Gettys: Time to lead on the desktop

Something like facebook does not require massive CPU unless you explicitly decide you want a centralized db. There are promising alternative protocols, like FOAF, OpenID, CalDAV, LDAP. We just need to follow through with them more. Where is the simple reliable OpenID implementation for the home user that is not just a sample app for one user? Where are the FOAF apps that allow me to crawl my friends and notice the friends they are linked to and easily add them to my list of friends? How can I get in touch with them using FOAF without them exposing their email to spam? How can I be informed when their address changes? How can I populate my personal LDAP server with my friends info I get from FOAF? How can I enable my friends to update their info in my (and everyone else's) LDAP db themselves easily?

Where are the easy home user, (non-corporate focused) contact (LDAP perhaps) synchronization apps? Why can't I sync my palm pilot/cell phone to an LDAP or other open backend that is not based on a desktop app? Why can't my desktop apps do the same, without using import/export functions which don't really sync?

Now scheduling: why can't I do the same with my calendaring apps? Why can't I have a backend (CalDAV perhaps) which keeps track of my whole family's schedules and syncs (not import/exports) with all of my apps, desktops, phones, and has web access for when I do not have my own phone/desktop with me so that I can use it from an internet cafe? Why can't this calendaring backend federate and conflict resolve my schedules with online calendars from other organizations that I deal with, my soccer team, my wife's baby sitting co-op, and, of course, my work calendar which is on a dreaded outlook server?

Instead, I have a disjoint maze of electronic calendar sources which make using calendar apps a nightmare and traditional paper easier and a relief. On that note, where are the simple calendar apps that can do incremental printing of monthly calendars? In other words, why can't I print out this months calendar so I can hang it on the wall and simply feed the same sheet of paper through my printer whenever it needs to be updated without reprinting everything already on my page or wasting a new sheet of paper?

Simple things that make technology useful, that's where freedom is important and we should lead. Nothing I ask for/describe above is high tech or difficult. It's not some sexy fancy GUI (they help too though), but we do need this integration and it's something that we are often best at in our world. We just need to complete many of the loops that have already been started and we will be infinitely more useful than any online cloud service. Every time we loose to an online cloud service it's because they complete an important link in our loops that we thought was not important, that we undervalued (like keeping track of our friends, not just our coworkers). I mean, look at FOAF, it's a great start, but where are the real tools for it?

Sometimes we undervalue it simply because it is so simple, we simply forget to do it. How many applications that use contacts can access and LDAP db but can't update it (do any)? How many emailers can store contacts, but do not track to see if they've changed with every incoming email (do any)? These are too simple important steps that would have put free software way ahead, but I have yet to see them anywhere!

On a completely non cloud related note, but more related to Getty's article, GUI advancement: it lies not in configurability or more eye candy (which are great also), but in the advancement in keyboard usage!! Yes, that's right, you read that right, make better use of the keyboard. Any command line user knows that the keyboard is much more efficient than a mouse. A mouse is great, but it is just too limiting for so many things. We inherently realize this and, so, of course, we will not give up the keyboard/command line. Sometimes we like to use our keyboard with our GUI, so we add shortcuts to our apps, and these can actually dramatically improve our productivity with GUIs, but only for the most commonly used tasks. Ultimately our GUIs are designed around a mouse with the keyboard added as an afterthought which truly limits their efficiency. Design GUIs with the keyboard in mind first and you will have an efficient GUI, the use of the mouse will just flow from there.

If you can design GUIs for a keyboard then you will revolutionize the desktop. Start with the difficult apps if you have to, take a graphical drawing program such as the gimp or a diagramming program and see if you can make one that is more efficient with a keyboard then with a mouse. If you can, you may be onto something revolutionary.

How many times have you made a diagram easily with a mouse and then just found it too tough to make major adjustments to it later on. If a diagram becomes as easy to adjust as a text file is with a text editor, then you will have improved the GUI dramatically. Imagine you have 10 boxes in a diagram that need to be tweaked. Even with a mouse, current GUIs fall apart with such tasks: double or right click on first object, bring up properties/attributes pop-up, click in appropriate attribute cell, type new value, hit apply to see if it's right, click pop-up close, repeat... Do it once, OK, do it 5 times and curse. Edit five lines in an editor, done. Edit 10 lines in an editor and you probably have an efficient mechanism such as search and replace to do it painlessly! Do this 10 times in the GUI and you are sure to be searching to export your diagram to some text format so that you can edit it with a text editor and then re-import it. If this app were designed for the keyboard it would likely be an easy painless task with a mouse or a keyboard!

Ever try using a web browser without a mouse? Such a basic app cannot handle a keyboard well. It can be done, but it sure is not efficient! No wonder people insist on using a mouse, not because it is better, but because it is catered to by app/GUI toolkit designers. Surely someone can figure out a good way to use a keyboard with GUIs? Maybe it's as simple as starting with a non graphical app that is efficient and adding graphics to it without adding a mouse until it already works well? I am not sure how it can be done, but I am sure that if it is done, that it will lead the desktop.

What a rant! Sorry. :( I hope that my somewhat poor attitude can be overlooked and that my complaints/suggestions may inspire someone to more creativity, or to add that simple hook to complete and important loop. :) As others have pointed out, there is plenty of innovation in free software, we simply don't have a monopoly on it!