New issue
Advanced search Search tips

Issue 13 link

Starred by 91 users

Issue metadata

Status: WontFix
Owner: ----
Closed: Sep 2008
EstimatedDays: ----
NextAction: ----
OS: All
Pri: 2
Type: Feature

  • Only users with Commit permission may comment.

Sign in to add a comment

Closing last tab also closes window

Reported by, Sep 2 2008

Issue description

Product Version      : (1583)
URLs (if applicable) :
Other browsers tested: n/a

What steps will reproduce the problem?
1. Have a window with one tab
2. Close that tab

What is the expected result?
The tab returns to the New Tab page or whatever the user set.

What happens instead?
The window closes.

Please make a setting that allows the user to keep the window open when 
closing last tab.
Showing comments 44 - 143 of 143 Older
Keep in mind that Chrome runs tabs in processes, hence, each tab has its own process 
or a process has a group of tabs. So if you move a tab out of the window, it will 
have its own  border like structure. What I am trying to say, a window with one tab 
is a single process, if you close that tab, you close that window. 

I believe the current functionality is correct, since  the whole tabbing architecture 
differs from other browsers. They should make a feature where if all tabs are closed, 
it will go back to the chromium home page.
Actually, Chrome uses 2 processes when only 1 tab is open (x tabs + 1 main process) 
so that isn't true.

Comment 46 by, Sep 6 2008

I really don't think this is any issue. If you accidentally close the last tab, just 
open Chrome again. It doesn't much time at all...
Close the last tab using the tab's X button, should open the start page, not close 
the browser. 

Comment 48 by, Sep 6 2008

Status: WontFix
Unlikely to implement this as described, because:

- overriding the meaning of standard controls is undesirable from an interaction 
point of view.
- we don't want to carry an option specifically to do this.

If anything, this is extension fodder.

Comment 49 by, Sep 6 2008

As long as this will eventually be possible through a future extension, I'm cool with 
"overriding the meaning of standard controls is undesirable from an interaction 
point of view." - this is not a standard control and there is no inherent meaning to
be overriden.

The thing is, the "X" button on the tab is NOT a standard control with a defined
meaning. It only exists in other web browsers, whose behavior is not the same as
Chrome's current behavior. In IE8 and Opera, the X button vanishes when there is only
one tab. In Firefox 3 (default settings) and Safari the tab bar vanishes when there
is only one tab. Firefox allows an option to instead replace the tab with a blank tab.

In NONE of these browsers is it possible to close the window by means of a button
located on the tab.

There is nothing to "override", and the meaning of this _non_-standard control should
still be negotiable at this stage.

Ctrl-W, on the other hand, _does_ close the browser in all of these. This should
continue doing what it does now.

Name one program other than Chrome/Chromium that behaves in this way.

Comment 51 by, Sep 6 2008

"overriding the meaning of standard controls"?

Does Windows shut down when the last application window is closed?  No; the user has
to click to the menu option that specifically says "Shut Down My Computer".  In the
same regard, the little "x" in my browser that means "Close This Tab" DOES NOT and
SHOULD NOT EQUAL "Terminate This Program".

Please,, explain how we are all mistaken, for every other
tabbed/windowed program in existence today will need to be updated to conform to your
new "meaning of standard controls".

Comment 52 by, Sep 6 2008

Thanks Ben. I support your views.  Users: stop using your mouse so much. Power users use keyboard shortcuts.
A browser should not be idented for "power users".

And again: The "close tab" button is not a standard control.

Comment 54 by, Sep 6 2008

Just my two cents.  This particular behavior does bother me.  I don't know how many 
times that I have closed the whole browser by mistake, simply because I was done with 
Gmail... didn't want to see it anymore... planned on visiting another site, but chose 
to close the Gmail tab first (which happened to be the last tab.)  In Firefox, it was 
configured to leave a blank tab if the last tab was closed.  I honestly don't know if 
this is the default behavior, or if this is the way I always configured it on all of 
my machines.

To me, the intuitive behavior matches the way all other Windows applications act when 
there are multiple sub-windows.  For instance, Word when you have multiple Word files 
open.  Closing the application closes all of the Word files.  Closing the Word files 
individually leaves the application open even when the last Word file is closed.

With this being said, I can still understand the decision made by the Chromium team 
here.  Their goal wasn't to make a browser that is just like any other Windows app.  
Their goal was to make a browser unlike any other browser.  One of their design 
philosophies was that there shouldn't be pop-ups or prompts... the user shouldn't be 
aware of the browser... the browser should just be a window to the web pages or web 
applications and stay out of the way.  So, if closing the last tab "prompts" you... 
it would break this design philosophy.  If closing one tab at a time suddenly takes 
away the close button of the last tab, although it would be behavior which might 
mimic other browser, it wouldn't exactly be persistent behavior.  If closing the last 
tab doesn't actually close the tab, but opens a "New Tab", this would be inconsistent 
with the act of closing a tab.  Finally, the browser has no real "purpose" to have 
zero tabs.  If the only thing you can interact with is the "New Tab" button or the 
settings menus, then the application isn't being transparent enough.  So, it only 
makes sense that it goes away at that point.

What would make me happy?  Personally, the only frustration with the whole browser 
closing when it does... is that I like for my task bar to auto hide... and for some 
reason, from time to time, Windows makes my quick links disappear and I have gotten 
tired of trying to bring them back each time.  Also, my Start Menu is so cluttered 
that it isn't a very fast way for me to run applications.  This means that I run 
applications in one of two ways:

1) Double-click an icon on the desktop.
2) Hit ctrl-ctrl to bring up the Google Desktop search box... start typing in the 
name of an app... wait for the name of that app to appear and be highlighted... and 
hit enter.

Option number one is often not very fast because I have several windows open that are 
blocking the desktop.  Since the quick links on the task bar are gone, I usually 
right-click on the task bar to show the desktop, then find my icon.  After 
accidentally closing the super-fast Chrome browser, this becomes a huge speed bump in 
my productivity.

Option number two is efficient most of the time, but sometimes it takes a while for 
the search box to recognize I am typing in the name of an application I wish to run.  
If I hit enter too quickly, it performs a search for that name instead.  An 
alternative way might be for me to simply perform a generic or random search using 
the Google Desktop quick search box which will inevitably bring up Chrome, but it 
wouldn't bring up the expected tab (the New Tab or my home page)... it would bring up 
a search result for something I didn't actually intend on searching.  So this isn't 

Ultimately, what would make me happy is if there were an easy-to-perform keyboard 
shortcut for running Chrome on my machine.  Something as easy as ctrl-ctrl is.  I 
don't want to set a keyboard shortcut for the desktop icon because I don't think 
ctrl-alt-[key] is necessarily intuitive.  Instead, it would really be awesome if a 
simple CTRL-T would run the Chrome application (if it isn't already running).  After 
all, ctrl-T is meant to bring up a new tab.  If it can't be this for conflict 
reasons, perhaps it could be something as simple as ctrl-ctrl or shift-shift which 
were already implemented in Google Desktop.  Maybe something like alt-alt, provided 
the alt key could be recognized in this manner?

