Linked by Kostis Kapelonis on Mon 29th Nov 2004 18:55 UTC
Editorial The IT sector today is a complete mess. The end-users rarely understand this, but most insiders reach a point when they realize that things should be different. The problems are numerous but they all reduce to a basic principle. IT and consumer electronics companies are interested more about money than helping people solve their problems. Of course companies need to make a profit and nobody denies that. They should however make money by helping people and not by creating more problems for them.
Permalink for comment
To read all comments associated with this story, please click here.
Bad examples
by drsmithy on Mon 29th Nov 2004 21:58 UTC

Even simple actions like opening a document in an application are more complex than they seem. One has to remember the exact location of the file (its position in the file system) in order to retrieve it. Search functions and "recently opened" lists solve partially this, but the problem never disappears.

Of course it doesn't, because the "problem" is inherent to the whole concept of storing things for future retrieval (be they computer information or physical items) - they have to be put somewhere.

Search functions _are_ a solution, because they make the location operation - inherent to the concept of store and retrieve - quicker (do you have a search function to help you find that book you left lying somewhere around the house ?).

In most cases however, the images will have cryptic names like DSC06458.JPG and the directories will show just the model of the camera. The user ends up manually searching all the images previewing them in one of the zillions programs that exist for this kind of boring operation.

This is why things like "thumbnail view" were put into file managers back in the mid 90s (if not earlier).

You're not identifying a "problem", you're describing aspects of the whole concept of storing and retrieving things. Moreover, you're not even offering possible solutions at a conceptual level, let alone a technical one.

The whole concept of different programs [...]

The concept of "programs" driving the UI was largely eliminated back in the mid 80s with the Mac and has been being refined since. That some (most) people insist on keeping it alive is not the fault of the computer.

MacOS, Windows and most others have had interfaces that stressed the concept of primarily operating on "documents" and "objects" (rather than opening programs and then "loading" files) for quite some time now. Obviously in some contexts this analogy can't work (eg: Calculators) but in most of the cases it can, it does.

[...] and windows that need to be resized and moved (isn't this the job of the window manager?).

No. Windows can be resized and moved by the user because that is a *good* thing. There might be an argument that window sizing and placement algorithms could use some improvement, but the concept of being allowed to move and resize windows is quite sound.

Yes I know that windows XP does this but still the process takes two steps (1.drop files 2.write CD). This is something completely strange for the naive user.

It's all in the explanation. The two step process is attributable to the physical limitations of the technology (CDs are WORM devices). If you explain to the end user that CDs can only be written to once [0] and that because of this they must first "copy" the files to CD and then "confirm" (or "commit") them, the confusion largely disappears.

In short, no amount of UI improvement or "automation" can circumvent physical limitations.

Why have a defragging application on its own? Why doesn't the computer defrag itself when idle? It is not a technical problem (see most UNIX filesystems).

Firstly, unix filesystems don't defragment themselves.
Secondly, for almost all users (certainly the ones being identified by this article) defragging is not something they have to worry about. Fragmentation is a vastly overblown "problem" that only has a meaningful impact in a very small list of corner cases.
Thirdly, defragging programs (and procedures, for those platforms that lack specific software tools) exist because in some situations it *is* useful to be able to do.

Again - like your "programs" distinction - this falls into the category of people making things more complex than they need to be, generally because they either don't know better, or can't be bothered learning.

For example, the new Start Menu in Windows XP is a vast improvement on the old one in terms of functionality and efficiency. Yet the number of *experienced users* who immediately change it back to the old way because the new way sucks [1] is enormous.

Automation is not something exotic or new. If you spend some time calculating how much time you spend in front of the computer doing actual work and how much time you spend on unneeded management, administration and maintenance you will be surprised.

The level of automation in OSes like OS X and Windows is _massive_. In many cases when people like you are complaining about things that should be automated, they often already are and you just don't want to use the new methods.

Why do we have a save function in Word 2003?

Because there needs to be a way for the end user to tell the computer to commit their changes.

The same function existed 9 years ago in Word 6.

Your reasoning is stupid. The Bold and Italics features were in Word 9 years ago as well, do they need to be removed as well ?

Saving can be automatic. I am not talking about partial solutions (like Vi and Emacs do) which protect the user from losing work. I am talking about the whole idea of saving. Why torture the user with the save function at all? The application should save the document at all times keeping different versions and revisions. The whole .doc file should contain all user actions on the document (think CVS in a single file).

Word already does this. Revision tracking is only enabled by default in Word 2003, but the facility has been around for a decade or more.

The big UI problem with your scheme is that it becomes frustrating to deal with "confirming" all those changes.

Opening the file would be a simple question. Open the latest version or edit the version of a specific date/time.

For someone who professes not to like the separate application paradigm, you seem to be advocating it here.

The Word application should not have a save menu/button anywhere on the interface. The user doesn't care about this.

The purpose of the "Save" function is to tell the machine "all the changes I have made up until this point I am happy with, please keep them".

(Ok, ok maybe a "save as... which just relocates the document file but you get the idea).

No. The "Save As" functionality should *NOT* be in programs at all. "Save As" is a *file management* operation and should be performed from the shell.

Why do I have to buy a "4 in 1" card reader? Why?

Because we operate (ostensibly) in a "free market". That means there isn't some central authority deciding which technology should be used. Most people consider this to be a good thing.

I wonder if you are similarly annoyed at being able to pick from a plethora of OSes to put on your computer as well ?

The same approach can be applied to user interface, consumer electronics,compression algorithms, image formats (do we need all of them?) and most other areas of IT and computing in general.Think the user first! I could write a load more about autonomous computing, the amount of money companies get from technical support or even several other interesting ideas that have recently appeared (ratpoison, ion and friends) but this is just a simple article and nothing more. Food for your brain....

I think you need to work through some of your examples and ideas first. In particular, UI paradigms with regards to "applications" and the problems of getting people to actually use the functionality they already have.

[0] Yes, I know about multisession disks. No, I don't think it's relevant to this context.

[1] 9/10 times people can't say why it sucks, they just say it does.