« first day (384 days earlier)      last day (4117 days later) » 

12:19 AM
Anybody else seeing an issue with the site's stylesheet?
Like, it's not loading.
 
12:45 AM
@jfklein13 Should be fixed now... apparently their CDN had some issues.
 
@Gabriel Try Edit > Paste Special and select Picture or PDF. PDF seems to give better results in my trial, but that's up to you. (Mma 9.0.0, Word 2011)
 
1:21 AM
Which systems don't support pipes? (regarding Open Read documentation "On systems that support pipes"
 
@sssch: I have wondered that myself, and I was working in that code last year. My guess is that comment no longer applies.
I can tell you that we make use of pipes in our daily tests on every $SystemID that Mathematica supports.
 
@jfklein13 Good, thanks!
 
J G
1:45 AM
anybody know perl here?
 
2:37 AM
@JG What for?
 
J G
@belisarius web scraping
 
@JG In Mathematica?
 
J G
perl
 
@JG This site is about Mathematica ...
 
2:50 AM
@JG you've caused a tiny incident in TeX.SE chat :)
though I still don't get how
 
I have yet to see a good coding style in the demos, granted I've only seen a few of them so far. Why don't MMA(that's the acronym for mathematica right?) users take advantage of whitespace, indents and newlines to make really readable code?

So far I've only seen code that is densely packed like the programmer had a Guillotine to their throat if they didn't get the code done as quick as possible.
I get the IDE is a little cumbersome, but with a little editing, I'm making the demonstrations I download really readable.
 
@AdamDreaver Try the workbench
 
@belisarius I don't know what that is, I'll google it
Oh i see
I'll give it a try, ty you belisarius
 
@AdamDreaver you're welcome
 
3:06 AM
belisarius did you see that fourier demo made with mathematica on reddit? I thought it was neat!
 
@AdamDreaver nope
 
code for is right on the side, but you seem like you could come up with in your sleep you pro :)
 
@Adam The front end doesn't make it easy to format code ... you can use the auto-formatting but if you want to change that beyond inserting a few newlines, it becomes painful. And even the newlines are too easily destroyed by copying into a plain text editor and then copying back.
Can be a bit annoying ...
But you can use code-style cells. Press Alt-8 then start typing
 
@Szabolcs Or forget about it and try to get used to Mma FE coding style
 
FE coding style? That code-style cell is much nicer than normal input for pure coding
 
3:17 AM
But coding is not what one usually does in Mathematica.
 
What does one usually do?
I thought code your math problems to be solved :(
 
I mean, it's not like writing C or Python: write the program, test the program, correct the program, then put it away for the future.
I usually work "real time", and 90% of what I do is thrown away. E.g. some data needs to be kneaded into a different shape, so I start transforming it by applying functions one by one until I like it, then I plot it, etc.
Can look like magic to those who're not used to it, not having to wait for me to write the program, just stand there and watch it come alive ;-)
@Adam If you write a package that you will re-use often, then I think it's often better to use a plain text editor or the Workbench. The workbench is much better suited for writing big things. The front end is better for quick and dirty stuff, playing, trying out things, and supports graphics of course :)
BTW the reason I love Mathematica is that it makes this type of work possible.
 
It takes a while for one to get used to the WB, because it is inert (i.e., non-interactive)
 
@Szabolcs you should do a screencast, I'd love to watch!
you make it sound so organic and awesome
 
If you get too used to Greek letters, fancy unicode, subscripts, superscripts, etc., you're in for a shock
 
3:24 AM
@Adam If you use the system daily for a while, you'll soon get used to it, and it'll be the same your you. I'm afraid I made it sound more interesting than it really is. :)
 
haha
 
OK, it's late here, I need to go. Good night everyone!
 
Good night
 
4:05 AM
Hi all, I'm having problems using Show and RegionPlot together
If I do Show[myregionpl, otherpl], it looks ugly and it seems to turn antialiasing off
If I do Show[otherpl, myregionpl], the region plot shows the mesh, which is also ugly
Heavily related to this question: mathematica.stackexchange.com/questions/1400/…
But it doesn't address how bloody ugly the plots are that show the RegionPlot first!
 