Sorry if i was misunderstood. When the last tab is closed, is not clear the browser
will go shutdown! The user spected close the browser clicking in the X-top-right corner.
Maybe, when close the last tab, must open the "default" chrome tab, with the
thumbnails of the most visited sites.
I agree with ben's decision.

I mean, it's a close button. What should a close button do ?
- Close the current tab
- Navigate to my home page

I think the choice is pretty easy.

Comment 57 by, Sep 11 2008

If you click the close button of the last active application in an operating system, 
how should the operating system react?

It should log you out, or should shut down, because clearly if you are finished with 
the last application, it is pretty obvious that you were finished with the entire 
operating system.

If you planned on running another application, just boot up the operating system 
again, or make sure that you run some other application before closing the last one.


Why should there even BE a close button on the last tab? IE doesn't. Opera doesn't
(by default). Firefox doesn't (by default), since it doesn't have a tab bar. Neither
does Safari.

Its behavior is clearly throwing people off (since in the only browsers [FF and
Opera] where you can get a close button on the only tab at all, and on those it's not
by default, it doesn't close the browser)

Comment 59 by, Sep 14 2008

I prefer the current behaviour.  I don't want to hit Alt-F4 instead of the usual 
Ctrl-W just because it's the last tab. Chrome loads fast enough anyway, so if you 
want a "New page" tab, just load Chrome from scratch.

Comment 60 by, Sep 15 2008

I had a bit of a revelation regarding this issue the other day and realized that it 
has nothing (or little) to do with the close button on the tab, or the fact that when 
the tab is closed the browser closes.  I may be wrong, but I BELIEVE that most people 
who are clicking this button are under the impression that the whole application will 
not close because this tab is so far to the left (and most people are used to 
clicking the close button far to the right to close a full application.)

So, I started thinking about tabs, and the position of this close button.  I also 
started thinking about why the whole browser would exist WITHOUT A TAB... there 
really isn't much reason, unless you accidentally closed the tab.  I thought about 
the last tab NOT having a close button... while this would be consistent with other 
browsers, it isn't very consistent behavior.  Here is what I ultimately came up with.

If the browser only has ONE TAB... why does this tab only have to take up a small 
portion of the title bar area of the browser?  Why doesn't the tab extend to the full 
width of the tab area?  Here is what this would do:

1) It will give the tab more room to show longer titles, rather than always 
truncating titles to a large degree.

2) It will put the close button of a SINGLE-TAB-BROWSER so far to the right that 
instinctively a user will better realize that this is much different than closing a 
single tab, and leaving the browser open.

3) It will eliminate a lot of wasted space in the browser when a single tab is open.

4) It will be more consistent with the behavior of the browser when many tabs are 
open.  Rather than using only a small section of the tab area when there are a few 
tabs open and using the full tab area when many tabs are open... it will always use 
the full tab area.  It will also always shrink tabs to be proportionate, rather than 
only implementing this behavior when enough tabs are open.

5) It will place the "new tab button" in a static location, rather than constantly 
moving it to the right until there are so many tabs that it stays to the right.  
Instead, the new tab button will always be to the right.


Comment 61 by Deleted ...@, Sep 16 2008


I totally agree that closing the last tab should NOT close the whole browser.

I expect a blank tab so I can start browsing all over again.
I think that clicking on the X of a tab closing the whole browser is counter-intuitive.  
The UI behavior of the X button on a tab is to close the tab, not to close the browser - 
if I wanted to do that, there's a standard way of doing that - the window's X button.

I think the current behavior is wrong.  Chrome should either:
a. Have no x on a tab if it's the last one
b. have an empty window with no tabs after the last tab is closed, or
c. The tab should close, then immediately open a blank tab.

My preference is c, and the wording is significant - the tab should definitely close, as 
that's the standard behavior for the button, but the standard behavior for an empty 
window should be to open a blank tab, on the grounds that an empty browser window is 

It can be made clear with the tab open/close animation that this is the sequence of 
I'd like to point out that closing a tab can be done in several ways, as can closing 
an application, if you abide by Microsoft's Guidelines.

I.e., Alt+F4 is a standard and well-known way to close an application.  Ctrl+F4 is 
also a well-known way to close an MDI window or tab. It would be very confusing to 
use a shortcut known to be 'close this tab', and as a result your browser is shut 

I agree with the fact that closing the last tab should not close the BROWSER. But 
what it should do is close the browser WINDOW as long as it is not the last open 
window. If you think about it, i.e. 'Open link in new window' can (should?) be read 
as 'Open link in a new tab in a new window'. If you close that tab, which also 
happens to be the last tab in that window, it would make sense to close that window. 
Just as long as there's another (non-incognito?) window left.
Good comment and I must say, that I'm the same opinion. So this should be a good 
option. Maybe it may be possible to let the user decide by himself if the 
browserwindow closes on closing the last Tab in the last open Window or if the option 
of erikheemskerk is chosen.

Comment 65 by, Oct 16 2008

I just noticed that Firefox 3.1 Beta 1 has a new (?) option
"browser.tabs.closeWindowWithLastTab" wich defaults to "true". So Firefox has at
least adapted Chrome's way of closing the last tab. But at least you can disable it. :)
Can we revive this? Seems like an important feature. Closing the last tab should at
most close the last tab, not the browser.

I'm used to this too and I've closed Chrome accidentally this way a few times too many.
There seems to be a lot of supporters for this behavior, so in the end it would make
users happier, so we should. Everyone was supporting this until comment #17 and then
it got mixed up but I still see more opinions for than against.

Sometimes I like that simply closing a tab quits the whole browser, however it causes 
a serious performance hit. Running Chrome (when there are no existing processes 
running) causes it to do all kinds of maintenance things like thrashing the HOSTS 
file, updating the Safe Browsing file, etc. Closing that browser because there are no 
open tabs means that all this thrashing and work happens unnecessarily the next time 
that you run Chrome.

One ideal solution would be that when the last tab is closed, Chrome opens a blank 
tab and minimizes to the system tray. That way it effectively shuts down, but remains 
alive so that you can quickly and easily open a new page without it re-doing the maintenance updates.

Comment 68 by, Jan 13 2009

I hate to be the devil's advocate, but the reason this feature is unlikely to change 
is because doing so would add inconsistency.  To the lay-person who has never seen a 
browser before, why would closing one tab (when others are open) close that tab... 
but closing the last tab "open a blank tab and minimize to the system tray"?  All of 
these sound like great alternatives, but if the alternative is "do something 
different when you close the last tab"... it isn't likely going to be implemented, 
because it is inconsistent.

Imagine this is the Windows operating system... and if you close an application, it 
always closes the application... UNLESS you happen to close the LAST application, and 
then it just opens up a blank document and minimizes that document to the system 
tray.  This is the type of inconsistency the developers are trying to avoid.

A possible alternative is this.  If you close the browser using the main close 
button, it closes immediately.  If, however, you close the browser by closing the 
last tab, it instantly hides the window, but the chrome.exe process stays open for 
about 5 seconds longer before doing anything.  If, within that 5 seconds, you run 
Chrome again, it simply re-activates that session, but with the startup tab.

