[CLOSED!] [Survey] Anyone really needing Auto-pairing of “ and ”?


Who should care

If you never type “ and ” , you don’t need to read on this topic.

If in your natural language, such as Chinese Simplified, Arabic, Korean, the standard symbols of quotation marks are “ and ” , you may probably face the same problem that I had.

What happens?

Note: This topic is a follow-up to my last one.

Here is a digest.

I’m a Chinese user.

When I input a Chinese opening quotation mark (“), Atom immediately adds a Chinese closing quotation mark (”) after it.

The problem is that the IME (Input Method Editor) I’m using will also give me a closing mark the next time I hit Shift+'.

So, I always get two closing marks and thus have to delete either of them. That is annoying.

Why it happens?

In Chinese, “ and ” are different marks, but they share the same key ". Similarly, ‘ and ’ share key '.

So, any mordern Chinese IME must have such a feature: It gives an opening mark first; from then on, it remembers the last mark it gave and gives the opposite one at next time.

In other words, odd for opening and even for closing.

It works perfect everywhere except in Atom, which adds redundant closing marks.

Thanks to some replies in my last topic, I’ve learnt that there is a core package of Atom, bracket-matcher, who is responsible.

Here’s what it says in the Readme:

You can toggle whether English/French style quotation marks ( “ ”, ‘ ’, « » and ‹ › ) are autocompleted via the Autocomplete Smart Quotes setting in the settings view.

After I disabled the option “Autocomplete Smart Quotes”, I am happy.

What not happens?

Don’t mistake.

I love auto-pairing of ascii closing marks ( " and ’ ) whose opening and closing symbols are the same.

IME doesn’t change that. Thus, no redundant and no annoyance.

The first and basic optioln of that package is about this feature. I just left it enabled as default.

What to survey?

Problem solved. Right?

However, since the option is enalbed by default, all users who are like me will be annoyed.

Some may help themselves; some may ask for help, just as I did, repetitively; the rest may simply give up Atom.

I am wondering: Is there anybody in the world really needing this “Autocomplete Smart Quotes” feature?

Let’s put aside «» and ‹›. Just talk about “” and ‘’.

I guest every IME should already handles “” and ‘’ well enough, so there should be no need to do an extra “smart” job in Atom.

There’s a Chinese saying: (Don’t) paint (more precisely, draw) a snake with legs. It means that a superfluous feature will spoil the job by overdoing it.

I high suspect that “Autocomplete Smart Quotes” of “” and ‘’ falls into that category.

I can bet on Chinese users. But how about others?

Would you kindly tell me about your case?

I’d like to know:

  1. What natural language is yours?
  2. Do you prefer that Autocomplete of standard quote marks be disabled by default?
  3. If not, how do you input standard quotation marks ( “” and ‘’ ) with your IME?

For what?

If everyone finds this feature superfluous, Atom team should disable it by default or even remove it.

(Reminder: just about “ ” and ‘ ’ , not ascii " and ’ or other brackets)


Well, I’ll bite.

  1. English. But maybe the better question is what programming languages do you use or what keyboard type do you use.
  2. No, I like it how it is. When I type a string, I need two quotation marks. There is never a time when I need one.
  3. I don’t really understand this question. I use the quotation mark key on the keyboard ("").

I can see that it’s annoying to you, but you can disable this feature. I don’t like every default setting, but Atom is set up so you can configure to your heart’s content. Or change the IME settings so it doesn’t add a closing quotation mark. If you really want to make this feature disabled by default, make a pull request to change it and see why the Atom team has it set up this way.


Just exactly as what I was afraid of!

Thanks for you reply, but you didn’t seemed to have read carefully and thus just missed the point.

I’m NOT talking about ASCII quote marks. And I said I LOVE auto-pairing of them. And it is “Autocomplete Brackets” in the settings that is responsible for that. I never need to change it.

I’m talking about the option “Autocomplete Smart Quotes”. I wonder whether it is really needed for anyone.

I had a problem when typing Chinese (simplified) quote marks ( “ ” and ‘ ’ ) which, as I learnt from Wikipedia, are also the standard quote mark symbols of many other natural languages.

If you, like me, need to type THOSE quote marks, you have ONE key to press but TWO symbols (the opening mark and the closing one are different) to output.

You must be using an IME which can help you. It already helps you nicely everywhere.

But in (and only in) Atom, the editor so-called smartly adds a closing mark, TOO! Then you will find youself deleting this redundant character.