 
1 hour later…
5:21 AM
if statements r cool functional languages, it's the do/whiles/and fors we want to avoid right?
^cool in functional languages
I guess they must, how else would do you branching code?
 
5:46 AM
@AdamDreaver Something like this? f[x_] := x^2 /; x > 0; f[x_] := -x^2 /; x < 0; Plot[f@x, {x, -1, 1}]
 
@belisarius what is the name of that operator, /;
 
@AdamDreaver Learn how to use the Help System. Highlight it with your mouse and press "F1". ... (Teaching to fish)
 
Ohh, ty
 
@AdamDreaver Anyway ... If[] can fit smoothly in a functional style
 
I think trying to learn functional programming with mathematica might be a mistake for me because of all the imperative syntax; I should of started with haskell or LISP.
 
5:53 AM
@AdamDreaver Mma isn't only functional. The problem with it is that it implements several paradigms. It isn't a language for learning a paradigm, that's for sure
 
I was trying to kill two birds with one stone, learn both MMA and functional programming; I'll just separate that learning process. /; is a nice and beatiful, I don't think using an if statement instead of /; would make your example a lot more ugly
 
Mma is a prgmatic language, not a pure one. It's designed to solve math (and related) problems, not to be a CS playground
@AdamDreaver Oh you're wrong!
 
not surprised
:)
this stupid wandering cat is distracting me from clear thought, damn it's cuteness
 
just imagine testing your x values each time you need to use f[x] at your program
@AdamDreaver In Spanish we call "cats" to those women going up and down the street at night
they're distracting also
 
This is an actual cat, I think I'm infected with that weird cat virus, cats are like sirens to me now. I can't imagine testing my x values. I'd do better thinking if I was in a strip bar.
 
6:02 AM
@AdamDreaver Actual cats are difficult to point out. How on earth could you know? They don't speak. Has it already declared its true self?
 
They meow... and I'm pretty sure I know what a cat looks and feels like.
 
@AdamDreaver That's wishful thinking
@AdamDreaver going back to the If[] problem.
you still here?
 
aye
All ears
 
Try thos
ClearAll[f];
f[x_?OddQ] := x;
f[3]
 
ok will do, I just tried this, f[x_] := If[x > 0, x^2, If[x < 0, -x^2]]; Plot[f@x, {x, -1, 1}], and it has the same results as your /; example, i'll now try what you just posted.
f1 and help IS USEFUL!
 
6:11 AM
hehe
yup
 
@belisarius Done and favorited! tyty
Ok I need to make myself a healthy organic dinner so I can think and be healthy, may you slay any difficult or frustrating work with ease my friend.
 
@AdamDreaver Go and kill your lettuce in peace
 
haha :)
 
 
3 hours later…
9:29 AM
Anyone know how to change the color of comments? it's a dull grey, i want C++ green like comments
Found it!
 
 
3 hours later…
12:52 PM
@AdamDreaver (Menu) Edit -> Preferences -> Appearance -> Syntax Coloring -> Other -> Cooments ... Check Show Color Spectraand select your lucky color
 
1:02 PM
I can't understand what he wants:
2
Q: axis label and ticks alignment of ListPlot3D and Histogram3D

dudeI have a question on axis label alignment of ListPlot3D and Histogram3D. These labels are not aligned with the axis of the view, but only horizontally. I have tried very hard to align them manually, but the method sucks and still the output lacks precision and neatness. I wonder if M 9.0.0 has th...

Anyone?
@P.Fonseca Do you?
 
1:23 PM
@belisarius I think he wants the text of the axis labels rotated in alignment with the axis themselves
 
@P.Fonseca Oh! great! Never thought of that
 
@P.Fonseca Mmm.. I still prefer the labels aligned with me :D
 
@P.Fonseca since you have generated the images that demonstrate the point exactly, could you edit them into the original question?
 
@Ajasja Good idea
@Ajasja , @P.Fonseca Done
 
1:41 PM
@belisarius thank you...
 
