3 ms·
In my mind, there are a handful of things that are essential to getting Thunderbird back to a usable state... some of these could be plugins... First, the exch
by tracker1 10y ago
In my mind, there are a handful of things that are essential to getting Thunderbird back to a usable state... some of these could be plugins...
First, the exchange/calendar integration options clearly suck... establishing a clear calendar interface as a built-in with extensible points for plugins for authentication/sync of calendars would be a good start.
Second, likewise with calendar auth/sync would be an extensible interface for folder sync and authentication, so that a cleaner integration for common providers based on an underlying IMAP can be used... this way the conventional "junk, spam, inbox, sent" folders could be presented the correct way as well as the underlying storage for a given provider.
Also, along with calendar/email would be more extension points for scheduling, contacts, etc...
As it stands, even if there were different plugins for a google calendar and an exchange/o365 calendar, contacts, etc... if the underlying pieces can be shared, it would be a better user experience.
Moreover would be some serious reconsideration regarding the UI/UX... I'm a big fan of material design, but some variation on that coming a lot closer to a Gmail app for desktop would be a really nice start... but getting a calendar/task/contacts integration points and primitives for extension would go a long way here. Having the core UI going the same direction as Servo, and having most of the UI/Extensions being HTML/JS based would be nice.
Likewise, NPM compatibility for extensions' modules would be nice as well.
- Psilidae 10y agoRegarding NPM compatibility: I'd be worried about potential security issues with something like npm. I recognize that one must already put a lot of trust on developers when using extensions, but I'd worry NPM would make it a lot easier for a nefarious developer to add malicious code to a really deeply buried dependency.
- tracker1 10y agoMaybe it's just a matter of using node/npm tooling to develop/package an extension... using a bundler. Of course, at this point, it might be worth re-doing the whole thing using electron as a base.. the more I use VS Code, the more inclined I am to feel that it may be good enough, and if it's a react app, maybe target react-native for phone/tablet platforms, and electron for desktop.