@C.E. Algorithmically, I consider this a good approach. A simple test e1==e2 is too cheap to make it considerably faster with parallelization. But actually, the Bag method contains already too much knowledge and care for memory. We know/hope that at least some elements will be removed from the list and therefore, we rather use a dynamically growing list with Bag.

This has it's own (small) overhead. So in this specific case, the dumbest implementation is probably one of the fastest: We create a second array as long as the input list and copy only the elements that are not equal our test-element. In the end, we return only as much elements as we have assigned:

deleteElementDumb = Compile[{{l, _Real, 1}, {elm, _Real, 0}},
  Module[{tmp = l, j = 1},
   Do[
    If[l[[i]] != elm,
     tmp[[j++]] = l[[i]]
     ], {i, Length[l]}
    ];
   tmp[[;; j - 1]]
   ], CompilationTarget -> "C", RuntimeOptions -> "Speed"
  ]

This gives a further speed-up

In[69]:= deleteElement[list, 0.01]; // AbsoluteTiming

Out[69]= {0.256218, Null}

In[71]:= deleteElementDumb[list, 0.01]; // AbsoluteTiming

Out[71]= {0.184316, Null}

Now let me test one last idea.

@halirutan Cool, makes me wonder if these "dumb" implementations are underappreciated in the Mathematica community. In this case it is clearly useful.

@C.E. OK, my last idea is a bit slower. The approach was to not copy number by number in the loop, but rather copy complete chunks that are free of the element. Usually, if the element is sparsely distributed in large lists, this should give an improvement. Unfortunately, it doesn't make up for the additional code that is required to track which portion should be copied.

@C.E. I'm not sure if these core things are underappreciated because usually such questions as the deletion of an element are more rethorical.

As we have seen, we can come up with several high-level implementations basing on set-theory, pattern matching, picking with equality, selecting conditionally, and so on.

Hey @halirutan I'm having some trouble getting the SE Tools uploader to work

more specifically it's not loading things to my clipboard

running over Ubuntu 16.04

Happy to help debug this

(though if it requires sudo access I can't do it on this machine)

@EmilioPisanty Can you tell me the SEUploader version you are using?

@halirutan Installed Version: 129

@EmilioPisanty Have you only "installed" it or did you download the GitHub code and have it on your machine?

@halirutan Originally I installed it by running one line of code on Mathematica, as indicated

I then updated it through the palette, relatively recently

@EmilioPisanty Would you do me a favor and try something? Download the complete zip of the repo

github.com/halirutan/Mathematica-SE-Tools

@halirutan done

@EmilioPisanty Ahh, sorry, you don't even need to :)

@halirutan Then I'm going to need to charge you for that use of the bandwidth =P

@EmilioPisanty Sure :)

OK, go to your $UserBaseDirectory/Applications

there you find the SETools directory

sure

Open SEUploader.m in an editor

done

@EmilioPisanty You need to change this line of code

to this

If[$VersionNumber >= 11,

After saving, I want you to load this .m file in Mathematica by

<< path/to/Applications/SETools/SEUploader.m

SEUploaderPalette[]

You will get a new SEUploader that should contain the changed line. Try it out and tell me if it works.

Import["http://halirutan.github.io/Mathematica-SE-Tools/decode.m"]["http://i.sta‌​ck.imgur.com/6eJZT.png"]

yes

works

@EmilioPisanty OK, when you close this very palette, you will be asked if you want to save it.

You want!

Save it to the SETools directory under FrontEnd/Palettes

overwriting the old one.

@EmilioPisanty I will fix this problem in the online version and publish a patch.

Until then you have a working palette

let's see

bingo

works perfectly, thanks =)

@EmilioPisanty No problem at all.

Have to run. See you all later

@halirutan run safe ;-)

The Tutorial links have been removed from the top of each documentation page. Reference pages still have the list of links at the end, so that is okay. But guide pages do not.

I also do not like that the fonts and the line spacing is too big, and less information fits on my screen. The changes are good (and necessary) for mobile. But they hurt desktop usability. And there is no reason why the advantages of the old design couldn't have been retained on desktop, while also making the pages usable on mobile.

Is there anyone else (other than me and @Nasser) who prefers the v11.0 documentation style to the v11.1 one? (After the freshness wears off.)

@Szabolcs Have you bugged Support about this already?

Yes, I have, multiple times. Actually, in the last prerelease the design was not finished yet, and I didn't expect that there would be no "final quality" prerelease for one last testing round with users ...

I send reports about this, but I think one person is not sufficient. Multiple people need to mention it.

And even then, it may be ascribed to the usual resistance to change that users tend to have.

@Szabolcs On Linux it is just devastating. I also though that the prerelease design is not finished as it wasn't on 10 either. Wanna see something cool, let make a gif.. one sec

Let's see it!

@Szabolcs I proudly present the Mathematica editor I have to work with. Unique features include that the code shakes around due to brace highlighting and some awesome font that was used

@halirutan I wouldn't be able to put up with that ... Is that new in 11.1?

@Szabolcs Yes, it is a new features.

I'm not even sure why this happens. Even if I change the font to Courier this happens.

Is there a function to test the state of a stream? Open or closed.

MemberQ[Streams[]]?

The context is that I am venturing into using CheckAbort and AbortProtect and want to close down the stream if it is still open.

Does anyone have experience with the control system framework in Mathematica? There are no tutorials in the documentation about control system.

I'm confused by the fundamental concepts, such as TransferFunctionModel, StateOutputEstimator, OutputResponse, etc.

We were discussing just the other day how big a difference it makes when a topic has a good tutorial to go with it. The last one that was added was the FEM one, and it is really helpful.

Even image processing has its tutorial, which explains the basic such as coordinate systems. There are some small things missing though. For example, an explanation of the various data types (bit, byte, integer16, real32, real64, etc.) and their differences. E.g. that the real ones do not clip beyond the max value, but the integer ones do.

@Szabolcs Would also be nice if they updated the tutorial to mention the new features in 11.1. Rewriting the example with ImageAdd and ImageMultiply at the very least.