12:34 AM
2 hours later…
2:20 AM
3:01 AM
Oh no yeah no doubt @b3m2a1, I had originally written “could you implement a”, and changed it from there—is it bad in your opinion to just quiet the messages somehow? Is it just parsing across the interval so quickly that it prints out enough outputs from
Part
that it ends up crashing? Or are the messages themselves containing too much data? They should autoiconize that sort of thing imo.
@CATrevillian it's unclear what is breaking the kernel, but it's clearly a lack of good unit testing or the like. These are pretty random, unexpected messages so the breakage can't be predicted. If you ask me WRI made a mistake a long time back by trying to encode data into messages rather than providing a general exception handling framework that would encode the data. Now that they kind of have that with
Failure
they should actually make use of it.
The old
Message
paradigm is clumsy, as is much of the system. Returning unevaluated is a terrible way to show an error happened. Robust, common Failure
s are the way to go, but as with much of Mathematica I think WRI is too committed to their decisions for it even to make sense to advocate for that even if I had a channel to advocate on.
16 hours later…
6:46 PM
0
I have a more general question regarding StackExchange (Sorry if this is not the right place to post such a "meta-question"): When I use a code posted here for, lets say, doing some research and I publish this research how do I give appropriate credit (attribution) for using this code? see here:...
7:02 PM
Dang @b3m2a1 ...I definitely enjoy using Mathematica, and I will for a long time, and I know you do too, but that doesn't make what you laid out here less true. I just can only really do what I want to do with Mathematica, but that may just be my lack of learning of other languages? Not entirely sure.
I'm currently struggling with something super basic, and it is rather frustrating, though I think I just have to use a bunch of messes of Tables, it's quite simple really, but the method escapes me....I've got sets of eigenvectors for a bunch of indices of an iteration parameter--its
Dimensions
are `{40,100001,40}, and I can easily split it into the upperhalf of the eigenvectors, that's just a choice of parts....but then I can't for the life of me figure out how to retrieve from each of the
7:34 PM
@CATrevillian What about
set[[All,All,;; ;;2]]
? The problem with separate Part
accesses (as in set[[21]][[1]][[;; ;;2]]
) is that they only work if every step removes the top dimension. If you use All
or any other specification that gives you multiple items, the depth of your structure stays the same, which means the next Part
access is applied to the wrong level
7:53 PM
8:26 PM
The following seems to be crashing the kernel for me, can anyone confirm this? (I'm aware that the syntax is wrong, but I'd expect a warning, not a silent death)
2 hours later…
10:20 PM
@LukasLang I think I might dance a happy jig!!!!!!!!! You rock, and I learned something, thank you immensely.
@LukasLang could it be that the
MakeExpression
is accessing some more core FE components and overloading them?
« first day (2771 days earlier) ← previous day next day → last day (1681 days later) »