For those who are bothered by the amount of system resources closing and re-opening 
Chrome, this should alleviate it for the most part.  And although the behavior in the 
background is inconsistent, the behavior from the user's perspective IS consistent.  
As far as the user is concerned, the browser closed.  While Chrome does load really 
quickly, it would load even more quickly in these instances where one is more likely 
to "accidentally" close Chrome and want to immediately re-open it again.

Getting a bit more technical, this would mean that when the last tab is still open, 
there will be two chrome.exe processes in memory; one for the tab and the other for 
the main Chrome window.  When the last tab is closed, that tab's chrome.exe process 
is killed and the main Chrome window's chrome.exe process remains live, but the 
window itself is hidden from view.  Then, a counter of 5 seconds (or some other well 
determined time) is started.  If Chrome is run by the user again, it simply runs a 
single chrome.exe tab process which attaches itself to the already existing 
chrome.exe main window in memory, which then shows itself again.  This would 
theoretically happen as quickly as one would open a new tab in an already running 
Chrome.  If, however, those 5 seconds elapse and Chrome is not run again, the usual 
shut-down processing can take place and the final chrome.exe process is purged.


There are a lot of special case behaviors implemented, but the most complex they are, 
the harder it is for people to understand them, so this minimize on last-tab-close 
behavior seems like a stretch.

The main point is that closing a tab must not close the browser, just like closing 
the last application does not shutdown the OS. The same thing with VC++, when the 
last file is closed, or the last project, or solution. VC does not close.

The special case comes from the what do we do when the browser is to get zero tabs: 
Either we simply support that case or we have to provide a blank tab. It seems that a 
blank tab is more useful than zero tabs and does not had much overhead, so I feel 
this is the preferred solution.

We already have a single tab workaround for dragging one. IE simply removes the close 
button from the last tab. Yet, a close button which clears out the last tab (as is 
done in Firefox and Seamonkey) is useful when the user feels one starts a new 
'browsing-task' since it clears the tab's history.

Comment 70 by, Jan 13 2009

One of the points that is being missed, and was something that I didn't grasp fully 
when I first fought for this change, is that the reason the Chromium developers are 
fighting against this change is because, in their view, Chromium itself is not 
important.  Chromium is meant only to be a development platform through which web 
applications or websites can be viewed.  In other words, if you are only viewing one 
tab... in their mind, the website or the web application you are viewing in that tab 
is the only thing that is important to you.  If you close that one tab, then you are 
closing the only thing you are concerned about.  Therefore, leaving open a "web 
browser" is not the proper behavior in their opinion, because the "web browser" is 
not what the user is interested in... it's the web application itself.

I had previously used the "closing the last application closes the operating system" 
analogy.  Except, an operating system without any windows open is still useful and 
important to the user.  To make a closer comparison, if the user opened up Windows 
Task Manager and started closing each process one at a time and killed all of them, 
including explorer.exe... the operating system itself WOULD essentially be closed.  
So, I think this is why they are taking the viewpoint that if all chrome.exe 
processes (tabs) are killed... there's no point in leaving a shell of an application 
open that cannot directly be used... and it would be inconsistent to re-open a blank 
tab... after all, you told it to CLOSE a tab, you never told it to OPEN a tab, even a 
blank one.

And I don't think the idea of telling them "but application X does this"... because 
Chromium was, from the start, built from the ground up, in its own philosophy.  It is 
specifically trying NOT to be like "any other application".

Put another way, when you first run Chromium... you aren't running a web browser, you 
are opening a TAB.  After all, what is loaded on your screen is a single tab.  If you 
close that tab, it should eliminate everything you just ran.  It is a philosophy I 
have grown used to, and I rarely ever "accidentally" close Chromium these days.  I 
turned on the "home" button and usually hit this when I am finished with a website.  
If I only had one tab open, it is the behavior most are wanting when they close this 
tab.  If I have other tabs open, and I don't want those open anymore, I will close 
those... but I only close the background tabs, leaving the still-facing tab open and 
in the "home" position.

At this point, the main reasoning I can see behind why someone would not want 
Chromium to close and then to have to be opened again (because Chromium DOES open 
rather quickly) is due to the extra thrashing that the system goes through when 
Chromium is completely closed... or when it is open for the first time.  My solution, 
I think, would eliminate that.  After that, the only other "beef" I could see is due 
to the learning curve... un-learning the Microsoft Windows way of things and re-
learning the Chromium way of things.
None of this changes the fact that "close tab" should never equal the completely 
different action "close application" without some other hint to the user.

I don't so much care about the time or "thrashing" it might take to re-open Chrome, I 
care about having to perform any action at all (finding the program icon or even 
pressing a keyboard shortcut) to reopen a program that I DID NOT WANT TO CLOSE.

How about a compromise?  (1) Closing the last tab, if it has been used, would bring 
up a fresh tab in it's place.  Almost like hitting the hidden "home" button, but with 
added benefit of fresh history and a new process.  (2) If the last tab closed *is* a 
new tab already, then close the whole thing.  This way I can close a single, stale 
tab and receive a fresh one in its place without being stuck in a counter-intuitive 
cycle of "why won't this last tab ever close?"

When closing tabs I shouldn't have to be "on alert" for the last one just so I can be 
careful to open a new one *before* closing it -- Chrome should understand I want that 
based on the fact that I closed the tab and NOT the browser.  I don't know how it 
could be more clear.

Comment 72 Deleted

Well, even the status of this issue is WontFix, close last tab really should open a
blank page. Is more like OpenOffice Writer, but instead close the child window (with
no documents open) will open a blank page. More intuitive. X's tab closes the tab,
X's window closes the main app. More simple, impossible.

Comment 74 Deleted

No other browser even allows you to close the last tab - the simplest solution would
be to not show the tab close box when there is only one tab.

Comment 76 by, Jan 13 2009

All valid arguments.  Since other browsers do things this way, Chromium should be 
made to do the same thing.  In time, once Chromium has become fully compliant with 
the ways of the other browsers, we will discuss changing the name of Chromium.  
Perhaps something more compliant with other browsers... like we could call it 
Internet Navigator or Chromium Explorer.


@iambob - the developers' perspective seems to be that they consider a tab close
button to be a "standard control" - if that is to be the case, it should act the same
in all browsers. Or maybe they should change Chromium to behave differently in other
ways? For example, maybe require double-clicking to open links, or put the scroll bar
on the opposite side of the window. Or reverse the meaning of the ctrl-mouse-wheel to
be down to increase text size, up to decrease it (just when the other browsers have
_finally_ standardized on the opposite). Or maybe get rid of the right-click menu.

There is no need to be gratuitously different, especially when it inconveniences
users. There are already two perfectly viables way to close the window: the close box
in the upper right corner, and CTRL-W [which no-one is advocating getting rid of,
@gaston @igitur...]. We don't need an extra button to do the same thing, particularly
when no other browser has it.

Comment 78 by, Jan 14 2009

No, from the developers' perspective, these two facts are true:

1) The Chromium browser without a tab open is useless and should never happen.
2) The close button on a tab should act "in a vacuum."  That is to say, it shouldn't 
need to be aware of any other tabs.  So, if it is the "last tab" or "one of seven" 
tabs should be irrelevant as to how the close button on the tab responds.