@P.Fonseca Edit it if my update doesn't fit your taste
 
1:55 PM
Hey, I have a expression like this:
List[Rule[PlotLabel,Style["{2004, 1, 1, 0, 0, 0.} - {2005, 7, 14, 0, 0, 0.}",Bold]]]

How can I get only this part {2004, 1, 1, 0, 0, 0.} - {2005, 7, 14, 0, 0, 0.}?
I tried ReplaceAll but it doesn't want to work.
 
@crobartie Try with Part
exp = List[
  Rule[PlotLabel,
   Style["{2004, 1, 1, 0, 0, 0.} - {2005, 7, 14, 0, 0, 0.}", Bold]]]
exp[[1, 2]]
 
Ok.
@Ajasja That work List[Rule[PlotLabel,Style["{2004, 1, 1, 0, 0, 0.} - {2005, 7, 14, 0, 0, 0.}",Bold]]][[1,2,1]] but I get "{2004, 1, 1, 0, 0, 0.} - {2005, 7, 14, 0, 0, 0.}"
@Ajasja I want 2 lists from that (not a String)
 
Well you said
>How can I get only this part {2004, 1, 1, 0, 0, 0.} - {2005, 7, 14, 0, 0, 0.}?
 
2:12 PM
@Ajasja My error :/
@Ajasja I think StringSplit@str will be ok..
 
@crobartie Yup. And you can use ToExpression on each part of the string to get the lists, if you need them as lists.
 
@Ajasja Thank you! :)
 
@crobartie You're welcome:)
 
2:48 PM
@YvesKlett Got windows and the brand new 9.0.1?
Try to define a function großÄ[x_]:=x and then start to write it and use autocompletion.
Here this ends UTF8 codes.
 
3:13 PM
@halirutan Autocompletion produces
gro \337 \304 which evaluates to Ä gro ß
how do you come across such things?
 
3:36 PM
@halirutan Working on the camel hump fix?
 
3:46 PM
@Rojo Hi red one
@Rojo My last question is a hook for you
 
@belisarius Whats up
 
@Rojo Will you bite?
 
@belisarius Let me search
 
1
Q: Performance: Collapsing repeated contiguous rows & cols from a matrix

belisariusIn this answer I needed to remove contiguous zero-valued cols and rows from a matrix, leaving only two of them in place, no matter what the original number was. I made up this code: m = RandomVariate[BinomialDistribution[1, 10^-3], {400, 400}]; rule = {h__, {0 ..}, w : {0 ..}, {0 ..}, t__} -...

 
@belisarius Oh, I recall that neat line
@belisarius Bit
 
3:55 PM
@YvesKlett I love finding bugs.
 
As long as they do not bite back
 
@Rojo Unfortunately I'm currently fighting with blender.
I just checked and saw that it's really not working anymore :-(
 
4:23 PM
@belisarius You don't care about the bordering 0 rows and cols?
 
@Rojo Ideally they should be treated as the rest
 
@belisarius Ah, so you should change h__ to h___ and t__ to t___ I beleive
 
@belisarius A single zero row/column is left alone, right?
 
@rm-rf yup
@Rojo Try it ... it failed to me
 
Hmm... I get different dimensions. Grr.
 
4:25 PM
@halirutan did you report this to support@wolfram.com ? They might do something if it is in their system.
 
@RolfMertig Hehe.. let me check whether it's fullmoon soon..
 
@halirutan WolframAlpha["nearest full moon leipzig"] was last month ("wolf moon" !), the next one is WolframAlpha["next full moon leipzig"] February 25th ("snow moon") , just before the Mathematica day here ...
 
@RolfMertig We maybe don't want to wait that long to disappoint the Wolfram crew about their update.
 
 
1 hour later…
5:58 PM
Pretty terrible selection of questions we got for the community review, I thought... given that it seems to be automated now, I can't understand why not pick 10 random questions for each reviewer? Better coverage and more representative that way; at least we wouldn't get (on average) one closed question and one that was about to be closed as TL
 
6:13 PM
@OleksandrR. I asked about that and the reason they gave was that it's harder to implement. Besides, showing everyone the same set allows you them to get an aggregate...
The closed question was not closed when it was picked... closing it was the outcome of my review =)
From what I understand, this was only meant to be a beta site feature and since it is automated now, they opened it to the graduated sites as well (except SO/SU/SF). They currently have no plans to do anything with the data from graduated sites (what can they do?)
FWIW, here are the review results (it'll end soon and probably won't change very much):
 
Hmm, is it possible that when I am working on Compress'ed files algorithm works slightly faster;)?
 
