Checkboxes that kill your product
limi.net
limi.net
Common sense: you don't know what this control is for? DON'T touch it. Or touch it and face the consequences (learning's got to start somewhere, right?).
This article seems to encourage apple's way of thinking ('simple', 'just works') and this is one of the main reasons for me (and many other people, I reckon) to ignore all their products completely. You want to go this way? fine. Don't expect anyone technically-literate to use your thingie though (because it becomes harder to use for debugging and all those things than, say, Chrome). Also:
> Is it really worth having a preference panel that benefits fewer than 2% of users overall?
and then
> Even 1% of a few hundred million isn’t a trivial number of people.
Decide on it already :)
I would say that since the code already exists to do these things, maybe the right way to deal with this is to move dangerous preferences to a sub-menu with warning signs, and then some sort of icon or notification to warn you that you've got a setting enabled that will break big swaths of the Internet.
Checkboxes have a clear path to resolution, you [un]check the box to return it to it's original state.
To prevent the user forgetting then perhaps configuration changes could be added to the history (or a different "undo-history") so the user can easily back-track.
Or disabling image loading, maybe it works fine on the first few sites you visit, but then you go to the google homepage and it stops working - what do you do then? What broke that functionality? What if you're not the person who [un]checked the box?
The perspective of the post is about how regular users might easily find themselves in a situation where "basic" settings broke core functionality in the browser unintentionally, without guidance on how to fix it.
I do like the "config change history" idea, though that might be Yet Another Panel(tm) that's too complicated for the regular user. But maybe there's a "Big Red Button" that firefox could have that sets things back to a "standard" or "known good" config. These types of things are great for novice users because you can tell someone, "Look, if you get into trouble, don't know what's going on, hit the big red button".
I've used stuff like this for certain family members who are not great with cmoputers - rather than trying to teach them all of the different knobs to twiddle, spend a lot of time teaching them how to get to a known good state ("Close all fo the windows and restart", etc.).
And the APIs to customize your desktop experience are much more permissive than what Windows allows. Or at least the Desktop customization apps do much more than their windows equivalents.
I'd say that having usable defaults is better than having easy customization. People who customize to that level don't care about it being "easy"(myself included).
On apple: look at the iphone vs android war. Android was and always will be a go-to for customization.
Now, yes, people will point out that OS X only runs on Apple's hardware and that is restrictive. However as a counterpoint I will say that OS X is free and it would be uneconomical without that tie-in. That's their business model effectively.
On one hand you have hundreds of millions of Internet users, including many from developing countries who have never used a PC in their lives. On the other you have Internet services that pile on endless 3rd party tech and scripts into their sites, most of which doesn't degrade properly.
The result is pretty much what the author highlighted. Sites will break. Users will be flummoxed.
I don't think we can wish away this problem by having users figure it out. I'm not saying Apple's shiny-button-abstraction is the answer, but there has to be a way for both users and websites to figure out non-standard browser config settings and pop up alerts.
You're saying, decide on the 2%, who need a rare feature, and who are technically capable enough to download an add-on... vs. the 1% of people who click the wrong thing, and are not technically capable to fix what happened. That's easy—pick the 1%, and move the browser-breaking things into an add-on, or something that a stray click won't trigger. Which is what the author is arguing.
Then later he says "it's 1% of people! that's a LOT of people!"
The point is that 1-2% of people is either a LOT of people and worth caring about, or it's not. If so, keep the options and make it harder to break. If not, delete it all and don't worry about it. It's self-contradictory to do otherwise.
1% * 10% problem severity = 0.1% people-problem severity
1% * 50% problem severity = 0.5% people-problem severity
If you do the math that way, sure.
Here's the thing, though: you don't (nor does the author) get to say "it's a minor inconvenience" because you're not the person being inconvenienced! Only the people who want the functionality and won't have it anymore can tell you the magnitude of the problem. And the guy writing the article clearly didn't ask any of the people who use the feature either.
When your revenue comes from people using your product so that you can collect advertising and search revenue, every percent counts. How much would the business side people be willing to spend to grow the user base 1%? Ask yourself that before you do something to alienate 1% of your users and then think "is there a way we can make this better without causing people to switch?"
All these options are under the configuration menu anyway so no inexperienced user are not going to go in this menu anyway... I don't really see the point this article is trying to make.
It's a problem if you're an armchair "usability engineer" who needs to write a blog post.
I'm guessing the word previously wasn't there back when the post was written.
These options are there because they've always been there.
Would you give users of the JVM (not program writers, mind you, but just people using the JVM) the ability to turn off the garbage collector?
>All these options are under the configuration menu anyway so no inexperienced user are not going to go in this menu anyway...
First rule of UX is not to assume you know how a user acts. Very inexperienced people can end up legitimately confused and do very irrational things. Probably for the best not to give them the gun to shoot their feet.
No, Google broke the Internet. If they've introduced a dependency between text boxes and images (?!), that's down to them. Web interface should work via progressive enhancement such that it's perfectly possible to browse and use a site without javascript, css, or even images.
There may be non-economic arguments, for example making a site more accessible. However, this needs to be balanced against practical concerns. For a government service then sure they should spend the money making it accessible to everyone; for the small startup I work for the cost and maintenance overhead would cripple us.
To be fair, I agree that having a text box which needs images is a little strange.
https://gds.blog.gov.uk/2013/10/21/how-many-people-are-missi...
> A spa is a location where mineral-rich spring water is used to give medicinal baths. Spa towns or spa resorts (including hot springs resorts) typically offer various health treatments, which are also known as balneotherapy. The belief in the curative powers of mineral waters goes back to prehistoric times.
Now I've got plenty of tech illiterate friends, and they've asked me many questions over the years, but never did it have anything to do with browser settings not loading pictures or javascript.
In fact, it's the one thing people tend not to mess with. They'll do tons of weird software, they'll even do regedits from a misinformed guide, but touching browser settings? I don't know about you, but none of my friends ever did.
So what's the problem here? Don't see one. What's the solution if there was a problem? Create a 'dev-mode' button, give the user a warning they shouldn't use it unless they're software developers, and put your tricky settings behind that wall.
The solution doesn't seem to me to require people to install add-ons for what is basic functionality to a decent amount of people, especially not in an addon store with lots of third-party crap that might work now, but doesn't work in combo with another add-on, or stops working after the browser updates and the add-on dev moved on to other things.
That's my preferred solution. On both Android (tap the build version 7 times) and Chrome (chrome://flags) it works extremely well, allowing you access to settings you wouldn't dream of giving the general public.
They could also move some of this from the general settings panel into the Developer Bar. Chrome allows you to disable JavaScript for your current page from there (although it still has the global setting too).
I'm all for Firefox reducing the complexity/gotchas within the general public's settings panel. Just allow advanced users to continue to access them via some means and everyone is happy. It wouldn't be a big deal to e.g. have to type in About:DeveloperSettings.
After looking through my settings, more or less all of them, with the exception of cache management, it seems.
But never once did I, or would I, blame the user.
These are just the decisions we have to make to get things done some days.
It turns out you can right-click on an empty space on the tab bar to reenable the menu bar, but this is far from obvious.
So in the future I'll have to browse through another 6 dozen 'classic options' add-ons until I find one that is maintained, compatible, does what it claims, and hopefully doesn't break my privacy?
Sounds great that the company that puts "Your Privacy First" finds that untrusted third parties are the best providers of core functionality that should have been in the browser to begin with.
For complicated products like Firefox, there can be a range of options based on how in your face they are:
1. wizard that demands a decision when you first install the app (last resort)
2. main pane of preferences (common decisions)
3. about:config (obscure use-cases, dangerous use cases)
4. only within the dev-tools (common case for a specific group of people)
5. recompiling (the other last resort)
The customization/safety tradeoff is real, but it's not zero-sum.Most of the options he's complaining about are not easily changed in my current Windows version of Firefox.
Only the certificates (which is a multi-click process to get to) and the cache management are located as described.
Although, looking around I found an option I never bothered to notice before. View -> Page Style seems to provide the ability to disable all CSS quite easily. What purpose does that serve for an everyday user?
For example if it takes 1 week to add support for JS disabled clients and they are 1% of your target market but you could spend a week on a feature that would likely add 2% growth, it's a pretty easy decision. The exact numbers are sometimes hard to work out, but the thought process can make the decision easier.
In some countries (in Europe mostly) there are certain legal requirements to make your sites accessible. And while these laws are infrequently enforced, you can be civilly sued if you upset enough people (governments have been sued for badly performing websites).
Instead you have to explicitly hit Work Offline to get into offline mode. That's totally bad for a modern application that can work both offline and online.
If your website needs with js, flash and need to load js plugins from 10 other ad domains to see, that's fine. I don't want to see it anyway.
I love to have choices.