12:38 AM
@Heike yup that me, second from the right. lol But in this case I actually meant outside: I've started rock climbing. So you'll know what happened if I suddenly and permanently disappear from StackExchange without notice. :-p

1:15 AM
@MrWizard Beware of your nearest rock!

@belisarius How do you find these things? :-)

@MrWizard I've read it many years ago in a wonderful book. Let me search for it.

I'd just like to point out that since Dec. 16, when we re-entered the definition phase, it has only been 187 days!

@MrWizard @MrWizard Here it is

@belisarius Thanks.

1:24 AM
@belisarius added to my amazon wish list.

It is only of historical interest. You'll not learn QM from it!

But, I'm interested in the history around the physics. That's one of the reasons I find the making of the atomic bomb so fascinating.

@rcollyer Did you see the Feynman's lectures at the Microsoft site?

@belisarius I saw one or two. I prefer his writing to his lectures.

@rcollyer Yeah, probably. But they are very nice historical docs.

1:28 AM
@belisarius that is very true.
But, he's an interesting character.

@rcollyer He was outrageous! So many anecdotes condensed in one person!

@belisarius He did a lot of work at a strip club.

@rcollyer He performed at a Scola do Samba at Rio carnival, and many, many, many more :)
@rcollyer I remember reading that he learned Spanish to be prepared for his sabbatical in Brazil

@belisarius Yep, and it was his example that led to the Brazilian Physical Society to give their presentations in Portuguese, not English as was previously done.
@belisarius :)
@belisarius And through, a lot of help, memorized his talk in Portuguese.
Looking at the answer I pointed out, I am trying to "improve" it. But, Apply doesn't seem to be doing what I want. For instance, f @@@ (3 x + 5 == 2) returns f[5, 3 x] == 2. Curious.
@belisarius It was his assessment of the problems leading to the Challenger disaster that truly cements his place in my heart, though.
Damnit. I understand why Apply isn't working: 2 is atomic, so Apply can't attach to anything. :P

1:48 AM
@rcollyer Curious. I love him since I learned he opened all the safe boxes in Los Alamos

Leonid recently posted a way to drastically improve the performance of the DayOfWeek function in the Calendar package. This is just one of many examples of less-than-optimum package functions. What do you all think about a community project to update and optimize some of the standard packages?
3

@belisarius that was phenomenal, but it was the fact he published his report despite the fact that the committee didn't want to.

@MrWizard I'd have to pull out the book (currently in storage), but if I remember correctly, the committee investigating the claim had their own report alongside Feynman's. It might have been included as an appendix, but I don't recall.
@MrWizard works for me. I think it would be a good project to go alongside our graduation. Can it be morphed into a contest? Maybe having a couple of trips to the developer's conference as prizes?

@rcollyer Maybe propose here (don't know why the feed hasn't picked it up yet)
0

We have recently learned that the site will graduate soon. That means we will soon have a nice new design, higher rep thresholds to do things, and a reason to celebrate. Should we try to do some more promotion of the site to coincide with graduation? Are there any events, "question parties" or ...

1:58 AM
@MrWizard Other swag is good, too. But, I think the conference is a jewel of a prize.
@Verbeia the feed has about a 1/2 lag (maybe longer) from meta.

That isn't really the direction I was thinking; I'd rather make it purely cooperative; I think we've got a lot of good Mathematica programmers here and it would be good to work together on something for a change.
4

Fine with me. That works, too.
Maybe set it up as a Town hall meeting wherein we discuss the project?

2:14 AM
1

We have recently learned that the site will graduate soon. That means we will soon have a nice new design, higher rep thresholds to do things, and a reason to celebrate. Should we try to do some more promotion of the site to coincide with graduation? Are there any events, "question parties" or ...

2:30 AM
@MrWizard Reminds me... I did do a lot of rewriting of built-in packages in old Mathematica back in the day, whenever I found the implementations clumsy or insufficiently general.

3:13 AM
Image processers
Any simple way to remove some straight lines from an image whose coordinates I've already found. The lines aren't 100% perfect

3:25 AM
@Rojo You may use Inpainting, for example, but let's see your image first

On it
That's a zoom to a little part of it
@belisarius
I did something already, but very manual brute force, and doesn't work too well. I had also done some hard work to detect the lines only to then see that ImageLines could do it. I'm a stranger to all the morphological image processing world

3:43 AM
@Rojo Ok. So you want to get rid of the horizontal AND vertical lines?

@belisarius Yeah

What else?
There is a lot of noise there

The blurred gray could go white, but those dots aren't noise
They are the data

OMG
Hmmm you better upload a full image to imgur. Meanwhile I'll see what I can do with it
@Rojo Around the horiz lines there are a lot of dots. Are they data too?