If you NEVER have this trouble and feel difficult to imagine the scene I was try to describle, I guess you are simply NOT the target of my survey.


I have to challenge some of the assumptions I feel are made in the construction of this survey:

  • A non-offensive set of defaults can be determined for all possible users
  • Changing the default behavior is a non-breaking change
  • You will get responses from people who are happy with the status quo
  • You can achieve a scientifically representative sample from a message board survey

It is literally impossible to find a set of defaults that all possible users will agree on. In the more than one year I have been involved with the Atom project, every time the Atom team has changed a default behavior, someone has gotten upset about it, questioned the motivation and advocated for the original default’s return. There is no set of defaults that everyone can agree on. See for example this discussion. This is a group of users that believe that Atom should not default to restoring the last session. This is after literal months of hearing that restoring whatever Atom had open last session was necessary feature that was preventing people from using Atom.

Everyone believes that their way is the way it should be by default. What they don’t realize is that when one advocates for a change to the default behavior, they are advocating for breaking the experience for other users. There is a topic on this board where changing the default theme was upsetting for someone. Adding options is not a breaking change, but changing the defaults is. At this point, any change to the defaults will upset someone. And they are right to be upset, something that once worked for them now doesn’t. In my opinion, this makes the burden not “Do some people find this feature superfluous?” but “The current default is so egregious that literally breaking Atom for some users is preferable to keeping things the way they are.”

In the two bugs I linked to above, one advocating for restoring the last session and the other for not restoring the last session, did you notice that the people advocating for not restoring the last session did not appear in the discussion that led to the change? Why is that? Because they were happy with the way things were. It is very, very rare for someone to get involved in a project unless it does something they don’t want it to or doesn’t do something they want it to. People who are happy with the status quo won’t even come to the board to see this survey. Which leads me to …

Message board surveys are by their nature not scientific for the reason stated above as well as the simple fact that some people don’t like message boards. Some people who use Atom only use Twitter or Stack Overflow or Reddit or Slack or GitHub Issues or don’t communicate with other Atom users at all. And the stated goal of the survey is:

(Emphasis mine.) It is going to be a lot of work to get a hold of every Atom user to get their opinion on this.

I understand that you were annoyed by the default and that others will also be annoyed by it until they find the method for disabling this feature just as you did. I feel there are other solutions that don’t involve breaking the experience for other users. Perhaps we could make the way for turning this feature off more obvious? Perhaps we could detect that someone often deletes the auto-added closing quote and offer to turn the feature off? Perhaps we could detect somehow that the user is Chinese or uses an IME that helps like yours does and turn the feature off in that case?

I feel that there are more constructive ways to solve this problem than changing the default.


Hi, leedoh.

I am glad to read a very serious reply from you.

However, sadly, again I am misunderstood.

Is this just yet another meaningless trivial problem?

You see, I’ve already learnt where to change the responsible option.

So I did. “Problem solved” and I could just stop there.

Then, why bother to continue discussing? Especially writing a long article in English which is not my native language?

Because I think this is a rare and typical issue worth talking about.

I must point out that this issue is very likely meaningful only to a minority.

Many may never use an IME to input “” and ‘’. To them this topic is completely irrelevant. They don’t care at all.

And they shouldn’t.

Changing the default value of the option of that feature, if happens, will never break anything for them.

But as to the few who use an IME, it probably matters.

And why I even think this can be discussed?

I can see you find discussions can easily deteriorate into messes and end (or do not end) with two (if not more) sides, either sticking to their own preference and failing to convince the other.

Believe me, my job is designing software products and I am too familiar with such fruitless disputes.

I hate them as much as you do, I assume.

So, did I start one myself, maybe out of emotional impulse?

Of course not.

We have to admit, there are other discussions which can, at high possibilities, reach convincing conclusions.

Otherwise, imagine how the world could go without anything worth discussing at all.

Why can there be a conclusion?

I am perfectionist and believe in logic. I never talk much about things in which I find little certainty.

In this topic, I have good reasoning.

Note: Afraid of being misunderstood, I have to repeat my statements as follows. I am talking only about the standard quote marks ( “ ” and ‘ ’ ) commonly used in Chinese, Korean, Ababic and some other languages. As to «» and ‹› or other brackets, as well as ACSII marks, they are off topic, and auto-pairing them seem to be very good.

As I said, there is only one key in the keyboard for two different symbols, one opening mark and one closing mark, of standard quotation marks in Chinese simplified (as well as a few other languages).

