Help us finish the version 0.8.0 release

[Kenny Evitt]

Magnus and I have been trying to figure out how to finish up the LT migration from node-webkit/NW.js to { Atom shell / Electron } that the core team had mostly finished.

We can build the *atom-shell* branch on Mac OS X, Windows, and I believe others have built it on Linux too. And the builds are working.

But we could use more 'eyes' – to create new issues for bugs, review old issues to determine if they can be reproduced in the new version, etc.

And we could use more 'hands' – to create pull requests with bug fixes, changes or additions to the documentation, etc.

[Simon E.]

Does that eyes and hands include the build process itself? I've finally managed to build the atom-shell branch using Windows/Cygwin64, but it had some hitches. I'll start using it and report any issues I find, in any case.

I've been using LightTable lately for some Clojure related stuff and think it's pretty awesome so I'd like to contribute if I can.

[Kenny Evitt]

Yes, build process included.

--
You received this message because you are subscribed to a topic in the Google Groups "Light Table Discussion" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/light-table-discussion/Qemk-Uet2y8/unsubscribe.
To unsubscribe from this group and all its topics, send an email to light-table-discu...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Fred Eisele]

So, what you are looking for is...

Build and install LightTable according to

https://github.com/LightTable/LightTable/blob/atom-shell/doc/developer-install.md

Go through old and pending tickets to verify that there are no regressions

https://github.com/LightTable/LightTable/issues

Use LightTable to do normal work type things, when a problem is found and it is not ticketed make a ticket

Select an outstanding ticket, fork the repo, fix the problem, make a pull request.

Does that about cover it?

[Kenny Evitt]

Fred, that all would be fantastic. But any combination of the above would be great too, so no pressure!

[Magnus Rundberget]

Appreciate the push and the effort Kenny !

I'll try to allocate some more time going forward to try and get this release done.

cheers
Magnus

To unsubscribe from this group and all its topics, send an email to light-table-discussion+unsub...@googlegroups.com.

[Simon E.]

Should we mention in issues that attempts to reproduce issues in the *atom-shell* branch are part of the effort for the version 0.8.0 release? Just asking because it's not immediately obvious without reading discussion that this verification effort is for a new release.

[Kenny Evitt]

Simon, that's a good idea; so yes.

On Thursday, June 25, 2015, Simon E. <sd.es...@gmail.com> wrote:

Should we mention in issues that attempts to reproduce issues in the *atom-shell* branch are part of the effort for the version 0.8.0 release? Just asking because it's not immediately obvious without reading discussion that this verification effort is for a new release.

--

You received this message because you are subscribed to a topic in the Google Groups "Light Table Discussion" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/light-table-discussion/Qemk-Uet2y8/unsubscribe.

To unsubscribe from this group and all its topics, send an email to light-table-discu...@googlegroups.com.

[Tarjei Knapstad]

I haven't gone through much detailed testing, but the atom-shell branch builds without problems on Fedora 22 x86_64, and I haven't detected during a short session of editing html/js sources.

--

Tarjei

[Fred Eisele]

Regarding the script/build.sh on Debian
Linux face-conform 3.16.0-4-amd64 #1 SMP Debian 3.16.7-ckt9-2 (2015-04-13) x86_64 GNU/Linux
The command for node is nodejs, I added a link
ln /usr/bin/nodjs /usr/local/bin/node
and LightTable built fine.

Seems pretty trivial, should there be an issue for this?

[Kenny Evitt]

Fred, an issue would be a fine way to document your workaround.

--

[Kenny Evitt]

Fred, I just ran into the same issue on one of my Ubuntu boxes. I created an LT GitHub issue. You may want to consider installing Node.js via the instructions to which I linked in the GH issue – the version available thru the package manager is older. Following those instructions also doesn't require adding a symlink for the Node.js binary.

[~]

Hey so I do not really have a background working with large projects or through GitHub but I figured incompetent eyes are better than none. However, I'm not sure if I have done something odd because I seem to have run into at least one previously resolved issue: https://github.com/LightTable/LightTable/issues/1230 and a number of the stock commands and keybindings are not cooperating with me. I was just going to comment on that issue but seeing as how I was having other problems I thought I would try here first.

I cloned the repository from and up-to-date copy of emacs, not sure if that would effect anything, and I did have to move the cloned directory because it was nested a layer deeper than I intended. Other than that the build was in line with the provided instructions. I booted up Light Table -- which informed me of 'webkitAudioContext' and 'window.webkitStorageInfo' being deprecated but otherwise threw no errors or anything -- and started looking over the changelog and documentation. I'm not sure if this as designed, but the changelog does not wrap text by default, though the documentation does. First keybindings I ran into issues with (and this could just be a matter of out-of-date documentation, since the command list does not seem to reflect this) were those for opening and creating files; the documentation claimed you could use ctrl or cmd, and only cmd seemed to work. This is in line with the command list though, so not a real issue.

Anyway, I opened a clojure file I had previously written (the first project euler problem), ran the instarepl command, got distracted by something in another window, then came back and toggled the "live" button a few times and started writing some expressions I wanted evaluated -- worked fine. Then I called a function defined in the program itself with some variables, and it worked as expected except that a couple of lines had multiple identically-valued watches that I could not account for. As I changed the values in the function call, watches continued to pile up on those same lines, and just those lines.

Here's a picture, I would appreciate if someone has insight into if this is a result of my build, a system-specific glitch, my sloppy code, or a previous issue re-opening. In any case, it is not the expected behaviour.

Thx.

-Anthony

[~]

I realise this may not be the place for the post so if so let me know and I can delete it and redistribute the content elsewhere -- perhaps to the gitter if it is relating to my individual setup.

-Anthony

[Kenny Evitt]

Hi Anthony,

The discussion group is the perfect venue for this kind of thing.

The changelog not wrapping versus the docs wrapping is ... apparently just an artifact of how both are being displayed. The changelog is being shown as a Markdown file whereas the docs are an iframe for the docs site. It's a very low relative priority, but I created a GitHub issue for it. Pull requests welcome!

I'm guessing you're running LT on a Mac, hence your mentioning the Command key. On Windows and Linux, generally Ctrl is used; on Mac it's generally Command. The docs could probably make that point more clearly (and explicitly). Pull requests welcome! Here's the GitHub repo for the docs site.

I'm pretty sure I've seen the same behavior with regard to the "watches" (REPL results); please submit an issue on GitHub for that. Please follow the guidelines for reporting bugs; basically, include the minimum steps to reproduce (and example code that others can run to do so), your operating system (and version), the LT version (e.g. 7.2 or source from master branch as-of 2015-07-08 09:21-05:00), any plugins installed (and their versions), and any relevant portions of your user.keymap and user.behaviors files. This isn't a high priority issue either but hopefully someone will take a look at it eventually.