first, great article, kurt! very enjoyable read indeed.
i also just realized that in my blog from earlier i forgot to mention a conversation i had with a columbian kde developer who will be teaching a university comp sci course in which they will be doing practical work using kde libraries. he wants to have the 17 students work during the practical section (some 3-4 months) for 8-12 hours a week on projects that can be useful to kde. neat! he asked for some input and help finding projects for the students to work on. i couldn't be happier to do so, really. i'll report more on this when there is more to report on =)
tonight's event that i attended was really great. an amazing open discussion between a number of people who represented a number of different political ideologies in their thinking. democratic discourse at its best.
thought about konqueror some more on the way home. i started with a question: if we took konqueror and created three refined interfaces for: web browsing, file management and content browsing/viewing (think "universal doc viewer + content manager"), what technologies that we currently use would be rendered unecessary, unuseful or even unused?
kparts? no. dcop? no. xmlgui? no. kio? no. in fact, i couldn't think of any. it occured to me that it would actually require refining each of these and granting them a bit more power. up to now we've somewhat randomly used each of them in forming konqueror. to their credit, this actually resulted in something very useful and very powerful.
now imagine if we used these technologies with purpose and clarity. perhaps we may even end up with a proper doc viewer that embeds the view but also sports a nice "edit this document" button prominently in the now cleared up UI so you can switch from reading to editing quickly.
heck, imagine if we had a universal editor that used kparts that were designed around consistent sets of functionality. the user may only perceive half a dozen apps. i think this is where koffice was always trying to head, just as konqueror has been striving towards unversal viewing.
and perhaps that is what matthias was trying to suggest. of course, i'm not in his head so i don't know =) maybe he'll expound more in his next blog.
i still believe in single click however ;)
Subscribe to:
Post Comments (Atom)
1 comment:
This is what I tried to convey to you a month or so back in #plasma. See the OSS model has an issue, they're emulating the islands that are applications. The thing is we don't have to define clearly delineated products, merely functionality, we're not tied to trying to "pitch products".
Instead, KDE should merely be providing interfaces to features that load and unload around a document centric model (this goes back to Xerox PARC). The only time one should choose "apps" (READ: interfaces) is if there is a matter of choice when it comes to interfaces and you can set those choices up in some control panel. Otherwise, I choose my document and go.
Not everything would fit in this model. All data would. The other model you'd need side by side would be a "portal" concept, that could encompass going on the web and IMing and so on.
Post a Comment