Firefox 3 Feature Planning: tell me about *your* problems.
Mike Beltzner
mozilla.dev.apps.firefox with a followup set to
mozilla.dev.apps.firefox; if you are reading this as an email list,
please make sure you reply to dev-apps...@lists.mozilla.org, and
make sure you are subscribed to that list so your reply doesn't
bounce!)
A few weeks (months?) ago on a Gecko 1.9 / Firefox 3 planning call, I
mentioned that one thing that I was never terribly thrilled about was
that when looking at planning out the features for Firefox 2, I didn't
push hard enough for every proposed feature to explain the problem that
it was attempting to solve. Without understanding the problem we
believe to be solving with a feature, it's extremely hard to get a
sense of when the feature is complete, how important the feature is,
what the impact of cutting the feature is, and what advantage we get by
including it in the final product.
In last week's meeting
(http://wiki.mozilla.org/Firefox3/StatusMeetings/2006-08-16#Firefox_3),
I started brainstorming up a list of problems that I felt users were
facing when using web browsers today. I don't think we should focus on
solutions or implementations just yet, but instead focus on answering
the following question: "What are the problems with the web that users
are currently facing, whether they realize it or not?"
So go ahead and throw a couple of ideas onto this thread!
cheers,
mike
beltzner
> the following question: "What are the problems with the web that users
> are currently facing, whether they realize it or not?"
>
> So go ahead and throw a couple of ideas onto this thread!
Some great stuff so far, with the right level of fidelity and detail;
one comment, though. Please make sure to think about not only problems
that you have experienced personally, but those that you've witnessed
in peer groups, families, with less savvy users, etc. People who do a
lot of reading through MozillaZine, work in tech support, act as tech
support for their families and friends, I'm looking for your input!
Also be sure to think not only of problems with existing features, but
problems that people don't even realize they're having yet. For
example: users who don't understand the web very well might not
realize that they're on a fraudulent site, or that they shouldn't
enter financial information online unless they are very sure who is
receiving that information.
cheers,
mike
--
/ mike beltzner / phenomenologist / mozilla corporation /
Martijn
internet thing), that they really have a hard time clicking on a link.
A lot of times they also move the mouse while clicking on the link,
which results in a dragging action (this is also a problem in IE).
So disabling any drag action would be a solution (since they would
never use that, anyway), or a keyboard equivalent for the left (and
right) mouse button.
Regards,
Martijn
On 8/23/06, beltzner <mbel...@gmail.com> wrote:
> On 22 Aug 2006 23:52:09 -0700, Mike Beltzner <mbel...@gmail.com> wrote:
> > the following question: "What are the problems with the web that users
> > are currently facing, whether they realize it or not?"
> >
> > So go ahead and throw a couple of ideas onto this thread!
>
> Some great stuff so far, with the right level of fidelity and detail;
> one comment, though. Please make sure to think about not only problems
> that you have experienced personally, but those that you've witnessed
> in peer groups, families, with less savvy users, etc. People who do a
> lot of reading through MozillaZine, work in tech support, act as tech
> support for their families and friends, I'm looking for your input!
>
> Also be sure to think not only of problems with existing features, but
> problems that people don't even realize they're having yet. For
> example: users who don't understand the web very well might not
> realize that they're on a fraudulent site, or that they shouldn't
> enter financial information online unless they are very sure who is
> receiving that information.
>
> cheers,
> mike
>
> --
> / mike beltzner / phenomenologist / mozilla corporation /
> _______________________________________________
> dev-planning mailing list
> dev-pl...@lists.mozilla.org
> https://lists.mozilla.org/listinfo/dev-planning
>
Peter Kasting
> So go ahead and throw a couple of ideas onto this thread!
Here's what I've seen from novice users:
* People don't understand URLs. They don't know what they mean, they
can't parse them, and they don't understand where to type them. This is
why they get repeatedly fooled by obvious phishing URLs, why they type
URLs into the Google search box instead of into the address bar, and why
a dedicated search box is friendly: it doesn't have that weird string in
it that takes a long time to delete.
* Related: people don't really get the difference between searching for
something and going somewhere. Even those who sort of know the URL for
where they want to go ("espn.com") type it into the google search box
and then click the first link that comes up, because they know how to
"search" for something, but not "go" somewhere. This is why people
REALLY like using the search boxes on browser homepages; give them a
blank homepage and they're unsure how to get anywhere.
* Related/clarification: People don't understand multiple distinct input
fields. The difference between the url bar and the search box, or the
boxes in chrome and on the page, are not clear to them.
So, with those three points in mind, I would suggest rethinking how URLs
and search are presented and handled. I have a lot of ideas about this
but I'd rather not pollute people's thinking. I WOULD suggest that the
current browser UI for navigation is an evolutionary modification of
designs originally made for people who understood the World Wide Web,
and that perhaps some rather drastic UI rethinking is warranted.