This is a fact, long existing before Atom was created. And the very source of our issue here.

Just a simple question: How could people, who need to type those marks, survive so far?

One way might be applications like Microsoft Office Word, which can automatically replace odd "s with “s and even "s with ”s.

However, obviously you cannot enjoy this feature in other applications. When you need it, you have to edit your articles in them and later copy the content out.

Anyway, It is irrelevant to my topic since features in other applications will never conflict one in Atom.

One other way, and the most common way nowadays, for Chinese, Japanese, Korean and more people, is to use IMEs. It is safe to say we cannot live without IMEs.

IME does the “odd and even” job (if Chinese mode is on) EVERYWHERE.

Now let me quote myself: It works perfect everywhere except in Atom, which adds redundant closing marks.

Every other application I’ve ever used doesn’t have this auto-pairing feature about Chinese quote marks, so it is OK.

Atom is an EXCEPTION. The only one I’ve met so far.

Can you see why I am able to be strong against it?

If there is a “best practice” in your industry, it is wise to adopt it in your own product, unless you are truly creative to surpass it with a better solution.

On the other side of the same coin, there are “non-practices” which no successful players in your industry are adopting. Some players might have tried but failed to keep it on, for reasons.

Now, if you come to a “good” idea and think others are just too stupid or too lazy to do it before you, in 99.9%, you are just overlooking the bad things about your “solution” and there is a lesson waiting for you to learn.

Yes, there is still a 0.1% chance that you are innovating something really awesome and pushing the boundary of your industry. Nothing is absulote.

leedoh, you think I am “advocating”.

Partly, yes.

But I am open-minded. I am just wondering if my story also applies (or more importantly, doesn’t apply) to others (claim again: the relevant people).

I can bet on all Simplified Chinese users.

But of course, the world is huge and anything is possible. There might be a guy saying “I type those Chinese marks and somehow I don’t have a problem like yours”. If that happens, I will be curious to learn, honestly.

And I am also prepared to hear from users in other lanuages like Korean and Arabic who might find the auto-pairing actually helpful. If so, I will agree that this feature is indeed necessary and should be enabled by default.

I created this topic as a “survey”, not a “proposal”, remember?

In fact, I did make a proposal, which was my last topic, “Please stop auto-pairing Chinese quote marks”.

I found it inappropriate and thus discontinued it and restarted in a different direction. So comes this topic.

Sum up

I want a survey.

I welcome disproofs against my assumption. I’m curious.

But you must be my target at first place.

If you are irrelevant, you won’t be able to see the real picture of this issue, let alone its importance and certainty within.

Yes, it is of zero meaning to you. If there is a change, you won’t be among the guys who will be affected.

To me, and probably quite a few users yet to come, it is a topic worth discussing.

It’s not about personal preferences.

I highly suspect that ALL relevant people will agree with a same conclusion. And I highly DOUBT there can be a second group who claim they really need that feature.

Tell me wrong and make my eyes wide open if you can. I will be glad to learn the beautiful diversity in the world.

Will Atom team react?

I agree that “some people don’t like message boards”.

Even if all users love them, it is still impossible to make any survey “scientific” in terms of statistics.

Sorry about my misleading words about “everyone”. As I explain above, this issue is rare for it has great certainty in its facts and logic. I didn’t mean to count numbers. I meant to collect ideas and stories.

I’d like to learn if there is a opposite user case beyond my imagination, which I think has less than a tiny chance.

Can this survey, or anything else at all, pursuade Atom team to change anything? How can I know?

It would be nice if the team give us some guidance about how to contribute. But it can hardly be definite and ultimate solutions.

The only thing I know is that when I design products, I will have to make a variety of decisions, from extremely easy to judge to impossible to predict. No matter what, I have to make a proportion of them at my humanly best, take my responsibilities when they turn to be successes or failure, and then I go on.

I just guess this specific topic, which I’ve spent so much time to discuss, seems to be at somewhere near the easy end of that spectrum. Otherwise, I would not wasted my time at the beginning.


I’ll answer this question first. The Atom team is very open to feedback and changing their position. As I stated above, they have changed defaults at least a few times already and continue to do so. They have even accepted an entirely new autocompletion engine created by the community. All one must do is open a Pull Request with the change that is desired and it will be reviewed by the Atom team when they can. Which brings me to:

They have. There is a CONTRIBUTING document with guidelines on what is required in order for anything to get merged. And, for the most part, it is pretty simple. Just include as much information as you can, document your code, follow the coding guidelines and include test specs to ensure that the feature doesn’t regress accidentally.

