[DISCONTINUED] MATLAB Answers Wish-list #5 (and bug reports)

This is the 5th installment of the wish-list and bug report thread.

This topic is the follow on to the first Wish-list for MATLAB Answer sections and second MATLAB Answers Wish-list #2 (and bug reports). The third started out as New design of the forum - grey on white and the fourth MATLAB Answers Wish-list #4 (and bug reports) is also growing so large it is slow to load and navigate.

Same idea as the previous ones: one wish (or bug report) per answer, so that people can vote their wishes.

What should you post where?

Wishlist threads (#1 #2 #3 #4 #5 #6): bugs and feature requests for Matlab Answers

Frustation threads (#1 #2): frustations about usage and capabilities of Matlab itself

Missing feature threads (#1 #2): features that you whish Matlab would have had

Next Gen threads (#1): features that would break compatibility with previous versions, but would be nice to have

@anyone posting a new thread when the last one gets too large (about 50 answers seems a reasonable limit per thread), please update this list in all last threads. (if you don't have editing privileges, just post a comment asking someone to do the edit)

Answer 1

[Adam Danz]

Spell check for questions / answers / comments

A spell check option would improve the quality of content, reduce misunderstandings, and reduce the need to edit our comments. The spell checker would ignore blocks of formatted code which could serve a second purpose of reminding people to format their code when their variable names are all detected as misspellings.

Browser extensions such as grammarly do not work in this forum even though they work in other forums.

[David]

Thanks Adam, we are aware of this limitation, which is due to the way the wysiwyg editor works. While providing more functionality that the previous editor since it renders complex HTML behind the scenes it cannot take advantage of the default spellchecker(s) within most browsers. That being said, there are ways to get a spellchecker that we're actively looking into.

[Adam Danz]

Thanks for the heads up, David.

Answer 2

[Image Analyst]

It would be really nice if we could double click a function name in our Answers response, click some toolbar icon, and have it convert that word in our response to the URL for the online MATLAB documentation for that function.

[Rik]

Any quick way to link to the doc would be really nice. IMHO the documentation and Answers are the biggest advantages Matlab has over its competitors, and currently they are poorly linked.

[Kent Millard]

This feature has been released.

Type @ or @doc:[function name] to search and add a Documentation link to your post.

Let us know what you think.

[Image Analyst]

Kent, it's an improvement, and works fine if I remember in advance to do it. However I'd still like a tool bar icon to doc it after I type the function name. Here is a common workflow for me

1. Type sentence, including some function name, for example bwlabel().
2. Remember that I can make the function name a link but I forgot to add @doc: in front of the name.
3. Return cursor to the beginning of the function and type @doc:
4. That does not bring up the popup list of functions, so I have to delete the function name I typed, and start all over again using @doc: in front and re-type the function name.

It would be easier if I could just double-click the work and click the toolbar button to turn it into the link.

[Kent Millard]

Thanks for describing your workflow @Image Analyst--this is very helpful. As you pointed out, step 3 does not trigger the dropdown--I agree with you that it should. I'll talk with the developer about this. Thanks again.

[Star Strider]

I just type ‘@doc:’ then type enough of the function name to bring up the appropriate function in the drop-down menu (that magickally appears at that point), and click on the one I want. I generally use bold+underline+monospace for function links so I have to go back and do that, however once I highlight the function name that’s straightforward.

[dpb]

Since smart-sense can put in the links, I wonder why it can't do something to automagically highlight them some how as well -- at least with settings here (that I've made no known changes to any defaults I'm aware of, anyway), they're essentially indistinguishable from ordinary text. I think there may be a color change, but it's not enough I can really even tell that for sure. I (sometimes) will use <> to surround, but it's more time/effort than typically take unless pasting a remote link with the CTRL_K route in which case have to edit the display text anyways, generally.

However, the doc link, while useful, doesn't have links to base functionality -- other than functions, it doesn't know about keywords where those could be the most helpful/pertinent to the poster/question raised. I tried to link to <for> the other day -- it didn't know anything about it.