Another set of points:
* People have a hard time reading pages.
* Fonts are often too small
* Text is all over the place and obscured/broken up by ads
* Scrolling makes it easy to lose one's place
* "Next page" links can be hard to find and click
* Pages are long, and "Find in page" isn't very discoverable (my wife
never knew it existed until I showed her), nor does it do a great job of
guiding people to the content they wanted, which is often "what I just
searched for in my search engine"
Browsers spend a lot of UI for navigation between pages, but making
pages more readable doesn't seem to be a high priority. What if you
could go to a page, and it would be big and easy to read, with
beautiful-looking fonts, all the content nicely laid out for you to read
with no distractions, and the important parts highlighted? Now, how can
we use the preferences of the user, and the intelligence of the
collected audience globally, to accomplish this?
Finally:
* Stuff should just work
I can't tell you how many times I've seen people go somewhere and not
understand what that puzzle piece icon means, or why their site looks
wrong, or why their browser locked up for 20 seconds (they don't notice
the little coffee cup that then appears in their systray). If they
can't depend on Firefox to always work, they won't use it at all. But
working means a lot of things. Maybe we ship some common plugins by
default. Maybe we implement a very limited set of ActiveX controls for
media player compatibility. Maybe we see what IE-isms we can
cherry-pick for increased site compatibility. Anything we can do to
make "the internet work" in our browser eliminates a reason people have
to switch away from it.
PK
Robert Sayre
>
> Some great stuff so far, with the right level of fidelity and detail;
> one comment, though. Please make sure to think about not only problems
> that you have experienced personally, but those that you've witnessed
> in peer groups, families, with less savvy users, etc.
My girlfriend hates the update notices, where we throw version numbers
at the user. I think she stops reading the number at "1.". She likes the
way Apple software update does it, by date.
-Rob
Greg
- Most people go to the same 10 sites and they don't know how to
bookmark so they type the same sites over and over. If they know about
the URL-bar that's great but many people will type web addresses
directly into a search box. Can't we make bookmarking more automatic?
Perhaps a new toolbar that shows most frequently used sites?
- Registering for a new site sucks. Users have to input a bunch of
repetitive information, enter a username/email, type a password, type
it again. Check email. Click a confirmation link. Any one of these
steps can cause users problems. Can't we create a mechanism to allow
instant registration? As a bonus can't we think of a way to uniquely
register people without necessarily divulging so much personal
information?
- People often mis-type URL's. This problem is so bad that many
companies squat typo addresses. There's already an auto-complete for
sites that are in my history but why can't the browser help me
auto-suggest the URL for a new website that I've started typing but
haven't yet visited?
If a website wants to provide standard links to sites they have to
create individual links to each site. For example "Add to MyYahoo, Add
to BlogLines, Add to FeedBurner, etc, etc." If I happen to use a
different RSS service I'm out of luck. The lack of standard service
link types makes it hard for services to soft-link to one another. The
dictionary of link types needs to grow. We have href and mailto links.
Where are the others? Phone, TV, Calendar, Contacts, RSS, etc.
Similarly we need a way for sites to request to become the default
providers for these types of services. This is the web equivalent of
registering for file-type extensions.
Other problems:
- As a designer it's really hard to design resolution independent
sites. I don't want to optimize for 800x600 or 1024x768, I want the
site to scale proportionally. This is easy in Flash, I've never seen
this done well in HTML.
- If a user has trouble seeing there is no way to zoom the page.
Increasing the font-size doesn't always work because this can break the
design of the site.
- Uploading more then one file is really awkward.
- The control set available for the web is really small. Web-designers
re-create standard controls (auto-complete, sliders, menus, tabs, etc).
All these controls have non-standard usage and they often don't have
accessibility. It would be great to have a richer more consistent
control experience across sites.
Justin Wood (Callek)
> Some problems I've seen...
And some (known) solutions to these problems, mostly future, with no ETA.
>
> - Most people go to the same 10 sites and they don't know how to
> bookmark so they type the same sites over and over. If they know about
> the URL-bar that's great but many people will type web addresses
> directly into a search box. Can't we make bookmarking more automatic?
> Perhaps a new toolbar that shows most frequently used sites?
FF3's "Places" should address this.
>
> - Registering for a new site sucks. Users have to input a bunch of
> repetitive information, enter a username/email, type a password, type
> it again. Check email. Click a confirmation link. Any one of these
> steps can cause users problems. Can't we create a mechanism to allow
> instant registration? As a bonus can't we think of a way to uniquely
> register people without necessarily divulging so much personal
> information?
A partial solution to this is the "Form History" in Firefox. Another
potential solution, pending application is "Google Account
Authentication" [1] this, of course, must be employed by the website
itself, rather than the UA (Firefox).
You can even then use "OpenID" [2] as an option.
