Alex Hudson

Thoughts on Technology, Product, & Strategy

Thunderbird & calendars

When the new “MailCo” (as it was at the time) was announced, the first thing they did was poll people on what changes they should be looking to make to Thunderbird as kick-ass as possible – you can see from the initial launch blog post that integrated calendaring was item #1 on a list of new features.

Now, only one year on, the word is that calendaring won’t be integrated or bundled at all – in fact, it gets worse, because the calendar developers have also announced that a number of full-time developers have been lost. Doesn’t say why, but I’m assuming you can blame the Global Downturn™.

I can’t express what a huge mistake I think this is for Thunderbird. One of the main reasons I got involved with Bongo was to provide a calendar server for my day job’s choice of mail client (Tbird), and when Jonny produced his CalDAV code things were really looking pretty sweet.

The problem for me comes down to basically one of maintenance. Extensions that you can install are great, but they have a tendency to bit-rot and suffer versioning issues. Even on Firefox you have issues with a version upgrade making an existing extension wobble or fall over; and with Tbird3 being developed at a more rapid pace, there is a real problem there.

The calendar devs can only concentrate on one thing, and it seems like that one thing has to be a plug-in. In my opinion, a non-working calendar in Tbird3 ought to be a release blocker, but that is now off the table. I have to say, if Evolution on w32 was a viable option, I think that’s where we’d be looking right now, but I’m going to reserve judgement until Tbird3 releases.

Previous

Maintaining bits in Fedora…

Next

Government Action Plan: Open Source, re-use, etc.

3 Comments

  1. All good points. One (annoying 😉 question: you say that a non-working calendar in Tb3 should be a release blocker. How long would you delay Tb3 by until the calendar capability was finished? 1 month? 3? 6? 12?

    Note as well that while there certainly are challenges to maintenance of add-ons, the calendar team has shown that they can move that plugin along and make it better while the Thunderbird 2 codebase wasn’t evolving. In other words, decoupling the two makes it more likely, IMO, that we’ll end up with a good calendar integration and a good email client than if we delayed the release of the latter until the former was ready.

    But yes, tough choices all around.

  2. Alex

    Honestly, I would delay it until it’s done, and if necessarily I would reduce the scope of what “done” means so that “until it’s done” doesn’t stretch off far into the distance. I don’t know what the plans on your road-map are, but honestly the other stuff I’ve seen about Tbird3 doesn’t give me any good reason to move up; UI changes are nice, autoconfiguration is nice, etc., but it’s all just “nice”. I have a deep, deep suspicion about Gloda because I don’t want to see large chunks of cached data in my user’s roaming profiles, so that’s probably going to be turned off, and aside from that I’m not sure there’s much else to talk about. Calendar, for me and I’m sure others, is the USP.

    If calendar integration can really work, then great. At the moment, I’m sitting on Tbird 2 with Lightning 0.8 installed. Most times I start it up it tells me 0.9 is available: great, but I can’t actually install it because for some reason it’s incompatible and doesn’t work. Same with EnigMail. In my experience, the better integrated the extension, the more brittle it is, particularly with the UI hooks points via overlay – the Mozilla UI extension mechanism leaves me cold because it basically turns UI into API and any small change can throw stuff into a spin.

    I think we both agree on how important the calendar is. I worry that being an extension necessarily metes out second-class treatment: e.g., you’re not going to hold back a security update if it breaks the calendar. You’re not going to spend time doing QA on it. You’re not going to ensure Tbird3 releases have working calendar from day one. (At least, if you are doing those things, there’s no point it being an extension). If we’ve learned anything from Linux, it’s that stuff “outside the tree” just rots. With a relatively part-time bunch of people on the calendar, I don’t see how it’s going to be any other way. I just don’t see what is going to make Tbird3 more compelling that Tbird2, because right now I don’t see any substantial difference.

  3. This really, really sucks.

    I’m trying to get by with t-bird in a corporate setting, but honestly the lack of calendar is crippling me. I’ve got a half way solution using Apple’s iCal, but iCal really wants Mail.app for sending invitations and confirmations (maybe because of bugs like https://bugzilla.mozilla.org/show_bug.cgi?id=472891 which make third-party mailers a risky dependency?)

    I agree that tbird3 without calendar is not worth shipping. Damn.

Leave a Reply