Answer 3

[Sindar]

Add a set of pre-written comments for common situations. Things like:

• "this is clearly a homework problem, please display some effort first"
• "god your formatting is terrible, here's how to do better..."
• "please give a minimal example, not 200 lines of code"
• "Saying 'it errored' is entirely useless, be more specific"
• "please go through the MATLAB onramp"

(The actual comment set should include complete instructions, and be slightly less hostile)

Having the ability to insert these comments would save us time, lead to friendlier interactions (we all get impatient), and give better info to questioners

[Rik]

I used to have a file from which I could copy-paste responses, but now with the new editor that doesn't really work well anymore. I now have alt+1 to alt+7 trigger an autohotkey script that types those responses. That way I still have the benefits of an automatic reply, as well as proper formatting. Guess what is under alt+1 (extra line breaks for readability):

Have a read [here](https://www.mathworks.com/matlabcentral/answers/6200-tutorial-how-to-ask-a-question-on-answers-and-get-a-fast-answer) 
and [here](https://www.mathworks.com/matlabcentral/answers/help/rtc#rtc_summary). 
It will greatly improve your chances of getting an answer.

[Guillaume]

This is actually a very good idea, Sindar. I'm not sure it could be implemented with the current interface, but certainly, it's something I'd like to see.

I always have to google the homework tutorial and the how to ask tutorial when I want to link it. Unfortunately, I'm too lazy/old to set up a system like Rik's.

[Rik]

@Guillaume: it's surprising how far you can get with a txt file in dropbox that you keep a bookmark for ;)

[Adam Danz]

I'd like to see these builtin templates as well.

[Walter Roberson]

I have a series of bookmarks in my browser menu bar that I copy out of the menu bar and paste in .

[Rena Berman]

This is actually already in consideration. Thanks for all of your feedback!!

Answer 4

[Adam]

Has there been any further thought on changing what is considered a compulsory field for users asking a question? Title, body and tags appear to be still the only compulsory ones and some users are too lazy to put something intelligent in the tag field which allows you to type literally anything anyway.

Product is not compulsory and, as discussed previously, Release is locked until you select a product. This was the reason given last time I asked about this:

"With respect to the release drop-down being filtered, we do think this is important as it provides users with a smaller, pre-filtered list that is easier to use."

That was over a year ago and it still makes no more sense now than it did then given the reality. As another quick small sample, from the first page of questions just before I typed this, just 8 questions (excluding a Mathworks Tech Support team question) contained product information other than 'Matlab' (for which no releases would be filtered out anyway) and of those, just 5 contain the release. So not only is release locked until you enter the (optional) product, but it isn't compulsory even when you do enter the product. We're simply talking about a list of release numbers that a user has to select from. Having it almost useless just to make it 'easier to use' by filtering out a few releases is creating a 'solution' to a problem that didn't exist.

The release field seems fundamental to me and trivial to include properly. It clearly isn't working as it currently is because very few people fill in the release field. Surely the purpose of Matlab answers is to allow people to ask questions and get answers to them as smoothly as possible. So making it more difficult for the people answering the questions by not having release information just seems totally counter-intuitive. Obviously having the product compulsory for all questions would be a big help too, though again some users just put totally inappropriate products (Matlab Coder and Matlab compiler seem very popular!).

[Rik]

I disagreed with the reasoning then and I disagree with the reasoning now. Especially with the larger language changes it is important to know the release. Otherwise users with older releases are getting errors or users of newer releases don't take advantage of new syntax options.

• HG2 in R2014b
• implicit expansion in R2016b
• dot indexing of temporary variables in R2019b (the syntax myFunction(data).fieldA is now valid)

[Walter Roberson]

It is a nuisance that if I add another product to the list that the release information is discarded.

[Guillaume]

You get the same nuisance when you raise a bug through their web interface and it's just as annoying there.

I've suggested before to the developers to allow users to enter just a version number and if no product is select to default to matlab. I do think that version should be a compulsory field (with maybe a Not Applicable option for where it doesn't matter like this answer).

For what it's worth we (the CAB) have had several discussions with Mathworks on how to improve answers.

[Sindar]

on the tag note: if requiring, can there at least be a character limit warning "this is a bit long for a tag, are you asking a question in a tag?"

[Rik]

And the release ambiguity struck again: I should have noticed that the screenshots in this question showed an unusually dark background, looking an awful lot like HG1. If the release was a compulsory field, a lot of back and forth (leading to the discovery that the release in question was R2013a) would have been avoided.

Answer 5

[Rik]

Another nice-to-have feature: being able to move threads between the answer section and comments without having to repost them. One example of when it would be useful to only change the post type: this. In that thread there are 6 answers, each with some comments. I am not going to spend time finding out in what order they should be reposted, but if I could move the comment-answers (and their comments) to the comment section of one answer (or the question), that would solve the issue.

That thread is a relatively extreme example, but the problem of posting a comment in the answer section is a common occurence. I believe it has even been suggested that people should get a warning pop-up that they are about to post an answer to their own question (with the default option being 'post as comment').

