« first day (1895 days earlier)      last day (2595 days later) » 

7:35 AM
@P.Fonseca Probably Mma kernel could be ported to WebAssembly eventually. I have no idea how large it would be, though...
But persistent/offline storage on modern browsers would probably resolve that issue to a large extent.
 
8:13 AM
Simplify is fickle. -Sin[2 ArcTan[2 Sqrt[7/3] - 5/Sqrt[3]]] // FullSimplify works but Sin[2 ArcTan[2 Sqrt[7/3] - 5/Sqrt[3]]] // FullSimplify does not. The only difference is the minus sign.
A TrigExpand thrown in the middle helps.
 
 
3 hours later…
10:57 AM
Excerpt from a (quickly deleted) Wolfram Community post:
> I was just going to pat you on the back and warn you that stack exchange is the #1 spreader of obstruction and bad advice - i've been kicked off stack exchange for correcting bad advice where i could %100 prove in court I had the right answer.
 
11:12 AM
wat.
 
This deserves more attention:
16
Q: How to estimate geodesics on discrete surfaces?

DunlopContinuing with my interest on curvature of discrete surfaces here and here I would like to also calculate and plot geodesics on discretised (triangulated) surfaces. Basically my long term idea would be to eventually estimate what path a particle would take if it is confined to a surface and move...

 
11:46 AM
@C.E. Can you check if you have a LinkSnooper entry in 11.1 in the kernel config options that you didn't add yourself?
@yode Can you tell me what that LinkSnooper config looks like? I can't tell if what I have is something I added myself previously, or the "official" version from 11.1.
 
 
1 hour later…
12:57 PM
May I have your attention please? As @yode pointed out in a comment to my answer, the LinkSnooper to monitor the datastream between front end and kernel is now available per default from the Evaluation -> (DefaultKernel | Notebooks Kernel | ...) menu.
@Szabolcs Hehe this is nice.. "TED ehmm SET -- Bad advice worth spreading." Hilarious.
I guess another lucky customer.
 
1:14 PM
@kirma Indeed. With the new compiler framework, if it ends up covering the entire language specification (which was still not the case the last time I heard), it would probably be feasible the compilation (in the packaging sense) of "expected" dependencies, eventually with the possibility of having the user add others not possible to identify automatically... But that will be in 2030, when the phone browsers will "eat" 1 GB is a fraction of a second...
 
1:40 PM
@halirutan But it doesn't work on Mac if the app name has a space in it ("Mathematica 11.1"). With the earlier solutions here, I also couldn't handle the spaces. What do you see in your kernel config? Is it the following?
-LinkMode Launch -LinkName {oq}`javaw` -classpath {iq}`jlinkjar`{iq} com.wolfram.jlink.util.LinkSnooper -noinit -kernelname {iq}`mathkernel`{iq}{oq}
 
@Szabolcs I see
 
Perhaps those {iq} and {oq} are substituted by the appropriate "inner" and "outer" quotation marks on each platform.
Wow, you have a weird looking window there
 
@Szabolcs Just made it wide enough to have no linebreak in it
 
@halirutan That looks more like something you might have entered yourself. It has the path hard-coded.
Are you sure this is the "official" version from 11.1?
 
@Szabolcs It's the default Kernel conf for link snooper
Let me look in the init.m
 
1:45 PM
Why does it refer to version 10.3.1 then?
 
@Szabolcs Yep, my mistake.
It is what I have put there.. it's still in the init.m and I thought it is the default that is new in 11
Sorry. But you have an entry, do you?
 
Yes, that's why I was asking these questions earlier. I wasn't sure if the version I saw was the default or something I put there earlier and I forgot about.
But I think the entry I have is the official one.
 
@Szabolcs I renamed my LinkSnooper but didn't get an additional one.
 
2:07 PM
@halirutan Do you have version-specific front end preferences turned on? If not, try to turn it on.
It's under Preferences -> Advanced
 
2:19 PM
@Szabolcs I haven't used LinkSnooper so I haven't added any options myself, I don't know how to find other LinkSnooper options, if they exist.
 
@halirutan I'm here now.
@Szabolcs Like this
 
3:05 PM
I meant the text in your kernel config options.
 
3:27 PM
I did not know that D had special handling of some Inactive heads.
In[11]:= D[
 Inactive[Sum][a[i] x^i, {i, Infinity}],
 x]

Out[11]= Inactive[Sum][i x^(-1 + i) a[i], {i, \[Infinity]}]
I still do not fully understand the intended purpose of Inactive. So it is really for calculations then, not just presentation.
There's a more complicated FourierTransform example in the D docs.
Are there any other operations that treat some Inactive heads specially?
I believe there is NDSolve. Anything else?
There is also DifferenceDelta.
 
3:58 PM
@Szabolcs You miss a comma here
And I don't know how to check my kernel config options..
 
4:52 PM
@Szabolcs The section "NeumannValue and Formal Partial Differential Equations" here explains one use case where Inactive is necessary, not for presentation and not because it is being treated specially.
 
@Szabolcs - do you know of any way to query for package dependencies? Say I've loaded some packages, and they are listed in $Packages, can I find out what contexts were called via Needs for any given member of that list?
 
 
2 hours later…
6:27 PM
@JasonB Not sure I understand. Internal`$PackageDependencies?
@yode Thank you!
 
7:05 PM
@Szabolcs - interesting, that does seem to capture the dependencies that are declared inside BeginPackage[_,{<second argument>}], but not those that are loaded via a Needs statement inside the package. I might resort to reading in the packages as text files and doing a StringCases
 
 
5 hours later…
11:44 PM
Just procrastinating a bit and watching some "ignorant America" clips on youtube where they question people on basic stuff. Beside that it is hilarious, the coolest question was "you're in a race and you overtake the second person, what position are you in?".
Just watch this little girl in the middle. She is awesome. About half a dozen grown-ups around making fools of themselves and she nailed it after not even a second but is too shy to say it out loud.
 

« first day (1895 days earlier)      last day (2595 days later) »