@Heike @Szabolcs, @yoda, Heike wants pink ;-) You can make your version pink easily: Download the script from here github.com/halirutan/Mathematica-Source-Highlighting/blob/… instead of installing it right away. Open the js file with an editor and go to line 42. There you see the color codes for strings, keywords, etc. When you are finished, just open this file with Chrome and it asks you whether you want to install it
@Heike Don't forget to go to Tools->Extensions and remove the old one first!
@yoda @yoda I would be disconcerted if the mods could see votes. I think only actual sysops should have that access. BTW, I have no strong desire to be a moderator and I would be quite happy to let it fall to someone else, but I also want to keep this community strong and if that's what's needed I'll do it.
It appears that if I link a comment to an earlier one the system adds the @name automatically; is that correct?
@Heike But the main point is, that colors used for code should fit into the whole page. Look for instance at Java.SO here stackoverflow.com/questions/9055728/… See how there are 3 different shades of blue which don't look good together? Keywords, classnames and the links in the right panel. Looks awful.
@halirutan I agree that the colours should match (I'm actually quite pedantic about that when I present stuff), but I guess at the moment it doesn't really matter until we know what the colour scheme of the site is going to be.
@Szabolcs These are predefined token names from prettify: tag - tags in html, xml, etc, lit - literal, pln - plain. I just used them as they were, for instance "tag" for braces
@Szabolcs When we would have undefined symbols colored and operators black, they cannot both be in the same coloring class. But there are lot of color-classes available.
@Szabolcs in your version maybe keywords and operators would be in the same class: simple black.
@halirutan yes. Is it easy to modify the script that way (i.e. does it detect words in any way?) I'd just like to try it here, and I see I need to modify more than the CSS
@Heike Instead of changing the style, I shuffled around the highlight groups. You can search for things like PR_PLAIN, PR_KEYWORD, PR_TAG, etc. I put the system symbols in PR_TAG, the operators I kept in PR_PLAIN and the undefined keywords I put in PR_KEYWORD. The styled. I'll upload the file, so you can play. Wait.
@halirutan I browsed some old code-heavy SO posts a bit, and they're great. I don't see any major issues that need fixing. Two minor things (i.e. maybe nice to have but no need to bother) are unnamed patterns such as _ or _Integer and a few common system keywords that are not highlighted. Do you think we can have an extra "common keyword" list that will manually be added to those that are in FUnctionInformation?
I'm speaking about things like $ContextPath, $Failed, $OpreatingSystem (okay, this last one is not common, so maybe leave it alone --- it')
it's just an example I saw
Or did you exclude these on purpose because they are not functions?
@halirutan Yet another minor thing: what do you think about highlighting True/False as literals? Well, maybe not because people will also ask about Null, None or even Automatic then, which would be way too much.
It's really great we have this, I think it's almost ready to use (or fully ready to use based on what I saw so far, but let's see if anything major turns up).
The nice thing about FunctionInformation.m is that it has info about all standard packages
Yet another minor thing is things like \[Transpose] or \[Alpha]. These are really a single character, but the highlighter doesn't know this. Again: when I say minor thing, I really mean minor: don't really bother fixing it unless you really feel like. Just posting here everything I notice.
No, I'm using names right now. I would use Names[RegularExpression["\$?[A-Z].*"]] which contains the $variables too. True and False should be highlighted correctly right now, since they are in this list.
@halirutan I can try to extract for you from reference.wolfram.com/mathematica/guide/… if you want, but how about auto-detecting it based on the \[ and ] delimiters?
I just looked into FunctionInformation.m and a lot of stuff is missing. I think this file is for correct highlighting of functions (i.e. what options they take, etc.) Maybe Names["..."] is better.
@halirutan note it's a bit tricky: some of those special characters will get displayed as special in the front end, even in strings. I used InputForm to prevent this. If you export to a text file, it'll look fine.
@Szabolcs this sentence makes me remembering this one "Theoretical math is cool as shit. You know what's not cool? You being an idiot. Math theory involves solving problems like finding patterns in prime numbers, topology and thinking about the different kinds of infinity. And yes, there are different kinds of infinities." from thebestpageintheuniverse.net/c.cgi?u=math
I'm trying to figure out where LaunchKernels[PBS[], 2] fails for me. No error messages whatsoever ... just returning $Failed. And I hate to Trace, especially in command line mode ...
@szabolcs @rcollyer @yoda @verbeia @acl @heike @mr.wizard @MikeHoneychurch and others. Do we have sufficient moderators at this time? If not, I'll step forward and volunteer as well.
well ... this package is full of inter DebugPrint symbols, so glad I read @Brett's post about these and know I need to do Block[{DebugPrint = Print}, ... ] now :-)
I don't know how many we need @Sjoerd We have three candidates. You have my support of course. You should talk to the other volunteers and see what they say. Especially to @rcollyer, I think he's a bit short on time
I've no idea. @yoda said he expects the mods will be chosen next week, based on his experience. I think SE is watching our meta thread. You can post there and volunteer --> meta.mathematica.stackexchange.com/questions/154/…
Let's just not create the impression that there is competition between the candidates because this is not the case at all
I'm not sure whether four moderators would be acceptable, but if so, I would like to offer my services as well. This in addition to, and expressly not in competition with, those already offered by rcollyer, Mr.wizard and J.M.
I'm a long time user of Mathematica (about 20 years). I participated i...
@halirutan I know I am in a good timezone, but my real-life work would preclude me from committing. I have been visiting every day lately, but things can change to make that impossible, especially when I am travelling. My job involves a lot of travel, just not in the December-January period.
@SjoerdCdeVries I was thinking about this one: "The major problem — one of the major problems, for there are several — one of the many major problems with governing people is that of whom you get to do it; or rather of who manages to get people to let them do it to them. To summarize: it is a well known fact that those people who most want to rule people are, ipso facto, those least suited to do it. To summarize the summary: anyone who is capable of getting themselves made President should on no account be allowed to do the job. To summarize the summary of the summary: people are a problem."
@Heike Even before that I read an SF short story about a planet where the president was elected from those that least wanted it (for a year). Might be Brown, but I forgot.
The "Deep Thought" episode in The Hitchhiker's Guide to the Galaxy by Douglas Adams appears to be a parody of his [Fredric Browns] short story "Answer".
@MrWizard haha, I was referring to the fact that you used to (perhaps in your early days) leave comments like "I'm not voting for you, and I'm letting you know because only moderators can see who voted for each post..." :)
@acl I'm not very familiar with the history of greece but I thought that in Athens those who were in the council were often severely punished for not doing their job.
@Heike neither am I to be honest. but from what little I remember from school yes, they seem to have gotten severely punished for not being perceived to do their job
In general, form in Trace[expr,form] is compared both with each complete expression that is evaluated, and with the tag associated with any transformation rule used in the evaluation.
