel DiPaolo May 4 '11 at 21:19
  • @Shog9: In the past you were also involved in similar stupid merge, which was later restored (partially) by the SO administration. Now I know who did this. No doubt... – Dimitre Novatchev May 4 '11 at 21:34
  • 1
    @Dimitre: If you're talking about the XSLT-XSL thing, I wasn't involved, apart from answering your question about the UI. And you argued for that merge... Trust and believe, I had nothing to do with this one, apart from stepping in twice to clean up the mess you've made of your answers here. I'm locking this answer - if you want to post yet another one, I suggest you answer the questions being asked here rather than abusing your privileges to silence other users. – Shog9 May 4 '11 at 23:39