Garbage software is one of the primary reasons I left my last job despite high pay. It just got too friggin annoying to use. They’d roll out a ‘hotfix’ to fix something they had broken 3 months earlier and they’d break 2 new things which previously had been working fine for years. The support was so bad I just bought a magic eight ball for our office and we’d ask it our support questions.
The Ribbon interface used on office products isn’t there because it’s good UX. It exists because there’s a software patent on it.
If office didn’t use a patented UI, someone could make office software that replicated the UI of MS Office which would allow companies to switch to other products without having to retrain staff.
Microsoft was enshittifying their software long before anyone else.
I used the ribbon API when building a C# GUI. It’s just part of the Microsoft application framework. Maybe they prevent other frameworks from using it? Underneath the fancy paint there’s not really much to it though, it’s just adding a tab bar to a tool bar.
I did an internship where I was creating a prototype UI for a Windows application, and used the ribbon API to build it. I thought it was a well thought out design, and was definitely an improvement over nested menus. A problem I’ve seen come up a lot though is shitty implementations where the pattern wasn’t followed correctly making it really hard to find things because the developers put items in dumb places.
Generally in UX you want often used buttons to always be in the same place to take advantage of muscle memory. Text is more intuitive than an icon, but an icon will use less screenspace, so once the user learns the icon, you can have an interface that’s more user friendly (though less intuitive) so that’s fine. Small amount of experience or training needed with the softwareresults in more buttons available at all times, so it’s worth the trade off to use one button bar. Less used items should be put into a menu because a) it’s not used often so it’s fine to be hidden away unless needed and b) it’s not used often so the user isn’t going to be familiar with an icon so text is preferable.
The ribbon is some weird combination between a menu and a bar with buttons on it. So all of the disadvantages of menu (buttons aren’t always on the screen) and all of the disadvantages of button panel (icons that have to be learned for nearly every single feature). The advantages of being able to access the most used features from muscle memory is lost, the advantage of being able to discover lesser used features by simply reading text is lost.
It’s just indecisive design. Not putting any thought about how the user actually uses the software, Just chuck some buttons onto a ribbon somewhere, make a pretty icon so it looks good and let the user click on various ribbons an click on random pretty buttons until they find the button that adds an attachment to an email in outlook. But when they find that button, make sure we default to OneDrive instead of the Documents folder because pushing cloud storage is currently the top priority as MS.
Sorry… bit of a rant there. But yeah, just put thought into which features will be used most often make them to be the buttons on the bar, put everything else into a menu. Worst case is the user has to click two things to use a feature, which is the same as using ribbons. Best case the user is clicking the same button they’ve clicked 100 times before and it’s in the exact same place as when they clicked it all of those times before.
I don’t disagree with a lot of what you said, but the applications that use the ribbon are just complex applications with too many commands and options, and while some commands are very common most of them end up having tons of middling use, where it is annoying to have to find them in a menu every time. I think the problem it’s trying to solve is a hard problem and all the other attempts at solving it have a lot of problems too, and I really don’t think menu based design is an improvement. I’m not saying ribbons are the perfect solution, just that the other solutions suck more.
Designed properly the ribbon should take advantage of repeat contexts, so if you’re doing one kind of task repeatedly, it’s likely you need commands from one tab over and over again which beats having to navigate through a menu multiple times. Of course it’s not always designed properly.
Frankly, menu navigation is probably the worst navigation ever devised in terms of hit targets and findability. It has the same organization hurdles of the ribbon, but worse navigation and sometimes nested menus which make things even harder to find. Really, all the ribbon is, is a tab bar on a tool bar, and power users can easily switch tabs with the number hotkeys. The power user option for menus is letter and arrow keys which suck.
Personally, I think the best option for productivity is the command pallet approach like in code editors, but the downside of that is that you need to be a power user to be effective at it.
Dude, I’m a surgical tech - my job is to stand in an OR and be a surgeon’s bitch while we’re flaying some fucker open. …and I still spend what feels like 90% of my day on Outlook -_-
I’ve been using Typst. Its (mostly) open source and much simpler than LaTeX. It’s still very new though, so it doesn’t have all of LaTeX’s features, but it’s making very steady progress.
Man I guess I’m spoiled. We get access to the top row except SolidWorks because we license an alternative. We use the entire MS suite too though but as a supplement. I don’t use excel hardly at all because JMP is superior in every single way, except for dashboards where we use PowerBI.
I want to love Julia so much, but it’s always something. The funky handling of scope in the REPL was the latest off-putting thing for me, but maybe I should give it a try again…
If you don’t like MATLAB your probably not the correct audience. It’s for people needing to do data analysis, simulation or control and have a lot of money to pay for the libraries. The things software developers hate about it tend to be what makes it better for statistics and modelling. Math works even suggest it isn’t appropriate for making software as the sell simulink coder that turns simulink models into c++ code.
Add comment