The end result (imho) is "Form History" where possible, and other more
enameled solutions would still require websites to comply, and be
cumbersome unless other UA's adopted the same mechanics. Which is hard,
given that a Login to site X, that you create, may allready be used on
site Y, and you may want two logins on Site Y, due to features it has,
etc. etc. etc.
>
> If a website wants to provide standard links to sites they have to
> create individual links to each site. For example "Add to MyYahoo, Add
> to BlogLines, Add to FeedBurner, etc, etc." If I happen to use a
> different RSS service I'm out of luck. The lack of standard service
> link types makes it hard for services to soft-link to one another. The
> dictionary of link types needs to grow. We have href and mailto links.
> Where are the others? Phone, TV, Calendar, Contacts, RSS, etc.
> Similarly we need a way for sites to request to become the default
> providers for these types of services. This is the web equivalent of
> registering for file-type extensions.
The WhatWG addresses some of these concerns with their
"registerContentHandler" and "registerProtocolHandler" interface for
UA's in the current Working Draft of the "Web Apps" spec. [3]
> Other problems:
> - As a designer it's really hard to design resolution independent
> sites. I don't want to optimize for 800x600 or 1024x768, I want the
> site to scale proportionally. This is easy in Flash, I've never seen
> this done well in HTML.
Full page scaling is a planned feature, (aiui)
> - If a user has trouble seeing there is no way to zoom the page.
> Increasing the font-size doesn't always work because this can break the
> design of the site.
see above
> - Uploading more then one file is really awkward.
Uploading, assuming you mean "in-web page" is addressed by the WhatWG's
"Web Forms" Working Draft [4]
> - The control set available for the web is really small. Web-designers
> re-create standard controls (auto-complete, sliders, menus, tabs, etc).
> All these controls have non-standard usage and they often don't have
> accessibility. It would be great to have a richer more consistent
> control experience across sites.
>
See the entire "Web Forms" Working Draft [5].
~Justin Wood (Callek)
[1] - http://code.google.com/apis/accounts/Authentication.html
[2] - http://openid.net/
[3] - http://www.whatwg.org/specs/web-apps/current-work/#browser
[4] - http://www.whatwg.org/specs/web-forms/current-work/#upload
[5] - http://www.whatwg.org/specs/web-forms/current-work/
Håkan Waara
I think all of your points are great, and I agree with them.
Now, regarding the URL bar and search boxes: I've seen the same
behavior. Actually, my whole family use the browser like that. They
type a common URL in the google search box, and click the first search
result. That is probably a very, very common real-world use case.
So, if normal people actually consider the location bar, and the google
search box the same thing -- if evolution is allowed to decide - it IS
the same thing!
Basically, what I propose is to merge them.
I realize that we might already do, in some sense. If you type a URL,
and we don't find it, we search for it on google, I think?
The looking glass icon could be added somewhere sensibly to the URL
bar, we rip out the search box, and we have further simplifed the user
interface, according to how a lot of users actually do use it.
I would like to hear your opinions on this.
/Håkan
Peter Kasting
> So, if normal people actually consider the location bar, and the google
> search box the same thing -- if evolution is allowed to decide - it IS
> the same thing!
>
> Basically, what I propose is to merge them.
>
I've actually done this in my configuration, and the configuration I set
up for end-users, for a couple of years. Once I set keyword.URL to do a
standard Google Search (much more predictable and therefore less
unfriendly), and set browser.urlbar.autoFill and
browser.urlbar.matchOnlyTyped to true (to try and tweak the autocomplete
behavior to be a bit more useful when people type a lot in the urlbar),
I can simply remove the search box from the Chrome entirely. My family
members, who are computer novices, all figured this out with a simple
"when you want something, type it in the box".
I'd certainly make improvements to that if I were to deploy it to the
world. Much better autocomplete heuristics, some integration of
suggestions/search results into the autocomplete results, icon and text
cues for what would happen when you hit enter, etc.
However, as beltzner has repeatedly said, this is a thread for problems,
not solutions. There are likely other good ways to solve this problem.
I encourage you to try using Fx 2 with the settings I give above and
see what kinds of _other_ user problems you encounter (there certainly
are some!).
PK
Jon Smirl
1) Fonts too small, sites that design for 800x600 look like a postage
stamp at 1600x1200. If I set the fonts big enough to make the problem
sites readable everything else is too big, plus their fixed layouts
get messed up. I see IE7 has a page zoom feature. History should
remember the zoom factor set.
Viewing Eweek email in gmail is a good example of this. If I set the
fonts small enough that everything lays out correctly I have to have a
magnifying glass to read it.
2) When I am composing an email in gmail and typing sometimes the edit
control goes unresponsive for several seconds. This is very annoying.
I believe this is occurring when gmail executes Javascript to save a
draft or poll for new messages and the server is slow to respond.
