An AMD-compatible build tool.
We are very excited about RaveJS, the successor to curl.js and cram.js. At its core, rave uses an ES6 Loader, rather than an AMD loader. However, rave can load many types of modules, including AMD. Therefore, rave will work with your application's AMD modules.
Rave's most exciting feature, however, is it's ability to eliminate the drudgery of bootstrapping and maintaining modern, modular web applications. If you've built a non-trivial AMD-based web app, you should be excited about rave, too.
Rave is definitely the future, so we're putting all of our effort into ensuring that RaveJS is as awesome as it can possibly be. Watch these RaveJS github projects to stay up to date:
This means that all development for curl.js and cram.js has stopped. For the foreseeable future, we will continue to respond to issues on github, as well as in the #cujojs room on freenode. However, keep in mind that there will be no further development, so some issues might not be resolved fully.
If you're as excited as we are about the future, consider helping us improve the curl-to-rave migration guide. There are many ways to architect curl+cram-based applications, so your experience migrating your app from curl+cram to rave is extremely valuable to other curl+cram users.
Also, if you are interested in becoming the lead maintainer of curl.js and/or cram.js, please let us know on #cujojs!
See the docs/ folder for instructions and useful information.
New! gulp integration via gulp-cram
To run cram from any folder:
npm install --global cram
- Add deprecation notice.
- Add links to Rave Starters.
- Add a link to the migration guide.
- Update links to gulp-cram and bclozel
- Fix module parsing code to stop prematurely detecting the end of an
AMD define when there are regexp literals inside parentheses.
- Only declare global define() during bundling to help isolate problems
with UMD modules running under node (as with gulp and grunt).
- Separate argument-parsing code from cramming code so logic to run
as a module or as a command-line tool can be done more sanely.
- Fix logic to detect whether cram is running as an npm task and ensure
that the promise returned from the API only resolves once the bundle
- Stop appending to the cached AST data. It should be overwritten.
- Correct regression:
--exclude CLI option now works again.
- Handle when factory args > dependent module ids when simplifying AMD defines.
For instance, the curl/loader/cjsm11 cram plugin adds an extra arg now.
- Stop failing when plugins output more than one
- Fix compile bug that prevented use of css! plugin.
- Start using when.js 2.x.
- Fix configuration merging errors when inspecting both html and run.js.
- Ensure that baseUrl and output are converted to absolute paths so
node doesn't try to resolve them using node_modules logic.
- Bug fix: plugins are now responsible for resolving url/filepath.
- Bug fix: Normalize r-value require()s when using loaders, too.
- Bug fix: Infer appRoot from run.js file, not just index.html.
Improve logged messages a bit, too.
- Bug fix: Excluded modules don't get compiled anymore.
Force 'curl' and 'curl/_privileged' to be excluded in all cases.
- AMD modules are now published to npm, so manually running bower is
no longer necessary.
- Failing to supply some AMD configuration properties no longer causes
cram.js to throw. (Thanks Blaine!)
- Lots of fixes to the code that groks html files.
- Internal refactoring and reorganization
- Use same logic as curl.js when assigning configuration to modules that
may be loaded by a plugin or module loader.
- Add a legacy loader example.
- Stop detecting code such as
goog.require('id') as a CommonJS-style
- Improve support for i18n and CommonJS modules via curl.js 0.8
- Improve module parsing (fixes some issues with lodash).
- Parse literal RegExps better.
- Include simple example apps.
- Find build override json files without absolute paths.
- Fix many other minor things.