Given these two facts, and accepting that breaking either of these rules is out of 
the question, how do you suppose solving this situation other than to say "throw out 
the rules" or "other browsers so such-and-such"?  This is the crux of the problem and  
the water is getting cloudy with talk that completely disregards these two 
facts/rules, which is exactly why suggestions that ignore these philosophies result 
in a "WontFix" response.
@iambob I did NOT just make that up. The one developer who has actually commented
here said exactly the same thing that I attributed to the developers in my post:
"overriding the meaning of standard controls is undesirable from an interaction point
of view." - the problem is, if the close button is a standard control, they are
overriding its meaning NOW.

Comment 80 by, Jan 14 2009

I'm not disagreeing with you.  I'm just trying to explain why the Chromium developers 
are ignoring the arguments being made.  According to their philosophical document 
which explains what Chromium is and what Chromium isn't, the two points I stated are 
simply fact.  If in trying to make your argument, your suggestion breaks one of those 
two rules, it will be disregarded entirely.

If you look closely to the beginning of this thread, I fought against this behavior.  
The problem is, my solutions which still followed those two rules were drowned out by 
people just saying things like "Firefox does such-and-such" or "in IE, it's expected 
behavior to do such-and-such"... so the short response of "overriding the meaning of 
standard controls is undesirable" and the quick dropping of this issue basically 
showed that no amount of suggestions is going to revive it, unless you can find a way 
to not break the two rules I stated.

Think of it this way.  The close button is a standard control, right?  But TABS in 
Firefox and IE are DIFFERENT than the tabs in Chromium.  In Firefox and IE, each tab 
is not its own process... and cannot be dragged outside of the browser (without a 
plugin.)  So, what this means is... a tab is not merely a "child window" of the 
overall application.  The tab IS the application.  Having a browser window with three 
tabs in IE or Firefox is like having three child windows inside of the parent 
application.  However, in Chromium, having three tabs means you have three 
applications which are all connected together in a neat little cluster.  So, once you 
close all three, there is nothing left to "be open"... and if clicking a CLOSE button 
also opens something (such as a blank tab, etc...) then the control does not have 
consistent behavior.

So, again, as long as everyone's argument for fixing this issue involves using terms 
like "Firefox" and "IE"... the arguments are going to be ignored, because Chromium 
was specifically developed to be different from those applications.  Yes, it has 
standard controls... like a close button closes... and a minimize button minimizes.  
But assuming that the "tab" within the window is "a child window" is making a false 
assumption.  To assume the "browser without tabs" is the "operating system" is also a 
false assumption.  Each tab is an application... all within an application space... 
once the last tab in that application space is closed, the application space ceases 
to exist.  The application space doesn't need to "remain" because the operating 
system (Windows, Linux, etc...) will do a fine enough job to allow you to run the 
application again.  And even in Windows, if you close the last Notepad... it doesn't 
arbitrarily close that Notepad document and then suddenly load a blank Notepad for 
you.  If you still wanted to type in a Notepad document, just open another one... or 
select File --> New from the last open one.
Except that Notepad does not perform a bunch of maintenance such as version checking, 
anti-phishing updating, HOSTS reading, etc. each time it starts up.

Comment 82 by, Jan 14 2009

Precisely.  Which is why my solution involved solving the maintenance and background 
process issues and not trying to change the way the behavior appears to the user.

Comment 83 by, Jan 14 2009

I will reiterate... my solution was a background solution only.  If the last tab 
(application) is closed... rather than the application space (the extra chrome.exe) 
closing immediately... it could remain running for 5 seconds or so.  If no action is 
taken before that time is up, all of the normal maintenance and background processes 
are run for closing... and the next time the application is run, it can do its normal 
start-up processing.  If, however, before that 5 second interval is up, a new 
Chromium application (tab/browser) is opened, it can attach to that existing 
chrome.exe process and you would not have unnecessary startup and shutdown processing 
happening every time the last tab is "accidentally" closed.

This would only work this way if the TAB's close button is clicked.  If the close 
button of the whole window is clicked, it could either perform as it currently does 
(nuking everything immediately)... or, if it would be more in line with consistency 
to follow this same rule, it could do this as well.  The only time it could become a 
problem is if someone were closing Chromium for a reason (memory leak?) and really 
wanted it completely flushed from memory.  To do this, they would need to wait out 
the 5 seconds or so, unless closing the window itself terminated everything 

I would go for that. I would be happy using a third party app to minimize it to the 
tray as long as Chrome would:

(1) start up faster (don’t automatically do all that stuff each and every time)
(2) allow blank new tabs

Of course it would still be better if a new, blank tab were opened when you close the 
last one since currently there isn’t a way to open a blank tab at all (ie start over 
with a fresh, empty browser).

This is a UI bug.  This is not a feature request.

There is confusion here between "tab" referring to a separate browsing process and "tab" referring to the 
graphical interface strip at the top of the screen.  This bug is specifically regarding the ergonomic deficiency 
of the close button on the tab-like graphical interface.

The graphic design of Chrome is no accident.  The placement of the "new tab" [+] button is is just as vital as 
the placement of the "close tab" (x) button; their locations relative to the tab UI are one of the only hints as 
to their function.  Labels would be obtrusive and a waste of space, sure, but even tooltips are absent here; the 
user is trusted to rely on graphic design cues as well as prior experience with UI in general.

The design cues are clear.  A plus sign next to open tabs will create a new tab, a red X inside a tab will close 
that tab.  However there is no indication to the user (and certainly little or no prior experience) that closing 
the last tab in the strip will cause all of Chrome to disappear.  This means either:

a)the design of the button is deficient, or
b)the function of the button is erroneous.

Comment 86 by, Jan 14 2009


c) Sometimes, someone has to take the lead in a new design methodology.

Whenever someone steps out of the confines of the design cue "box"... they are going 
to take heat for coloring outside of the lines.  They are only confusing users... 
causing unexpected behavior... mass confusion... world wide panic.  Then, as time 
goes on, and all others copy the genius of the new design, it suddenly becomes the 
new "way"... and anyone who strays from THAT design is the next "mistake."

Although I dislike the behavior (since I am used to different behavior from other 
systems,) in principle I like the new methodology because it is consistent.  You 
might feel it is inconsistent because it does something different from all other 
systems, leaving users to be baffled by the behavior.  However, I see the existing 
functionality of current systems to be inconsistent because, logically, it makes no 
sense.  Is a close button only a close button, or a button that does one thing one 
moment (closes the tab) and another thing the next moment (replaces the current tab 
with a blank tab.)  If you're going to go that route, you might as well change the 
icon of the last tab to a new icon representing "change to blank"... but then, in 
doing so, you cause a close button to suddenly vanish and turn into something else.