6:50 PM
@rm-rf well, I had hoped the review was also meant to be useful to us, the site members... in the sense that we all seem to more or less agree on what constitutes a good or bad question for this poor selection I suppose it was somewhat. Actually I was thinking of starting a thread on meta to discuss what seems like a lot of mediocre questions lately and not enough voting. I get the impression that most people are ignoring most questions now.
Pretty much all of the votes come from people on pages 1 and 2 (out of 10.3) of the voters' league... although it certainly agrees with the Pareto distribution, I'm not sure how sustainable this is if the number of questions continues to increase. It'd be a shame if more questions were to go ignored and attract no votes.
Most questions lately seem to be not hugely compelling, although good ones still garner 10 upvotes within the first hour or so. In the interest of finding something for everyone, I'm thinking more of the less experienced users who might have learnt something from the "average" type of question and who could perhaps be encouraged to vote more.
 
@OleksandrR. Yes, we are free to use the data as we fit for our own improvement (which is actually the main purpose of these). I agree with you re: the voting, but I don't know what can be done (or should be done)
I certainly have slowed down compared to this time last year and tend to do more janitorial work than anything else, and perhaps others too are feeling this "fatigue" (boredom might be a better word).
That period between last march and july was when we had the most activity (pretty much everyone on page1 repwise had a field day)... although regulars have slowed down, the decrease is not dramatic, which shows that more of the newer users are voting. Nevertheless, a meta discussion might be interesting, so please do post if you have something to share
 
7:36 PM
@OleksandrR. Very good idea. I have the same feeling for a while now.
 
@rm-rf thanks for the plot. Well, clearly voting is still proceeding apace, but I had hoped that the number of votes cast would have risen more given that we've got many more questions/day and many new members since the middle of last year.
 
Alex Miller on February 05, 2013

Welcome to Stack Exchange Podcast #42 – it’s our usual gang back this week with Joel, Jay, David, and Producer Alex.  There’s plenty of inside baseball, so put on your rally caps and make sure to stick it through to the end!

David Mamet, apparently. Jay was a drama major.

Michael forgot to pay the Google bill, so our hangouts are back down to 10 person limits (but its fixed now!)

We have one big thing to talk about that made a change and generated controversy. Joel correctly guesses what it is: we no longer display your accept rate (the percentage of questions you asked that you accepted an answer for). …

 
8:28 PM
@OleksandrR. Same here.
 
Does anyone know how to add keyboard shortcuts without touching KeyEventTranslations.tr in the installation directory? I'm tired of having to re-add them every time I upgrade. The only ones I use are for entering [[ and ]] btw
 
8:52 PM
Hey is it just me or does <key>Esc</key>co<key>Esc</key> look exactly like nothing?
a<esc>co<esc> should be Conjugate[a] but before evaluation it jsut looks like a
 
It looks like a star to me. Maybe your default font is missing the star for some reason?
 
@Xerxes Was worried that was the case, thanks
 
@ssch You're on Linux, right? I think I heard people on Linux complain about this before.
But I may be mistaken.
 
@Szabolcs Yea
@Szabolcs What's your [[ event?
 
@ssch Command-[ or Control-[
 
8:57 PM
@Szabolcs: I just added that trick under my local directory in Linux in the .Mathematica directory.
 
ah, was hoping for auto rewrite of [[ to that compact form
 
@ssch that's a difficult problem because of bracket matching. [[ can be rewritten but what about ]]?
 
oh right
 
@Xerxes You copied the .tr file there? Doesn't feel safe ...
 
@Szabolcs: You're thinking that the default KeyEventTranslations.tr might change during an upgrade? I hadn't thought of that. Perhaps it reads both during startup?
It doesn't. :-(
I guess you're down to something horrible like adding a tiny script to the Mathematica startup to check the installed version to make sure it's older than the one in local or patch as necessary.
 
