New Tab Prototype
Daniel Mills
Last week Labs released a little prototype for a possible "new tab"
page:
http://labs.mozilla.com/2009/03/new-tab-page-proposed-design-principles-and-prototype/
There's a ton of discussion there, but I wanted to make sure everyone
here who might not follow the Labs blog had a chance to check it out
as well!
Here's the wiki page where you can find some additional links as well:
https://wiki.mozilla.org/About:newtab
Dan
Peter Lairo
> Hi,
>
> Last week Labs released a little prototype for a possible "new tab" page:
>
> http://labs.mozilla.com/2009/03/new-tab-page-proposed-design-principles-and-prototype/
I've installed the test-add-on. A good start. Does it auto-update?
--
Regards,
Peter Lairo
The browser you can trust: www.Firefox.com
Reclaim Your Inbox: www.GetThunderbird.com
Islam - Myths & Facts: http://www.jihadwatch.org/islam101/
Israel - Myths & Facts: http://www.JewishVirtualLibrary.org/
Church of the Flying Spaghetti Monster: http://www.venganza.org/
"So, why don't we ever talk about the sun's contribution to global
warming? Well, because we can't regulate it, tax it, or make it feel
guilty for what it's doing" (www.WhatYouOughtToKnow.com)
Mike Beltzner
> I've installed the test-add-on. A good start. Does it auto-update?
Of course. The latest version is 0.0.18. You can see some of the
latest thoughts from the people working on the development here:
http://ed.agadak.net/2009/03/abouttab-sprint-weekly-200910-2
But really, I think Dan, Aza, Ed and I would love to hear what people
here think after using it for a few hours, days, or whatever. I've
been running with it for a week, for example, and haven't fully
composed my thoughts, but have several! :)
cheers,
mike
Daniel Mills
> On 10.03.2009 22:22, Daniel Mills wrote:
>> Hi,
>>
>> Last week Labs released a little prototype for a possible "new tab"
>> page:
>>
>> http://labs.mozilla.com/2009/03/new-tab-page-proposed-design-principles-and-prototype/
>
> I've installed the test-add-on. A good start. Does it auto-update?
Yes, in fact I just pushed a minor update up (0.0.19).
Dan
Axel Hecht
I had given it a try for a few days, but I found it to be not adding
value to the habits I have. I want to the awesomebar if a open a new
tab, and the sites it offered rarely match what I need.
I ended up disabling the extension when I thought "wow, that's still in
your clipboard?" one time too often. Not that there'd be anyone in my
flat to look over my shoulder about that, it just became something I
didn't want to see myself.
I don't think I ended up clicking on one of the suggested pages once, too.
Axel
Daniel Mills
> I had given it a try for a few days, but I found it to be not adding
> value to the habits I have. I want to the awesomebar if a open a new
> tab, and the sites it offered rarely match what I need.
If they matched, do you think you would you click on them?
> I ended up disabling the extension when I thought "wow, that's still
> in your clipboard?" one time too often. Not that there'd be anyone
> in my flat to look over my shoulder about that, it just became
> something I didn't want to see myself.
Clipboard entries don't expire, so yeah. I actually wanted it to show
it every time, but the extension only shows them once (if you open
another tab it won't show the search suggestion).
Also, do you find it actually makes using just the awesomebar (or
search box) worse in any way? I don't think it's distracting enough
that when I open a new tab I feel like I must look at the content
there--but maybe that's just me.
Dan
Blair McBride (Unfocused)
I have something in the clipboard, I compulsively put a single white
space in the clipboard to overwrite it.
- Blair
Axel Hecht
> On Mar 10, 2009, at 5:30 PM, Axel Hecht wrote:
>
>> I had given it a try for a few days, but I found it to be not adding
>> value to the habits I have. I want to the awesomebar if a open a new
>> tab, and the sites it offered rarely match what I need.
>
> If they matched, do you think you would you click on them?
I don't know, tbh. I usually open a new tab via the keyboard, and I can
type quickly, so clicking is a pretty large context switch away.
I wonder, could you give a console snippet like for the places db
collection that would expose the top-20 or so sites in your metric plus
the value? I confess, I haven't run the new tab prototype on my main
profile, as that's still on BonEcho (I should look at extension compat
again). That would give me a better idea if it's just my odd profile
that I feel fine with the awesomebar there and didn't feel that the
choice of sites in new-tab hit the nail.
>> I ended up disabling the extension when I thought "wow, that's still
>> in your clipboard?" one time too often. Not that there'd be anyone in
>> my flat to look over my shoulder about that, it just became something
>> I didn't want to see myself.
>
> Clipboard entries don't expire, so yeah. I actually wanted it to show it
> every time, but the extension only shows them once (if you open another
> tab it won't show the search suggestion).
I guess the trigger was when I saw it mentioning something that I had
copied and pasted from Terminal to Aquamacs a while ago. No idea whether
that's a bug or a feature.
> Also, do you find it actually makes using just the awesomebar (or search
> box) worse in any way? I don't think it's distracting enough that when I
> open a new tab I feel like I must look at the content there--but maybe
> that's just me.