No matter how you slice it, if you ignore "this is how everyone does it" mentality... 
and try to assume, for a moment, that the way everyone has done it so far may not be 
100% logical nor the way it should be done... and apply only logic (based purely on 
assuming an alien visited our planet for the first time and you are trying to explain 
in simple terms what a close button does without having to talk about exceptions or 
case scenarios)... then you would realize that a lot of the ways applications act 
these days, there are inconsistencies.  From one version to the next, the concept 
changes.  One moment, closing all tabs but one makes the close button vanish from the 
last tab, forcing a close from the window only.  The next moment, the close button 
remains on the last tab but clicking it will close that tab but immediately replace 
it with a new blank tab.  In word processors, closing the last "sub window" will 
leave a blank word processor open with essentially no document windows open.

Three different ways of doing things from two different types of applications.  
Rather than seeing anyone complain about the way word processors work, or how the 
inconsistencies in the other browsers bothered them at first, it's always about how 
the "new guy" is just trying to be too new... like complaining about labels in Gmail 
and trying to bring back "folders".  It's a moot point.

If you want Chromium to act a different way for yourself, compile your own version or 
wait for an add-on system which lets you override these settings.  In the meantime, 
be understanding of the fact that Chromium isn't just trying to be a web browser.  
It's trying to be much more... and application browser, of sorts.  The fact that you 
can drag tabs out into their own windows seamlessly adds complexity that web browsers 
don't need to be faced with.

"The tab IS the application" is inaccurate anyway - being a process doesn't mean it's
an application. It's not really even an independent process; or all those startup
tasks people are complaining about would happen whenever someone opens a tab. A tab
cannot exist independently. UI-wise and in terms of user perception it is no
different at all from tabs in any other browser (dragging outside the window - to put
it in another window, mind you, not to create a 'tab outside of a window' - does not
change that). How it's implemented 'under the hood' doesn't matter at all, this is
PURELY a UI issue.

Comment 88 by, Feb 8 2009

I didn't say the tab IS the application because it is a process.  I said it because 
this is the philosophy the Chromium developers are taking.  So, before you think I'm 
just making things up, feel free to read about the design plans for Chromium's 
future.  At some point, nearly all of Google's installed "applications" will 
essentially be Chromium tabs.  Again, Chrome/Chromium is NOT just a "web browser."  
It is an application frame-set disguised at first as a web browser.

Imagine, for instance, you decide to run Google Earth (not Google Maps, but the 3-D 
Google Earth) and it runs as a Chrome tab.  In this case, the address bar would no 
longer be there, because that is only relevant to browsing the web.  Then you run 
Google Picasa (not the web-based Picasa Web, but the installed application) and it 
also runs as a Chrome tab.  You can drag these tabs out as their own stand-alone 
apps, or can let them nest with each other in the tabbed interface.

This is the future of Chromium/Chrome.  In the meantime, it "appears" to be just a 
web browser, but at its heart it is much more.  Everyone is spending so much time 
trying to focus on the web browser aspect of Chrome/Chromium that they don't 
understand why there is certain resistance for such common features found in other 
web browsers.  The reason is because those other web browser are just that... just 
web browsers.  Chromium/Chrome was intended for so much more.
Closing last tab also closes Window is a good idea, user can use middle button of 
mouse to close Chrome, no need to move the cursor to the upper-right corner anymore.
 Issue 2788  has been merged into this issue.

Comment 91 by, Aug 4 2009

This is the most starred WontFix

Comment 92 Deleted

Comment 93 by Deleted ...@, Aug 23 2009

Why not add the feature to choose if you want a blank tab open?  As a user, it would 
make the product more user friendly to me and perhaps the others that have 
opened/commented on this issue.

Comment 94 by, Aug 24 2009

Samantha, I totally agree with your suggestion!  However, to answer your question about 
"why not", it is because the Chrome devs are entirely biased against configuration 
options and actually making Chrome more user-friendly.
 Issue 312  has been merged into this issue.
 Issue 20434  has been merged into this issue.
I see almost consensus here about going to the start page. Keeping an X on the last tab 
is consistent with tabs acting the same regardless of context; the browser shell can 
choose to reopen a start tab on close of the last tab rather than shutdown. Keeping 
options to a minimum is a winning design aspect of Chrome in the same minimalistic 
spirit as Google search. I can also understand that this is not a high priority as many 
here have agreed too.

Comment 98 by Deleted ...@, Dec 2 2009

It'd be nice if the Chrome devs weren't biased against user-configuration - people are 
different after all. Perhaps they could apply some of their considerable talent to 
creating a user-friendly configuration system. It's a total cop-out to say that 
'giving users options == not user-friendly'.

But regardless, as many others have said, closing the last tab clearly shouldn't close 
the application.

Comment 99 by, Dec 2 2009

I've gotten used to it.

#1 issue I have with Chrome.  And don't give me "it opens so fast!"  The point is it 
shouldn't have closed in the first place.  This is called overloading a UI element and 
is one of the worst offenses in recent history when it comes to browsers.  The X on 
the tab CLOSES THE TAB.  The X in the upper left?  Closes the window.  2 different 
actions, 2 different UI elements.  Don't mix them and don't patronize people who don't 
like needlessly complex UI overloading, especially one that is so annoying.

Comment 101 by Deleted ...@, Feb 11 2010

What gets to me is that this issue has been presented to Chrome since at least 2008 and 
this is 2010 and zilch!  I am hunting all over the web for a fix to this and see that 
myriads feel the same - doesn't Chrome listen??  

Well, look no further!

"Keep Last Two Tabs"
This extension keep last two tabs to prevent Chrome quits when last tab is closed.
The major problem I have with this is not the fact that the browser closes, if that is 
decided to be the way it should be, then so be it.  But what has to change in my 
opinion is the fact that I am CLOSING a tab and the tab that I CLOSED reappears the 
next time I open the browser.  My setting says "Reopen the pages that were OPEN last".  
Not reopen the last tab I closed.  I think this is what the entire discussion should 
be around and as far as I'm concerned there is no discussion.

Comment 104 by, Feb 12 2010


I think I see what you're talking about and I completely agree with you. If the 
setting is to reopen the pages that were open last, it should either:

1) Open the last tab you had open if you closed the browser in the upper-right-hand-
2) Open with the new-tab-page if you closed the browser by closing the last tab.

However, this is a different issue. I would recommend you create a new issue for that 
particular aspect of functionality. If you read the original issue that this 
discussion thread is connected to, it simply has to do with whether or not closing 
the last tab closes the browser.

Comment 105 by Deleted ...@, Feb 12 2010

I find "clicking X on the last tab closes the entire browser" not appropriate. 
Clicking X on the browser window should close the browser window, and clicking X on 
any tab (including the last one) should close the window. If that is to be done, then 
one way is to compare it to an OS: when we close the last app in the oS, it doesn't 
close the OS (or it's desktop). If our NTP is considered analogous to the desktop, 
and tabs analogous to apps, then it's probably a good idea to show some default 
screen (NTP is one obvious choice) when last tab is closed.

One argument is that opening a tab (to show NTP) when user closes last tab might be 
weird. One way to handle that could be to not show NTP in a tab-like UI. Show it more 
like Chrome's desktop background. Then it doesn't look weird that closing a tab 
closed that tab and opened another one. And then Chrome can also have a way to Show 
Desktop :)

Comment 106 by, Feb 12 2010


You make sense. However, this assumes that the NTP is analogous to the desktop and 
that Chrome is analogous to an operating system. (Sure, Chrome OS is an operating 
system... but if closing the last tab in Chrome OS shuts down the operating system, I 
will most certainly complain.)