9:09 PM
@Szabolcs Why do you add it to the installation directory? Copy the file (mimic the directory tree) to $UserBaseDirectory and you don't have to change anything after installing a new version
I think the rules are applied in this order: $InstallationDirectory --> $BaseDirectory --> $UserBaseDirectory, so whatever is in the last one can override the previous ones.
 
@rm-rf: Is there a notation for appending to the EventTranslations list instead of overwriting it? My first test did not work.
 
10:10 PM
@Rojo online?
 
@halirutan Yeah
JUst back
 
@Rojo CamelHump completion is dead :-(
 
@halirutan I saw that
The question is, can it be reborn?
 
No, I mean it's dead until Wolfram decides to make it built-in.
@Rojo Have you tried to turn on LinkSnooper in 9.0.1?
 
@halirutan nice @ smooth ColorFunction via Texture, that ruggedness has made me push plotpoints to absurd numbers a few times
 
10:15 PM
@ssch Yep, that's the usual trick
 
@halirutan No. From what you say I-m guessing it is hopeless?
 
@Rojo They do everything in the front end without asking the kernel. I cannot intercept anymore.
 
@halirutan Oh, cr%^%^
@halirutan We could consider using 9.0.0's front end with 9.0.1's kernel
 
What happened to the completion? I was looking forward to that
 
@halirutan Now that 9.0.1 is out, shall we compile a new symbol list for the site's syntax highlighter?
 
10:21 PM
@Rojo That is maybe possible, but it's not an option for everyone.
@ssch They changed how it worked in the last patch. Now it's gone forever.
 
:(
 
@Szabolcs Yes, we should do that.
 
I noticed that there are several things it doesn't complete in 9.0.1, not sure why.
 
@halirutan Try in a clean linksnooper krenel ToExpression["woa=8"];
and start typing "w"
 
acl
all, has this been fixed in 9.0.1? I guess not, but can't hurt to ask
haven't yet switched to it
 
10:27 PM
It does autocomplete. How did the FE know?
 
@Rojo FE has it's own symbol list ;-)
I assume
 
@halirutan I said nothing. i just saw the packet where the kernel tells the fe about the new symbol
@halirutan I thought for a second that it wasn't there, so how did it learn
 
@Rojo Only the Fuz can tell.
 
@halirutan but we are good at guessing, or your camel humps would have never lived
 
@Szabolcs Yes, and there are other things
@Szabolcs Look for instance at this
8 hours ago, by halirutan
Try to define a function großÄ[x_]:=x and then start to write it and use autocompletion.
 
10:33 PM
@halirutan Yeah, that doesn't work properly
neither does stuff from the Developer` context. Autocomplete removes the explicit context
Also, the fix to the custom functions Ctrl+shift+k is quite lame
 
Exactly. The number of votes is the same, so but with many more questions, the votes get diluted. This has the side effect that new users gain rep more slowly.
Basically @Leonid's predictions were spot on: the inflow of new users was not necessarily only positive, as there are now much more of the mediocre questions. Or to put it another way: we are losing the signal to noise ratio:)
 
It probably works, but you need special string representation of boxes, and not the usual usage definitions,m for it to work properly
However, it seems like you can add multiple definitions if your usage message has mutliple lines... They should document that properly
 
acl
@Rojo what fix?
 
@acl f::usage="f[x_] blabla", in v8 and earlier that made autocomplete with ctrl+shift+k work
in v9.0.0 it was broken
Now its "kinda back"
 
acl
@Rojo I know
 
10:35 PM
Back but it changed the way to specify the template, and it doesn't look simpler
 
acl
so what do you mean "kinda back"? that's really making my everyday coding on 9.0.0 much more tedious than before...
 