I guess I optimized the "don't look at it" to the extreme by disabling it.
Axel
Jesper Kristensen
I tried it for a few days, and I ended up disabling it for some reasons:
* The tab thumbnails was slow to load, which annoyed my eyes.
* I never clicked anything on that page. I guess it is because when I
open a new tab, I know where I want to go. I also know that there is one
place where I can get to it relatively quick: The location bar. It takes
time to look through the suggestions on the new tab page, and I don't
know if it suggests what I want, whereas the location bar always takes
me to the place I want.
* I usually open a new tab using Ctrl+T, and the focus is immediately in
the location bar. It is much faster to type something in there than it
is to read the suggestions, move my hand to the mouse, find out where
the cursor is on the screen and click somewhere.
* The addon broke another tab-related addon which I use heavily.
Daniel Mills
> I tried it for a few days, and I ended up disabling it for some
> reasons:
>
> * The tab thumbnails was slow to load, which annoyed my eyes.
It's only slow the first time you open it. In a less prototypey
implementation, though, there are several ways of mitigating that
problem.
> * I never clicked anything on that page. I guess it is because when
> I open a new tab, I know where I want to go. I also know that there
> is one place where I can get to it relatively quick: The location
> bar. It takes time to look through the suggestions on the new tab
> page, and I don't know if it suggests what I want, whereas the
> location bar always takes me to the place I want.
> * I usually open a new tab using Ctrl+T, and the focus is
> immediately in the location bar. It is much faster to type something
> in there than it is to read the suggestions, move my hand to the
> mouse, find out where the cursor is on the screen and click somewhere.
These are problems I've had myself, though I wonder if the new tab
prototype is more useful for a different (perhaps less technical)
audience.
> * The addon broke another tab-related addon which I use heavily.
Which one?
Dan
Daniel Mills
> I wonder, could you give a console snippet like for the places db
> collection that would expose the top-20 or so sites in your metric
> plus the value? I confess, I haven't run the new tab prototype on my
> main profile, as that's still on BonEcho (I should look at extension
> compat again). That would give me a better idea if it's just my odd
> profile that I feel fine with the awesomebar there and didn't feel
> that the choice of sites in new-tab hit the nail.
We've played with the query a bit. Something I thought of today is
that it's possible that "sites that you go to often" is actually
different from "sites you spend a lot of time on". That is, perhaps I
open gmail in the morning and just leave it up, but maybe I keep going
to twitter and closing the tab. In that example "twitter" would sort
much higher than "gmail", when really they are both probably as
important. In the awesomebar this is less of a problem because it's
triggered by user input.
> I guess the trigger was when I saw it mentioning something that I
> had copied and pasted from Terminal to Aquamacs a while ago. No idea
> whether that's a bug or a feature.
Aza is hacking up a new version that will monitor the clipboard and
only show clipboard entries if you copied the text recently. That
should remove the surprise factor of seeing an old entry without
impairing the "copy, new tab, act" use-case.
>> Also, do you find it actually makes using just the awesomebar (or
>> search
>> box) worse in any way? I don't think it's distracting enough that
>> when I
>> open a new tab I feel like I must look at the content there--but
>> maybe
>> that's just me.
>
> I guess I optimized the "don't look at it" to the extreme by
> disabling it.
Yeah, ideally we get to a point where it's useful for some set of
people, and it's not annoying to anyone else.
Dan
Johnathan Nightingale
> On Mar 11, 2009, at 3:47 AM, Axel Hecht wrote:
>
>> collection that would expose the top-20 or so sites in your metric
>> plus the value? I confess, I haven't run the new tab prototype on
>> my main profile, as that's still on BonEcho (I should look at
>> extension compat again). That would give me a better idea if it's
>> just my odd profile that I feel fine with the awesomebar there and
>> didn't feel that the choice of sites in new-tab hit the nail.
>
> that it's possible that "sites that you go to often" is actually
> different from "sites you spend a lot of time on". That is, perhaps
> I open gmail in the morning and just leave it up, but maybe I keep
> going to twitter and closing the tab. In that example "twitter"
> would sort much higher than "gmail", when really they are both
> probably as important. In the awesomebar this is less of a problem
> because it's triggered by user input.
I suspect there would be value in, if we aren't doing so already,
annotating history entries which were the first navigation in a new
tab, and then using frecency on *that* list to populate the new tab
experience. Using sites that start a browsing-chain (which I
understand to be the current approach?) is a good idea, better than
raw frecency I think, but is still not quite a direct match for the
situation here.
Or am I talking crazy talk/already tried-and-abandoned ideas?
J
---
Johnathan Nightingale
Human Shield
joh...@mozilla.com
David McRitchie
> > https://wiki.mozilla.org/About:newtab
This is probably the shortest duration testing of any extension, I immediately
uninstalled it. Nothing but a blank tab should be created with a New Tab
button or Ctrl+T, and this extension is 52KB which is fairly large. AutoHide
which I liked the appearance of (impress onlookers with a chrome shortcut)
but rejected when it became tied to a new tab is 6KB download by comparison
and it did noticeable take a while to populate certainly not something I would
want builtin to Firefox.
