Beefy Boxes and Bandwidth Generously Provided by pair Networks
The stupid question is the question not asked

Considering nodes for re-titling

by VSarkiss (Monsignor)
on Jan 05, 2005 at 18:25 UTC ( #419689=monkdiscuss: print w/replies, xml ) Need Help??

There are already many fine nodes On Responsible Considerations, but I just want to add a clarification, specific to the topic of considerations for title changes.

Editing and consideration is a touchy topic, and I want to clarify my opinions first. Though I didn't reply to Don't Retitle This Node, I'm firmly in the camp that says the title is part of the author's submission, and changing it should not be undertaken lightly, but with Care and Discretion.

Generally, the only time I think a node should be put on NTC for retitle is one with a single-word title in SOPW, or one that's a close duplicate of a well-known node (like fullpage chat). The reason has to do with how the Search box in the top bar works. To see this for yourself, try typing each of the following in the search box, and note the results:

  • VSarkiss
  • question
  • while
In other words, if there's a single node title matching the word, it's pulled up immediately. If more than one node has the same matching title, you get the "Multiple nodes were drunk" page. And if there's not a complete match, you get a list of matching titles. (Let's leave out the discussion of what happens when you search for a number.)

This behavior is the reason why I think certain single-word titles are poor choices; there's nothing wrong with one-word titles per se. Most nicks (and hence homenodes) are a single word, as are many obfuscations. But it's unlikely that a newcomer, who most likely isn't clear on exactly how the search bar works, will type !1 in there, or search for Saturn. It's more likely that they'll search for dbi or loop, and it's better if they end up at a list of titles rather than a single node. (Old timers know to search for Saturn asdf if they want a list.)

That's the motivation. Now to repeat some guidelines:

  • If you put up a node on NTC for retitle, always add a suggestion for what it change it to.
  • If at all possible, try to preserve most or all of the original title. This makes it easier for the OP to find the node, and is also less obtrusive IMO. (In other words, if the node was titled "DBI", you may want to change it to "DBI: how to use fetchall" or some such.)

The janitors and newbies thank you.

Replies are listed 'Best First'.
Re: Considering nodes for re-titling
by davido (Archbishop) on Jan 05, 2005 at 19:55 UTC

    Well said, VSarkiss. Let me enumerate some of the criteria that I look for before acting upon a retitle consideration. I generally look for one or more of the following:

    • Is the node title a single word? These can be a problem for the search (find) button.
    • Does the node title fail to describe the node's content? Titles like "Help please" may satisfy the 'more than one word' criteria, but fall desperately short of the 'be descriptive' criteria. If a node title completely fails to make the point I'll think about retitling.
    • Is the title abusive? (This is pretty rare, enough said).
    • Is the title really misleading? In other words, if the title is "DBI question" and the topic of the node is actually about regular expressions, it might warrant a change. Or it might warrant a little embellishment to something like "Using regular expressions to filter results of DBI queries."
    • Is the consideration accurate, well thought out, and complete? I can't stress this enough. If someone considers a node for "Retitle", or "One word title", or "Not a DBI question", etc., I'm going to vote keep, because the consideration is not giving me something to vote 'edit' on. If someone feels strongly enough about retitling to consider a node, that person should give it enough thought that the consideration suggests a reasonable course of action. It's not that I disagree that something needs to be done, but you haven't given me an action to agree with. I don't sign blank checks.
    • Finally, as a Janitor, when I look at the vote to determine if I'm going to act on it or not, when it comes to retitling I look for about a 2:1 vote in favor of retitling... plus I'm looking for it to meet some of the criteria listed above. If people don't feel strongly enough to vote predominantly in unison (with a few dissenters), I don't feel a mandate to take action.

    I think it might also be time to remind people of two important nodes:

    Above all, please think before considering, and remember to suggest a course of action. If there's a surefire way to get an uncompelling vote on a consideration, it's to say something like "Maybe retitle, or edit or something". Each consideration should address one problem, and should suggest one definate course of action.

    And let's err on the side of inaction. Unless there really is a need, we should just leave things alone. The Janitors Thread Retitler should be looked upon like antibiotics; they work, but overuse can be bad for everyone.


Re: Considering nodes for re-titling
by borisz (Canon) on Jan 06, 2005 at 01:14 UTC
    Since my english is somewhat bad, from time to time someone retitle my nodes. I noticed that the new title is in better english, but did not reflect my question anymore.

    I suggest, that a retitled node must have the original title at the top of the question.
      Agreed. This should satisfy the search mechanism's ability to find your original node.

      Give me strength for today.. I will not talk it away..
      Just for a moment.. It will burn through the clouds.. and shine down on me.

Re: Considering nodes for re-titling
by calin (Deacon) on Jan 05, 2005 at 20:25 UTC

    Agreed with the intent.

    I'd like to see moderators and editors making an effort to shake off this "kindergarten principal" kind of attitude. Most of the people who post here are grown-ups, and nobody wants to find, at the end of a day of hard work, that some clever boys and girs are messing with their posts.

    I'm no saint, I've made some mistakes with excessive moderation in the past. Most of these were driven by the ire provoked by what I perceived as abuse -- but sometimes, particularly at the end of some stressful days, the distinction between real abuse and the feedback of your own frustrations tends to become blurry. I think this is the moment you need to pay attention to -- take a deep breath and reconsider.

    One single criterion that I think it's always worth following is the criterion of good faith. If you can find good faith between the lines, have the strength to leave the post alone.

    If you don't find good faith, it should be further considered case by case (with the option of course of no action / involvement). My opinion these days is that off-topic etc. posts should be left standing unless they're quantitatively abusive -- many abusive posts in a row by the same author / posts of excessive length lacking justification ("code dump") / an emergence of widespread bad behaviour on the site which demands strong action to restore order and sanity etc.