Background Javascript shouldn't impact editing.
3) I would really like to see an option for disabling Flash animation
until the mouse is over the Flash object. Dozens of people have asked
me how to stop these animations and I install FlashBlock for them.
These people don't mind the ads, its the animations that annoy them.
4) I would like to see an icon on the status line for pages that have
parsing errors or that are violating standards. Maybe that would shame
the author into fixing them.
5) I'm using a spellcheck add-in. This is probably common enough to
warrant being built in. Provide a toolbar button for it to help new
users.
6) When using the mouse to switch tabs quickly I often trigger an
unwanted drag operation. This bothers other people I work with too.
7) We need a Flash update for Linux. I am having to switch to WIndows
to access several sites I commonly use.
8) Playing multimedia formats on Linux is a major pain. I still can't
figure out how to make financial sites using Microsoft formats play on
Linux. Again I have to switch to Windows.
--
Jon Smirl
jons...@gmail.com
Peter Kasting
> 5) I'm using a spellcheck add-in. This is probably common enough to
> warrant being built in. Provide a toolbar button for it to help new
> users.
Automatic spellchecking as-you-type is built in to Firefox 2.
PK
Jon Smirl
I'll check it out, I'm still using 1.5.
On the multimedia point, how about a 'Verify Installation' choice on
the Help menu? It would take you to a series of pages for checking all
of the common media types. If one doesn't work it would provide
working instructions for fixing it. Working instructions is a key
point, Real Player has told me that my format is unsupported and I
need to upgrade hundreds of times, and after the upgrade the format
still doesn't work.
Tighter integration with gstreamer might solve this. If gstreamer is
used to play all of the format then the only thing that needs to be
checked for is the existence of the codec library. Can FIrefox borrow
code from Songbird to make this easier?
My preference would be to get rid of all the video player plugins and
more to one internal player that can use multiple codecs.
--
Jon Smirl
jons...@gmail.com
Robert Kaiser
> I've actually done this in my configuration, and the configuration I set
> up for end-users, for a couple of years.
Sounds quite similar to how the SeaMonkey URLbar behaves, I think ;-)
Robert Kaiser
Robert Kaiser
> 5) I'm using a spellcheck add-in. This is probably common enough to
> warrant being built in.
FIXED in FF2 already, IIRC. ;-)
Robert Kaiser
Martijn
Except with Peter's setup I don't have to arrow down to get to the
"Search Google for" thingie. This is something I alway found extremely
annoying in Seamonkey (and it is something I often made mistakes with)
With Peter's setup, I can just type some text, press enter and do a search.
Regards,
Martijn
> Robert Kaiser
Boris Zbarsky
> 7) We need a Flash update for Linux.
Too bad we don't have the source to it.
> 8) Playing multimedia formats on Linux is a major pain. I still can't
> figure out how to make financial sites using Microsoft formats play on
> Linux. Again I have to switch to Windows.
Again, too bad we don't have the source to those.
-Boris
Boris Zbarsky
> * People don't understand URLs. They don't know what they mean, they
> can't parse them, and they don't understand where to type them.
> I WOULD suggest that the
> current browser UI for navigation is an evolutionary modification of
> designs originally made for people who understood the World Wide Web,
> and that perhaps some rather drastic UI rethinking is warranted.
I have to admit that it took me a bit to convince myself of this; it doesn't
help that most people I know are not typical users. That said, it seems to me
that we should consider something along the following lines as our default UI:
1) Have a single text entry field in the UI; have this be a
search-and-url-entry widget.
2) Do not show the current URL in this widget. Show either nothing or the last
thing searched for.
3) Show the current _hostname_ (or some other "what site this is" indicator) to
prevent phishing.
4) Have the current URL bar available via toolbar customization so that we can
use the browser as dogfood. ;)
This does, of course, raise the issue of us actually getting sufficient testing
on the "default" UI if a significant fraction of our testing community
customizes it per #4... But I think this setup would be something that someone
like my grandmother would find much more useful than what we have right now,
especially since so many URIs out there suck so much (complete with cgi-bin
directories, GET queries, etc).
-Boris
Peter Kasting
> 2) Do not show the current URL in this widget. Show either nothing or
> the last
> thing searched for.
> 3) Show the current _hostname_ (or some other "what site this is"
> indicator) to
> prevent phishing.
Yeah, as I considered the issue over the last several weeks, I decided
this almost has to be the case. The box doesn't appear to be somewhere
you can search, or be very inviting, if it's showing a big URL. And to
most users, the hostname is much more informative than the full URL
anyway (see Spoofstick).
However, I still wanted a place to show the full URL, and to copy it
from if you wanted to make a small modification. I was undecided
whether this should hide in something like Page Info, be shown along
with the full page title on a small strip at the top of the current
page, show up in the status bar somewhere, or be available as a toolbar
widget for customization purposes.
