A web server in your web browser.
Running a local web server allows for easier testing and development without the need for a network. The nohost web server takes this a step futher, and allows web sites to be hosted inside the browser without any server or network.
nohost was created in order to make it easier to do in browser, live previews of complex web sites and applications that span many files.
You can try a demo here.
Like any web server, you must first install what you want to serve. The nohost server uses Filer in order to create a filesystem within the browser.
Files are then installed in the filesystem using the
?install boot option, and choosing a disk
image (*.zip). A demo disk image is included, which can be installed by doing the following:
Once installed, the filesystem will survive the server being started and stopped (i.e., closing the window).
If you want to clear the filesystem, use the
?reset boot option:
Now that files are installed into the fileystem, you can browse them by adding a path to the query string:
- Web Root - http://humphd.github.io/nohost?/
- An HTML page - http://humphd.github.io/nohost/?/webmaker-kits-gh-pages/kit.html
Any path not found in the filesystem will produce a 404:
How it Works
nohost boots the web server whenever you load its
index.html page. Boot options are read from the query string
and then actions are taken using the filesystem. Since the filesystem survives between executions of the page,
there is stability between various requests.
Each request for a path involves rewriting. The file at a given path is read and parsed, and processed so as to
inline any external resources (e.g., images, stylesheets, etc). When the document is done being processed,
it is again turned into text, and
document.write is used to replace the web server's boot page with the new content. Some special pages are synthesized, similar to how a normal web server works. For example: directory listings and images.