• 1 Post
  • 48 Comments
Joined 1 year ago
cake
Cake day: June 11th, 2023

help-circle


  • If a user speaks a different language, good usability knowledge will tell you, change the software to help the user. Not change the user to help the software. The software is only there to make things easier for people.

    As I said for many people, the tasks they do are not always possible or not easy with the CLI. Try drawing a curve, try moving an object from bottom left to a position higher up to the right. Even navigating a tree structure, common in many apps, it’s easy to click on a chosen branch directly. Even with CLI options, more people, including CLI users, feel it’s natural to use a GUI app to do their email, manage files or browse the web. There is a lot of learnability built in. Discovering new things by accident is a natural benefit. And a big downside of the CLI. Which is not THE natural way at all.




  • Sure there are some people who can’t do anything. But there are a large number of full time computer users not in IT who know their GUIs really well. These are candidates to switch to Linux.

    If you give someone a text string to paste in, chances are they won’t be able to tell if it worked. They might need another command for that. And how can they undo that command? And the next time they need that command they’ll have to have stored that command string somewhere! Which is why it is better to show them the option in their application GUI, as the GUI will provide feedback on the status. And makes it obvious how to undo the change, and they know where to go next time. Otherwise they are dependent on you forever. Also, I doubt if there are any text commands for most things I do on a computer.

    You don’t design a UI around the relatively few occasions when GUI help is too hard for some helper.




  • The CLI is not faster for everyone. Especially non IT users who don’t know any of the options and would have to search the web every time. And there are so many commands the CLI is useless at. Graphics and presentations a big fail. And few people use CLI for email or spreadsheets.

    GUI users have a chance to work out how to use a new GUI 99% of the time, as it is all familiar. No exact syntax to type perfectly or it will fail. On other occasions, web help tells you where an option or command is. There is no need to entirety change the way you work, just for a few times when learning new software.



  • The CLI has lots of hurdles. Such having a blank screen with no prompts. Where the GUI shows the options you have. And 1 click to set the option. And how to unset the option is obvious. You only need to half remember a feature. Not precisely memorise and type command exactly or it will fail. Or worse, delete something you need. The GUI is preferred by the vast majority for good reason.


  • Yes. I’ve been using Linux for over 10 years without touching the command line. I used Ubuntu up to Unity, then switched to Kubuntu and Plasma. I’m not in IT, so I don’t need IT stuff. It all works by GUI. People who haven’t tried it might say it’s not possible. But they are not speaking from experience. Some others, not interested usability, don’t understand why GUIs are so successful and dominant. Which is absolutely fine, as long as they don’t try think they are suddenly knowledgeable in usability, and have tried 10+ years of GUI only.


  • The Registry Editor is a GUI that doesn’t rely on memorising commands. This is home territory for users. A tree to navigate, similar to a file manager is very familiar. Even the first time. Just give me the path. It is also easy to work out how to edit entries, and to revert changes. Sure, you might not like the looks of it. But it lies in the ‘normal’ world. Not in some strange world. People not into usability sometimes don’t get that.


  • Terrible usability will be the right term, if someone suggests applying one type of UI to an inappropriate situation/user/task. Such suggestions sadly seem to happen a lot in the Linux space. And saying CLI is easier is a sweeping proclamation. Whereas I’ve avoided making sweeping proclamations, repeatedly describing the many cases where CLI is poor. Usability analysis needs to know about the user and the situation. It’s not one size fits all.

    I’ve used various command line systems a lot in the past.

    I’m saying it’s more productive for many to invest in extending learning their home environment than learning a completely unfamiliar and inappropriate environment.


  • I launch favourite commands with 2 clicks. Once on the App launcher button, and once on the App itself. My hand is on the mouse anyway. So it’s fast. Way faster than typing a whole bunch of characters. For less used apps It’s 3 clicks as I’d open a category like “Media” or “Games”. And doing that, I get to see what I have in there. This builds up a picture in the users head for future use. Learning “Add to favourites” is time well spent. It can even be called “Pin to Start” or “Bookmark on Launcher” it doesn’t matter. You don’t need to memorise that exactly like the CLI. And right-clicking things is already second nature to huge numbers of users.

    So I have no incentive to use text commands. It’s not faster. My hand is on the mouse for my apps anyway. And the CLI has terrible usability, via poor learnability, zero tolerance, and poor visual feedback. And completely useless for most things I do, like working with 3D models, images or drawings. I’m not a “text-worker” like IT tend to be. Plus, I want more non-IT people to use Linux, so discovering the easy ways to do things can help spread the word to them.

    For me it would be like stepping off a high-speed train and walking over uneven ground instead.


  • Memorising does not need to be precise with a GUI, as you are given visual cues and can see the next step to click. You don’t need to remember precisely every letter or it fails. You don’t even need to remember the name of an application. The desktop app launcher shows you which apps you have installed. I often pin apps to favourites as a reminder. Some Appimage apps don’t appear in the launcher. I forget I have them installed and they don’t get used.

    Differences between Desktop Environments are easily found when you change. As GUIs are in many users comfort zone. We use them all the time. People know their home environment, and differences need only just that discovering. Not a whole new environment.




  • Yes. I’ve been using Ubuntu and now Kubuntu for about 12 years and I don’t use the CLI. I don’t play computer maintenance guy, so don’t need any weird hacks. I just use my applications, which all have GUIs. I don’t need the CLI despite people telling me I need to use it. They have never tried GUI only. So they don’t know what they are talking about. The next lot, who typically have no idea about usability, tell me I’m missing out on something. But it’s always something I’ve never needed. If I were to use the CLI, I would need to spend ages researching not just some command, but a whole lot of other concepts that I have no clue about, only to forget it all if I ever need that again. So not as fast as people claim. Luckily, Desktop Environment developers know this and put a lot of effort into making them user friendly. They understand usability. And that different users have different needs.


  • “It is not strange”

    The terminal is nothing like familiar GUIs people use daily. To most, the terminal is strange and full of opportunities to mess up.

    “so much faster by copy/paste”

    It’s not faster when the user needs to go to a separate application first to find the instructions, then find the text to copy. And also search for how to use the terminal, and that it might be called Konsole confusingly. And also to understand if the command did anything. Does it print anything that you need to read? Should you close the terminal afterwards? Should you trust copy pasting from Internet strangers? All this is missing from online help, where they assume people have used the command line before.

    Such a GUI app could be launched from Dolphin by right-clicking the share, and selecting a new option “Mount” that would help discoverability. A standalone app would need:

    • A clear title. Like “Mount Share”
    • A ‘Share selecter’ Browser.
    • The ‘mount point’ definition should have both:
      • ‘Folder Browser’ for manual selection.
      • ‘Default path’ option. Most don’t know where it should be mounted.
    • ‘Credential definition’ needs a ‘Guest/Anonymous’ option for when there is no account defined.
    • Feedback message on success or failure
    • Easy to install.