[Guillaume]

This is something that we've been bugging Mathworks about for a long time and I know they're very aware of the need for it. Of course, the implementation have to be balanced with the need of more important features.

Answer 6

[madhan ravi]

There is a feature called "Managed Followed Content" , would like to see a feature to be able to "bookmark/pin" questions within MATLAB Answers.

[David]

Thanks for sharing your idea Madhan. Can you say more about what you're thinking? What do you mean by bookmarking or pinning questions?

[Image Analyst]

I just add a bookmark to my Firefox bookmarks. That Firefox allows bookmarks while Chrome is sorely lacking in that capability is why I choose Firefox over Chrome.

[madhan ravi]

Hello David , thank you for the response.

Illustration:

Imagine if I find "Matlab answers wish list 5 and bug reports" question interesting. If I follow it I won't see this thread ever again if nobody answers it right? So what I do is now I pin this thread which option "would be" near to the follow button. One may ask how's it different than following it , my point would be that it would be far much easier than a followed thread. While I follow multiple threads per day which makes it hard for me to find the exact thread which drew my attention to. This pinned threads could be found next to "Managed Followed Content".

Sir Image Analyst: Your point is also valid but it would never equalize bookmarking/pinning within MATLAB Answers.

[Adam Danz]

I get email notifications on followed threads which can really flood the inbox after a while. Using gmail (and many other providers), a filter can be set to send the notifications to the trash unless content from their subject lines is listed in the filter. That's a relatively easy workaround to the problem.

But I can see how pinned content could be useful. For example, 'Pinned content' could be added to the dropdown menu at the top of the 'Follow' tab within the profile page.

[Sindar]

Agreed, this would be very useful. I prefer keeping my bookmarks relatively sparse so autocomplete works well, but could see myself favoriting several items a week. (Once you've got this, can you add "REALLY Favorites" so I can filter to the stuff I know I'll be using for months-years to come?)

[Adam Danz]

And then "Super Duper, Seriously, For-Real Favorites".

That highlights the limits of utility for the inevitable accumulation of pinned posts.

[David]

Thanks for the clarifications

Answer 7

[Mehmed Saad]

Someone is Typing (Whistlist)

This feature can be very usefull not for the questioner but for contributors. For example

1. Walter Roberson is typing a comment
2. Image analyist is typing an Answer

Pros:it can save contributors time

Cons: Dont know xD

Like this or a popup for someone who is answering the question (that two contributors are typing the answer)

[Adam Danz]

+ inf

[Rik]

Apart from the privacy-concerns, a good idea, but difficult to implement well.

[Adam Danz]

Well, if you're going to leave a comment, it's going to be public anyway. But I guess it would reveal the amount of time spent on a comment if anyone wanted to manually track that.

Some other potential Cons

• It may prevent people from adding their own answer if they see someone else is already doing so. This reduces the number and variety of answers.
• It may motivate people to rush their answer so it appears first, if they see that someone else is also typing. This may reduce qualtify of answers.

But I agree with the Pro that Mehmed noted: it would save time that could be devoted to an answer that no one is working on. Another pro: sometimes I'm writing a response while the OP is also writing a comment that answers the questions I'm about to ask. Those types of threads can become difficult to follow. If I see the OP is typing, I'd likely wait.

[Walter Roberson]

If I am typing something and I got notice that the original poster was also typing something, then I would want to be able to see what they wrote (when it gets posted) without losing what I have written (as would happen if I refreshed). Perhaps a "Save Draft" feature?

[Adam Danz]

@Walter Roberson, sometimes I get a notification that OP left a comment while I'm constructing a response. In those cases, I usually either

1. Open the thread in a new tab to see what OP wrote while my content is preserved in the original tab.
2. Copy-paste what I've written into a text doc so that I can paste it back into the forum when needed.

[madhan ravi]

@Adam I usually use the first point you mentioned.

Answer 8

[Sindar]

Being able to collapse items in the activity feed would be useful. Fairly regularly, I have 3-7 items for "X commented and edited their comment 5 times." Occasionally, I'll have a full page of items related to a longer discussion (usually on one of these meta-threads). It would be useful if I could manually (or Mathworks automatically) collapse large sets into single concise items ("X,Y,Z commented 30 times on an answer to the question What frustrates you about MATLAB? #2"). This would make it easier to scan my log without worrying about missing a more relevant item in the noise

[Walter Roberson]

Yes, I agree. I know this has been proposed to Mathworks more than once.

