one-more-minute/Jewel

Mike J Innes

@MikeInnes

Woops, you're right – well spotted! Well, it'd still be nice to have a rainbow-vars plugin at some point, but that does solve some issues.

I was actually wondering if match-highlighter uses the DOM to find tokens – if it does then it will support unicode once the parser does with no extra effort, which would be nice

Mike J Innes

@MikeInnes

Ok, I got rid of the rainbow vars in the light theme, that wasn't working – but I had a go at a more pastel-y/zenburnesque dark theme which I think works quite nicely with it.

Sebastian Pfitzner

@pfitzseb

Just one random thought I had while staring at some code:
Wouldn't it be pretty cool if some plugin not only highlights the exact same variable but also any it is assigned to? So you'd have something like a common thread running through your code, branching when necessary? I think it's pretty important to be able to immediatly see data flow...

Although that should probably be a mode in addition to conventional syntax highlighting... But still, I somehow like it. Thoughts?

Sebastian Pfitzner

@pfitzseb

While playing a bit with your css I noticed that it doesn't look half bad if function calls are a bit transparent and bold (e.g. .cm-s-june-night span.cm-variable-2 {color: #ffc; font-weight: bold; opacity: .55;}).
And btw, thanks for fixing that backwards-compatibility-thing!

Mike J Innes

@MikeInnes

Some kind of data-flow-based highlighting is an interesting idea – you could take x = f(y, z) and colour x as a mix of y and z's colours, too. It would be fairly hard to do well, I think, but it would definitely be worth a shot.

Reducing the opacity is a nice way to distinguish functions from variables without making them really pop out – I'll play around with it a bit. For the record, I do think that bold functions work nicely with your theme. Oh, and no problem!

Mike J Innes

@MikeInnes

Backtraces should be much nicer now, let me know if there's any problems with the new code. Also, the highlighter now supports splicing ($(...)), which I think is pretty neat :)

Sebastian Pfitzner

@pfitzseb

About the data-flow-highlighting: While it does sound pretty neat in theory, I'm not at all sure if it can be implemented in any reasonable fashion. Might have to look into that... I think I'll just throw together a sketch for something a bit more feasible (hopefully).

About backtraces: Sadly, there still are some cases where it's not terribly helpful yet. One case that is pretty annoying for my style of development:
There's a function defined in some file and I try to call it within that same file. But if I've made a mistake in the function definition the line numbers displayed in the error message is about the representation of that function in init.jl and the line numbers displayed aren't the ones I would want to see.

Sebastian Pfitzner

@pfitzseb

What I had in mind was something similar to the picture above. Whenever your cursor is above some variable (and/or some key is pressed), that variable and it's "first-order-children" (or whatever those might be called) are highlighted while the rest of the code is a bit dimmed. Otherwise it looks just like the version to the left.
That seems to be a compromise between rainbow vars and conventional highlighting - which of course doesn't have to be a good thing ;).

Mike J Innes

@MikeInnes

I like that idea for highlighting matching variables. It'd probably be fairly straightforward to prototype something like that if you're interested in having a go.

Could you give me an example of the kind of backtraces you're getting? That's exactly what I had tried to fix, on my system backtraces are very to-the-point. Are you sure you have the latest Jewel.jl code (the fix isn't in a released version yet)?

Sebastian Pfitzner

@pfitzseb

Ugh, stupid me... Works like a charm! No I finally don't have to debug in the REPL because I won't get proper backtraces in LT. Thanks!

Mike J Innes

@MikeInnes

Great!

Mike J Innes

@MikeInnes

Just so everyone knows – I'm going to change Jewel's URL to one-more-minute/Jewel-LT. gitter.im doesn't handle renames AFAIK so there won't be a redirect, but the URL of the chat room will change accordingly.

Mike J Innes

@MikeInnes

0.5.0 released yesterday, with much better highlighting/indentation support, new commands to interrupt and and kill the Julia process and improved error messages

Ethan Sherbondy

@sherbondy

Loving the project so far. Is there a means for the inline evaluator to run relative to the path of the file being evaluated rather than the LightTable directory? e.g. run(pwd) yields /Applications/LightTable.app/Contents/Resources/app.nw. Is the expectation to launch a terminal client instead?

Mike J Innes

@MikeInnes

Sure, you can change the working directory at any time with cd – in your case you probably want to put cd(dirname(@__FILE__)) at the top of the file and evaluate it before anything else. You can also add a cd call to ~/juliarc.jl to have it run every time Julia is booted.

Oh, and thanks!

Mike J Innes

@MikeInnes

(also, just in case you don't know, include statements will always work relative to the path of the file they are called from whether you cd or not)

Mike J Innes

@MikeInnes

If anyone on master is using the #jewel module ... directives, they're no longer necessary – you should see the current module come up automatically in the status bar

Sebastian Pfitzner

@pfitzseb

Hey, I just had a look at this after some time - the changes and added functionality are very impressive! Just one minor nitpick: I think it'd be more consistent if any css changes were to be done by June and not by Jewel itself. Except for that fancy error background (.CodeMirror-linebackground.error) that isn't implemented anywhere else.