Meanwhile Cat Thief's "New Tab" extension (6KB) appears to be still required for 3.1,
seems the new tab button remains now and works, but the the extension is still
needed to eliminate the one on the tab bar which destroys usage.
http://www.tom-cat.com/mozilla/firefox/extensions-newtabbutton.html
I can't understand the goal of breaking simple functional design. If you want to
see how to really break something consider all of the work that went into the
Microsoft Office Ribbon and how well designed plans can utterly destroy
usage of the software. It is the ultimate in poor ergonomic design.
Nothing is gained by making something more easily
found by someone who is unfamiliar with a browser, the learning time is generally
the same for such things, it would be far better to train people to use existing
features than to create confusing features that will always take longer for an
experienced person.
You create a new tab it is empty. Two things you can do, put something into it
or get rid it. You have something already in it when you create a new tab
that you don't ask for and now you have confusion /noise -- so I definitely would
never want to see that as a built-in feature.
--
HTH,
David McRitchie, extensions I use are briefly documented on my site
Firefox Custom: http://www.mvps.org/dmcritchie/firefox/firefox.htm
Jesper Kristensen
>> * The addon broke another tab-related addon which I use heavily.
>
> Which one?
It is one of my own addons which I haven't published yet. It allows me
to have the browser open without having any tabs open.
Daniel Mills
On Windows? Nice.
More to the point, how does it conflict with it? Is there something
about:tab is doing that makes breaking other addons likely? (globals,
etc.)
Dan
Axel Hecht
> On Mar 11, 2009, at 3:47 AM, Axel Hecht wrote:
>
>> I wonder, could you give a console snippet like for the places db
>> collection that would expose the top-20 or so sites in your metric
>> plus the value? I confess, I haven't run the new tab prototype on my
>> main profile, as that's still on BonEcho (I should look at extension
>> compat again). That would give me a better idea if it's just my odd
>> profile that I feel fine with the awesomebar there and didn't feel
>> that the choice of sites in new-tab hit the nail.
>
> We've played with the query a bit. Something I thought of today is that
> it's possible that "sites that you go to often" is actually different
> from "sites you spend a lot of time on". That is, perhaps I open gmail
> in the morning and just leave it up, but maybe I keep going to twitter
> and closing the tab. In that example "twitter" would sort much higher
> than "gmail", when really they are both probably as important. In the
> awesomebar this is less of a problem because it's triggered by user input.
I guess there are tons of pattern that one could use, it's hard to
evaluate from my point without actually seeing some concrete metrics and
the numbers they produce.
Most of the ideas I had before my morning tea work along detecting
patterns between chosen new tabs and other tabs/windows. Like "this is
the window with my daily comics, 90% of the new tabs in that scenario
end up in searches on dict.leo.org". And clipboard was never on them, as
the words I looked up were in jpegs ;-).
Wow, now I get a cookie, with herbs, to get really weird fantasies.
Axel
Thomas Stache
I like the extension for one thing: casually check a couple of sites
like Engadget or DPReview for new content. (I opt to not subscribe to
them because their high-volume feeds would constantly distract me)
To really be useful though, I'd need to be able to remove several sites
currently shown from the list, and promote others on the New Tab Page.
Additionally the feed autoselection is failing for me e.g. on
http://www.dpreview.com, a site that offers multiple feeds, but Ambient
News and about:tab pick the wrong one (not the main news content).
The one Major Annoyance of the current prototype is focus mangling,
Ctrl-T puts focus in the URL bar, but after a couple of characters the
focus goes elsewhere and the user is interrupted. (Same for Ctrl-T,
Ctrl-B and typing in the sidebar search field, focus is lost after ~1s.)
Thomas
Jesper Kristensen
I don't know. I don't blame about:tab for the fault. It could just as
well be mine. My addon looks for page loads in a newly created tab. It
ignores the first load, as that is normally about:blank, which I don't
want to consider a load. I would guess that about:tab creates more than
none of the events I am listening for, which tricks my extension to
believe that the tab is being used for something. But that is pure
guesswork.
hold...@gmail.com
better to write something from the start that's really easily
extensible (easier than writing an extension)? Sorta like iGoogle
crossed with Ubiquity. Create some really simple command utils/xbl
bindings that let you run simple places queries, create thumbnails,
etc and let people dig into XPCOM/Canvas/Etc. for other things they
want. Maybe an API that made forced elements to load ASYNC or in
threads so that performance wouldn't have to take a huge hit. You
could get something out that was comparable to what other people are
trying, and within a year of first release you'd probably have a set
of new ideas that were even more in line with what users actually want
(plus the additional 1000 things that only a small subset of people
care about).
Maybe that's way to dangerous, or maybe it would be (risking) to much
bloat in a page that should load quickly. Still Firefox is known/loved/
hated for extensibility and creating a simple way for people to add
new "panes" seems right in line. And it seems better to use these
original "widgets" (frequent tabs, undo closed tabs, etc) to create an
API than to try and mush it in later.