[Author Prev][Author Next][Thread Prev][Thread Next][Author Index][Thread Index]

Re: gEDA-user: PCB: Connecting Nets by hand



On 6/26/05, bumpelo <bumpelo@xxxxxxxxxxx> wrote:
> If you think the arrow tool violates every GUI on the planet, I can't
> help you.

This isn't just about the arrow tool.  It's about the entire user
interface as a whole.

I'll admit that I haven't read the documentation to PCB
cover-to-cover.  But I'm also going to say that as someone who hasn't,
I think it is a good idea for those involved with its development to
listen to the issues that are being raised by those who are having
some difficulty with it.  "You didn't read the docs, so you're input
is worthless."  I didn't read the docs because I didn't formerly
expect that the UI was 100% bass-ackwards to every other GUI program
I've used.  Sorry!  Maybe that is a good indication that a
modernization of the GUI is in order?  Who knows.  I feel it is. 
Others here obviously don't.

> There are features
> which are only accessible via the keyboard. There are many powerful
> programs where that is
> true, like every spreadsheet I've seen requires that you type data into
> the cells.

Apples and oranges.  Data entry != command invokation.

> Spreadsheets are
> also famous for "hidden" features like that you type "=B1 + B2" to add
> cells.

*EVERY* spreadsheet I've seen very clearly has a bar on the top or on
the bottom which clearly shows the user the equivalent
keyboard/data-entry equivalent.  So if I select a series of rows or
columns, and click on the summation button, I see the equivalent
command in the result cell.  It is patently not hidden.

> If you never bother
> to read the documentation, calculations in the cells will indeed be a
> "hidden" feature. That said we

No, it isn't.  However, that being said, I patently do not subscribe
to the Microsoftian philosophy of "hide-n-seek" and "protocol of
discovery" that is regularly espoused in Microsoft style guidelines,
which are implicitly accepted as "the truth" of GUI design today.  I'm
a fan of Jef Raskin, personally -- make everything as explicit,
obvious, and *habitual* as possible.

I was unaware of the quick reference information that you mentioned. 
I only found it by walking through the file hierarchy in my
geda-installation/ directory.  That's not obvious.  The command-set
for it, after reviewing it, doesn't appear to be very regular.  The
only thing it is is "explicit", since it's one big list of commands
and their equivalent functions.

> should probably proceed to a maze of menu items for those that want to
> use the tool in a more
> cumbersome way.

Actually, this is a good idea.  Well organized menus are quite a
valuable resource.

Menus are great for telling the user what is/is not possible to do
with an object.  Each menu item can also have a hotkey documented next
to it as well.  For the clueless newbie, menus serve a nice reminder. 
After a while, the newbie associates the hotkeys, and starts to use
the hotkeys for everything.  It's a natural migration path from
inexperienced to experienced.

> Finally, pcb came first (before gEDA) so maybe gEDA should bend to pcb's
> UI. I certainly find
> pcb's to be more natural and less "thick" i.e. fewer keystrokes or other
> operations to accomplish
> something.

Relative to what?  To gschem?  Allow me to posit the argument that you
don't know gschem very well then, because I'm at least 4x to 5x faster
in gschem at getting work done than I am in pcb, easily.

--
Samuel A. Falvo II