This would benefit greatly from a couple of months of testing various
configurations so we had a better feel for what sorts of use cases it
does and does not serve.
PK
Andrew Schultz
> Except with Peter's setup I don't have to arrow down to get to the
> "Search Google for" thingie. This is something I alway found extremely
> annoying in Seamonkey (and it is something I often made mistakes with)
> With Peter's setup, I can just type some text, press enter and do a search.
You can, although not by default. Enable "Internet Keywords" in the
smart browser pref panel and it works as you describe (this has been
broken on trunk and branch for a bit but recently got fixed).
--
Andrew Schultz
aj...@buffalo.edu
http://www.sens.buffalo.edu/~ajs42/
hold...@gmail.com
ago, with IE4 or 5 if I remember. At least they were combined in some
sense back then, partly for expert web users and partly for novices.
The idea of combining them sounds better in promise than it really is.
At least, to me, it becomes more confusing what exactly is going on
when I do type something there.
I do like the idea of parsing URL's more though, and presenting them in
more readable fashions. Perhaps with the root shown prominently, and
subdirectories, subfiles and cgi codes somehow offset from it. Or maybe
give the page title more UI space in the address bar somehow. I don't
know if that's ever been looked into or not.
Peter Lairo
> Mike Beltzner wrote:
>> So go ahead and throw a couple of ideas onto this thread!
>
> Here's what I've seen from novice users:
>
> * People don't understand URLs. They don't know what they mean, they
> can't parse them, and they don't understand where to type them. This is
> why they get repeatedly fooled by obvious phishing URLs, why they type
> URLs into the Google search box instead of into the address bar, and why
> a dedicated search box is friendly: it doesn't have that weird string in
> it that takes a long time to delete.
This "weird string" is *the* *unique* identifyer of a page, and
sometimes the only way to know where on a large site one is. It's
possibly the one piece of UI that a browser must have. And to make it
read-only (as in the status-bar) would seem like a "castrated" (AOL-ish)
browser.
Please don't remove it!
PS. Sorry if I'm ranting. It just seems like some seemingly good ideas
quickly gain a "me too" snowball momentum.
--
Regards,
Peter Lairo
The browser you can trust: www.GetFirefox.com
Reclaim Your Inbox: www.GetThunderbird.com
Boris Zbarsky
> This "weird string" is *the* *unique* identifyer of a page
You know that. I know that. Most people don't know that.
> and sometimes the only way to know where on a large site one is.
Very rarely do I care about this, especially because most large sites have
completely inane structure. For example, consider the following URI in one of
my browser windows right now:
What does this URI tell me? Absolutely nothing past the fact that I'm dealing
with a Williams-Sonoma registry, and telling this involves understanding what
the different parts of the URI mean. There's no indication that this is the
registry-editing page. Even the indication that this is Williams-Sonoma is
weak, due to the weird "www1" thing.
> It's possibly the one piece of UI that a browser must have.
For me and you, yes. In general, I'm not convinced.
Again, I'd support having the option to show and allow editing the full URI.
The question is whether most people would make use of this option.
This is really a case where a usability study would be nice so we have some hard
data.
> PS. Sorry if I'm ranting. It just seems like some seemingly good ideas
> quickly gain a "me too" snowball momentum.
And it seems to me that there is a strong tendency to object to things "because
it's not what I'm used to." I've been guilty of that myself, of course.
-Boris
Sander
> Peter Lairo wrote:
>> This "weird string" is *the* *unique* identifyer of a page
>
> You know that. I know that. Most people don't know that.
_Is_ there any solution which can really be good both for those who know
and those who don't? Hasn't the gap simply grown too large?
I can see only more and more compromises ahead, making things "not
quite" what they could be to be perfect for novice users, and ever less
attractive to expert users.
Or to couch it as an answer to the original question of problems facing
users: There is a really wide range of understanding by users of each
separate bit of functionality that is provided by the browser and by
websites. The user is facing the problem that virtually none of this
functionality is exactly as simple or as complex as it'd need to be for
him to complete his tasks most usefully/effectively.
How can we manage to make available all parts of the functionality, each
at the right level of complexity and usefulness for the individual user?
(Without falling in the trap of pushing all the hard work of solving
this to the user himself?)
Sander
Boris Zbarsky
> Whatever is done with it, I think it must be very easily
> discoverable. I see addresses in print media and hear them spoken
> on the radio very often. Usually, they aren't any larger than
> something like www.example.com/foo, but I think many
> far-from-expert users do need an easy place to type such a thing.
There's a difference between a field to type URIs and a place to edit whatever
URI you're at right now. We need the former. I'm not sure typical users need
the latter.
-Boris
Adam Smith
> So, if normal people actually consider the location bar, and the google
> search box the same thing -- if evolution is allowed to decide - it IS
> the same thing!
