Does anyone have ideas on how to solve/simplify Reduce[(1 + k)^k (1 + k)^(1 + k)^(3 + k) == (1 + k) (8 k + 1)^k, k, Modulus -> 32768] ?

I feel like this should be doable (solving for k), even by brute force, but the numbers are just too large to execute in reasonable time

Hmm interesting discussion going on over on meta about licensing SE code:

@SjoerdC.deVries likewise!

@blochwave I was about to post the same link. I have no problem with it (my code is already available under the MIT licence [not license...]) but clearly most people who have an opinion about it seem to object strongly

you can understand why. I certainly would not be happy if I hadn't already made this decision for myself

@OleksandrR. indeed - especially with only 2 weeks to consider the change!

@jtbandes (1) I get Overflow[], which suggest the numbers are too large for Mathematica. (2) Exponential functions f[k_] := a^k in general are not well-defined mod n. E.g. Mod[(1 + k)^(3 + k), 32768] /. k -> {1, 1 + 32768}.

Hey, 10.3.0 is now available from my IT dept!!! Oh wait...Wolfram is up to 10.3.1...Rrrrr. :/ ("I don't understand why we don't do this through the user portal." "Power! Control!")

@OleksandrR. Waiting for my chocolate bar to cool! We made dark chocolate so no worries about the Hershey process. Had to grind the roasted beans by hand though so will be a bit grainy.

@MichaelHale Felchlin makes an unconched chocolate Centenario Crudo (felchlin.com/en/product/cacao-centenario) that has a wonderful aroma and flavor. It's gritty but good. We used it to make chocolate spice cookies for office xmas gifts (we were out of the smooth dark chocolate). The chocolate, almonds, spices are ground in a food processor and bound with eggs. This chocolate made noticeably better cookies.

@MichaelE2 Thanks, I'll check that out. I've found the grittiness has increased my appreciation of the raw bean flavor. Also got to suck on a fresh bean straight from the fruit. The coating tastes a bit like skittles.

10.3.1 seems to be available for download. Has it been there for long?

Once again, a real changelog would be fancy.

Neither of bugs I have reported on 10.3 are fixed, at least. One is a kernel hang, another a clearly incorrect result from RegionBounds.

This one should be obvious, but I'm too tired. Anyone?

The compositional one takes 3x the time of the other

@Dr.belisarius If I replace k by kd and ti by kc, I get similar timings. V10.2, Mac.

@MichaelE2 Hi!

Nope, the same with AbsoluteTiming ...and sorry for the naming mixup

In fact the difference is visible with "the naked watch"

Strange. Windoze and V9 here

@Dr.belisarius I still have a V9...let me check

@MichaelE2 OK!

@Dr.belisarius 2.29 times slower with Composition.

@MichaelE2 Ha!

No so obvious then

@Dr.belisarius Try it with On["Packing"]

Probably a bug someone caught...

@MichaelE2 Most probably, yes. Thanks a lot for your help!

@Dr.belisarius De nada.

(Of course, I learned that on american tv...)

@MichaelE2 He! My English comes from there too :=)

Has anyone else seen their PC appear as a "4-node homogeneous cluster" before?

I mean, it's just basic Windows...and I have an i5 processor, yet it shows up much better than the i7...?

I'm very confused

@kirma The RegionBounds bug is on my radar. The poor result comes from a numerical optimization method where NMaximize is getting stuck at a local maximum. Developers are looking into it and hopefully will make improvements, but in general always finding the global maximum is not guaranteed.

@ilian Nice to hear! :)

It being on your radar part, I mean. Of course the problem itself is one that can turn nasty quite quickly.

The primary concern for me was that it returned insufficient bounding box - instead of just being smart enough to return unevaluated, or whatever would be the natural approach to indicating Mma wasn't capable of acquiring an useful result.

Hi @kirma and all!

Hi!

Hi!

Do you happen to know how to use SystemOpen to edit the file in the system-default application instead of opening it with?

Hi @Szabolcs

Is there a difference between editing and opening?

Yes, I think. I remember seeing something in the Windows file-association manager where you can set things like: Open, Edit, Open with... to be the default applications.