Mike J Innes

@MikeInnes

Thanks!
Unfortunately a couple of patches are currently necessary (at least until I get some PRs merged into LT) and it'll always be good to have some defaults for things like data frames. (Although I notice there's a font change there that really shouldn't be.) I'm going to do a lot more work on graphical output and CSS so I'll start making sure things are a bit more organised then.

Ismael Venegas Castelló

@Ismael-VC

Hello! I'm using Arch Linux and I just instaled julia from the repos, then Pkg.add("Jewel"), finally I install the Jupiter plugin in Light Table, but I get this error: http://bit.ly/UXFvG8

I mean I get that error when I try to evaluate some Julia code with ctrl+enter.

Ismael Venegas Castelló

@Ismael-VC

Julia Version 0.3.0-prerelease+2954
Commit 3985890 (2014-05-08 04:14 UTC)

Mike J Innes

@MikeInnes

Look like gotos were only added to Julia recently, so the slightly older version you've got doesn't have them. I'd better fix that so it doesn't cause problems.

Sebastian Pfitzner

@pfitzseb

I ran into that problem too, but with the latest prerelease it works flawlessly. Not sure if you really need to do something about it...

Ismael Venegas Castelló

@Ismael-VC

ok I will compile julia insted of using the one provided by the Arch repos and test again then, thanks!

Mike J Innes

@MikeInnes

@Ismael-VC Great, let me know if you have any other issues!

@Varanas TBH it's a tiny fix, and it might be a few weeks before all the prebuilt copies of Julia are up to date, so it's probably worthwhile

Ismael Venegas Castelló

@Ismael-VC

@one-more-minute Sure I will certainly do so!

Sebastian Pfitzner

@pfitzseb

Oh, and those inline plots are pretty neat! Although I'm getting an AssertionError: path must be a string when using Gadfly (the plot is still displayed). Anyone else seeing that?

Mike J Innes

@MikeInnes

Try Pkg.clone("Gadfly")

I've got a patch for the JS but it's not out yet. If you get the master version you'll get interactivity

Mike J Innes

@MikeInnes

@Ismael-VC If you have problems doing the build then Pkg.update() should fix the Jewel problem for you

Andre Pemmelaar

@Andy-P

I upgraded the light table plugin today. seem to have lost line-wrapping of error msgs. That's making it hard to debug. Did something change? If not, how do we force line-wrapping of error msgs?

Mike J Innes

@MikeInnes

I have changed error messages a little but didn't intend to remove line wrapping. Would it be possible to send me a small snippet that causes this kind of long error message? Or perhaps a screenshot if I've misunderstood you? This should be fixable anyway.

Andre Pemmelaar

@Andy-P

I'm not at the machine that caused the error, but as I recall it was a @test macro used for unit testing. The problem was on a Windows 64 bit machine. I also use light table on a mac. I upgraded that as well, but am not experiencing the problem there. I will send the snippet tomorrow when I get to work.

Try @test 1 == 2 to see if that generates the problem

Mike J Innes

@MikeInnes

Ok, I think I see what you mean – if you evaluate `@test` with a long expression the header trails off the screen, right? (I'm on a mac too, but I can test on Windows though if that's not it). When I can I'll patch this anyway

Ismael Venegas Castelló

@Ismael-VC

@one-more-minute I've just compiled from source, and updted everything, now I get this: http://bit.ly/1x6hsCF

Mike J Innes

@MikeInnes

Not sure about the first one but the second one looks to be from an outdated plugin – do you definitely see Jewel 0.6.3 in the plugin manager?

Ismael Venegas Castelló

@Ismael-VC

julia> pkgs = Pkg.installed(); pkgs["Jewel"]
v"0.6.3"

Ismael Venegas Castelló

@Ismael-VC

ismaelvc@toybox ~ % julia -v
julia version 0.3.0-prerelease+3956
ismaelvc@toybox ~ % pacman -Qs lighttable                                                                   
local/lighttable 0.6.6-2

The Jupiter plugin is also updated.

Andre Pemmelaar

@Andy-P

Screen Shot 2014-07-01 at 7.12.30 AM.png

@one-more-minute I think you have already identified the error, but just in case, here is a screen shot of the problem I described earlier.

Mike J Innes

@MikeInnes

Great, thanks @Andy-P, I'll see what I can do

Andre Pemmelaar

@Andy-P

The problem is now on both my mac and windows after upgrading. Also,,The font for error msg seems to have changed.

Mike J Innes

@MikeInnes

Oh yes, there's a CSS definition in Jewel which shouldn't be there – I'll fix that too

Andre Pemmelaar

@Andy-P

Despite this minor problem, I think lightTables + Julia is brilliant!

Finding it very productive!

Where communities thrive

People

Repo info

Activity