« first day (2624 days earlier)      last day (1861 days later) » 

2:14 AM
@Kuba I can't reproduce it anymore -___- but I think it was basically that I had accidentally written Cell[StyleData[StyleDefinitions->"StylesheetFormatting.nb"]] to a notebook with StyleDefinitions->"PrivateStylsheetFormatting.nb" set amongst other things.
@Kuba do you think it'd be faster to reimplement FrontEndTokenExecute[nb, "Save", {fileName, "Package"}] as a Kernel function that'd have to call into FrontEndExecute@ExportPacket[..., "InputText"]?
Currently saving my files before switching tabs is my biggest performance lag in this little IDE
Well along with the 5000 Dynamic objects included in the FileViewer (but those are only slow on the first render)
 
3:08 AM
@Kuba should work now
Here's a preview of what the IDE looks like right now:
You can see the tab lag is the worst part
Oh somehow I also forgot to show a regular notebook and a text file there... it deals well with both of those and changes up the stylesheet for them automatically.
It's currently only a light-mode thing but I designed everything so when I have the time I can easily make a dark-mode stylesheet to plug in where desired
 
 
6 hours later…
9:37 AM
@Szabolcs Nice job with the BoolEval package! What did you use to make the docs?
 
10:14 AM
Time for a competing BooleVal package to make people confused...
 
10:47 AM
@ChrisK Workbench + post-processing. I do not trust the doc tools enough to use them for a large package, but this one had only 3 functions.
 
@Szabolcs Thanks, I'll have a look at you post-processor later.
One little suggestion for the docs: all of the examples in the tutorial work on only one array - showing off the fact that you can use conditions on one array to get parts of another would be nice here
 
@ChrisK It's kind of a mess. If you are interested in what it does or how it works, better ask me. It is based on a personal package that will only work on Mac (without modifications), and is a bit messy.
 
@Szabolcs Will do!
 
 
1 hour later…
12:19 PM
@Szabolcs Okay, just tested it on my friend's installation of 11.2; it looks okay.
Also had a look at your code; for greater flexibility/generality with conditions (so it can handle e.g. Implies[], Majority[], Equivalent[], etc.), you can preprocess with BooleanConvert[] (and prolly BooleanMinimize[] for simplification) and then the rest of your rules should be able to handle the result.
 
 
2 hours later…
2:03 PM
@Szabolcs I see that your post-processor removes the erroneous URLs at the top of every page; could it also automatically open the "Details" section?
 
2:25 PM
Hi, I'm having trouble with the updating of variables inside a while loop. Mathematica claims the variables are global, but the updating does not seem to be.
I don't think I can avoid while, since I'm trying to optimize a functional such that some equality is satisfied
Any hints/
My loop looks like this
While[Abs[Subscript[\[CapitalDelta], \[CapitalNu]]] > \[Epsilon],
Subscript[\[Lambda], F] += Subscript[\[CapitalDelta], F];
Subscript[\[Rho], Th] =
FredholmKind2[{-Subscript[\[Lambda], F], Subscript[\[Lambda], F],
eigenval, Kpart, Gpart}, intervals, Method -> NIntegrate];
Subscript[\[CapitalDelta], \[CapitalNu]] = \[ScriptCapitalD] -
NIntegrate[
Subscript[\[Rho], Th][\[Mu]], {\[Mu], -Subscript[\[Lambda], F],
Subscript[\[Lambda], F]}]
]
And I've previously run the calculation inside while
Evaluating the [CapitalDelta]_[CapitalNu] gives the same value as before while, so I suspect that the updated variables does not survive the closing of the while loop
 
@ChrisK Yes, in principle it could.
 
 
2 hours later…
4:51 PM
default = <|  "hts" -> <|"nb" -> <|"a" -> 1, "b" -> 2|>,   "nb2" -> <|"a" -> 1, "b" -> 2|>|>   |>


asso = default;
asso[["hts", All, "new"]] =  asso // Query["hts", All /* Values, #a + #b &];
asso

asso = default;
asso["hts"] = <|#, "new" -> (#a + #b)|> & /@ asso["hts"];
asso

asso = default;
asso = MapAt[  <|#, "new" -> (#a + #b)|> &,  asso,  {"hts", All}  ]
asso
Here are 3 ways to add a key to a slice of association. Any comments on which is better in terms of speed, flexibility etc?
Or maybe there is something better? I hope so because I find them ugly. The best is the 3rd one but I am afraid it can be too expensive, right?
 
 
1 hour later…
5:56 PM
fyi, an interesting interview with Wolfram at lifehacker.com/how-to-think-like-stephen-wolfram-1833462019 published today
 
6:56 PM
@Nasser I'm amused by the "some of our websites mysteriously lost their CSS" line because I reported that bug like 8 times to them all to get the response: "there's no bug you're crazy" and then it happened on a livestream once and then the next time I reported it they said "we're already tracking that one"
 
7:09 PM
For those who are making publication figures around this time and already have MaTeX installed: Any feedback on this (old) content ("Preparing Figures to Size") is welcome: mathematica.stackexchange.com/a/193996/12
 
7:40 PM
Reflection and Anamorphosis in a Spherical Mirror
https://community.wolfram.com/groups/-/m/t/1640552
 
@VitaliyKaurov Very cool. Those without the means to 3D print could render it with Blender or another reflection-capable raytracer.
 
8:05 PM
@Szabolcs nice idea :)
 

« first day (2624 days earlier)      last day (1861 days later) »