[Adam Danz]

Yep, I'd like that too.

[Rik]

This can happen especially when some users many several smaller edits in a row to the same comment/answer. It would be nice if those were merged in the first place.

Answer 9

[per isakson]

I wish Answers would automatically reject comments like

"I did that but it didnt work"

Answer 10

[madhan ravi]

It would be better if there is a feature that warns when the first word of the poster is Write. Because at the end the poster edits away the question making the thread look pointless wasting the answerers time.

[Rik]

Alternatively it would be a solution if there is a restore point created each time an answer is posted. Then it would be easy to undo such an edit.

[madhan ravi]

Yes that what would be better too!

[Rik]

I have the feeling this problem is getting worse, so a capture of the question at the time it is answered would really be nice. I hope it gets better when exam season is over.

[madhan ravi]

Exactly!

[Adam Danz]

This problem is definitly becoming more frequent. Here are some other suggestions.

1. Users with sufficient reputation can flag other users who delete their questions. This flag will be visible in the stack of questions or when the thread is opened.
2. A change-detection could detect when a certain percentage of content from a question is edited or removed which would trigger a confirmation dialog that requires the user to confirm the change and would warn the user that their profile will be flagged if the content is edited-away.
3. First-time-posters or first-time-in-a-while posters have to accept a displayed agreement that their question-content could be edited but not removed or obfuscated so that users are aware that they are giving up their rights to privacy.

[Sindar]

Additional idea for deletion issue: keep a record of deleted questions in users' Contributions log

For identifying low-quality or suspicious questions: don't allow questions that are just an image

[Rik]

Related to this: I just wrote a guide how to restore questions from Google cache. It is not a perfect solution, but at least this can partly solve the symptoms.

[madhan ravi]

Thanks Rik!

Answer 11

[per isakson]

Don't encourage the use of "matlab" and "simulink", by showing them as popular tags. Instead, encourage the use of the Product field.

[Image Analyst]

Yes, I bring that up constantly at board meetings. Should be really easy to fix but they just never seem to do it. Plus on the list is a better curated list of prepopulated tag suggestions. I've been bringing these up for literally like 5 or 6 years.

In fact every day or two I go in and remove the totally useless "matlab code" from 10 to 30 posts. Plus remove the "matlab coder" tag from posts that have nothing whatsoever to do with the coder toolbox. In fact, usually posters will put matlab, matlab code, matlab coder, and code ALL in there so I have to remove 4 tags on hundreds of posts per week. And often they'll throw in "help, help me, and problem" just to pile on the useless tags.

[Paul]

Fully agree that the use of the Product field should be encouraged. Way too many questions post w/o the relevant product(s) tagged, which is bad for the questioner who might not get answers from product subject matter experts because the question is buried on the main page, and makes it more difficult for those same experts to answer questions for the same reason.

Answer 12

[Image Analyst]

In Firefox v67, if you click on the m-file attached to a question, it offers to open it in MATLAB. But if I say to open it in MATLAB, it opens a new instance of MATLAB instead of using the existing, already open instance. Plus to make things worse, it does not even open the m-file at all. It just launches MATLAB with all your old m-files open.

[Walter Roberson]

This is not happening for me on Mac.

[Image Analyst]

It does however set the current folder to the temporary folder where the m-file gets downloaded to, however it doesn't go so far as to actually open the file in an editor window.

Answer 13

[Adam Danz]

Expand/Collapse all comments

This goes back to Matt Tearle's request in the first wish list (2011). An expand/collapes toggle button would be useful so we can search a page without having to expand all of the comments. For example, find "lOVIT" in this thread.

This feature already exists in the MathWorks documentation:

[Guillaume]

FWIW, with the new 'less blank space' layout recently introduced, I suggested that all comments should be expanded by default. It's very rare that the first I do not do is expand the comments when looking at a post.

[Walter Roberson]

Well, I don't want all the comments expanded when I read this post. Or the one about error -9. Lots of history that is mostly irrelevant to the immediate concerns.

[Adam Danz]

I would favor load-time over comment-accessibility. Some pages with lots of comments take quite a while to load. Even in the documentation, I often "collapse all in page" before scrolling through the page so I can find a section faster.

Answer 14

[Mario Malic]

When user sorts questions for which answer has not been accepted, the questions are found in two different sections. I believe this lowers the visibility of questions and adds an extra effort for the users who would like to help/contribute due to the fact that they have to filter by both statuses.