Ok, maybe not.

But then how can I force SystemOpen to open the file in a certain system-default application. Like when you put mailto: before a mail address, it knows how to start up your client.

I hoped that there is an edit: or similar prefix for such things, but no.

Doesn't it open it in the system default application already? On OS X it seems to work.

If I SystemOpen a .txt, it uses the default text editor, if I SystemOpen an .xls, it uses Excel.

But the problem is, that an image opens in the WIndows Image Viewer whereas I want to send it to e.g. Photoshop

It even honours file-unique associations (on OS X it is possible to set an association for a single file).

I can't try on Windows right now, unfortunately.

@IstvánZachar Does Run["c:\\path\\to\\photoshop\\Photoshop.exe myimage.jpg"] work? Just pass it as a command line argument. That's what Explorer does when we double click a file, I think.

Yes, it will but my problem is, that parhaps the end-user doesn't have Photoshop, so I have to rely on the system default app for editing images (if there is any).

As there is an "View image" and "Edit image" in your OS shell context menu with set defaults, I want to use those from Mathematica.

Maybe this functionality is only existing in my mind...

@IstvánZachar Ah, I understand now. My wife is using her computer now so I can't use Windows at the moment. Is there always an Edit option in the context menu for images? Or only sometimes? Could it be possible to find what program the edit option launches through some registry keys? The registry can be accessed with some Developer context functions.

I'm using OS X now, I don't think there's any distinction between edit and open on OS X.

@OleksandrR. @MichaelE2

@Szabolcs Thansk for looking into it! I gave up for the moment, simply opening is enough for now.

Hi @Szabolcs

@MichaelHale looks great! What went into it apart from cocoa mass and cocoa butter? Some sugar and perhaps milk/cream?

@OleksandrR. Just the cocoa beans and butter, then sugar to taste. Condensed milk optional.

@MichaelE2 I had no idea one can make unconched chocolate and have it hold together. Does it taste a lot different? Less creamy I guess?

@MichaelE2 same problem for me. In fact WRI wouldn't even accept my bug report because the licence is not in my name...

@blochwave this will happen when you have the Parallel` package loaded and some kernels started. The benchmark is different and the way the score is added up differs as well. (Versus the standard benchmark.)

@OleksandrR. Ha! Catch-22 bug reporting procedure

@ilian some improved differential evolution mutators would go a long way toward helping that. :)

@ilian if you care a lot about this then see my Python differential evolution here. Some of the methods are much more effective than the classical rand/1/bin (and thus it is much more effective than NMinimize on the same problems)

MDE5 is very effective a lot of the time.

@OleksandrR. Lindt is the one who discovered conching, which results in a very smooth chocolate. Felchlin also adds sugar crystals after the basic grinding of the beans which contributes to the graininess. I suspect that the sugar is added is to control/mask any unpleasantness in the crystal structure of the chocolate and the larger bits of cacao nibs that remain. The sugar grains, which are tiny, are easily crushed or dissolved, but I do notice some little bits of nibs. About the flavor...

@OleksandrR. does that load up automatically? Because I didn't do anything...and it didn't happen before my upgrade!

How weird.

@blochwave shouldn't do, but who knows

@OleksandrR. Chocolates can have a wide range of subtle aromas and flavors, that come from the variety/varieties of cacao beans, the fermentation process, etc. The language for these, like that for wine, is a bit beyond me. It does not have that satisfying mouth-feel of regular chocolate, but it seems more fresh/fruity/floral than most chocolate. I quite like it, but sometimes I just have to have the smooth kind. I haven't tried making a ganache or mousse with it. I tend to think it would fail.

@MichaelE2 I think the Lindt 99% and 100% cacao chocolates (which are quite nice) would probably be impossible without conching

The mysteries of Mathematica strike again.

@MichaelE2 I'll make a point to see if I can find unconched chocolate

@OleksandrR. Those very dark chocolates sometimes have a similar flavor profile, imo, but the one I'm talking about is 70%. It's the only unconched one I know about, since the conched, smooth kind has become the expected product and is very popular. The smoother, the better the quality, in the opinion of most.