With Chrome the browser, however, it is not an operating system. It is an 

I think the best way to view Chrome (and how it was likely intended to be seen) is 
this. Each tab is an application. The wrapper around Chrome is really just a place 
for the tabs to live. If no tabs are alive, there is no need for a place for the tabs 
to live. So, it should only exist when one or more tabs exist.

Imagine is all tabs were always detached. Every "window" was a one-tab window. 
Closing the tab closes that window. It makes sense. At that point, it only seems 
weird that there are two close buttons.

However, give the ability to combine tabs together, and now having a single close 
button to close all tabs makes sense.

Put another way. The close button in Chrome in the upper-left-hand-corner is merely a 
quick way to "close all tabs". That's it. You aren't "closing Chrome"... you're 
"closing all tabs". When no tabs are open, Chrome really doesn't exist.

Chrome isn't "a web browser with tabs"... it is an environment within an OS that 
allows tabs to live and interact with each other. There is no reason for that 
environment to exist in a tabless situation.

The fact that Chrome loads super-fast means you are moments away from opening a tab 

The only thing that makes all of this complicated is that annoying "operating system" 
layer that we still have to deal with. Once you can get rid of that (and you're left 
with just Chrome OS), then it will all make sense.

Of course, if closing the last tab in Chrome OS shuts down the operating system, then 
you will most definitely have a point. I don't care if it only takes 7 seconds to 

Y'know, it finally makes sense, given the way iambob described it. If you think of the 
X button in the upper-right corner as "Close all tabs", rather than "Close 
application", then even when there's only one tab to close, the behavior is consistent. 
The problem is that the X button is designed to look like a standard Windows close 
button, and this confuses a lot of users. If Chrome isn't an MDI interface, maybe it 
shouldn't suggest that it is to the user. I'm not sure exactly what kind of change 
would clarify this, however.

Comment 108 by, Feb 12 2010

I think the catch-22 Chrome has found itself in... is that it is attempting to not be 
specially a "Windows application" because, at its heart, it is the foundation of its 
own OS. It is OS independent.

However, if they were missing key features that other OSes share... people would be 
confused. i.e., it is already "missing" a standard title bar... but people don't seem 
to miss this. Missing a typical "close", "maximize", and "minimize" button that 
everyone comes to expect in the upper-right-hand corner would turn people away, 
thinking it is "broken" or "out of touch".

Since Chrome is trying to move people from one OS paradigm to a completely different 
OS paradigm, until the transition is complete... and as long as it is still available 
on other OSes, it will always be stuck in purgatory in certain respects. I believe 
the close/max/min is one of these gray areas. Another has been the layout of popup 
windows, which need to follow the rules in regards to JavaScript's ability to create 
these popup windows with certain typical OS features.

Virtually everything else within Chrome is nearly non-standard when it comes to 
typical OS behavior... and this is to slowly wean people off of thinking about 
typical OS behavior and to grow accustomed to the transition to Chrome OS. :)

How about a slightly different highlight for the last tab's close button? (images 
attached)  If not square-vs-round then maybe an extra ring around the circle or a 
slight flash (similar to the "Close other tabs" / "Close tabs to the right" pulsing). 
Even smart users need hints occasionally!
26.3 KB View Download
30.4 KB View Download
30.1 KB View Download
The alternate close button does not really help when using keyboard shortcuts though.

Hi guys!

I also request more good tab managment. My brain was collapsed when I close last tab 
and app also was closed %-| 
I agree with most people here.  Either;

1. Open a new tab with the start page/dashboard when the last tab is closed
2. Disable the close button when only one tab is left.

It's so annoying to have to re-open the app.  I also use Opera & Firefox on a regular 
basis (different Gmail accounts tied to each one) and they do not act this way.  I'm 
ready to switch my least used Gmail account to Chrome so I don't have to deal with 
this as often.

Comment 113 by, May 26 2010

So, this was closed WontFix, suggesting that an extension should solve this, but 
extensions are gimped in Chrome to the point where modifying this sort of behaviour 
is impossible?

And, with no shortcut to close all tabs, ctrl-w killing the entire window is painful 
and confusing.

Honestly, people objecting to this on the basis that architecturally the tab is a 
core application component, and that closing the last tab removes the need for the 
window to exist should consider: applications are for serving the user, not the other 
way around.

This issue should certainly be re-examined - either the extension API needs to be 
vastly expanded, or there needs to be an option added to enable this behaviour.  Even 
a startup switch would be sufficient if the aesthetics of an additional checkbox are 
too offensive to consider.
CTRL+SHIFT+W (and ALT+F4) are Chrome keyboard shortcuts to close all tabs.

And I found at least 4 extensions that keep Chrome from closing via just CTRL+W... with 
"NeverClose" and "KeepLastTab" working the smoothest.

Even the "LockTab" extension (or similar) could be used as yet another way to keep it 
from closing.  Just lock a tab down and then it'll confirm if you try to 
close that 1 (or all) tab(s).

Comment 115 Deleted

@spoidar: “people objecting to this on the basis that architecturally the tab is a 
core application component, and that closing the last tab removes the need for the 
window to exist”

You are correct. The purpose of Chrome was to be a framework for running web-apps 
rather than just being a web-browser. However there are two things that confound that 
excuse for this issue:

(1) Regardless of the original intention, Chromium (and its derived browsers) are 
indeed ed facto web-browsers. Chrome(ium)OS is now the de jure framework for running 

(2) I cannot think of any OSes that shutdown once there are no apps running. In fact, 
it would make no sense since there would be no way to run an app in the first place if 
you cannot start the OS without an app. Windows, OSX, Linux, DOS, etc. all remain 
running once all apps are exited. Chromium should be the same. It is wrong to force the 
user to keep one web-app or web-page open in order to be able to start another one.
Regarding Chrome OS - remember that its boot time (two seconds) renders this argument 
sort of useless.
It is energy efficient, also. ;)
@phistuck, except that this is not ChromeOS, this is Chromium. Running 
Chromium/Chrome on Windows, etc. takes more than two seconds for most people. It has 
to re-read the HOSTS file, the whole history file, and other startup routines that 
make it useless for more than two seconds. You may have a super-fast computer with 
nothing installed on a fresh copy of Windows where it starts quickly, but not 
everybody else does.

Plus even shutting down has a bunch of cleanup routines that take time. I have seen 
people having problems due to corruption of their profile/preferences that was caused 
by Chrome.exe continuing to run in order to cleanup stuff after they exited Chrome 
only to end up running it again right away. The new instance tried to load while the 
old instance was still writing to the files, thus causing corruption and/or crashes. Regarding the corruption - nice to know, file a new issue with reproduction steps so it will not happen anymore. This should 
definitely be fixed and it will be fixed. You just have to tell them about it. Showing them exactly how to reproduce will make the fix come even 
sooner. (but please search the database at for existing issues regarding this and star them instead of 
creating a new one, if you find one)

I do not have a super fast computer, neither do I have a fresh Windows installation with nothing installed. I have a few computers, one of which 
is Pentium III, less than 400 MBs of free disk space and 384 MB of RAM and a 64 MB graphic adapter and Windows XP Professional full of useless 
applications (can we agree that it is old and slow?) and it takes Chrome about a second or two to load.
So perhaps you are having issues with Chrome (or with your computer in general), which should definitely be reported or determined, because it 
does start quickly, even on slow and old computers.

