Features and priorities

[Rafael Bugajewski]

Hello,

As I understand the feature set is already written down and
prioritized. But I wanted to propose a possible workflow for
developing a couple of “basic” features.

E.g. I can't use a browser without bookmarklets accessible via
keyboard shortcuts (command-n for n=0..9) and without 1Password. It
just doesn't work and makes me completely unproductive. Everyone has
such (subjectively) major features. I want to propose to let the
specific people implement what they want and to coordinate it somehow
so it doesn't get too messy. I don't I we should discuss it on this
list or write it down on a single wiki page.

In the end we still can remove or rewrite a feature.

I don't know if I have the time and experience to implement this basic
bookmarking behavior and don't know how many changes would be
necessary from the 1Password folks, but if I take over some dev work
this is definitely my 1st priority.

What do you think about this approach? Is it to chaotic?


--
Rafael Bugajewski (via iPhone)

--
Subscription settings: http://groups.google.com/group/rouse-dev/subscribe?hl=en

[Jesper]

> As I understand the feature set is already written down and prioritized. But
> I wanted to propose a possible workflow for developing a couple of “basic”
> features.
>
> E.g. I can't use a browser without bookmarklets accessible via keyboard
> shortcuts (command-n for n=0..9) and without 1Password. It just doesn't work
> and makes me completely unproductive. Everyone has such (subjectively) major
> features. I want to propose to let the specific people implement what they
> want and to coordinate it somehow so it doesn't get too messy. I don't I we
> should discuss it on this list or write it down on a single wiki page.

I use Command+digit myself heavily in OmniWeb.

If someone feels strongly about a feature (will) and has the technical
skill to implement it, it's probably worth looking into the feature.
There's gotta be some coordination or everything gets a preference
pane (for example), but if it actually turns out to work, the bar will
be set high to take it out of the product.

There's a reason I'm not running around, Letters-esque, saying "let's
just implement everything in plugins" because that *is* fucking hard.
I don't aim for that level of modularity. But there has to be some way
of weeding out the
necessary-for-just-me-and-fun-thing-to-have-for-everyone-else from the
actually-necessary-for-everyone.

> In the end we still can remove or rewrite a feature.

This feels like saying "Drink as much as you want and feel free to do
what you will on the carpet; there are always dry-cleaners." It should
be a last resort rather than a comfort. That said, you should
certainly be able to make your case and implement your feature. You
can't funnel people into clean, well-written code, but the
architecture should be such that it encourages a clean implementation.
If you've got a great idea, I don't want to stand in anyone's way, but
I don't want to unconditionally accept every scrap of Objective-C
either.

> I don't know if I have the time and experience to implement this basic
> bookmarking behavior and don't know how many changes would be necessary from
> the 1Password folks, but if I take over some dev work this is definitely my
> 1st priority.
>
> What do you think about this approach? Is it to chaotic?

Partially yes, but mostly no. There are two ways of dealing with
things like 1Passwd or Gears (which I know is now no longer being
developed) or Evernote: to not (in which case they might brute force
it anyway), or to provide good tools for integration. I would love to
be able to provide stable and thought-out support. What could best be
done at this stage might mean finding out how the integration should
work.

/Jesper

[Jesper]

Let me respond to this again. I still stand by what I said, but I
didn't include everything that I had planned to say.

> As I understand the feature set is already written down and prioritized. But
> I wanted to propose a possible workflow for developing a couple of “basic”
> features.

The gist is written down, and some specifics are even defined and
prioritized (like tabs). But certainly there's room to define what's
going to be available of the actual features. The reason I want to
avoid building some features at first isn't because I don't ultimately
want them, but because if it come to that I'd have to build a lot of
browser infrastructure before I could say "tier one is done, *now*
let's look at this and see what we can do". If you or someone else
would like to help implementing a bona fide feature instead of the
placeholder I'm afraid you might be stuck with instead, I'm game.

The command-digit shortcuts for "favorites bar" bookmarks and
bookmarklets sounds like a perfect opportunity to do something like
that. 1Passwd sounds like it could be one API call, it could be just
as much work as autofill or it could be one order of magnitude larger
than actually implementing something like bookmarks. Define the
feature deeper and it'll be a lot more easier to discuss.

/Jesper

[Mark Norman Francis]

> There's a reason I'm not running around, Letters-esque, saying "let's
> just implement everything in plugins" because that *is* fucking hard.

It also smacks of "everything is a preference" too. Given how many browsers exist already, I'd say there's room for an opinionated power-user browser that is less crazy-configurable and more just-gets-on-with-it.

-- Norm.

[Rafael Bugajewski]

On 04/20/2010, at 01:53 AM, Jesper wrote:

> The command-digit shortcuts for "favorites bar" bookmarks and
> bookmarklets sounds like a perfect opportunity to do something like
> that. 1Passwd sounds like it could be one API call, it could be just
> as much work as autofill or it could be one order of magnitude larger
> than actually implementing something like bookmarks. Define the
> feature deeper and it'll be a lot more easier to discuss.

OK, so here it is:

Feature 1: Command-digit
------------------------

1) Add (max. 10 bookmarks)
2) Make them accessible via command-digit
3) Bookmarklets like Instapaper’s Read Later, Readability or Shortware should work (this is really unknown territory for me)

What I don’t mind in the first version(s):

* Removing bookmarks
* Reordering bookmarks
* Managing bookmarks at all
* Adding more than 10 bookmarks
* Visual bookmark bar

In the end I don’t even know if bookmarks integration would be necessary for me if the tab implementation will be so sophisticated like you discussed before.

Feature 2: 1Password integration
--------------------------------

1) Press a keyboard shortcut (preferable 1Passwords keyboard shortcut) and let 1Password fill out the form for you (text fields, password fields and checkmarks)
2) Strong password generator integration, so I can at least generate a password from within the browser and copy it to the clipboard

What I don’t mind in the first version(s):

* 1Password menu with the whole management features
* Auto recognition of newly entered forms
* All other 1Password related stuff

Is this detailed enough so we can go on with the discussion?

BTW: Thanks for starting this project Jesper.

Rafael