Konqueror going the way of the Dodo ?
I’ve read ettrich’s rant about possible improvments to KDE 4 over the current status.
Some things stated are very important, and some I think are way overrated and are mostly just “usability hype”. One of the issues for which ettrich’s got the most flak was his rants about “embedding is the Wrong Thing(tm)”.
I understand where people are coming from. The reason why KDE is currently the number 1 desktop environment in Linux is not (only) because its the closest thing in Linux to the MS-Windows explorer shell (look-wise), but mostly because it gives people – both end-users, power users and developers – the power of integration. Embedding is only part of it (I love the ability to browse the web and then click on a link to a PDF and view the PDF in the same windows – a natural extension to the browsing experience. The Mozilla/IE way of either launching an external application which takes years to load or embedding an additional interface inside the browser view port is annoying), but integration is more important – I can drag stuff from a web browser to my desktop, to a network share, to an email composer, to an editor, to the address book, to a media player and it does the right thing every time.
The problem is that this behavior isn’t pollished enough and new users can get confused by the enourmous amount of options they are offered. I’m personally not in favour of the current trend of truncating RMB menu enteries: when I right click a link in the web browser (Konqueror), I get the options of “sending to bluetooth”, “compress as
When things like that are being discussed, I’m reminded of the “bug” opened in the KDE Bugzilla about redundant enteries in KHTML RMB menu. One of the enteries that were quickly discarded (thankfully to be reinstated not much later) was the “Set character encoding..” menu. This item is a must for non-native english speakers who use langauge native pages on the web. Yet it was quickly discarded because “it clutters the menu”. We have to be extra careful not to cut away functionality that people actually use. Even if that means that 90% of the users only use 10% of the features (as the famous gurnisht law clearly and wrongly states), we don’t want to make the other 10% leave us because they can’t get any work done.
All in all, I’m confident that the KDE developers will know how to maintain the balance, and am not too worried. And if things break – we will always have Bugzilla 🙂
It is my suspicion that people such as yourself who want all of those options constantly available are in the minority. What is really needed is an easy way to edit the right-click context menu. Then we could have a default, uncluttered menu of commonly-used options, while power users such as yourself could add whatever additional options you wanted. This is fairly common-sense; it’s how most toolbars work, actually. It’s just that the right-click menu as a primary interface is only just reaching maturity.
As for Konqueror integration, I’d favor a two-toolbar approach: One constant and one contextual. Something akin to the “two-interface” approach you see with embedding in web browsers, except far better integrated.
a) Only if you consider native iso-8859-1(15) “speakers” people a majority – every one else require these additional interfaces. I hate to go that way, but fact is that Internet Explorer still have these features in its RMB menu – they target more then just North America/Wetern Europe, and I believe they know their target market.
There are other RMB menu features I enjoy – and that I discussed above – that may not be used by a lot of people at all, maybe most of the people, but still – If they weren’t in the RMB menu then there would have been no way to get at them.
I don’t think that editing the RMB menu is a good choice – its far too complicated and most likely more complicated then the people, who use these features, can handle – hence you’d be effectivly removing them from the menu.
Possibly we can have a collapsing menu, similar to what Windows 2000 pioneered, where the less used (or more advanced) options are hidden behind a “collapsed” marker. When you hover or click the expander, the entire menu comes into view – while you can still clearly see which enteries will be hidden next time. The interface can also be intelligent and remember often clicked enteries and bring them out of the collapsed collection.
b) I though this is what Konqueror currently does ? play around with KParts and you can see how different options are added and removed from the toolbars as the views change. You can edit the toolbar and check the “merge” entry. If you add another toolbar and put the merge entry there (and remove it from everywhere else), you’d effectivly created your “tow toolbar approach”. I personally don’t like it – I think that having more then one toolbar really hurts usability and we shouldn’t go that way.