Filter by Status needs an addition or change: unsolved and solved. Furthermore, in unsolved status, there can be another one: answered/unanswered.

[Image Analyst]

How are you doing the initial filtering? Is it by these links:

or some other way? How would you change the above? Do you want "Answer Accepted" to be called "Solved" and "Answered" to be called "Answered and Solved"? Some posts that have not been officially "Accepted" are considered solved by the person(s) posting the answer(s), it's just that the original poster never accepted any answer.

[Mario Malic]

Yes, those links.

Maybe something like this, where 'Answered' and 'Unanswered' would be grouped under one status, with the possibility of further filtering it. If questions are sorted by 'Date updated (Newest-Oldest)', all relevant questions will be higher in the feed. The questions that have received an answer that solves the issue, but have not been accepted yet, will not be updated and they will be lower in the feed.

• Solved (Answer accepted)
• Unsolved
• - Answered
• - Not answered

[Image Analyst]

I agree that would be more explicit and intuitive

Answer 15

[Mario Malic]

Users who edit away the questions get an explicit warning not to do it again. Freeze account/remove rights to post questions/answers/comments for some period of time. Might be a bit harsh, so some alterations are welcome.

[DGM]

I kind of agree with this, but I think most of the offenders wouldn't be affected by a temporary suspension. It seems they tend to purge the evidence only after they've decided they're done exploiting the forum.

[Rik]

Also, it is trivial to create a new account, so a suspension might not be effective enough.

Maybe a 'first question' badge would be more effective as a warning signal.

Answer 16

[DGM]

Now that moving answers/comments is possible, we kind of need these objects to have accessible timestamps with a resolution better than 24h. When multiple people are posting comments on each other's comments-as-answers, it gets a bit difficult to untangle the thread without knowing the order of events. The sequential identifier numbers are useful to some degree, but comments and answers are not comparable in that manner; furthermore, once an object is moved, it's no longer in-sequence.

A lot of other sites will show rough timestamps, but a mouseover will reveal the full timestamp. That'd be fine.

[Rena Berman]

(Answers dev) Thanks @DGM for your feedback. I will capture it for consideration.

Answer 17

[Image Analyst]

When we insert a hyperlink into Answers from chain link button tool on the tool ribbon, the focus/cursor is not in the "Display Text" or "Target URL" field. It should be in one of them, preferably the Target URL field because I have a link in the clipboard all ready to be pasted in and I'd like to do that as soon as possible. Save us a step and put the cursor in one of the edit fields.

[Guillaume]

Yes, as I complained early on, I find the hyperlink dialog completely useless.

On the other hand, I don't bother with it. It's much easier to insert a url by typing [text to display ](actual url) which gets autoconverted into a url when you type the closing ). Much faster as it doesn't require reaching for the mouse.

[Jan]

When I press the icon for insterting a link in Firefox/Windows, I get a greyed out page only, but not any dialog. I have to click in the background and press escape to get control over the page again. Which browser is working better?

[Walter Roberson]

Jan: Interestingly, I do not encounter that problem on Firefox for Mac.

[Guillaume]

On the rare occasion that I do use the dialog, I've not encountered that problem either. Firefox on Windows (various FF versions on various computers with either win7 or win10).

[Walter Roberson]

These days the focus is in the target URL section -- that is, this is implemented.

Answer 18

[Image Analyst]

Firefox scrolling issue not fixed. I don't know if this happens in other browsers but if I go to comment or answer, the top of the edit box is off the top of the screen, so if I start typing immediately I can't see what I typed. I need to scroll down. It updates the screen with a nice sized edit box, but the top of it should not be above the viewable area.

[Rik]

Sometime on Chrome I scroll up to read a part of the question. Normally if I start typing again, the browser will scroll me to the text box, but under some circumstances which I can't seem to reproduce, this doesn't happen correctly. Maybe the two are related, although browsers being browsers they probably aren't.

[Bob Thompson]

I think the issue is similar on Explorer (not sure if that is even still supported these days). Specifically, if I type a long response, it seems like once I reach a certain length then I get the same typing below my screen issue. I can scroll down to view what I have typed, but as soon as I enter or remove a new character then the view jumps back up too high.

I don't remember if viewing other tabs, or scrolling up to not be looking at the edit window is also a factor, but I know that length is 'consistent.'

[Guillaume]

I used to run in the scrolling bug fairly often and it really made writing answers a pain (I gave up a few times). However, since we were told that the bug had been fully fixed, I've not seen it even once.