No, those aren't

3:49 AM
For example around the sixth line counting from below

The image isi an image with dots, that has been drawn, on top, those dotty lines. The dots that make up the lines should go, and those that are near should stay. The image sucks, it has noise, but that's the general idea.
It doesn't need to be perfect at all
but what I did after detecting where the lines were, so far, is not quite good. Estimating the width of the line, taking samples from above and below, averaging and replacing the line with that, windowed with something
It didn't work well at all in those parts were there are thicks lines of data crossing the grid

Removing the lines is trivial. Selecting what part of the lines should remain isn't

That's right

let me work on it for a few mins
first result

4:11 AM
Detected :)
Thanks a lot... Btw, rui.rojo@gmail.com is my email if you don't mind sending me the code of what you're playing around with then

wait, more to come
i = Import["http://i.stack.imgur.com/4VRfG.png"];
j = MorphologicalBinarize[i];
k = EdgeDetect[j, 7];
lines = ImageLines[k, .15];
jw = Image[ImageData[j] //. 0 -> 1];

sh = Show[jw, Graphics[{Thick, Black, Line /@ lines}]];
er = Erosion[sh, 4];
ColorNegate@Closing[ColorNegate[er], 5]

So that gives me the neat black lines

Yeap, we must manage to get the clean graph with that mask

I see... That's the common way or processing parts of images, right? First finding a binary image with black in the parts you want to process...

then you should use morphologicalcomponents to rebuild the original image, after
cn1 = ColorNegate@Closing[ColorNegate[er], 5]
Ups. sorry, I must leave. Post a question or lets continue with it tomorrow. Perhaps is wiser to post a q, since somebody could come up with csomething better

4:28 AM
Ok, thanks a lot

good night, and sorry for not being able to finish taht

Night!

4 hours later…
8:27 AM
Good news :-)

3 hours later…
11:07 AM
Dammit I capped again... :(

11:51 AM
@JM Ahh, poor you. ;-)

@Heike Yes, such are the problems we have... :D

Are you trying to reach 10000 before we graduate?

"This person has already capped, are you sure you still want to vote this post?"
@Heike I'm not sure it's feasible; I have a lot of dry days...

@JM For a little sum ...

@belisarius Is $\varepsilon$ "little" enough? :)

11:58 AM
@JM you just need to cap for the next 6 days.

@JM If we can write an $\infinite$ over the sigma

@Heike Still sounds like a tall order, sadly. :)

@JM You're probably right but it's not impossible. I think Mr. Wizard capped about once every three days on average.

1:01 PM
Mooorning

@Rojo Buenos días. :)

@JM, ;)
I'm at work
in a nice (more or less) morning in Buenos Aires
with view to the Rio de la Plata
starting my birthday
reaching 9K rep

"nice (more or less) morning" - In that case, why are you not outside enjoying your birthday? (Happy birthday, BTW.)
Oh right, work. Pfft.

No, I'm inside with a big window showing me the ontside
outside
Some people running
training
Others walking
A ship on the horizon

Being cooped up inside doesn't sound like a proper birthday to me... :(

1:05 PM
C'est la vie, that's as good as it will get, so
I'll see the third glass full

@Rojo Now if only you had a camera with you (hint, hint).

:)
I have my cell phone
I also have a biiig chunk of laziness

@Rojo So that's why you're here instead of working? ;P

I'm here because, I just arrived, many co-workers haven't, and I'm about to get interruptoed and invited to a nice breakfast :P
@belisarius, how would you say "fiaca", in the context of "me da fiaca eso", or more precisely, how would you translate "straw" in this context en->sp->en?
Anyway, what's up with you @JM? What time is it there

@Rojo NLMPI

1:10 PM
@belisarius AOSFDIJ

@Rojo 21:10. Reminds me, I should be cooking dinner. Back in a bit...

@JM later

They just changed all the telephone wiring from our modem down to where it goes underground two blocks down ... hope that we'll have a connection reliable enough to post an answer without disconnecting

1:37 PM
On any other day, I'd be happy to get 100 rep. Today, it's about the amount of rep I could have gotten from votes after I hit the cap... :D

2:14 PM

It seems I'm connected via a cable just hanging down from the 4th floor, to the ground level. In the windiest island of the Mediterranean ... it's a miracle it lasted so long. Anyway, they changed the cable.

1 hour later…
3:29 PM
Is anyone here familiar with .NET and COM Automation?

3:59 PM
How can I prevent the DialogInput from failing as soon as I close the "Close this" dialog?

1 hour later…
5:25 PM
We have this tag which seems to be somewhere between and . Anyone has any specific thing to say about ? The tag wiki is empty, and it is a bit hard to guess what it is about.
I would rather remove it and retag questions either as or .

5:43 PM
I have an image and a graphics such that Show[img, graph] shows them nicely. I want the image version of graph such that they overlap just like Show shows, but there's always some offset or reflection with what I'm trying. I'm sure this is an easy one...

2 hours later…
7:53 PM
Good morning, does anyone here uses LevelScheme? I have used in in the past in Mma version 5 or 4 and wonder if it still around and working in Mma version 8?

1 hour later…
8:53 PM
@Matariki I think Szabolcs does use LevelScheme, at least according to this post.

@IstvánZachar Thanks, I'll have a look it.
@IstvánZachar Great news, thanks for the link. It is still free software by the looks of it.

@IstvánZachar I agree. And there are only 7 questions to retag.

9:51 PM