And as I wrote, "Regarding Chrome OS", not regarding Chrome, since you mentioned Chrome OS and said that it should be like Windows\Linux, but 
these are two different types of operating systems - and this one loads 'too quickly' to let this get in its way.
@phistuck, the race condition I described is not a new issue; I’m sure I’ve seen it 
listed somewhere (including in the Chrome forums as a possible cause of the loss of 

How is your usage of Chromium on said system? Of course Chromium starts up quickly with 
a fresh install and/or lightly used Chromium. However once you use it regularly enough 
and your history files start getting to be >100MB, and Chromium wants to update its 
safe-browsing filters, and it reads in a large HOSTS file, and loads all the custom 
search engines that you configure in the Options menu, and it loads a couple of small 
extensions, then it takes a while to load.

(I should clarify my definition of load: the window pops up quickly, but it is 
relatively useless since Chromium thrashes the disk for ~10 seconds while it loads the 
primary history file; and using custom-engine shortcuts don’t work until it gets around 
to reading them (apparently *after* it reads in the history file). So by load, I mean 
that it is actually usable where you can type addresses in using custom-engine 
shortcuts, or the first few characters of a URL that you frequent and have Chromium fill 
the rest in based on your history.)

As for ChromeOS, I didn’t say that it should be exactly like Windows, just that when you 
close the last web-app, the computer/OS should not automatically shut down since you may 
want to open another one. You should not have to keep one open before you can open 
another one, especially if you don’t want to open another app until a few seconds after 
you closed the last app (have you never remembered something you needed to do two 
seconds after clicking shutdown?)

synetech, can you try to determine what is causing your slow shutdown/startup? You 
could perhaps try starting chrome with some command line flags to get a lighter load. 
Perhaps you have a misbehaving extension?
Chrome does try to use a lot of memory; could it be possible that you have declared 
to have lots of swap memory but very little physical memory, and that this confuses 
chrome to try to use the swap virtual memory as if it was real memory? 

(I've used some versions of chromeOS where closing the last tab made the chrome 
window to disappear and come back again. It was almost instantaneous on an Asus Eee 
with Atom N270, but also there wasn't any profile data and probably ~1GB memory free)

Comment 122 Deleted

I know exactly why Chromium starts up so slowly: file access. Like I said, every time 
it runs, it reads in a bunch of files that grow over time. Currently (less than a 
week into the current month) I have:

  Archived History: 49.2MB
  History: 34.4MB
  History Index 2010-06: 10.2MB
  Thumbnails: 50.6 (I stopped deleting this file less than a month ago)
  Visited Links: 0.99MB
  Web Data: 2.00MB
  Safe Browsing Bloom: 11.7MB (The Safe Browsing files are less than a month old)
  Safe Browsing Bloom Filter 2: 1.16MB
  Local Storage: 7.33MB
  HOSTS: 6.47MB

That’s 174MB, and that’s with all the other monthly history files archived away, 
otherwise it would add another 1.02GB (thank goodness it presumably only ever reads 
those files when you do a history search).

One way to speed things up a little is to completely remove all unused extensions 
(it’s not enough to simply disable them since Chromium reads them lal in anyway, so 
either delete them or move them to another folder).

However when Chromium is used regularly for a while, regardless of extensions, it 
accumulates quite a lot, quite quickly. I’ve been using Chromium since September 2008 
(the month it was released to the public), so it’s gotten much more loaded than a 
fresh install or empty profile that would be present in a test system. All those 
files and all that data (again >150MB) have to be read each time Chromium runs (and 
worse, it seems that some of them are read more than once on startup!). That bogs it 

This thread is at risk for being jacked, so it’s better to just search for other 
reports of Chromium taking more than “two seconds” to start up because I know that 
the latest versions do (yes the original Chrome was amazingly fast, but that was two 
years ago).

Startup time aside, the argument remains that it makes no sense to be required to 
keep an app open in order to open a new one. The framework should not close until the 
user decides that they have no more use for it; the framework is taking too many 
liberties of thinking for the user.

I don’t see why there cannot be an option to have the engine itself remain active in 
the tray, and have the apps/pages open a new rendering process, and then close that 
rendering process when they are done. I tried to simulate this with a minimize-to-
tray and a keep-a-tab-open-to-prevent-Chromium-from-quitting-on-last-tab-close 
extension, but that did not work.

Absolutely support the suggestion!!
Please implement it  as soon as possible.
Open a start page or a new tab instead of quiting application when the last tab was closed.

Comment 125 by Deleted ...@, Aug 22 2010

Just thought I'd add this here.  For those of you who want persistent last tab functionality, I wrote this extension:

I think it works better than what's currently available; let me know what you think :)
I've installed luckygalleon's extension, and it works pretty well. Not as good as in Firefox, because of Chrome's script limitation of acting on built in pages, but it does well on general daily navigating.
> overriding the meaning of standard controls is undesirable from an interaction 
point of view.


- People are closing the last TAB, NOT the entire browser.
- Your controls are custom anyway.
- No, extensions don't solve this issue, as they introduce puzzling behaviour which is quite inconsistent with what user comes to expect from other browsers
- for some odd reson the browser doesn't close in MacOS when you cose the last tab. You didn't think it was overriding behaviour there for some odd reason.
> for some odd reson the browser doesn't close in MacOS when you cose the last tab. You didn't think it was overriding behaviour there for some odd reason.

Chromium on Mac closes the window, but doesn't quit the application. Standard behaviour on Mac.

Comment 129 by Deleted ...@, Dec 17 2011

For years now I have rejected Chrome every 2 or 3 months after trying it out again to see if it the whole browser still closes down when I close the last tab.  I have read innumerable pleas and complaints about this and Chrome is apparently deaf to what many want.  How much trouble would it be to have an option that all other browsers have by default?  Until Chrome solves this issue, gives us the option, it will not be on my system for long.  Also, I want the option to open up new tabs in my same Google home page.

Comment 130 by Deleted ...@, Dec 17 2011

Also, how do I save a link to a webpage to my desktop? 
@braintree - 
@129 - In case you are on Windows, the Chrome Toolbox extension (by Google) implements this feature (and a lot more) -

@130 - This is not a forum. Do not ask this type of questions here.

Comment 132 Deleted

@132, please add some explanations/experience on what problems you get when Chrome closes?
Of those who still want this fixed (91 people starring this, you still there?) how many have problems with the keyboard command (Ctrl W) and how many have problems from clicking and closing the last tab? For mouse-users, it is more common that the big red X in the outer edge of the browser window closes the browser software, while the smaller (x) to close a tab would still leave the software running after closing the last "document" inside.