Nonetheless, I've alerted the developers of this discussion. It would be nice if they could tell us the cause of the bug before, so we know if it's still the same bug.

[Image Analyst]

They fixed the one bug where every time you typed a letter it would scroll to the top if your message got longer than the viewing window, but there is still a bug where right when you start a reply, the top of the edit box is above the viewing window in the browser. Not sure if they're related or if it's a new bug.

[Rik]

Thanks, I thought something was crawling over my screen XD

[madhan ravi]

I just panicked for a moment when I saw the comment with a bug crawling ;)

[Image Analyst]

I'm not experiencing the scrolling off the top of the screen issue anymore. I think I've maybe updated Firefox since then. So maybe it was a Firefox issue that's been fixed. I'm now using Firefox version 68.0. I'll continue to see if the issue arises again.

Answer 19

[madhan ravi]

Whenever a text is edited by mobile [safari browser] , the click produces superfluous three letters " ltr " , which is really frustrating.

[Rik]

I think it is odd how untested this editor feels for mobile. They probably put hours into testing for mobile as well, but as an end-user it doesn't feel that way.

[madhan ravi]

True ;), hope it gets fixed soon as well.

Answer 20

[Rik]

Maybe this idea has been suggested before, but maybe it would be good to have a down vote button. I sometimes see answers of questionable quality, where I don't know where to start to explain why it is such a bad idea. I don't think they should be outright deleted, and a flag doesn't really seem the solution either.

Is this something that I will just have to learn to deal with? Or is there some feature that would solve this?

(obviously I will not link to any recent examples, but you could look up some of my embarrassing first answers)

[Walter Roberson]

If I recall correctly we decided against downvotes early on, feeling that it would be better to comment or to post better answers (that might include material showing that the alternative would not work)

Answer 21

[darova]

I want a button for spoiler

Can't believe there is no such option on this forum

[Rik]

What would be your use case on this forum?

[Adam]

Would that just be for making answers look shorter initially, like collapsing code or a tree? I don't use all that many forums so apart from obvious purposes where the term 'spoiler' makes more sense I'm not too familiar with its usage. I wouldn't have thought spoilers are really a concept in a Matlab forum?!

[Guillaume]

Whether it would be useful or not is debatable and considering that lots of people struggle to use the formatting tools that are already there, I fear it would be another thing we (editors) would spend time fixing.

In any case, the answer editor is using the matlab online editor, so the feature would have to be implemented by the matlab online team first. I don't see that happening.

[darova]

The asnwer could be shorter and clearer

Some comment1

[spoiler]LONG CODE[/spoiler]

Some comment2

[spoiler]LONG CODE[/spoiler]

[Rik]

Automatically collapsing code blocks that contain more than 10 lines or so might indeed be a useful thing. I would argue against something like that being user-controlable because than we get the scenario Guillaume is describing.

[Image Analyst]

It might be used for homework answers where you give a hint first, then the hidden spoiler below it. However I don't see any student who would just take the hint and not reveal the answer hidden by the spoiler button.

Or maybe for riddles/jokes??

[darova]

Can someone please vote for my answer?

Answer 22

[Adam]

I realised I posted this in the ancient wishlist thread before.

Can we not have some white space back between answers? Since the last update all answers just roll into each other which looks frankly awful, almost to the point of being unusable for threads with multiple answers. White space is a good thing, in a forum as well as in code!

Different answers can often be significantly different. They shouldn't just flow straight into each other as though the 2nd was just a comment or continuation of a conversation in the 1st answer.

[Guillaume]

Yes, several of us CAB members have already asked for this to be changed. Hopefully, this will be included in the next update.

[Kent Millard]

Thanks for your feedback Adam. We’ve received several comments on the redesign and are actively collecting feedback. If you’re available, we’d like to contact you directly to get more specific information from you. Thanks again.

Best – Kent (User Experience for Answers)

[per isakson]

Yes, better separation between answers would be good. However, more white space would not, since less visible text and more scrolling.

Better separation may be better achieved by putting the header/title stuff on a pale tint-plate (background color rectangle).

[Adam]

My profile is set to not let random people contact me, but Mathworks should have my address (I receive e-mails for the board editors) and are welcome to get in touch using that for any information you wish..

[Adam]

The location of the 'Answer this Question' button is annoying too.

Probably about 80-90% of my contributions are comments, usually on the question itself, but when you type your comment (assuming you managed to click the link to comment rather than answer in the first place), you have a small 'Submit' button and a much bigger 'Answer this Question' button right underneath it, shouting to be clicked. So far I have accidentally hit that button instead of the 'Submit' button numerous times.

