Known Issues

Regressions we're working on

The following are issues the DocPad team are aware of and will fix shortly:

It just hangs after accepting the TOS or subscribing to the newsletter

It seems that this is associated to being behind a firewall or a proxy. You can apply this workaround.

I got “npm ERR! Failed to parse json”

Check your project's package.json file with JSONLint, to ensure it does not contain any errors, such as missing semicolons, quotes or commas.

I am getting permission errors after I install things

Chances are this isn't a problem within DocPad, but rather one of your node/npm installations. Run the following in Terminal, once done, try your original action again:

sudo chown -R $USER /usr/local ~/.npm
chmod -R 755 ~/.npm

Be sure that your tmp dir is writable for the current user:

sudo chown -R $USER:$GROUPS ~/tmp

If that fails, we'd recommend either:

When I run 'npm install' on windows, I get 'gyp ERR! configure error'

When I run docpad run on Windows, it asks me what program I would like to open the file in

It turns out that Windows will prefer to treat the local docpad.js file as the executable versus the global docpad.cmd file. To get around this, type docpad.cmd run instead in projects that have a docpad.js file. More information here.

Watching doesn't work; works only some of the time; I get EISDIR errors

File watching is a pretty timid thing that we are currently working on a lot to make better. There are currently two methods we can watch files, watch and watchFile. watch is the default and uses the operating system's watching mechanisms, however sometimes the operating system's watching mechanisms may not be the best. If that is the case, we'd like to switch our watching method to watchFile which is a slower, but does work when watch doesn't. To do this, add the following to your docpad configuration file:

watchOptions: preferredMethods: ['watchFile','watch']

Mac OS X users are most likely to encounter this issue.

Watching is very slow to notice changes

Add the following to your docpad configuration file:

regenerateDelay: 0
watchOptions: catchupDelay: 0

This will introduce problems however if you have previously customised watchOptions, or if your editor uses swapfiles. More info here.

I got “We couldn't find an existing DocPad project inside your current directory…”

This occurs when you run docpad run inside a directory that already has existing files, but doesn't have a structure that resembles a DocPad project. We can't directly ask you if you would like to use an existing skeleton for the basis of your new website, as pulling in a skeleton inside a non-empty directory may overwrite your existing files. If would like to still use a skeleton for the basis of your new website, you will have to run docpad inside a new empty directory. If you would like to start your website from scratch (not use an existing skeleton) then you can follow the Getting Started guide. Hope that helps :) If you need more help then check out our Support Channels.

I'm getting “Could not locate git binary”

This happens when the git installation is not exposed to your PATH variable. You can solve this in either of these two ways:

  • Reinstall git and make sure to select the option during installation that asks if you would like to add it to your PATH variable (may also be called, would you like git to be available to the command line)
  • Manually add the location your git binary resides in to your PATH variable

More information about this here.

I'm getting “EMFILE, too many open files”

As node.js is like an octopus, able to do many things at the same time, sometimes node.js will always try to do too many things than the operating system will let it. In which case, you can increase the amount of files allowed at the same time by running ulimit -n 8192 in your terminal.

How can I make DocPad go even faster?

There are a few things you can do:

I upgraded, and it doesn't work

Check out the Upgrade Guides here

Whenever I output a variable (like content) it is escaped (< rendered as &lt;)?

Template engines by default escape all variable output. Escaping is when we turn things like the open bracket < into it's html entity equivalent &lt;. This helps prevent malicious code accidentally being injected into your website which can open the door to XSS attacks. As such, we have to use a special syntax to keep the variable unescaped when outputted. The special syntax is different for the templating engine your using, so here are the ways we know:

  • Eco: <%- content %> instead of <%= content %>
  • Jade: != content instead of = content
  • Haml: != content instead of = content

The output of a variable (like document.title) is empty or null

Be sure that you use the correct syntax for your template language, read the documentation of your chosen language.

For example: When you want to put the output of a variable into the content of an (HTML) element in Jade, you must not write a whitespace between the element and the =. So this is wrong: title = document.title and that is correct: title= document.title

I get a whole bunch of npm / missing module/package / installation failed errors

If you are using Dropbox (an online syncing & backup tool) and your project is inside your DropBox folder, then click the dropbox menu icon and select “Pause Syncing”. Once this is done, try whatever you were doing again, you may need to run rm -Rf node_modules; npm install as well. Once it's all working, then you're free to resume dropbox syncing.

If you're still experiencing issues, then be sure to post about it on the issue tracker.

The growl notifications aren't displaying?

This means you need to download and install the growlnotify extra from the growl website. What this package does it provides command line applications the ability to call growl which is needed as docpad is a command line application.

The exception raised by the jade plug-in during documents generation makes no sense

The Jade compiler uses the full file content on the disk to show where the parsing error is. But since Docpad strips the meta header before submitting the data to the jade compiler, you must add the number of lines of this header to get the right error spot in your code.

Need more help?

IRC Chat