I'm concerned too. I think WRI is mashing and corrupting one of their primary assets, a clean notebook interface.
The principle here is that a user should be able to obtain help WHEN AND WHERE HE ASKS FOR IT and not otherwise.
The primary point in the interface is the new cell insertion point. This should be completely unadorned. There should be nothing there until the writer asks for it or types something. Similarly as a user types there should be nothing else unless and until he asks for it. The writer should not be constantly presented with "in your face" requests for which the answers are overwhelmingly no.
WolframAlpha is a nice feature but I think it would be only occasionally used in most notebooks. It can be obtained by using "=" as a shortcut, or by typing WolframAlpha[...]. I don't see why it couldn't be added to the context menu. We don't need a constant little advertisement for it at the primary point of interface. It's like walking into a store when you know exactly what you want but first having to fend off a sales person offering you various items of information or specials and reappearing before each new purchase.
The Input Assistant at first seems like a good idea but it has a way of turning into a huge dynamic distraction. I also find some difficulty in completing the symbol entry, sometimes getting the wrong symbol and sometimes having to grab the mouse and click in the notebook. You may complete the command you want but the list will still be there needing to be dealt with. Also, in the instructions when WRI says "Tab" I'm not certain if they mean the arrow keys instead. Also the command completion sometimes does not appear to work and other times does. The entire thing is a bit more complex than is convenient. I think the pre-Mathematica 9 method was better. It only appeared by request, using Ctrl+K, and then often completed at that point. If M9 gives a better context sensitive list of completions then that could still be incorporated.
Templates should be obtained only if one uses Ctrl+Shift+K. It is just as easy to type that as it is to grab the mouse and click a temporary button. It appears now that in M9 WRI is obtaining the templates from the multiple entries in the Function page Usage cell, with perhaps some method to add additional ones. But they use this only in their in-house Build process and don't yet make it available to developers. WRI really has to work on updating Workbench. I don't think it's been updated for over two years.
Many of these things, including the Predictive interface, might be better if they were on a context menu, always available but otherwise invisible. Or they might be buttons in the docked Toolbar at the top of the notebook. There seems to be plenty of room there and they might be more useful than the justification buttons. One of the problems with Mathematica is the great number of commands. Rather than writing code, I would think a more useful Predictive interface feature would be to give direct links to Guide pages that had routines related to the context of one's work.
Keep the notebook interface clean, like a blank piece of paper, and leave beginners` aids to tutorials, context menus, optional Toolbars or palettes.
So much about the (rare?) blue screens. Now I am more concerned about the route Wolfram indicates with the new Mathematica version. Each previous version of Mathematica has represented important steps forward, improving well-known features and adding others. Even those versions that gave us loads of work to do, rewriting older notebooks, have been appreciated, increasing the net benefit of the software. I am not equally enthusiastic about version 9 (up to now). Does it represent a turning point in the development of the Mathematica software? It seems to me that the integration of the Wolfram-Alpha ideas into the Mathematica environment is changing the simple and attractive structure of Mathematica. Do we have to move into the Math kernel shell to get back to the original Mathematica experience? I hope you will maintain Mathematica as the unique tool it was and still is and let Wolfram-Alpha develop alongside, maybe also as a first Mathematica experience. Co-existence instead of a pre dator-prey relation.
Best regards, Arne
On Monday, 3 December 2012 09:30:48 UTC+1, John Fultz wrote: > Some users who upgraded to Mathematica 9 for Windows have reported > > encountering blue-screen system crashes. The system crashes seem to > be > > caused by a Windows bug in handling how we update fonts, and can > > potentially happen even when not running Mathematica (for example, > > opening the Fonts control panel might cause a blue screen). The issue > > only affects Windows machines which had older versions of Mathematica > > installed. > > > > The issue is rare and was not found during extensive internal and > > external prerelease testing. However, we're taking the issue very > > seriously, and we recommend that you don't install Mathematica 9 for > > Windows until we can provide you with a patched installer, which > should > > be available for download from http://user.wolfram.com in the next few > > days. > > > > If you have already installed Mathematica 9 for Windows and have not > > encountered this issue there is nothing you need to do. > > > > If you have installed Mathematica 9 for Windows and have encountered > the > > issue, you can find a solution for the problem here: > > > > http://support.wolfram.com/kb/11160 > > > > Or, feel free to contact Technical Support and they can help you walk > > through deploying the fix. > > > > We apologize for any inconvenience this may cause you. > > > > Sincerely, > > > > John Fultz, Director of User Interface Technology > > Arnoud Buzing, Director of Quality Assurance > > Wolfram Research, Inc.