@acl You know ;D
 
At least there is one upside of the "not hugely compelling" questions: My addiction to mma.se has lessened and my PHD productivity has risen sharply :)
 
@acl It means that now it works, but not exactly the same as in v8
 
acl
@Rojo oh. how do I find out what the changes are? (without cycling back to work to play with 9.0.1)
@Ajasja that's one positive :)
 
10:37 PM
@acl Asking someone to play with 9.0.1
@acl What do you have?
@acl Nothing there? v8, v9.0.0?
 
acl
@Rojo 9.0.0 right now. I'll update to 9.0.1 tomorrow maybe
 
@acl Oh, ok
then, you can try this
@acl CellPrint@TextCell@Plot::usage
and enjoy ;D
 
@acl Yup, sure is:)
 
@acl Ok, this is less scary probably CellPrint@ExpressionCell[Plot::usage, "Input"]
 
acl
@Rojo can't really parse either but thanks :)
 
10:39 PM
@acl Well, that's the problem. It seems complicated. However
@acl it "kinda works" anyway with the usual syntax
You get the autocomplete, but you don't get the proper placeholders. Your patters are just text
 
acl
@Rojo oh right I see. That's terrible. I have all these packages with functions of 20 arguments. Now I need to switch my code to options (code which I no longer really understand unless I concentrate hard for a few days)
oh well
 
@acl I wouldn't do that
@acl Soon enough, someone will figure out the new ugly format, and we could programatically fix that
I hope
@acl OK, I'm going to give it a shot now, but NO MORE THAN 20 MINUTES
 
@Rojo me either. Although I would (and do) use for any new projects. It makes my life sweeter:)
 
acl
@Rojo not soon enough for me though... I have a number of concurrent projects and can't really drop everything until someone does! although I could switch most of the heavy lifting to python/scipy, if I'm rewriting everything...
@Rojo oh don't spend your time if you don't want to!
 
Well, bed time. Bye all!
 
10:43 PM
@acl The problem is that I want to, but I shouldn't
Useful for me too anyway
20 minutes. Don't let me go over that
 
acl
@Rojo yeah, what motivation do I have to let you, right?
:)
@Ajasja see you
 
Anyone knows how to turn string rep of boxes in to boxes?
 
@WRI people... I just noticed, you left the Prerelease Dashboard palette in the finished product :)
 
@Rojo Got a sample string?
 
10:58 PM
@rm-rf Plot::usage
 
@Rojo Evaluate@foo?
 
@rm-rf ...and then?
 
lol... I'm still not sure what you were trying to do. You want a formatted expression of all the \*RowBox[... stuff, right?
 
@Rojo TIME'S UP!
 
@rm-rf I want the Mathematica expression of the boxes, like the one you get when you do MakeBoxes[f[x^8]]
@OleksandrR. Haha, I only got one attempt and failed miserably
 
11:06 PM
"\!\(\*RowBox[{\"Plot\", \"[\", RowBox[{StyleBox[\"f\", \"TI\"], \
\",\", RowBox[{\"{\", RowBox[{StyleBox[\"x\", \"TI\"], \",\", \
SubscriptBox[StyleBox[\"x\", \"TI\"], StyleBox[\"min\", \"TI\"]], \",\
\", SubscriptBox[StyleBox[\"x\", \"TI\"], StyleBox[\"max\", \
\"TI\"]]}], \"}\"}]}], \"]\"}]\) generates a plot of \
\!\(\*StyleBox[\"f\", \"TI\"]\) as a function of \
\!\(\*StyleBox[\"x\", \"TI\"]\) from \
\!\(\*SubscriptBox[StyleBox[\"x\", \"TI\"], StyleBox[\"min\", \
\"TI\"]]\) to \!\(\*SubscriptBox[StyleBox[\"x\", \"TI\"], \
@Rojo You mean like that?
 
@rm-rf No, compare this to this
b = RowBox[{a, 4}]
c = ToString[b, InputForm]
I want the first, from the last
 
@Rojo But this is a string...
I thought styled strings were always of the form "\!\(...
 
@rm-rf Yea, but I don't want a string
I have a string
I have c, and want b
 
@Rojo But the c here and the Plot::usage you mentioned earlier are different...
For this case, you can just do ToExpression@c
ToExpression@c == b
(* True *)
 
@rm-rf I see
Well, if you people can decypher Plot::usage, and so make some function to turn a typical "f[x_, y_] this function does bla" into Plot::usage's format, it would be helpful
@OleksandrR. As you pointed out, I must be strong and leave this at the hands of all of you
;D
 
11:12 PM
@Rojo There is nothing to decipher there... it literally is the stuff I posted above. I think all those additional stuff come from using the WB to define the styled usage message.
 