I have understood everything you’re saying @GrandSong and I appreciate the time you’ve taken to explain to me your point of view. I am sure that you have run into many situations where trade offs are required and decisions have to be made. You see it as a situation where Atom is ignoring the best practices of an entire industry by doing something of minimal, if any, utility. I am happy that you are open minded to acknowledge that there may be a tiny chance that there is a use case you haven’t considered or encountered.

I see this as a form of the classic interview question given to Quality Assurance candidates:

There is a bug in the application that affects 10% of the current users of the application, making the application unusable for them. A fix has been created that will repair this bug, but a different 1% of users will suffer a one-time complete data loss. Do you advocate for or against releasing the fix?

There is no right answer necessarily. There isn’t enough information in the question to come to a definitive answer for all cases. I understand that you are asking questions, as any good developer would, to determine the correct answer to this particular conundrum.

I am not advocating for no action. I am suggesting that there are many more options in the beautiful diversity of the world than the one that it appears the survey is designed to address. As my Kung Fu instructor once asked of a student, “There are three birds on a fence. You take a slingshot and knock one of them off. How many are left?” The student replied, “There are two birds left, Teacher.” The instructor smiled and said, “No, there are none left. The slingshot scared the other two away.” As I am sure you understand, sometimes one focuses too much on the problem and the true answer is obscured.

It may be also that I should, as is said here in America, “practice what I preach” and step back to let your questions take their natural course. I would only ask that you consider as well the other options I have suggested or to submit a Pull Request to change the functionality you find objectionable and let the Atom team review it.


Just two hours ago, I visited the GitHub site of Atom for my first time, and found that that is the place where I can talk to dev team.

I will do so (after reading the helpful Contributing doc you refered, thanks) when I am ready.

Now, I need to gather more information about this issue, which is what this topic is for.

I still feel you have some misunderstandings. Not big deals. But if you like, I can try to restate better.

Atom is not “ignoring the best practices”. Instead, it created (or broke?) a “non-practice” and thus makes itself an exception.

In my working experience, whenever I or my colleauges discover such an obvious overdoing, we can always decide to remove it without any hesitation. Otherwise, if this kind of issue could make us hesitate, we would be unable to handle more difficult problems at all.

As to your example about QA, I modify it to fit this particular issue as follows:

There is a bug. 98% of our users are never affected. 2% are. If we fix it, 98% wiil stay unaffected while the 2% will all be happy.

However, we cannot rule out the possibility that n% among the 2% actually don’t want the change. The value of “n” might be from 0 to 1%. 0 is very likely but we are not sure.

In fact, before the bug was reported, we had assumed that n is the entire 2%. That was why the bug had been created at first place. Unfortunately, we were wrong. Even more unfortunately, we cannot say we were completely wrong. If we were, we can simply fix the bug. Well, what shall we do now?

An interviewee may ask: “Is this a QA thing? Why not let the designer do some user study?”

Well, a common beginning of user study is survey. Not ending, just beginning.

Ironically, it is theorectially never possible to prove anything by any number of proofs, but every easy to disprove by a single disproof. That is the foundation of science, expounded in a book titled “Impossibilities” which is one of my favorites.

Read the title of this topic again, and you can see that I am actually looking forward to disproofs as well as, if not more than, proofs.

I have to admit, it sounds just a rhetorical question. Sorry. But sincerely, it is not.

If a single disproof appears, I will probably close the topic because the issue becomes more complicated than I expected. I don’t like a quagmire.

I will still be satisfied. Although my proposal goes down, I will learn something (like another means of typing thoes marks without IME) beyond my imagination in surprise.

So, either way, I will gain.

PS: Glad to hear you are/were learning Kong Fu. Are you also interested in Chinese wisdom?


If you use a lot of words, that doesn’t mean you’re paying attention. If you’re so scientific and perfectionist, you should make a serious effort of trying to disprove your hypothesis that everyone agrees with you. Lee already told you this message board will not do that for you.

If you’ve just discovered the docs and repos, I suggest spending more time there, get a PR going, and take it from there.

Just for convenience sake: I am used to how quotes work and Atom and don’t want it to change. So there you go.