>
> Basically, what I propose is to merge them.
This is how I use FireFox *all the time*. I have kept things bookmarked,
sure, but most things I'm searching for are going to be the top hit in
Google anyway, so it's just easier for me to type, say "nine planets" in
the URL bar and Google's 'I'm Feeling Lucky' search will take me to
www.thenineplanets.org.
It's *so fast* that I don't even notice on my broadband connection.
This is a winning feature of Firefox and if it can be expanded or grown,
then fantastic.
My contributions:
- Integrate 'Adblock Plus' or write your own and make it easy to use.
Have it download 'community' ad blocking lists and have the option to
contribute back custom entries (with some way of preventing abuse.) I like
being able to block components and really tidy up those websites I
frequently use!
- The installation of browser plugins can be sucky, at least on a Mac. For
example, if you don't have Windows Media Player installed (on a Mac,
anyway) you get sent to the plugin finder but it quite often doesn't
recognise the MIME type and cracks the shits.
- When an xpi can't be installed because it's not from a known domain,
once the domain has been added to the whitelist, you have to re-click
the link to start the process again, which is really annoying.
- Let me right click on embedded components (flash, video, etc) or help me
grab their direct URLs easily without having to go through source code
or the page info window. I kinda just want to right click on them and
"Copy element URL to clipboard"
- Let me specify a custom User Agent like Konqueror does. Some websites
are really annoying.
- Firefox + BitTorrent!
- Use Aqua form elements, please! Safari's rendering of forms looks so
much nicer than Firefox on the Mac.
- This might be more of a bug than anything else, but try going to the
Google customised home page (http://www.google.com.au/ig) and then
enable the preference that lets you specify that a website must follow your
link colour scheme (ie the one that makes browsed links purple by default).
The rendering of the site (and probably many others) suddenly becomes
hideous. I've noticed that most websites I visit these days override
everything with CSS and your browser won't display browsed links in a
different colour anymore (the web, she is a changing!) I think this lovely
little forgotten feature has been left behind as the web got larger.
- Make the passwords exportable/importable. There's an add-on out there
that does this but I was surprised to find that the browser didn't do it
by default. Also, aren't they stored in an unencrypted or easily
decryptable fashion? Can this be fixed in the longer term? Usernames and
passwords are plentiful on the Internet and they need to be protected!
- Stop people from using the same usernames/passwords everywhere! I tend
to use the same username/password combination on most generic websites
because it's easy to remember and I can recall it from my brain when I want
to log in to gmail or youtube from my friend's house. But what happens if
youtube gets exploited and one day someone dumps their authentication
database to the web? My solution would be: Make Firefox generate a
password for me. Store that password in my password library but then make
my password library accessible somehow. Let me bring it to his house on my
USB key. But ultimately, make Firefox help me choose better passwords for
my websites so that I don't keep using the same ones over and over.
RSI setting in, so I'll end it here. Haha. Cheers.
--
Adam Smith
Corporate Support Team Leader
Internode
P (08) 8228 2999, F (08) 8235 6999
E ad...@internode.com.au, W http://www.internode.on.net
Robert Kaiser
> This is how I use FireFox *all the time*. I have kept things bookmarked,
> sure, but most things I'm searching for are going to be the top hit in
> Google anyway, so it's just easier for me to type, say "nine planets" in
> the URL bar and Google's 'I'm Feeling Lucky' search will take me to
> www.thenineplanets.org.
Pity that our solar system now only has eight planets ;-)
Robert Kaiser
Robert Kaiser
> I have to admit that it took me a bit to convince myself of this; it
> doesn't help that most people I know are not typical users. That said,
> it seems to me that we should consider something along the following
> lines as our default UI:
>
> 1) Have a single text entry field in the UI; have this be a
> search-and-url-entry widget.
> 2) Do not show the current URL in this widget. Show either nothing or
> the last
> thing searched for.
> 3) Show the current _hostname_ (or some other "what site this is"
> indicator) to
> prevent phishing.
> 4) Have the current URL bar available via toolbar customization so that
> we can
> use the browser as dogfood. ;)
Hmm, reading everything in here, an idea pooped up in my head (yes, one
of those nasty pop-ups):
What about making what's currently the URL widget a tabbed widget, i.e.
a text entry box with a small tab strip above it that lets you switch
between "Web Search", "Page Search", and "(Page) Location"?
It would look like this:
+------------+ +-------------+ +---------------+
| Web Search | | Page Search | | Page Location |
+-+ +-+-------------+-+---------------+-------------------+
| +-----------------------------------------------------+ +--------+ |
| | | | Search | |
| +-----------------------------------------------------+ +--------+ |
+--------------------------------------------------------------------+
The tab strip should probably be very narrow and in a relatively tiny
font. The Search button would change to "Go" in the "Page Location" view
- and of course, this bar can contain the search engine dropdown the
current search bar has, I also have left out stuff like RSS icon here,
it's just the basic idea.
