Python usually, as far as I can tell. Either language is fine, I just think these things should be done in something that comes pre-installed on more systems.
I mean, I knew Chris years ago and his language of choice was Ruby, and I don't think it's changed.
Sure basic install-ability is great, but I bet you this tool was made to satiate an itch he had, and he used the tools he's used to. Little apps like these aren't engineered. Just hacks, for hacks sake.
You know that once a user script is installed in chrome you can edit the script.js file in the extension's directory on the fly. It's a little more work than simply putting a file in ~/.js but it doesn't require a HTTP server to function.
True, and this is why I love dotjs. Thanks for making it!
Maybe a good followup project would be to use some of the dotjs code to help make a filesystem-based GreaseMonkey system for Chrome, which would be just as easy to manage as dotjs scripts are.
Unless you don't use Firefox and you tend to have a lot of dot files that you hack around with in your home directory. Then it starts to seem pretty awesome.
Did you investigate requesting the files directly from the filesystem, like "file:///home/monkey/.js/google.com.js".
You'd need to tell users to enable file access for the extension though chrome://extensions/, since that isn't enabled by default and users have to opt into it. I also don't know if there is a way to get the current user's homedir without hardcoding it.
Also: you're being super careful to not allow arbitrary reading of the filesystem, right? A good sanity check would be to ensure that realpath() of any file served is under ~/.js before handing it off.
ahahahahah what? Dude, do you work for Microsoft? "Well see, we just have to run MS-SQL server (DESKTOP OMG EXPRESS ENGINE) here, and then we run IIS on top of it to serve this file listing..."
It's a limitation of Chrome. This isn't the first time that someone tried to use a local webserver to do something that Chrome doesn't allow. IIRC, there was someone that used a webserver to implement a hacky version of It'sAllText! for Chrome.
Twitter previously used some stock bird art, but the logo was just the bubble letters. Now they have an original bird logo. But 3rd party Twitter stuff can use arbitrary bird art, or maybe the old stock one since it's unclear who owns it.
I think a good solution would be a new open license logo, like the Minefield logo as opposed to the restricted Firefox logo.
I've been wanting OAuth for stuff like http://calendaraboutnothing.com/. There's no way we should advocate entering GitHub passwords or API tokens into a 3rd party web apps.
Agreed! I want to build our entire community site around Github, but I don't want people to provide their usernames/passwords, and I'd prefer a way to specify whether or not they're providing access to SSH keys, which I don't need.
You read me correctly, I don't see why load balancing should be taking place down at the app server level.
Aren't you going to have multiple machines each with their own blessing of unicorns? You're still going to have to use some kind of load balancer in front of independent sockets.
https://github.com/defunkt/dotjs/blob/master/bin/djsd#L26
(Also, Ruby comes with OS X (just like Python) so it seemed a sane choice.)