@rm-rf Would you know how to create such function?
I got no clue
 
Aha... I think I see what you mean. All those TI there are placeholders for the template
brb, installing 9.0.1
 
@OleksandrR. I got curious, where? :) Can't find it.
 
@Szabolcs SystemFiles\FrontEnd\Palettes\PrereleaseDashboard.nb
 
@OleksandrR. Looks like it's meant for beta testers. Is the "Dynamic Highlighting" one half-checked for you as well?
 
11:23 PM
@Szabolcs yeah
 
@Rojo What do you mean it "kinda works"? It seems to autocomplete and have proper placeholders for me... cc @acl
 
@rm-rf REALLY?
nOT FOR ME
@rm-rf Could you give me a small example that works for you?
 
If the message is not long enough, it looks kinda awkward, but the nice thing is that you can have multiple templates
 
@rm-rf Wow, I couldn't be dumber
@rm-rf Perhaps I dreamed it but I have always had the habit of writing patterns in the templates!
so I tried stuff like f::usage="f[x_] bla" and not "f[x]"
@acl great news! I'm dumb :D
 
@Rojo Hehe... yeah, it's always the actual usage, not the pattern in the template
 
11:29 PM
@Rojo to convert a string box like "\!\(a+b*c-3\)" to fully parsed boxes you can strip off the leading \! and then call MakeExpression on it
 
@rm-rf Glad to learn that
@rm-rf Btw, we scare Leonid
 
JF mentioned that the delimiter between successive messages is a newline, so you can't have paragraphs in usage messages
 
@Rojo you scare me as well. When this site first opened I remember thinking that you and I were of a similar level of skill. Now you've far surpassed me.
 
@Rojo What happened now?
 
@OleksandrR. Nonsense
@rm-rf He said it in the comments yesterday
 
11:35 PM
Maybe if we scare him away really well, he'll go and write part 2 of an awesome book :D So I say, keep on scaring!
 
@rm-rf Room for improvement there's still plenty
at least for me
 
Same here
I should make a sincere effort to learn dynamic and friends really well... at the very least, you can "wow" and convert people you know a lot easily with them than with some clever functional coding which impresses no one.
 
@rm-rf it impresses me. I find Dynamics boring.
 
But you're not someone I'm trying to convert
 
The argument for ImageMargins is { left, right, !!?BOTTOM?!!, top }. I've worked with the worst apis ever, WINAPI, CSS, xaml, and I've never seen such a bizarre organization of parameters. Does the mma api team intentionally make their function arguments non intuitive?
 
acl
11:44 PM
@Rojo fantastic! @rm-rf
@Rojo (fantastic that it works, not that you're dumb--you're not dumb anyway :) )
 
@AdamDreaver Intuition is overrated
 
@rm-rf noooes! it's my strongest trait :(
 
@rm-rf With that objective, I would have to focus on graphics stuff and image processing. I handle dynamics pretty not bad compared to the avarege, but when its useless if I can't do pretty static stuff right :)
I anticipate that tomorrow I will be struggling to export a slideshow to pdf properly. I'm not good at that either
 
@Rojo You're not alone.. .no one has managed to export a presentation to pdf correctly. Unless of course, you make it with beamer
 
acl
@rm-rf typeset mma notebooks to latex would be nice
 
11:48 PM
We really need some sharing mechanisms. Packages, stylesheets
Oh, CDFs too. Dynamic stuff rarely works properly when I save as cdf and show it to someone. Annoying
 
@acl Actually it wouldn't be very hard. In fact, in my SENotebook, I can handle conversion to markdown for the most part. I could also convert to latex (although it's a bit more work), but one could also try the various markdown to tex parsers outthere
 
ConversionRules
cell option
 
Yes, that's one way
I think that's how mma does its conversion to HTML
 
acl
@rm-rf not very hard is relative....
 
..adn tex
 
acl
11:51 PM
@rm-rf markdown to tex is indeed not hard; I didn't think of it.
 
Although, I'm thinking of moving it to bitbucket, since github removed the ability to upload files
 
:)
 
acl
random but interesting article: arstechnica.com/science/2013/02/…
 

« first day (384 days earlier)      last day (4117 days later) »