Would users understand that? Would more professional users find it
convient enough to use?
Perhaps the "Web Search" mode should do some simple detection and use
the input as a URL for loading the page if it starts with a protocol
identifier or "www."
We also could merge the find bar into this as "Page Search" (or "Find in
Page"), as I still think the bar at the bottom of the screen is quite
unintuitive.
This is just an idea, but I've had this some days ago, revisited it
every time I read this thread and I still think it might be an
interesting way to go.
Robert Kaiser
Matthias Wallner
> [URLs]
> Very rarely do I care about this, especially because most large sites
> have completely inane structure.
just an idea:
what about changing the URL bar to something like the new Vista Explorer
does with File Paths
(see
http://common.ziffdavisinternet.com/util_get_image/10/0,1425,sz=1&i=109466,00.jpg
for an example)
you'll see
ww1.williams-sonoma.com > reg > (next part, shortened) > ...
and when you click into this, you'll get the text field with the URL.
It'd be a fun thing, if i'd could click the ">" on any part of the url
and see other parts of this page that i had allready visited.
But for this to really work, the URLs has to be changed by site owners
to actually mean something.
E.g. I'm now on
cnn.com > worldwide > europe > austria > politics > [something]
then click on the ">" between "europe" and "austria" and see the other
sections i've visited like "germany", "UK",...
i'd really like something like this.
Daniel Brooks
Boris Zbarsky <bzba...@mit.edu> writes:
>> Maybe instead of all that we should just have a big full screen narrated
>> tutorial that educates the user on what a url is, how phishing usually works,
>> why not all urls are very helpful, the difference between navigation and
>> searching, etc.
>
> I think users don't care, nor should they have to. Really, URLs are
> not integral to the web's user-facing aspect; they're an
> implementation detail. Users are interested in much more high-level
> concepts like "Who's site am I looking at" and "How do I get back to
> this state later?"
>
> -Boris
Well, I agree that urls are an implementation detail. However, understanding what a url is and how you use it is the key to understanding how a web browser works. Also, having a tutorial doesn't actually preclude changing the UI in any way to present the current url better, though I think a tutorial might get better results. I'm thinking of a tutorial that would begin with a brief explanation of what a url is, then go on to why we think that's too complicated for most people and either introduces any changes we've made to the UI, or explains how to get the most out of the current UI (because it's not actually that difficult, once you know what to do). Because this is the key way of interacting with a browser, a few minutes of education will go a long way.
Also, we could use an svg animation, since that can be done at a higher quality for the same number of bits per second.
db48x
Shadow Skill
While I did not originally suggest Go as You Type or GAYT as a method
of dealing with the confusion some (many) have with respect to urls I
would argue that if users tend to type urls into the google search
engine field the presence of the urlbar field inside the browser itself
should be dependant upon the act of typing first and foremost. If the
user doesn't understand what the url bar is they probably do have some
idea that you must type stuff in order to get anywhere.
GAYT is nothing more than type ahead find or Find As You Type (FAYT)
directed at the urlbar instead of directed at searching webpages for
strings. Quite frankly I think its backwards that FAYT exists yet a
feature like GAYT has yet to be implemented in anything other than file
managers.
I would also argue that the urlbar should possess the same behavior as
the string search field that appears when you use the find in page
functionality in that if the urlbar is positioned at the top or bottom
of the screen it "slides" in and out.
I think it would be a good idea to place the page load meter into the
urlbar like Opera. Its much easier to tell what a page is doing by
looking at the urlbar is rather than taking one's eye off of the url
bar to focus on the status bar that is usually at the bottom of the
screen.
I also find it difficult to focus on text in web pages when the mouse
changes into an I-bar, Opera does not change the pointer into an I-bar
unless the cursor is in a form it would be nice if there was an option
to do the same in Firefox. One other possible solution is to add an
option to the caret browsing mode that highlights the active line like
in the text editor show here http://www.putfile.com/pic.php?img=1576643
the line highlight colour should be user defineable.
Stylish support should be integrated into firefox or Firefox should not
allow webpages to inherit colours from the system theme. Stylish
provides a method to change various parts of a website via css and it
allows one to download pre made styles for various sites. I use it to
force all text fields to have a black background and white text
circumventing any stupidity propagated by the web browser or the theme
engine. Nor do I have to busy myself sending out emails to a bazillion
webmasters telling them that they need to explicitly declare their
colours. Do note that Opera already has 99% of the functionality of
the stylish extension built right in since at least Opera 8.5.
Ronin
1. Download Manager behavior: I should be able to resume download
across sessions otherwise "Pause" really is not of much use. I also
should ideally be able to schedule downloads. (Use case being where I
would like to schedule downloads during the day time while I am at
office and not using my browser).
