Re: Devin's efforts

Tanstaafl,

It is clear you really like TB and it is true that most of the operational 
basics can be achieved using plugins.  I switched to TB when I couldn't get 
MB to work.  It was my own error, one I've since, apparently, rectified, 
but as such I came to TB as a very happy MB user and saw the interface and 
thought "WTF?  Why would anyone do this to email?"

Of course the answer is "because they like it like that."   I'm still not 
sure why, but so it goes.  I managed to get TB to behave more like MB by 
seeking out plugins written by people who shared my preferences, perhaps in 
the minority though they may be.

I use "Folderpane Tools" to reorganize the folders from the default layout 
so I can get important information where i need it.

I use "compact header," "change quote and reply format," "smart template," 
and "toggle compose headers" to get the headers, both displayed and as 
sent, into a format that works for me.

I use "Sieve 0.2.3d" to manage server-side sieve filters and it works 
pretty well.

I use "select addressbook text" to fix a weird artifact in a disabled copy 
function.

I use "Enigmail" to enable GPG functionality, though I hate that the devs 
there completely change the UI every release and completely borked the 
plugin without even the slightest informative error message when they 
decided that if you have an MD5 hashed key in your keyring, you're not 
serious enough about encryption to use Enigmail.

I use "keyconfig" to get the keyboard commands I am used to in MB back.

I use "send and file" to get outbound messages to the right folders.

I use "search results sort by date not relevance" to rationalize the search 
results from the insanely bad default.

These all just give me near equivalent usability to MB.  There a few things 
I still miss daily: paste as quote in MB is awesome.  I haven't found a 
plugin to do that.  And there's an MB plugin to reduce quote levels, which 
is occasionally critical (I use Notepad++ and a saved regexp to remvoe 
quotes and then reflow the text if I am on TB).

But as I've noted, the main functional difference to me is the consequences 
of the out of control memory and processor footprint for TB and some of the 
weird (to me) ways that TB handles background sync (I particularly 
appreciate John Klensin's explanation of online/offline/disconnected mode, 
that was very informative and well-written: thanks!).  I had intended to do 
some tests with Claws and the fancy rendering (webkit/GTK) plugin, which I 
suspect is far more efficient than TB, but I haven't got it working on 
Windows yet and while it is easy on Linux, it isn't a direct test.

Having watched TB's development cycle for a few years now, I'm frankly not 
hopeful that the key annoyances will be fixed.  Seeing a lot of dev cycles 
allocated to things like WebRTC rather than fixing basic usability issues 
that have been open bugs for many years now is not heartening, I've filed 
quite a few bugs over the years, but the new bugzilla security placebo 
password requirement (any algorithmic complexity rule is just as easily 
brute forced as none, so they're all pointless) has also broken my access 
(the password token never arrives, could be too spam-like perhaps).

I do have renewed hope for the new TB team.  They seem committed and solid 
and I'll keep using it, especially after having invested so much time in 
unborking the UI with plugins.

As for MB, I really, really hope the energy displayed in defending it 
translates to some effort in reviving it and updating it.  As excellent and 
near perfect as it is, there are a few things that could be improved, most 
of all confidence that it will at least continue to operate as well as it 
does as the operating systems and standards evolve.  A new build and 
perhaps inclusion into a repository for Debian or the FreeBSD ports system 
would be lovely and might lead to greater usage and inspire a broader 
developer engagement. At the very least, familiarity with MB's efficient, 
well-written code and clean, well-designed UI should be a prerequisite for 
anyone contributing to email client development.

When I started this thread ("Devin's efforts"), I meant only to endorse 
that work.  I really like MB and want to continue using it.