Comment 15 for bug 402767

On Fri, Jul 24, 2009 at 02:35:16AM -0000, Jess Miller wrote:
> The problems I have with Multisearch:
>
> 1) The name of the extension is not easily related to its
> effect. "Multisearch" does affect searches, but only Google
> searches. To me, it sounds more like "something that adds a pack of
> search options" than "something that changes a standard Google
> search into a branded one and changes your new tab start page."

Names are always difficult and are non-trivial to decide on; and we
didn't spend much time thinking about the name for something that is
just used during alpha3-4 cycle to understand better how our users
(would) use the various search places.

> 2) Multisearch makes it impossible, while enabled, to do a "clean"
> Google search from the search bar and takes over the Google icon
> while not allowing access to Google functionality that users
> expect. If the Ubuntu custom search was installed as one option
> among others, this would be less frustrating. It is also frustrating
> because determining the cause is the "Multisearch" add-on is
> difficult because of its unhelpful name, and because it is not a
> user-installed add-on so a user does not expect to need to look
> there.

The custom search thing is currently the easiest way for us to gather
the data we want to get from this experiment.

> 3) Replacing the about:blank page with the Ubuntu custom search when
> a new tab is opened is simply tacky. More often than not, I have a
> particular website in mind for my new tab. Placing the custom search
> there is unexpected and disruptive. It is unnecessary visual
> clutter. It also reads as pushy, in-your-face marketing for the
> Google custom search. I don't see this particular action taken by
> Multisearch as morally wrong (though many others do, and I would
> argue that removing Google functionality while still identifying a
> search with the Google icon IS wrong), simply irritating and
> tacky. Personally, pushy marketing tactics as those used by Ubuntu
> in the Multisearch add-on serve to drive me away immediately,
> without any particular consideration as to the pros and cons of
> something.

A newtab implementation - if it was done - would be configurable and
far more powerful than what you see now. The current state is just a
quick pitch for the test we wanted to do in alpha3.

To get a better idea on what a newtab could look like, check out
mozilla's newtab effort:

http://labs.mozilla.com/2009/03/new-tab-page-proposed-design-principles-and-prototype/
http://labs.mozilla.com/2009/03/firefox-new-tab-next-iteration/

 - Alexander