It's all part of the same idea of needing to squash everything up rather than just spread things out so question, comments and answer clearly have their own obvious space.

We already used to have too many new people adding answers instead of comments.

[Guillaume]

Yes, it is also something we've complained about. Indeed with the new design I've accidentally posted an answer when I meant to post a comment, something I'd never done before. My suggestion was to move the button to the right. Yair's suggestion was to get rid of it altogether, not a bad idea as it forces potential answerers to actually scroll past (and thus read) answers.

Some things work very well with the new design (highlighted accepted answer, highlighted OP posts, structure of comments, less whitespace), others not so much. The whitespace cull may have been too aggressive.

[Walter Roberson]

With the new arrangement I have sometimes been accidentally posting as a Comment to the Question when I meant to post as an Answer.

Answer 23

[madhan ravi]

Why is there a lag in the recent activity page? This has never happened before.

[Walter Roberson]

lags in Recent Activity and Activity Feed have been known to happen; sometimes the feeding process crashes.

That said, I am not observing any lag myself.

[madhan ravi]

:/ I seem to have ~ 1 min lag

[Rik]

I haven't noticed any particular lag, but a 1 minute lag is not unusual for me. If I recall correctly there was a several hour long lag for multiple people one time a while back, which I only noticed after it was resolved.

[madhan ravi]

Hmm didn’t know at all.

Answer 24

[Rik]

A soft-lock for questions, similar to StackExchange.

This soft-lock would prevent low/no-reputation users (e.g. <5) from posting answers in threads that have the same answer over and over again, like the date validation assignment, the school admission assignment, and the picker function assignment (each of these probably have multiple incarnations). One example of a non-homework thread with this issue is this thread, which contains a stream of complaints and single-word answers.

That last one is probably due to people not being allowed to post a comment to the question, so the only button is 'answer this question', which seems like poor design IMHO.

[Rik]

I'm now starting to tag these threads with "soft-lock" in case this ever becomes possible.

[Rik]

This soft-lock would also be usefull for thread like this one, to force people to post new answers in the new threads.

[Adam Danz]

Large threads like this also need a mechanism to speed up the page loading time.

I prefer a mechanims that shows collapsed answers that require an expansion to view nested comments after the number of answers exceeds a threshold and the nested comments would load only when expanded. Additionally, a second greater threshold could be set to initialize a "view more" option such that only the first 'n' answers would appear in order of votes. These larger threads often consume so much load-time that I'm reluctant to view new answers/comments.

[Rik]

Until such features as you suggest are implemented, splitting the threads seems the only viable option. I share your point, especially on mobile I don't bother to open up the long threads.

Answer 25

[madhan ravi]

Not mandatory but a replying feature to a specific comment would be nice.

Answer 26

[per isakson]

Activity Feed: "See more" asks for "See less".

[Image Analyst]

Not sure what you mean by this. When I click it, it seems to display more items. It doesn't ask for "See less" or ask me for anything else. You probably need to explain more explicitly, probably with screenshots inserted, so we know what you're talking about.

[Walter Roberson]

I think per is saying that if you are going to have a "See more" then you ought to also have a "See less"

[Adam Danz]

I'd like to be able to collapse the group back to 1 again rather than see fewer.

Also, I'd like to see a number indicating the number of new threads since our last visit to that page, if possible. Something like this below.

[per isakson]

Yes, I intended to say that there is a need for an undo button that reverses the effect of "See more".

[Rik]

Excellent suggestions. I think this new page has a lot of potential, but it will take some time to get used to this.

Answer 27

[Paul]

After I click the "Follow" link on a question, I'd like that question to show up in my Managed Followed Content list immediately (with some reasonable delay for the system to catch up). As it stands now, it only shows up in my list after someone else subsequently answers or comments.

[Josh Natanson]

Thanks for sharing this frustration Paul. Your Manage Followed Content page should immediately show any item you've followed. I think the issue may be the default sort value of "Last Activity". The system only considers activity you've been notified of for this value, so it starts off as a null value, and is sorted to the end (probably on the last page of your follows, if you follow enough to paginate). One simple workaround would be to sort by the "Date Followed" header if you want to see the most recent item.

I'm curious what your workflow is in this case. What is your goal when looking for something on this page just after following?

[Adam Danz]

I was also not aware that newly-followed links would appear in that list without any recent activity. The sort solution is good to know.

