« first day (325 days earlier)      last day (4165 days later) » 

Szabolcs
Szabolcs
12:34 AM
There is a reference to NDSolve`FEM` in the installation dir, and there are a larger number of symbols in the context. Definitely not an abandoned project
 
rm -rf
rm -rf
12:58 AM
 
acl
acl
1:15 AM
@rm-rf had missed that
thanks
what happened to @belisarius? busy? I feel unentertained without him around
 
rm -rf
rm -rf
2:04 AM
@acl @nasserm.abbasi btw, I got a response to that MATLAB puzzler from a few days ago. Apparently, if you use a temp variable, then M parses x*x', identifies it as a matrix outer product and uses optimized algorithms to guarantee a Hermitian result. If it can't determine what x is when it parses, as was the case in squeeze(...)*squeeze(...)', then it computes the matrix product normally, thus accumulating FP errors, resulting in non-Hermitian outputs.
Who'da thunk that... oh, well. Seems like this is MATLAB's equivalent of a Mathematica's unpacking sneaking up on you...
 
Mike Honeychurch
Mike Honeychurch
2:27 AM
@Rojo I answered a question on embedding stylesheets a few months ago. let me find it ...
 
Rojo
Rojo
3:05 AM
@MikeHoneychurch Thanks, I'm taking a look to your self answer. That's exactly what I was hoping to find!
 
 
7 hours later…
Nasser M. Abbasi
Nasser M. Abbasi
10:03 AM
@rm-rf, ok about the Matlab puzzler. I found another Matlab puzzler the other day:

clear all;
A = {'A' 'a';'B' 'c'};
s = struct('data',{A(:,1)'});

Now I typed:

r1 = {A(:,1)'}
r2 = s.data

now you'd think r1 and r2 are the same. right? After all that is what I typed in the data field of the struct s. but they are not the same

r1 = {1x2 cell}
r2 = 'A' 'B'

And I used to think that Mathematica lists can get confusing until I started to look at Matlab cells :)
As I said before, Matlab is nice when working with matrices and vectors only. Beyond that, it is a mess actually.
 
 
2 hours later…
Nasser M. Abbasi
Nasser M. Abbasi
12:20 PM
any one see the error: "the following plug-in is unresponsive: Wolfram Mathematica, would you like to stop it?" under Chrom browser on window 7 running a CDF file ? I can reproduce it all the time with CDF generated under V9.
 
Nasser M. Abbasi
Nasser M. Abbasi
12:37 PM
i've just send a bug report to WRI just in case for support to look into it. This message shows up only under Chrome. I could not reproduce it under Firefox.
 
 
2 hours later…
rm -rf
rm -rf
2:56 PM
@NasserM.Abbasi No, I wouldn't think they're the same, because r1 is a cell inside a cell, whereas r2 is just a cell
Here's something to remember: indexing anything by using parentheses will give you the same type of object as the original. If you index a matrix, you get a matrix. If you index a cell, you get back a cell.
If you index a cell by {}, then it returns the contents of the cell, whatever that may be.
So to get something that's equal to r2, you'd have to do r3={A{:,1}}. So A{:,1} throws out the contents and you then put that inside a cell.
 
Nasser M. Abbasi
Nasser M. Abbasi
3:49 PM
@rm-rf sorry, I am not following you. but that may be becuase I need more coffee.

I am simply putting an X inside the struct, and then reading this X back. I compare X before I put it in the struct, with the X in the struct. I would think they would be the same thing, right?

I will re-write the same thing different way, I do not think I wrote things before in a clear way:

clear all;
A = {'A' 'a';'B' 'c'};
X = {A(:,1)'};

Now I will put this X in a struct. Then compare this X once it is in the struct to see if it changed:
 
rm -rf
rm -rf
4:24 PM
@NasserM.Abbasi but that is explicitly what the struct does! It takes the contents in the cell X and assigns it to the field 'X'. This is what the doc says:
> The size of the resulting structure is the same size as the value cell arrays[...]. Elements of the value array inputs are placed into corresponding structure array elements
 
Nasser M. Abbasi
Nasser M. Abbasi
4:39 PM
@rm-rf, OK. Thanks. So it works as documented. But it is still not logical nor intuitive. After all, when I write the number 5 into a record, and read it back again, I expect to get the same as I put in, which is 5. So it works as designed, but my point is that it is not what one would expect. I like languages that follow the principle of least surprises (sometimes I wonder then why I program in Mathematica, since Mathematica surprises me sometimes more than Matlab does actually :)
 
Szabolcs
Szabolcs
5:37 PM
Can you try the following please?
1. Start Mathematica
2. Needs["JLink`"]; InstallJava[]
3. open the palette from the palettes menu
4. evaluate SEUploader`onlineUpdate[]
Do you get any errors? Did it update to 107?
 
ssch
ssch
5:49 PM
In computer graphics, Is there a name for an object that can be seen behind non-transparent objects and can I give an object this property in mathematica?
 
 
2 hours later…
magma
magma
8:09 PM
@Szabolcs Hi, well...I do not get any obvious errors. How ever I get now a new , smaller palette , with just 3 buttons "Upload", "Upload with pp" and "History". No more update button and no indication of the current version number. This palette looks less perfected than the previous one. Is this the new palette?
The other one looked better in my opinion
 
Mathematica StackExchange (meta)
Mathematica StackExchange (meta)
8:37 PM
0
Q: How do I communicate about a rejected edit?

raxacoricofallapatoriusI have a proposed edit that was misunderstood and rejected. How do I communicate about that?

discussion
 
Eric Brown
Eric Brown
9:14 PM
I have been going through many of the notebooks in the Documentation center, and evaluating them. There are some serious bugs in many of them. As a practical matter, is it useful to send each and every one to Wolfram, or is it safe to assume that they know about them, and they will patch them in 9.0.1 ?
Two examples: IndependenceTest and PairedHistogram each have half a dozen errors, as confirmed on two separate computers
Also, any news on when a 9.0 compatible CDF player will be released? I made a CDF with ListPicker and was embarrassed at a presentation when it wouldn't play on a fresh installed CDF player.
 
Brett Champion
Brett Champion
9:55 PM
@EricBrown I would send them in. For PairedHistogram did you find anything other than problems with Ticks->None?
 
Eric Brown
Eric Brown
@BrettChampion Yeah, that looks like the common bug.
 
 
1 hour later…
acl
acl
11:33 PM
hmm.. not only does the frontend often spit out lots of messages involving XML when I copy stuff to the clipboard, but today it managed to mangle my code when I selected a cell and pressed cmd-/ (ended up with an uninterpretable cell)
nice
 

« first day (325 days earlier)      last day (4165 days later) »