Installation on Mac

edit

The "app" version is a simplified way of using Sage. To use it you just click on the application, just like most others on a Mac, and it will start a Sage instance in the background, open a browser window, and launch the Sage Notebook.

The "non-app" version is a directory containing the Sage binaries and necessary libraries. The main binary opens an instance of Terminal and runs Sage in the terminal. It's less user-friendly but it allows you to examine all of the components of Sage and (possibly) develop for Sage as well. (I say "possibly" because I've only developed for Sage via a source installation.)

Now, the "app" version also contains the usual Sage directories, binaries, etc. but they're just slightly harder to find. So basically, the "app" version is the same thing as the "non-app" version except it's more user-friendly.

I'm biased because I wrote the app, but somewhat in response to @niles answer, I would say try it unless you only ever use the commandline version. I'm a guy who spends most of his time in either Emacs or a iterm2, but I really like the app.

You can still run sage from the command line (it will be easier after trac #10556 is merged) and it comes with a nice "menu extra" (like spotlight menu at far right) that I find quite useful. You can use it with any sage tree, so e.g. you can move the sage tree from the app to somewhere else or just use it with a 'non-app' version that you have hanging around.

In short, it's my goal to make sure than everything is as easy with the app version as with the non-app version. Suggestions or complaints are always welcome.

UPDATE: I had tried the app version before, but trying it again recently is starting to change my mind -- it is pretty useful, and gives menu access to all kinds of different functionality (including Sage in the Terminal). My original answer is below.

To add to @cswiercz's answer, I would say that the "app" version is friendly to certain kinds of users. Running sage from the Terminal becomes useful when you're working on a large body of code, and you want to test using the code as you write. It's slightly more responsive than the notebook for things like docstrings and tab-completion, and I find the up-arrow command history very useful. The Terminal version is also nice when you're going to be printing out a lot of data which you want to skim, but not necessarily store.

I would say that if you're comfortable with command-line interfaces, try the "non-app" version. You can still run the notebook (just type notebook() to open it), but when you want to switch to the command line, it will be there waiting for you :)

Also, I've been doing development lately from a binary install without any problems.