Announcement: working towards Guard 3.x - what do you think?

[Cezary Baginski]

Guard 2.x has been around for quite a while (almost 2 years). 

There have been lots of fixes, bugs, issues and work to make it as stable as possible.

A lot of lessons were learned too, and a lot of design mistakes were made (and lots of things constantly broken by "yours truly").

More importantly, the technology, tools and people's needs have drastically changed since.

New ways to do things and "better" ways to do things have appeared.

I'm curious - what would you like to see changed/added/removed most in Guard 3.x?

• What frustrates you the most?
  
• What do you like the most?
  
• What do you constantly trip over the most?
  
• Are there any bugs/issues you'd like to see "finally" resolved or implemented?
  
• Is the documentation accessible? Or do you need to rely on "cheatsheets"? Or do you frequently copy&paste from existing projects?
  
• Is setting up new projects convenient? Or is it a drudgery?
• Do you use many of Guard's features? Or do you wish it was slimmer?
• Would you prefer Guard to be more of a "background daemon"? Or would you prefer "more diagnostics"?
• Would you prefer Guard to be "feature packed"? Or more "minimalistic" and "out of the way"?
• Is the DSL nice to work with? Or would you prefer "plain Ruby" without "the magic"? Would you prefer regexp, globs or smarter plugins?
• Are the events accurate? Or are you have having issues you've "learned to deal with"?
• Is resolving problems easy? Or do you find yourself debugging the codebase too often?
• As a plugin author - are the changes annoying and causing unnecessary maintenance? Or are you glad your plugins still work without issues you have to fix yourself?
• Is Guard's support good? Or do you wish it sucked less?
• Do you find Guard superior or inferior to alternatives? (see: https://github.com/guard/guard/wiki/Guard-alternatives)
• Do you still use Guard? Or have you "moved on" to something else? (If so, is there anything Guard could do to "win you back"?)
• Do you still use Ruby nowadays? Or have you jumped ship?
• How do you feel about the project? Does it seem "dead" or "stagnant" or "no longer relevant"?
• Is it easy to find plugins for features you want? Or would you prefer a "list of recommended plugins"?
• Does Guard properly support your platform/tools? (Listen, OS, Readline, plugins)?
• Is Guard "too slow" or "too resource hungry"?
• Is guard "easy to master" or is it "intimidating" beyond simple use?
• Does Guard feel "bloated" or "too minimalistic" (compared to an IDE)?

Here are some initial plans for Guard 3.x:

• remove deprecated functionality and simplify codebase
  
• implement globs along side regexps (https://github.com/guard/guard/issues/649)
  
• simplified documentation (easier to quickly find the info you need)
  
• more flexible DSL (custom matching, fined-grained config, "less magic", better errors)
  
• better info about what's happening (events received, how they were matched and handled)
  
• separate the interactive session from the output (to help avoid Pry/JRuby issue)
  
• support sockets and custom events to help define workflows and communicate with editors better (jump to error, etc.)
  
• allow using an plain Ruby API instead of the Guardfile DSL
  
• support smarter plugins (with a simple "auto-setup" with good defaults)
  
• better support for "one-shot" tasks, plugins and tools without specific Guard plugins
  
• integration with non-Ruby tools and backends
  
• possibly manage tasks from a Gemfile (instead of needing to add templates to the Guardfile) for lightning-fast setup
  
• making it easier to define complex workflows and sequences in a Guardfile
• more control over configuring watch rules and settings
• plugin version detection in Guardfile

If you have any ideas or issues - please add them and/or vote for them.

Thanks!