How to specify date ranges in interfaces?

    The surface search did not give an unambiguous and comprehensive answer to the question, reference books of an academic nature most often give results for publications (including with options for using Roman numerals, which is not very suitable for interfaces), so I want to understand how it is better to specify exactly ranges dates and it is in the interfaces and try to formulate a rule or identify patterns. To do this, I remembered everything I could, the cases, and arranged them in a table — a case, a numerical example, a full and short format, for days, weeks, months, quarters, semesters, and years (inside the post).

    I’ll explain the problem: for example, you need to display a summary of expenses for a certain period in a mobile application and formulate a headline that a user can understand with a selected date range. So that it’s not a set of numbers, but that it’s humanly understandable.

    The screen width of a mobile device is often small, so there is a need to reduce it. At the same time, in addition to the technical width, I also want to take into account aesthetic perception and not load the interface with unnecessary entities. A similar situation with web input. The situation is aggravated on small devices such as watches and various kinds of small displays.

    For example, we get such strange things:



    Numbers or text?


    It would seem that everyone just read the range 10.07 - 12.07, and no problems. But we often start to think in our minds what month it is - June or July? And if you write in words, it will be more convenient to read, and more understandable: "July 10 - July 12". But here we include the range within one month, and if the screen is small, then I want to reduce this range. The numbers are shorter, but not so convenient, so it comes to mind the general part (month) to indicate once: "July 10-12." Both clear and compact. Another of the arguments against: if each time you display different headers, and not of the same format, this can confuse the user. My answer is: the decision depends on how human you want to get an interface. The more human (up to the format of communication between living people), the more normal the different alerts will look.

    There are still purely mental states of perception such as “this year” versus “last year”, “now” versus “then”, “today” versus “yesterday” / “past”, etc. They can be taken into account in order to omit situations that are clear from the context: for example, the current year.

    Suppose two user situations:


    1. The user has just specified a range and understands what he sees in the header. In this case, our task (or the header task) is to unambiguously confirm to the user the correctness of the range indicated by him. Unambiguously in the literal sense - that is, that there were no other meanings (what was written was interpreted correctly).
    2. The user went to the section and saw in the heading the default date range. This range must be clearly read.


    DD.MM.YYYY - DD.MM.YYYY.


    Technically, the date range could be written as: DD.MM.YYYY - DD.MM.YYYY. All clear.

    When we come to real dates, the following can happen:

    03/09/2014 - 09/09/2014, that is, the range (for example, expenses, or exercises in the gym) is indicated on one day. Technically, everything is again correct, but for a person it is:
    • one day (day);
    • it is one specific day (which is important - it is a specific day / day);
    • if this day is current (today), then the word "today" will be clearer to the user than the date. And if yesterday - then "yesterday" is also good. How much did I spend yesterday? - So much;
    • finally (?), if this range = date (day = range) is a specific day of the past / forecasted year, then it’s more humane to show it in the format “March 9, 2014”, and if it doesn’t fit (large print or hours), then "Mar 9 2014 ", or" Mar. 9 2014 ", it is doubtful already -" 09.03.2014 "and at worst -" 9.3.2014 ".


    ! That is, the range of one whole (or current) day or a whole week / month / quarter / half year / year / decade / century is logical to reduce to one entity (decades and weeks cause doubts - they rarely think of the numbers of calendar weeks).

    Another story is intersections. Everything, too, is clear there: the general greatest importance is to make it to the right (March 1-3, March 3 - April 15, 2015, etc.). But 2010-15. (the millennium is still on the left).

    It is felt that you can omit values ​​such as the current (current) year. For example, if today is January 15, 2016, then with the selected, for example, January 10, writing simply “January 10”. A month to lower somehow the hand does not rise :-)

    Nameplate v1.0: an attempt to systematize these very abbreviations




    I considered options for days (I have missed weeks yet, but have plans), months, quarters, semesters and years, they are on the tablet .

    Explanations for the table:
    • Italics highlight the confusing moments.
    • Red - that which seems to me unacceptable or incomprehensible (raises the most questions).
    • You cannot edit the table, but you can specify comments on cells. I will study them with pleasure.


    About what and how competently:


    • Set city and city for year and range of years. 2015 and 2000-2005 On the topic: new.gramota.ru/spravka/letters/22-spravka/letters/90-data
    • Write the name of the month with a capital letter at the beginning of the line / after the period and with a small letter after the year digit (i.e., if not at the beginning). For example: Noah. - Dec
    • Reduce the numbers correctly: ilyabirman.ru/meanwhile/all/o-naraschenii-okonchaniy-chislitelnyh
    • Ranges with names, SURE, are correctly divided by Dashes (-), and numbers by DEPHIS (-). For example: April 12 - October 12, but April 12-17 and 1998-2000. I wonder what you think on this topic.


    By the way about “today” and “yesterday” in the tape


    Similarly, Apple’s mail is used to indicate specific dates (and probably a lot more where, because it’s normally accepted): when did the letter arrive? - today, yesterday, and if “today” = “Friday”, then “the day before yesterday” and before the beginning of the week will be shown as “Wednesday”, “Tuesday”, “Monday”, and then by digital dates. It’s a kind of conditional principle of attenuation (it’s logical to assume “last month”, “last year”, “decade”, “century”, “millennium” - but this study will be left out of the picture for now). At first - humanly, then, the further into the past - more in numbers. This must be separately described for tapes and similar records.

    More interesting points that colleagues noticed


    What is the "last day"? - there is such a wording in the applications. They also say "the last 24 hours." Depending on, for example, billing, it can be 24 hours ago from the current moment, or there can be a calendar day that starts at 00:00. It seems to me that if the system considers it exactly 24 hours ago from the current moment, then this must be indicated specifically and unambiguously.

    If we expand the topic, then in this context we can also take into account time zones. Actual for active travelers. For example, I am in the USA, the difference with Moscow is from 8 hours. If I arrived for two weeks, the details of the statement inside the application will be interesting to me in local time, regardless of what time zone the same Moscow banking application considers. After all, on the phone’s watch, I most likely automatically have local time (for example, New York).

    Intervals "from - to" and "from - to."It is about when to include the phrase “inclusively”, and whether it is necessary to use different approaches within the same interface (when the last day is included in the range or not). It seems to me that different approaches cannot be applied - there must be one. It seems that by default it is logical to use "inclusively" (while there is no need to write the word "inclusively" everywhere).

    PS


    Surely someone has combed this matter and put it in order. Maybe even one of the developers came up with a scheme and implemented all this. If so, then this is wonderful and I will be grateful to the colleagues who will outline these decisions. If not, I will be grateful for your thoughts on the topic and suggestions for its development. I will also be grateful to the pedants and experts for any information on the recording formats of such pieces.

    Also popular now: