ThinkingRock Forum

In another thread on KISS and feature overkill Vauha27 cautioned against adding too many features. I agree. The trick is getting it right for the right number of people.

Example: Someone posted they want the duration of an event. I don't want that cluttering up my screen. Next someone will want recurring events. And it becomes a calendar.

Would the duration feature be a deal killer for me? I doubt it (unless I was forced to put it in each time).

I'm not saying they should or shouldn't add duration just because of my preferences - rather I'm pointing out how tricky the choices are in design.

This is complicated by a choice of business model. Right now TR is free so developers are not guided by what sells. They are guided by - what THEY like? What generates the most downloads? The most comments here in the forum? I'm not sure. Maybe they aren't either. You can do that when you're working for free.

Then you have to pay the rent. So you have to abandon efforts, or charge for them. I'm encouraged by that because I want to see the product improved and supported for a long time. How they choose the business model affects their design choices.

If they go with a subscription model, that encourages them to design for more and longer subscriptions (not necessarily add more features). If they go with a sales model, that encourages them to design for continual upgrading. That I think is a major cause of bloat-ware. I think I'd favor a free trial, subscriptions after that and a way to keep using the software (but no upgrades or support) if you stopped paying your subscription. Maybe there's an even better model I haven't considered.

It's a dilemma but one I'd like to see solved in a way that that the product improves, does not get bloated and still brings the developers much success both in happiness and riches. And I want a genie to grant me 3 wishes.

- John Seiffer, Business Coach

Agreed.

What I had in mind was my experience with mlo getting more and more features by the minute and more and more options as well.

Imho Microsoft did a great job with introducing that ribbon system into office 2007. Once I got used to it, I recognized that I am much more at ease using the new office version. Its UI is much more user friendly and intuitive than the one of the previous. I have to use help and F1 much less to find any given feature.

Jeremy and Claire are certainly doing a wonderful job and they have my utmost respect. They don't have the means to do a developing similar to that of MS. Making TR more intutive is maybe an even bigger task than adding new features. And less promising or rewarding as well, because those power users and geeks who are way beyond the average user (I count myself being part of the second group) may make the impression that adding a ton of new features is the way to go.

I'd very much prefer to be offered different ways to do the job. One by clicking on plenty of self explaining buttons and tabs and dropdowns and all those GUI gimmicks and another one doing the same job using the keyboard. Prominent buttons remind me that there is a feature that may be useful to me. Having a keyboard only solution only does not offer the same help and intuitivity. I may even not know that there is this given feature. My personal learning curve is that I always start using the GUI and once I get familiar I start looking for keyboard alternatives to the same job in a quicker and shorter way.

This is by no means an attack on TR being user unfriendly or unintuitve. But every feature that is self explaining and makes me more at ease using a software is a good feature.

Well I hope I didn't offend anyone and was able to make my point of view clear. Others may have other opinions. So be it.

And of course you are absolutely right that it is not me developing TR so I can only offer my thoughts and wishes.

Volker

By far the best feature in the program by my taste is the first page with the GTD schematic you can click through. All other programs I tried I stopped using because it took way too much time to get used to the program, to see how it worked and what worked for me. It took me all of a five minute period beofre i started entering thoughts into it and all of a 30 minute period to know I would keep using it, because of it's simplicity.

I would strongly advice that more possibilities be added in such a way that the simplicity it has now isn't lost.

This can actually be done quite simply by putting more advanced features under an advanced button, so it is invisible.
Or create more schematics like the one on the first page.

Just my 2 cents.

I totally agree with this, if more features are added, it should stay simple or there needs to be an option to enable & disable more advanced features.
Ideally per feature.

What inevitable would lead to an feature overkill in the options section

I thought 2.0 was going to be modular. If you don't want advanced event scheduling, you don't install that module.

I don't think 'modular' is how it has been described, more 'plugable'. As far as I understand the signals so far the new version will aim towards being a platform that can be extended with plugins or modules, but what's in there will be in there. Although Claire mentioned several times that new functionality can be discarded if you like, i.e. if you don't want to use priority (for example) then just don't use it as it won't be mandatory to the overall TR workflow.

I'm sure Claire can elaborate more, but with the new release imminent now (I hope) I think we can just wait to see for ourselves

Most new fields, you can check if you want to use them or not. If you don't, they will not appear on the screen.

Softechmatrix was correct regarding plugable. It is for after 2.0.

_________________
Claire
ThinkingRock Analyst and Tester

Every option needs a decision to be made.
Every activation or deactivation needs a decision to be made.

What is it why people love their ipods and imacs? Because they have plenty of choices which accessories from which producers they can use and how they can tweak their system in endless ways? Or because the use of their devices is intuitive and easy going?

I just read an article in a local business magazine about the differences between the philosophies that divide Microsoft and Apple. Windows is technical and directed toward as much options and tweaks as possible. OSX is (although getting more and more complicated) for the user and the people making things easy going, intended to get as less options and decisions into the way of the user letting him do things easy going. I never used an imac or ibook but my experience with my ipod really makes the wonder if i should quit the Windows world and enter the OSX universe.

I for myself would prefer a tool which guides me and helps me achieving my goals without having to think what I have to do to set it up or keep it running. Example: I sit in my car, turn the key and start driving. I am not interested in looking under the hood, fiddling around with any wires, switches or changing anything, I just want to drive it and have fun driving it.

I hope that Claire and Jeremy will always try to make TR as user friendly and intuitive as possible even if that means, they would sacrifce some additional features for that.


Volker

Of course, your mileage may be total different

Actually, that isn't quite true; but you hear it a lot. It probably comes from the days of the old Mac OS, and since the covering surface of OS X bears a resemblence to the older system, one can be fooled into thinking it is still only surface deep. In actuality, OS X is extremely complicated, and can be tweaked and customised to an incredible degree at every level of depth.

In practice, using OS X feels much, much more simple than using Windows; but that does not mean that it is a simple operating system, or that advanced and specialist features have been deliberately removed for the sake of KISS or some other philosophy. The secret to OS X is that it's surface layer is so deceptively smooth, you need rarely ever know its full depth. Its so good, it evidently fooled the author of the article you read. Make no mistake, in the hands of an expert, OS X is more powerful than a Linux workstation because it can do everything Linux can do, plus everything a Mac can do, plus everything a Windows machine can do (using Parallels).

I didnt think people used pixel perfect collision detection since the early 90s?

And other than that, quadtrees are what you need to get performance when you start getting hundreds of collidable objects per frame.

Cas
_____________________________________
Website I designed for pay day loans edmonton company.

Don’t make a fool of yourself. By that logic, every OS can do everything, since it can run every other OS in a VM.

OS X has its strengths and its weaknesses compared to Linux. But it’s definitely not more powerful.
The only thing that’s true, is that once you have actually used one of them (not just played with colorful clickables (German: “Klickibunti”) , Windows feels like a total joke. Like a toy OS. Compared to them.

The strength of OS X is its GUI, which received lots of polishing and is very consistent because all of it it came from “THE ONE TRUE GOD”.
The weakness of OS X is one, that is just as much true for Windows: Lack of integration because of its closed-source-ness and being commercial.
Which is the main reason I see Linux as more powerful.
Just as Windows users can’t fathom, why OS X or Linux would be more powerful (they don’t get that you can do so much under the hood, or the point of shell scripting), OS X users can’t fathom the power of being able to use all the code that is available freely.
This allows very different projects to use the same libraries or even just rip out code snippets and files to be hacked into the own code.
It’s a very different culture. You just don’t program anything twice, unless you really really disagree with the existing one. (Example: KDE and Gnome: They agree on everything in freedesktop.org, but disagree in their core philosophy [Gnome is infected with KISS, KDE is infected with wanting to be Windows]) But the GUI is not the point of Linux anyway.
Automation of your work, that fits you like a glove, is.
That’s why Linux is not supposed to become the desktop OS for the average user (who for a weird reason prefers appliances and doing the same simple thing over and over again like an idiot). That would be just plain wrong. And it’s good that way.

The depth of integration is the power of Linux.

Luckily, you can at least partially do that on OS X too, since it supports most Linux stuff. It’s just pretty badly integrated into that slick GUI. And hence a bit pointless. But still that does not make OS X that bad.

IMPORTANT: DON’T USE KISS!

I know it sounds right. But the philosophy has a huge problem, that only comes out with time. But many people are too short-sighted (most businesses unfortunately), and will not notice it, before it is too late.

See, the original basic goal is/was, to make it as efficient to use as possible. Meaning that you can do the most with the least effort.
The problem is, that some people oversimplified this, by saying “simple” instead of “efficient”.
The key difference is, that higher efficiency is always better, while higher simplicity can make it worse too.
That’s because you can simplify things, while making them less efficient!

See your users as a Gaussian distribution curve on the intelligence scale.
Now psychologically, the more intelligent people simulate more possible outcomes in a shorter time, and hence are not so sure of themselves. They know in how many ways the could be wrong. While the less intelligent ones usually are very sure of themselves, despite not having any clue at all. They need to, to be able to survive at all. So it’s not anybody’s fault.
But people who are very sure of themselves, also act more confident. Or in other words: They are louder! (Which intelligent people are too, if they are very sure of something even after all the thoughts.)

This results in the developer getting a lot louder and more feedback on the negative side from the lower end of the Gaussian curve.
And here is where the error in most people’s behavior happens: They think that if they just make it simpler, those people will be happy.
But what they don’t see, are the side effects of this:
1. It will become less efficient (=slower, more dumbed down), and hence harder to use, for the upper end of the curve.
2. People who were even below the lower end, now start using it, and complain too. (“Nature just invents better idiots.” )
So in essence, the whole curve moves downwards, and the amount and source of complaining stays the same.

What that results in, in the long downwards spiral, are horrible things like the well-known Clippy, the hated “Assistant” from MS Office and MS Bob. But also things like the iPad not even having a file manager, and Windows hiding all its system folders etc. Which works, until something happens that you did not expect. Then they are completely stuck, since they never understood what they are actually doing.

That doesn’t mean that interfaces like VIM or Emacs are any better. They are just as horrible for “normal people”. Only in the other direction.

The simple thing is, that the interface has to match the users’ experience. And transform itself over time, when the skill of the user grows and decays again. But no matter where in the curve (and even beyond) the user is: The interface should always be the most efficient one possible.
Especially for experienced people (remember: they are underrepresented), who no longer want to click trough a dozen stupid wizards to do, what would only require two shortcuts, filling in a couple of values in a ini text file, and no mouse at all.

The solution is, so have multiple levels of functionality and options. Where there are always 3 levels „active“:
• One level as the current used level.
• One more experienced level, that as only hinted and does not clutter the interface. So the user can pick up on it, in case he needs it. (A good example are: Shortcuts everywhere! Show them in the menus etc!)
• One simpler level, that is visually a bit separated (e.g. more colorful and bigger than the rest) and prominently placed. (A good example are big buttons that run wizards. Another good rule is: Explanatory (longer!) tooltips everywhere!)
But that is only half the deal. What’s required on top, is a mechanism that helps the user, to switch between different levels. Preferably allowing different levels for different sets of tasks that your app offers.

Conclusion: Keep It Efficient, Not Stupid!

P.S.: I’m sorry that this landed in a separate thread previously. Fixed now.