[Design] What is the minimum feature set we can put up with in an early release of Chandler?

[Ted Leung]

On Jul 14, 2006, at 3:08 PM, Katie Capps Parlante wrote:

> Mikeal Rogers wrote:
>> I guess I just took for granted that one of our end goals was a  
>> complete replacement for people's current email clients.
>> I think our long term goals should be as ambitious as possible and  
>> I hope that one of our long term goals is a full email client  
>> replacement. I assumed Sheila's email was more about short term  
>> email in Chandler, and on that I agree that we should be trying to  
>> integrate with existing clients rather than trying to replace them  
>> while I do believe it is important that we have enough of our own  
>> email functionality in Chandler that we can show off the other  
>> integrated features that tie in to email as an input and  
>> communication medium. Email is just too integrated a part of  
>> peoples lives that we can expect them to move completely to  
>> Chandler for email unless we have everything they already use, but  
>> we'd loose out on a lot of valuable users if we waited to try and  
>> attract them until we were totally finished with email in Chandler.
>
> I think we're all in agreement about the long run -- email is very  
> important to the full vision of Chandler.

Yes, yes, yes.   And RSS too, but don't get me started on that...

>
> The short run is the harder problem.
>
> The question I'm asking: Is there a bare bones feature set that  
> would allow early adopters to use Chandler as an email client? What  
> is that feature set?

Ok, so after yesterday's discussion, I started to think about exactly  
what the feature set would be for me.     When I "came in" this  
morning, I watched all the operations that I did as I processed all  
three of my IMAP accounts.

Here are the actions that I actually did:

* view messages by thread
* delete a message
* select a message and  execute an Applescript/zsh script suite to  
kill comment spam postings off my blog - I think we can safely call  
this an outlier feature
* select  messages and use Mail Act-On! to execute an applescript  
that examines each message and files it correctly by looking at  
combinations of the mail headers - call this "filing"
* click on a URL embedded in a browser
* mark some read messages as unread for later re-processing - give me  
a triage based workflow for this
* reply to messages, including editing the to/cc lines because our  
default list reply to is wrong, i did work on several replies  
simultaneously.
* send new e-mail messages (to remembered previous recipients)

That's it.

There are a few more actions which I do less frequently, which I  
didn't do this morning:
* paste text from the clipboard into a composition/reply window
* verify/decrypt a PGP signed/encrypted message.   Bear and Heikki, I  
blame you... - consider this one very optional
* search for a particular message
* look at the raw source of a message - as a way of doing an  
antiphishing check before deleting
* work offline
* send again
* download an attachment
* attach a file via a standard file dialog (I don't do drag and drop  
attachments)

Things I don't do:

* compose HTML mail

>
> We've asked this question before, and the general answer was that  
> email clients have to be quite good before people will switch. The  
> reason I'm bringing it up again is that Philippe had made noises  
> about being willing to use Chandler as an email client even if it  
> had few features and paled in comparison to other existing email  
> clients. We certainly couldn't rely on all users being willing to  
> do this, hence the "bridging the gap" thread.

For the work that I am doing, being able to triage and effectively  
organize my mail would be a big enough gain to compensate for the  
loss of "full featuredness".   I know that I am not everyone, but I  
am saying that my information management problems are pressing enough  
that I would be willing to try before we could match Mail.app or  
Thunderbird on feature checklist scorecard.

>
> I agree with you that there is a real benefit to having 'plausible'  
> email features in the short run, to give a taste of where the  
> project is headed. It would be great if those features that  
> demonstrated plausibility actually enabled some early adopters to  
> use Chandler as a primary email client. (Perhaps this is wishful  
> thinking).

Well, now you have one person's list...

Ted