Any extended discussion of this is probably more appropriate for the lounge than for this thread – but I did want to add a brief note here since the preceding postings touched on one of my main worries about Alice – and one that I don’t really see being addresses.
Personally, I often have major problems with
both Alice and with the JRE for non-Alice applications – to the extent that I won’t use either if there is a better alternative. (I no longer have Open Office, a JRE based alternative to Microsoft office, on any of my machines just for that reason and despite the fact that it does some things better than the Microsoft application.)
Similarly, I tend to move out of Alice when things get “difficult” – it’s quite a tribute to the programming skills of the CMU team that a “teaching tool” can be as useful as it is.
My concerns are twofold:
First that students will get the idea that this type of reliability is the norm for true high rel application software. I certainly don’t want to fly in an aircraft where the flight control system was written in Alice. (Actually I am still a bit nervous with the 7xx system which is (or at least was) only quad redundant with two different OS’s operating on the hardware.) Even for games, I think Alice’s level of reliability is “iffy.”
Second, that when students “graduate” from Alice, the next level of software complexity may come as a real – and quite discouraging - shock. Getting students interested through Alice (or Scratch, for younger students) is great
in terms of increasing computer literacy – but it is only a first step towards increasing the supply of really competent computer engineers and programmers.