@GrandSong I just want to say I completely agree with @braver and @leedohm. I am really impressed with Lee because he took the time to write a really good and empathetic reply to you. Along with @braver:

I think you should be reading the documentation and working with the repositories, you can submit a fully tested pull request just like anyone else. That is where your attention needs to go.

I am used to how quotes work, and don’t want Atom to change it either. Thank you.



Off topic …

This is called the Trolley Problem (http://en.wikipedia.org/wiki/Trolley_problem). An interesting part of this problem is that self-driving cars will face this moral dillemma.


Hi @GrandSong, I think the problem is that many users will create the smart quotes using an IME, but many will not.

I use OSX and I set my keyboard to the “International English” variant. This means I can type Option-[ to get an opening English smart quote (“) and Option-Shift-[ to get a closing English smart quote (”).

For me, Atom does exactly the right thing: I hit Option-[ and Atom inserts “” into the text, with the cursor being between them. If I then hit Option-Shift-[, then the cursor skips over the closing quotation mark.

I know that there are more Chinese speakers than English speakers, but unfortunately the Atom development team has a bias to support English by default.

(An honest question: What about all the other Chinese quotation marks like ⌈ and ⌋ and similar? I got the impression there is a pretty wide variety of them. (I used technical symbols here, couldn’t find the actual Chinese characters. I hope you understand what I mean. I don’t really know Chinese.))


Thank you very much, kgrossjo!

Finally, here comes one who really solves my question!

My reply to kgrossjo

When I created this topic, I searched Wikipedia for quotation marks and found that the standard symbols in English are “”‘’ too.

When I replied to leedoh, I recalled how Microsoft Word handles those marks: an odd=and-even auto-replacing solution, which is originally for English users.

Then I become more curious: How do English non-programmers survive so far? And that would be a disproof which I was looking forward to.

Now I see how.

You have different key combinations for the two (opening and closing) symbols. And since the combination for the closing one is not easy to press, the autocompletion in Atom does you quite a favor indeed!

I can close this survey with satisfaction.

It was kind of you to say that.

In fact, I think Chinese (and Korean, Arabic, etc) users of Atom probably are not and won’t be a big group. Moreover, Atom is still in its infancy and ever changing quickly. It is reasonable and actually wise for the team to focus on English users.

I won’t bother the team until I believe an issue 1) affects only people like me and 2) all of us need a fix.

In comparison, there is another issue which falls into this category:

The “soft wrap” feature fails to work properly when there are Chinese, Japanese, etc characters in lines of Atom, while other good editors do well without this problem.

English users never care but all the people like me do. I am certain and thus don’t need to do a survey about that.

I have confidence to report it as a bug (or at least a deficiency) to Atom team. I will after I learn how to.

As an side, as I understand it, you are actually also using an invisible IME (with “International English” variant, integrated in OSX as a module or something).

When you need to type some characters the keyboard doesn’t have direct keymaps, your IME provides you with key combinations.

Note: The major difference between yours and mine is that mine will pop up a small window for me to select Chineses character candidates of the same pinyin (Chinese pronunciation) I input while yours doesn’t need to. But the pop-up window is just a visible UI component, not IME itself.

My reply to braver and BobCochran

I won’t thank you two guys.

I can afford misunderstandings. I am willing to improve my expressions and update my views.

But you criticize me with nothing really meaningful. I have to doubt that you have ever read through my words which may be too long for your precious patience.

And both of you ended your replies with “I don’t want change” but just that. Not another single sentence explaining why or how the change can truly affect you. Do you exactly know what I was discussing to change? And how can I learn anything new from your mere statements of preference?

I show gratitude to anyone helpful. If someone tried to help, I am also thankful.

But if you don’t want to, for the sake of everyone, please just ignore the topic and walk away.


There is already an Issue open for improving the soft wrap for Chinese, Japanese and Korean:

I am interested in wisdom from all sources, Chinese or otherwise, and endeavor to study the best of all cultures. I spent some time in Taiwan, Republic of China, on business many years ago and was able to spend a lot of time in the National Palace Museum there. I could have spent much more time there :grinning:

I took Kong Fu for a short time when I was younger. The experience had a lasting impact because of the teacher.



Sorry that I forgot to search already exisiting issues before I post a new one. I will keep this in mind from now on.

You are amazing.

I can feel you have benefited a lot from your experience in different cultures.

I a bit envy you. I haven’t been abroad yet.

This topic is closed. I am looking forward to sharing or exchanging with you later. See you around.


Closing by request.

closed #16