2. Tab Grouping: I really don't have a solution in mind, but I would
love to be able to group my Tabs. When I am browsing with 40-50 tabs
open, grouping them into categories like Travel, News, Mozilla etc
would be very helpful. Hint: See the new split screen functionality in
Maxthon 2.
3. Tab Opening Behavior: One of my pet peeves with the current behavior
is that if Bloglines is my first tab in the row and I have 40 tabs
open, the link from Bloglines opens in the 41st tab. What kind of a
usability is that? Child tabs must always open next to parent tab. New
tabs must open at the end (ideally user should be given a choice
between opening a new tab next to current or at the end).
4. Take Notes from the Web/Save Web page: This is self explanatory and
basically required for offline browsing especially since not everyone
is lucky enough to be online all the time. Some basic functionality to
save web pages for offline browsing is definitely required.
Problems observed with people not too experienced or keen (busy
executives/ parents etc):
1. Maintaining Add Ons: Philosophy behind keeping Firefox lean and mean
was that features could be added via Addons, but maintaining Addons in
a working condition beyond 6 months on an average is a real pain. If
the author of the addon does not update it, things go for a toss. Most
common scenario being where I set up Addons for my parents only to find
them not working 6 months later, either because Firefox version was
updated or because the author decided not to continue to update it. I
really have no solution for this but I hope something can be figured
out.
Deriving from the last point, while I do understand that Firefox should
be kept as basic as possible, but the definition of 'basic' keeps
changing and so should Firefox. A few features I would love to see
included are given below, coz depending on extensions for such basic
browser functionality does not really make too much of sense:
1. More Tab Options: Open New Tab from Address
Bar/Bookmarks/Links/History etc, Open Tabs Next to Current and so on.
2. Very Basic Super Drag and Drop: (Open links in new background tab
when dragged; Search for selected text in background tab when dragged):
Very minimal and basic coding would be required for this (see 'Drag De
Go'/'Super DragnDrop' extensions). Use it once and you will know how it
is really a breakthrough in user experience. Simple and very very
intuitive.
Thanks. I hope some of that makes a little sense. I would very much
love to see that Firefox is kept as light as possible but not at the
cost of very basic user needs and browser standards (I don't mean CSS;
I mean usability standards, Firefox is already the God of CSS).
Greg
number of UI explorations that looked at some of the issues mentioned
in this thread.
- URL / URI confusion
- Search / Location confusion
- History and Bookmarking
- Plus some others.
I got the green light to post some of these, I think they may get some
ideas flowing.
http://www.raizlabs.com/blog/2006/08/explorations-in-firefox-ui.html
Greg Raiz
Raizlabs.com
Kindhuman.org
Benjy Grogan
I was thinking more on Bookmarks and how to improve them and it
occurred to me that a system like GMail's Star label would be useful.
If you're going through all of your emails and you come across a
conversation you want to put aside because it has important
information then you simply click on the Star to the left of the
sender's address. If every web page/tab had an unhighlighted Star
then if you wanted to keep a bookmark on a web page you would simply
click on that Star, and the web page would be stored in a revamped
Bookmark system.
It's an interesting way to think about tabs to begin with. Maybe all
web pages could be thought of as incoming emails, and you could label
them beforehand based on certain cues.
Benjy
On 22 Aug 2006 23:52:09 -0700, Mike Beltzner <mbel...@gmail.com> wrote:
> (note: this message is being cross-posted to mozilla.dev.planning and
> mozilla.dev.apps.firefox with a followup set to
> mozilla.dev.apps.firefox; if you are reading this as an email list,
> please make sure you reply to dev-apps...@lists.mozilla.org, and
> make sure you are subscribed to that list so your reply doesn't
> bounce!)
>
> A few weeks (months?) ago on a Gecko 1.9 / Firefox 3 planning call, I
> mentioned that one thing that I was never terribly thrilled about was
> that when looking at planning out the features for Firefox 2, I didn't
> push hard enough for every proposed feature to explain the problem that
> it was attempting to solve. Without understanding the problem we
> believe to be solving with a feature, it's extremely hard to get a
> sense of when the feature is complete, how important the feature is,
> what the impact of cutting the feature is, and what advantage we get by
> including it in the final product.
>
> In last week's meeting
> (http://wiki.mozilla.org/Firefox3/StatusMeetings/2006-08-16#Firefox_3),
> I started brainstorming up a list of problems that I felt users were
> facing when using web browsers today. I don't think we should focus on
> solutions or implementations just yet, but instead focus on answering
> the following question: "What are the problems with the web that users
> are currently facing, whether they realize it or not?"
>
> So go ahead and throw a couple of ideas onto this thread!
>
> cheers,
> mike
mihai.d....@gmail.com
them refer to specific/ particular website design, not to the browser.