@Szabolcs I'm curious if something like this can exist at all. On Linux, the $BaseDirectory is per default a place a normal user has no access to. The minimum that is required is that Mathematica pops up a dialog asking for the sudo password. That, on the other hand, is something I have never seen so far.
So simply Blocking something as @b3m2a1 suggests is definitely not going to work on Linux. I just don't have write permissions in /usr/local/... as normal user.
@b3m2a1 From everything I've seen so far, even WRI never install a paclet in a different root than my home-directory.
The only process where they install something for "all users" is when you install Mathematica itself. And there, you do it from command line with root-access granted.
Other than that, blocking $userRepositoryDir seems the right choice.
Would it be useful to people to have a simple system to build out GitHub pages? Since it turns out we can uses GitHub as a paclet server and I already have both packages for building static sites and working with the GitHub API I was realizing it'd be easy enough to make a pages-integrated site / server builder. I'm not sure if I'd use it myself, though, so I'd want to know it'd be useful / interesting to others before sinking the half-hour or so into it.
On another note, it's fun to see just how much they've packed into "GeneralUtilities`". I auto-built (slash am building) some docs for convenience here. Most of it's simple of course, but it's still cool to see it all there.
@halirutan That's solvable, e.g. temporarily run Mathematica as root. This is a one-time installation anyway, done by a sysadmin.
@halirutan Possibly the best approach is to write a sysadmin guide telling them to rename the .paclet to .zip and simply extract it into $BaseDirectory/Applications, i.e. not install as paclet.
Background
DynamicLocation can be very useful:
LocatorPane[Dynamic@x,
Graphics[
{ EdgeForm @ Thick, FaceForm @ None, Rectangle[BoxID -> "box"]
, Arrow[{Dynamic[x], DynamicLocation["box", Automatic]}]
}
, PlotRange -> 2
]
]
Question
But what arguments does it accept and ...
@Kuba Did you see the GitHub issues I opened for MPM? Any comments? Accepting pull requests?
@Kuba I can send fixed for three issues: 1. M version check 2. M10.0 compatibility 3. Support installing prereleases, but for this we must agree on a syntax first, e.g. what option name to use to turn this on
I need an easy way to install packages. I strongly prefer a small and focused installer that can be loaded directly form GitHub, and installs directly form published releases. I want to use this in installation instructions for packages.
I'm aware of this WL tutorial for programmers and the equivalent for math students; but I haven't seen any for science students. Suggestions?
I'm looking for other material to complement what I've put together for my students, in the (likely?) event that I make assumptions they are too afraid to admit are over their heads.
@bobthechemist Even though it says for "math students" I think it is just as good for students who do math, e.g. science students.
Or I would maybe even say that someone studying e.g. physics is a math student (student of math, a person taking courses in math). I don't see the need to make a distinction even at the semantic level, even though "math student" is more commonly used for someone having this as his major.
But to answer the actual question; I haven't seen any either.
@C.E. I agree - it's satisfactory for getting up to speed on the interface and breadth of features. I envision a science-focused tutorial as one that is more data driven, and in my particular case, I'd like to have something that walks students through descriptive/comparative/modeling statistics since those are the topics my students tend to spend the most time on.
I'm trying to ignore all terms that are higher than second order in any combination of elements of some array b. So far, the size of b is small, so I've naively been using someExpression /. {b[_]^n_ /; n > 2 -> 0, b[x_]^m_ b[y_]^m_ /; m + n > 2 -> 0, b[x_] b[y_]^n_ /; n > 1 -> 0}, etc., but this is obviously pretty unweildy
are there any ways to match on the total combined power of multiples of elements of b?