[horde] [imp] default reply behavior in dimp/imp

[D G Teed]

On Tue, Sep 6, 2011 at 7:30 PM, Eric Jon Rostetter
<eric.rostetter at physics.utexas.edu> wrote:

> IMHO the problem is not that the proper solution isn't provided, but that
> the proper solution is not obvious to the end-user in DIMP like it is in
> IMP.  So the question becomes, how can we make the DIMP reply functionality
> easier to see and use, e.g., more obvious to the end-user.

I don't want it easier to find. I want it to be there as an option
plain as day.  I don't want people finding Reply and doing Reply
to All without realizing it, as has actually happened.

A Reply which is always defaulting to Reply All with mailman
is a recipe for disaster.

That some people don't get this just baffles me.  I suspect it
is because they are thinking of themselves working
within the interface, while I am thinking about 7000 users
and a trouble ticket I saw while things were still quiet.

> Why not just remove DIMP, and only use IMP?  No one is forcing you to
> use DIMP.  If DIMP isn't appropriate for your needs, don't use it.
> It provides no additional functionality to IMP, so there is no need to use
> it.  So simply don't use it.

Enforcing traditional (IMP) is the workaround.  It causes the detection
of mobile clients to also be lost.  The new mobile interface
has an awesome look.  Feels like an iPod app on the iPod.
So it is sad our new horde 4 looks and acts like horde 3.

Unlike Horde 3, I think dimp is integrated into imp (no separate dimp
app directory anyway).

> The point is, you raised a valid concern, and it is being discussed.
> Your point of view was rejected, and while the discussion continues
> and other ideas are explored, you seem to refused to participate in
> the process and keep harping on the already rejected solution as the
> only solution.  Instead of trying to be part of the solution, you
> are being part of the problem...  Again, that is your right.  But
> it isn't usually productive.

If your stance is "no drugs", and people counter propose
yes to drugs but how do we help drug users keep
their jobs and drive while drugged, do you drop your
"no drugs" belief?  There are times when it makes
no sense to discuss solutions which are a fix
for a problem we helped create in the first place.

> I know someone suggested you give an example of a problem mail from your
> mailman installation, and I've not seen you reply to it.

I'm puzzled why they can't see Reply All happening to mailman
emails, but at the same time I'm convinced 3 different problems
like it are solved with the simple all in one solution available
in existing code: using an interface like IMP.

> I've seen a
> couple other proposals, which you ignore or reject but don't say why they
> wouldn't work or why your way is truely better.  I've seen you make
> the same assertions over and over (like almost everyone knows outlook)
> but not replied when someone refuted it (by offering statistics about
> it).

What, like Outlook has 30% market share?  It doesn't matter.
What market share does gmail, hotmail, thunderbird,
apple mail, roundcube, imp, squirrelmail and outlook have?
They all have dual reply options, and this has been common
for decades.  It really is a commonly expected interface,
but developers here like to call it "incorrect".

They call it incorrect because they believe they can make
a good guess as to what type of reply is natural.

My response is, this does not prevent the problem of errant
Reply All, and until you do, this is not a solution.  In fact
you have introduced a problem by having one button.
Is it not natural to back out that change to prevent
the problem?

>  Yet you continue the same argument, without participating in the
> discussion or responding to the replies.  And you expect people to listen?

I didn't see anyone acknowledge the problem is real and
there is damage which can come out of it.  I saw refutation
of the importance of the issue.  I saw weak and desperate comparisons
to attempt to shore up alignment (Kontact, suggestion that gmail has
one button for reply (not quite true - has two links as well)).  I saw
theoretical discussions on how yet another new interface
default might partially resolve the issue.

Maybe old fashioned thinking, but when none of the new solutions is
good, you should consider tried, tested, and true until you can
really provide a foolproof alternative.