In addition to hearing Paul's workflow, I'd like to offer mine. The majority of followed threads are auto-followed after I contribute to them but some threads I follow without contirbuting to them because I'm interested in seeing the solutions by other people. Until now, I didn't realize we could see links to followed-threads that didn't have any recent activity to push them to the top of the list.

This partially fulfills a feature I wish we had, although I'd still like to have the feature. Sometimes I'd like to make a contribution to a thread but I think others in the forum could answer it better so I'll follow it the thread. But sometimes those questions never get answered and they get lost in the list of followed threads (~6000 followed threads for me at this time). I'd like to receive a notification when a followed thread does not receive an answer within 24 hours. This would only apply to newly asked questions (~48 hrs?). That way I can be reminded to answer the thread since no one else gave it a shot.

This would have two benefits:

1. The unanswered question stays in the unanswered-list longer than if an answer is added knowing that it may not be the best answer. Of course a comment could be added in this case but sometimes answers are a lot of work that we only want to invest if needed - ie, no one else answered it.
2. If we have intentions of making a contribution if no one else tries to answer it, there's less of a chance that we'll forget to revisit the question to check on it later.

[Paul]

It never occurred to me that it would be at the end of the list and that sorting by Date Followed would be the workaround. Good to know. But I like my list sorted by Last Activity, so it would be great if me clicking the "Follow" link was considered activity, thereby pushing it to the top of the Last Activity list.

As for workflow, sometimes I click on "follow" with the intent of coming back to the question, either to review it in more detail, answer, or comment. Of course, when someone else answers or comments it bubbles to the top of my list. But if nobody does, it gets lost in the shuffle and I'll forget about it. So I guess my use case is similar if not the same as Adam's use case.

[Walter Roberson]

Over 33000 threads followed for me ;-)

[Adam Danz]

But you have ~20K more answers than that. Was the follow-feature added later in the forum's development?

[Walter Roberson]

October 11, 2017.

As per the default, following is on for any Question I ask, Answer, or Comment on. The system thinks I have contributed to over 83000 Questions.

[Adam Danz]

I see my first follow is on that date too but I couldn't remember when I actually started contributing here. The system says 2014 but that was probably a single question/comment and I may not have followed it.

I'm surprised that I follow twice as many threads as the number of my contributions. I thought the number of follows would be slighly higher but not double.

[Josh Natanson]

@Paul, Thanks for the additional info. Defaulting the latest activity to the time of follow seems like reasonable behavior, we can consider that for a future update.

Answer 28

[per isakson]

Background: I recently put some effort into answering How to store and reuse coefficients in a for loop. My answer together with a comment includes a discussion on "column-major order" and "performance". I searched in vain for references in the documentation and in Answers regarding the significance of taking "column-major" into account. OP has read my answer an comment. So far so good. Now I would like a few more users to read my responses. To make my responses to OP's comments more visible, I sometimes add them at the end of my answer. And I add tags, in this case I added "column-major"..

Proposals:

• support multiple code blocks/wells in one answer. In the current case my answer contained one code block and my comment another. Adding the comment to the answer causes "Run in 2021a" to throw an error for the second code block.
• support searches (Cntr-F) in the current thread to include hidden comments. Currently, "Search Answers" for "tag:"column-major" singleton" returns the question ("singleton" is in a hidden comment), but Ctrl-F doesn't find "singleton". (I use Firefox.)

[Walter Roberson]

I suggested distinct code-wells to Mathworks a number of months ago, along with automatic numbering of lines in code wells, and control over which code wells restarted the line numbering .

It is common that we want to quote particular user code to discuss it. We might not want to quote the entire code before that point (readability) so variables might not have been defined that are needed -- and of course the reason we might have quoted a line is that it might have a syntax error that we want to discuss. And we might want to then show correct code. Thus there might be code wells that we do not want to have executed at all.

[Walter Roberson]

Mathworks did some work, at least for power users, to make comment expansion easier. Now that I think of it, though, it might possibly only be in place on mobile: on mobile, if you expand older comments, then it expands all of the comments in the entire post and it remembers that you want comments expanded.

[Kent Millard]

Thanks Per.

• For mulitple code wells, I'll refer that to @Tushal Desai--he implemented the Run feature and can respond to your suggestion.
• For supporting search of hidden comments, to Walter's point, we released a "Show/hide older comments" that appears at the top of every question thread. The toggle is persistent so if you prefer to see or, in your case, search threads with CTRL-F, you now have that ability.

Thanks as always Per for all your work in Answers.