    Another development that I don't like is the "Move to tutorials" consideration trend (on Meditation posts). You shoud assume that if an OP chose to post in the Meditation section he intended his post as a Meditation. You might discuss this with the author, but don't consider -- forceful consideration is the dysfunctional mode of operation of this site, it's the last resort in the case of broken communication.

    Now let's go back to the topic of this thread -- retitling. I'm thinking of an algorithm but I don't have the time now to elaborate -- I need to catch the last train. Some short ideas before I go: Talk with the autor first. Respect the tradition of authorship (written word should not be altered). Take forceful action only in case of real abuse.

Re: Considering nodes for re-titling
by Errto (Vicar) on Jan 06, 2005 at 00:41 UTC

    As a relatively recent arrival to Level 6: Friar I've noticed one particularly unhealthy pattern: requests to retitle SoPW nodes because the OP had some problem code but misunderstood where the problem is. In other words, the suggested title reflects the answers to the question, rather than the question itself as the OP understood it.

    To me, the importance of this goes beyond respecting authorship. The reason for preserving such titles is that a future seeker with a similar problem is probably more likely to frame their query in the way the OP did than in what the responses reflected. Please take care not to change the meaning of titles in this way.

Re: Considering nodes for re-titling
by ysth (Canon) on Jan 05, 2005 at 20:17 UTC
    Reminder to janitors: /msg the author of the considered node when you retitle.

    I do see a distinction between titles and body: misspelled words in titles that are likely to be valuable search terms should IMO be corrected.

      Yes, titles are different. There are a lot more ways that they can fall into the category of "hazard to navigation", which I still find to be a very good phrase for defining what (and only what) janitors should mess with.

      Note that I'm not disagreeing with much of the rest of this thread. For example, if a new title is suggested that is clearer than the current title but has no other advantages, then I would leave the title unchanged. Level of clarity is not much of a navigation issue and is also a rather subjective criterion and so should mostly be avoided (shoot for clarity when changing if there are other, more compelling reasons to make a change, of course).

      In contrast, a spelling error is often a hazard to navigation (as you noted, it makes the node harder to find in a careful search) and usually isn't subjective. So most spelling errors in titles should probably be corrected. However...

      But let's also avoid wrangling over contraversial spelling choices. If the spelling is either common or correct, then leave it be. If what to change it to isn't rather obvious (there being no clear alternative that is both common and correct), then leave it be. For example, just about any pluralization of "regex" should probably just be left alone.

      And if the word is spelled correctly out of context (that is, the wrong but related word was chosen), then we are back to clarity so leave it be. So seeing "data" when you think1 "datum" is what 'should' have been used is not a reason to consider a node for title change.

      Finally, a brief word on creative control: Let people have it. A word may be mispelt for humorous or other creative effect. And just because you don't get the joke doesn't mean that everyone fails to get it. So before considering a node, try to consider whether there might be some meaning (perhaps a playful one) that is escaping you. If you aren't sure, perhaps you should ask (such as in the CB) or just spend that time doing something else.

      That last point is even more important for janitors. If there is subtle meaning, then many will likely miss that and some such nodes will get considered (and get a lot of 'edit' votes just because that is the nature of beast). It is your duty to filter requests against principles and so you might need to unconsider a node instead of editting it.

      - tye        

      1 I use the word "think" because I believe it is more accurate even though I'm quite sure that many people who need to hear this advise will be using the word "know" instead.

      Reminder to janitors: /msg the author of the considered node when you retitle.
      Unless the author is yours truly!
Re: Considering nodes for re-titling
by demerphq (Chancellor) on Jan 06, 2005 at 12:30 UTC

    It occurs to me that the preview page could have a set of links "The following nodes have a similar title to yours, perhaps your should have a look at them first: ...." Only on root nodes, and user configurable of course...


Re: Considering nodes for re-titling
by Anonymous Monk on Feb 23, 2005 at 11:43 UTC
    This behavior is the reason why I think certain single-word titles are poor choices; there's nothing wrong with one-word titles per se.

    I agree with the first and the last part of the above statement - I disagree with the conclusion in the middle. It's the behavior that causes the problem - the one-word title is merely the trigger. So, fix the behavior. Searching with a single keyword is common. So, if it's a problem that a search with a single keyword jumps directly to a page instead of showing a list of alternatives - don't jump. Show the list!

    This reminds me about the story of the man who went to the doctor saying his eye hurts when drinking tea. And instead of suggesting to take out the spoon, the doctor tells the patient to stop taking sugar in his tea.

      To my mind, the "Search" box at the top of the page is poorly named. It isn't very useful for searching and is very useful for jumping to a specific node for which you know the title or the node ID or even just much of the title. I've long felt that the "Search" button should be renamed "Find". If you want to search, even with a single keyword, you are better off using Super Search or the google alternative it links to.

      Anyway, this is why the "fix" you propose hasn't happened. It would solve a problem with the misuse of the mis-named Search feature while simultaneously breaking one of the most important features of the site.

      Update: But doing just that isn't a very appealing "solution" either. How about s/Search/Goto/ and add another top-line input box for "Search" that actually uses Super Search to search node contents easily. Perhaps also move the links to Super Search into the top bar instead of with all the links just under it?

      - tye        

Log In?

What's my password?
Create A New User
Node Status?
node history
Node Type: monkdiscuss [id://419689]
Approved by bmann
Front-paged by gmax
and all is quiet...

How do I use this? | Other CB clients
Other Users?
Others making s'mores by the fire in the courtyard of the Monastery: (7)
As of 2017-01-17 15:38 GMT
Find Nodes?
    Voting Booth?
    Do you watch meteor showers?

    Results (157 votes). Check out past polls.