email rss
What's coming up in imag (39)
May 11, 2019
2 minutes read

This is the 39th iteration on what happened in the last four weeks in the imag project, the text based personal information management suite for the commandline.

This is the first update on what is happening in imag in this year. It took me rather long to get back to hacking, but I’m now back and am looking forward to a hacky hacky year!

So this is a short update on what is happening right now.

What happenend?

On November 10th, we released imag 0.9.0. A lot of things happened since, despite my vacation on the other side of the planet. I will not go into detail about each branch we merged into master, but rather sum it all up in short:

  • “imag-header” was added, a tool to modify headers of entries as well as query information and check values of headers
  • The StoreId type was redefined. This makes handling of objects of that type a bit smoother and easier.
  • libimagentryref was rewritten (almost from scratch). Now we can handle if a collection is located in different pathes on different devices where the store is synced to. (The commit message of the merge has more information if you’re interested in the details)
  • The code for all “mail” stuff was rewritten because of the rewrite of the libimagentryref.
  • The “Module entry path” infrastructure was rewritten and simplified.
  • libimagcontact was rewritten for the libimagentryref rewrite.
  • The imag binary ignores all .d binaries in $PATH if compiled in debug mode. This is a convenience thing for developers.

What will happen

I’m still working on the icalendar/vcard things. No good libraries available still, though. Same old story, really.

One thing I’m looking into a bit more lately is testing and coverage. I am thinking about running some coverage statistics tool on the imag codebase and publish the reports for each release starting with the next one. That would visualize how badly imag is tested at the moment and maybe drives me to add more tests. I’m also thinking about adding some more sophisticated testing framework in the codebase, so that we can test with quickcheck or other cool tools that are out there already.

As always: Patches would be really welcome!


Back to posts