On the other hand, maybe Chrome devs are very positive towards making people close the browser now and then, to get a chance to clean up any memory leaks, do software upgrade, get more data on how long time browser startup takes, etc. Making the browser close less often can have other effects that make it unwanted...

To solve your personal problem, you can simply install a browser extension. I'm a happy user of Window Close Protector  since over a year I guess. It just pops up a warning when I'm about to close a window completely. I needed it to prevent the list of "recently closed tabs" from getting lost, but after the recent redesign of the handling for recently closed I could probably uninstall that extension. I'll do just that now, and will report back if I come into any situation when I'd need the protection from window closing. Others, please write about what actual problems you see. This issue won't go anywhere unless you give really good explanations for why it should have to. And see comment @48 for an idea of how good explanations you'll need :-).
Yes it is Ctrl W that is the problem for me and yes I'd still love for
this to be fixed.

That's an odd one - it says it will warn you if you close a tab that
is not visible - but in this scenario surely we are closing the final
tab that *is* visible, if only for a millisecond.  Anyway, I am
currently using:
which forces the last two tabs to be kept open.  This seems to be the
way that all the extensions get around this problem - rather clumsy -
but it saves my sanity.
@133, the problem is very clear, I think. I believe that whether using Ctrl-W (my choice) or the smaller (x) means that we want to close the CURRENT TAB we are working on. Sometimes this reflex is so fast that we don't stop to think whether it's the last tab or not. And then the window closes. And then I have to open it again.

@48 "overriding the meaning of standard controls is undesirable from an interaction 
point of view". Well, Windows UI convention: close the window on the big red X button on the upper right! Some browser genius invented that closing the last tab should close the whole window because, you know, there's an x, it should close the tab, it's the last tab, we can't live without a tab, let's close it all. REALLY? What other software closes the whole window from an internal UI element that's not a big "You are going to close the whole window" sign? Can you imagine this in Photoshop for instance? You close your last document and BAM, you have to load it all over again...

"Chrome devs are very positive towards MAKING people close the browser now and then" WHAT? Shouldn't this be a user's choice? Couldn't you ASK them IF they want to close it for cleaning, upgrading, or even giving you "more data on how long time browser startup takes"?

My personal problem is solved of course since I'm a heavy user and I looked for an extension, but my GUESS is there are a lot of people that doesn't even know what is an extension is that suffer form this annoyance. A popup would solve it if you don't want to make it an option. If I really want to close it all, I'd still have the big old red X button just like every other software...

Thank you much for paying attention to this matter.
I recently downloaded the new Chrome Beta for Android ICS and noticed that when the last tab on it is closed, it leaves the Chrome Window open with a blank workspace.  Another observation from Android is that the default Android browser reopens the home page if the last "window" is closed.  I think this is the type of behavior that most people in this thread are after from Chrome. 
And on ChromeOS, when the last tab is closed it is replaced by a New Tab with no close button.  Why not log out then?  Wouldn't that be more consistent?

I still say if this is how it's going to be then you need to give the user some hint in the UI that something different is about to happen.  See #c109 above for an unobtrusive example.

Comment 138 Deleted

> please add some explanations/experience on what problems you get when Chrome closes?… problems with the keyboard command (Ctrl W)… the big red X in the outer edge

The problem is not as often due to closing the *last* tab, but rather *all* tabs. When there are a bunch of tabs, navigating between them and closing some becomes a common task. Having to press Ctrl+Shift+Tab to navigate backwards is an expected event in this case. When this is happening, it is not uncommon for the user to release the the Tab key and then press W to close the previous tab. But, having forgotten to also release the SHIFT key, they end up closing ALL tabs instead of just that one.

Also, if they are using a wireless or dirty keyboard, they keys/signals may get stuck and end up repeating. With no prompt, the browser would happily close. Imagine if your word processor did not bother to prompt you to save your file!

And yes, pressing the [x] in the corner is another cause of this. Most people maximize windows. As such, it is not uncommon to click the [x] of one window and accidentally click again (eg a wireless mouse repeated the signal, the window did not close fast enough, so the user clicked again, etc.) Once the other window has been closed, the second click goes to the [x] in Chrome, thus closing the whole browser accidentally.

> On the other hand, maybe Chrome devs are very positive towards making people close the browser now and then, to get a chance to clean up any memory leaks, do software upgrade, get more data on how long time browser startup takes, etc. Making the browser close less often can have other effects that make it unwanted...

No, no. No.

As a software developer I can tell you (and users) that his is absolutely wrong. No.

Imagine if your car manufacturer purposely left problems in your car. Imagine if they chose leave bad brakes in order to “increase the speed of the car by preventing the breaks from slowing it down as often” or “gather statistics on the top speed people usually drive (even if they specifically OPTED OUT of that option!), and so on. No.

What if your phone or television manufacturer left in a bug that makes it easy to accidentally change channels or drop calls so that “you don’t spend too much time on just one show/call for your own/the tv/phone’s good.” No.

What if your doctor chose to leave a tumor in because he felt it “makes you live life to the fullest”. No.

If a word processor or other program did not include a prompt to save your file or crashed every now and then, you would be livid! What if their excuse for CHOOSING NOT TO FIX IT was “it’s good for it to close now and then to address a bunch of other bugs that we didn’t bother fix that cause memory leaks, slow-downs, etc.”? NO!

To be honest, that part of the excuse was one of the dumbest things I’ve read in all of programming-dom.


> To solve your personal problem, you can simply install a browser extension.

Again, no. Why should users have to install yet another extension with yet another process using up yet more memory just for one, single function that should have been built-in from the start, the way it is with EVERY OTHER BROWSER?

Comment 140 Deleted

Comment 141 by Deleted ...@, Mar 11 2012

I do confirm that closing the application with the last tab is stupid.

IF there were separate windows for 
1. the browsing activity with the web page and the URL bar, in one or multiple tabbed window
2. the main application with the menus, settings, bookmarks, search bar, help, whatever.
THEN it would make sense to close the browsing activity window when closing the last tab and return the focus to the main application with the search bar and the bookmarks.

But with the current design packing the application and the browsing activity on the same window, there is more in the window you are closing than just the web application that you want to leave.  There must be a way to leave a web site and not the browser.

Today, more and more of a user's activity happens on the Internet, and the idea that the user wants to stop using the web and not just one web application is increasingly wrong.

As it happnes, I took the habit of opening a new tab before closing the previous one in order work around that unfortunate design choice.

If you close the last document in a word processor, does the word processor exit?

If you close the last application running on your computer, does the computer shut down?

So please let me close the last tab and stay in the browser.
After running into the same problem (and unable to find a satisfying solution), I wrote a Chrome extension for this. Basically, it creates a new tab automatically when the last tab is closed.

Feel free to try it out:
Project Member

Comment 143 by, Oct 13 2012

Labels: Restrict-AddIssueComment-Commit
Owner: ----
This issue has been closed for some time. No one will pay attention to new comments.
If you are seeing this bug or have new data, please click New Issue to start a new bug.
Showing comments 44 - 143 of 143 Older

Sign in to add a comment