Wolfram Mathematica

2:15 AM

@ArnoudBuzing Wouldn't this be something that can be changed with a fair amount of work? Right now, you have to handle your internal database and you are tracking issues reported through SE too. Isn't there someone who could convince Steven Wolfram to make a step into using a public tracking system?

With this, everyone could lookup whether an issue was already reported, developers could comment on their issues. It opens even the opportunity that developers can communicate with guys that know their stuff so that they can help to track down hard bugs.

@ArnoudBuzing I hope you (and I don't mean you..) are aware of how much support you would get from the community!

Meta

3:54 AM

0

Q: Mathematica SE Popularity

I saw this plot the other day and was amazed at how much of an outlier Mathematica is. Not necessarily in how little it is on GitHub, but how popular it is on the y-axis. Which got me wondering... why? Is the language that hard to learn? Is the language so powerful that questions keep on popping...

discussion

Guess who it is.

6:03 AM

@rcollyer, daaaang…

Patrick Stevens

8:37 AM

@Pickett although WolframAlpha will do it: wolframalpha.com/input/…

blochwave

9:28 AM

Is it just me, or does the third form of ImageAlign[] not work? i.e. ImageAlign[{image1,…,imagen}] - I just get an error message :-(

Pickett

11:15 AM

@PatrickStevens Questions with Wolfram Alpha answers are explicitly off topic, so it doesn't matter :)

halirutan

11:35 AM

Wrong in so many ways :-)

rcollyer

12:29 PM

@Guesswhoitis. I was on while it was in progress. So, every few seconds I got a notice that I was being upvoted. :P

Szabolcs

@rcollyer Do you know that problem where a MatrixPlot or ArrayPlot renders "fuzzy" in some PDF readers when it's exported to PDF format? Is there any chance this will be considered by WRI? I have the feeling they don't really treat it as a real bug.

rcollyer

Yep, 180 points gone. Oh well.

@Szabolcs interesting. I've not heard of it internally, but output formats are not my thing.

Which readers?

Szabolcs

@rcollyer What OS are you using?

rcollyer

mac

Szabolcs

@rcollyer Preview.app mostly, but I've seen it in Acrobat occasionally ... though Acrobat is fine most of the time. Let me show you:

Export["~/Desktop/mat.pdf",
 MatrixPlot[AdjacencyMatrix@RandomGraph[{10, 20}], Mesh -> All]
 ]

rcollyer

12:33 PM

ok, one sec.

that's interesting. I wonder if its bad pdf code or bad apple code.

I'll pass it along to the appropriate tester, so he can poke the appropriate people.

@Guesswhoitis. Hey, don't mess with my lord Baal! :)

Szabolcs

This is not actually wrong. The underlying Graphics primitive is Raster. It is exported in a way that is suitable for images, so some PDF readers will upscale the image with (apparently) linear interpolation instead of nearest neighbour. But in this particular case (and in most cases that are important to me) we need to have sharp rectangles, without any interpolation.

So it's not a bug, because in some cases it's preferable (images).

But it would still be very very useful to be able to choose to export a sharp image.

Also, what we get in the exported PDF depends on whether we use Export or Save As..., and sometimes even on the presence of frame labels or other seemingly random factors. Save As... usually produces sharp results, but not with frame labels.

rcollyer

that would be a nice to have. I'll pass it along to see what I can stir up.

Szabolcs

Sorry for going missing suddenly, something came up here, so I need to leave for a while ... will be back in some hours

rcollyer

ok. later

email me, if you need more.

halirutan

12:57 PM

@Szabolcs It looks fine in Adobe Products. When you use the Preview on Mac, you can go to Preferences->PDF and turn off the smoothing:

it's the checkbox "Smooth text and line art"

After that your blurry version looks like this:

it doesn't solve the problem because now you lost antialiasing of text and lines, but at least it's a start. I'm not sure whether this is really a problem of Mathematica or the Preview itself.

bobthechemist

1:14 PM

@halirutan I like that, thanks.

Szabolcs

1:34 PM

@halirutan Nevertheless, such a PDF figure is not suitable for publication, nor for sending to colleagues. You can't tell your audience that if they use Preview they have to adjust the settings before they can read your paper.

Well, I've completely forgotten about it:

7

Q: PDF exports of ArrayPlot's are fuzzy (OS X)

[The following problem occurs on OS X; I don't know if it also occurs with other OS's.] If I produce a Graphics object with ArrayPlot, right-click on it, choose "Save Graphic As...", and save it as a PDF file, the resulting PDF graphic is absurdly fuzzy/blurry (which is particularly perverse, co...

graphics export pdf-format

halirutan

@Szabolcs Yes, this is for sure.

@Szabolcs but I see a valid point for WRI here in using Raster. These functions have to work on very large matrices where Raster (or an Image) is the correct thing to do (instead of using many polygons or rectangles). Interpolation is then not a problem any more since you have many points

although... blurryness seems to be worse then I expected it, even for 200x200 matrices..

Szabolcs

2:19 PM

@halirutan I agree that Raster should be used. What I am wishing for is some Raster option that will control how it's displayed (and exported): to use or not to use interpolation.

Transcript for