JavaScript disabled - how likely?

[Mason A. Clark]

If I use javascript on my page, how likely is it that the
viewer will not have javascript? Anyone have data?

Mason C

 

[Thomas 'PointedEars' Lahn]

If I use javascript on my page, how likely is it that the viewer will
not have javascript?

It can happen and it will happen, so you should be prepared for it and
provide alternatives.

Anyone have data?

Anyone who claims to have reliable data (so-called "statistics") on this
subject is telling you plain lies. The Web's user structure is far too
volatile to make even reliable projections. What you can know for sure:
There are more than 0% and less than 100% of users who have support for
client-side scripting disabled, restricted to a certain extent, or use
UAs where that feature is not even present. YMMV.


PointedEars

 

[Thomas 'PointedEars' Lahn]

I don't know how accurate this is:

http://www.thecounter.com/stats/2004/February/javas.php

JavaScript Stats

Sun Feb 1 00:05:02 2004 - Wed Feb 25 20:55:03 2004 24.9 Days

Javascript 1.2+: 262730395 (94%)
Javascript <1.2: 439369 (0%)
Javascript false: 14202948 (5%)

There is no reliable way to determine the used JavaScript version,
and simple *access* data which claims to be statistics (but lacks
the most important features of such a work) for *one* site makes
exactly *none*, not even the slightest approximate argument for or
against JavaScript support, especially not for the whole World Wide
Web and all of its users. The above wannabe-stats are utter nonsense
and people who think that they can even draw conclusions from such
data, and plan implementation strategies that match reality to the
slightest, are fools.


PointedEars

 

[Thomas 'PointedEars' Lahn]

I don't know how accurate this is:

http://www.thecounter.com/stats/2004/February/javas.php

JavaScript Stats

Sun Feb 1 00:05:02 2004 - Wed Feb 25 20:55:03 2004 24.9 Days

Javascript 1.2+: 262730395 (94%)
Javascript <1.2: 439369 (0%)
Javascript false: 14202948 (5%)

There is no reliable way to determine the used JavaScript version,
and simple *access* data which claims to be statistics (but lacks
the most important features of such a work) makes exactly *none*,
not even the slightest approximate argument for or against JavaScript
support, especially not for the whole World Wide Web and all of its
users. The above wannabe-stats are utter nonsense and people who
think that they can even draw conclusions from such data, and plan
implementation strategies that match reality to the slightest, are fools.


PointedEars

 

[Ivo]

I don't know how accurate this is:

http://www.thecounter.com/stats/2004/February/javas.php

JavaScript Stats

Sun Feb 1 00:05:02 2004 - Wed Feb 25 20:55:03 2004 24.9 Days

Javascript 1.2+: 262730395 (94%)
Javascript <1.2: 439369 (0%)
Javascript false: 14202948 (5%)

This number of 94 bears a close resemblance to my server log files.
Ivo

 

[Kevin Scholl]

If I use javascript on my page, how likely is it that the
viewer will not have javascript? Anyone have data?

There really is no such thing as reliable statistics of any kind for the
Web, at least not on any far-reaching scale. The best you can do is
monitor the client's logs, and adjust accordingly on a project by
project basis.

That said, my views on the use of Javascript are thus:

Some people are vehemently against using Javascript of any kind,
particularly for navigation. However, with the proliferation of DHTML
and similar enhancements which utilize JS, I'm of the firm belief that
anyone who turns it off (it is on by default in all remotely significant
browsers) misses out on a lot the Web has to offer. The purported
security danger of Javascript is in and of itself largely nonsense, with
no technological basis. The only remotely logical reason to turn JS off
is to avoid pop-ups, but all major browsers now have the ability to do
so without compromising Javascript (such controls are built-in to the
user preferences of most the browsers, and available as small, simple
add-ons for others).

The important issue is accessibility; that is, if the user doesn't have
Javascript on, can they still get around the site (albeit perhaps not as
conveniently). I've found that simple text links, or even just a link to
an all-inclusive site map, in the footer of each page suffices. Another
means I use is an index page which notifies a user if they do not have
scripting turned on, and explains to them the benefits and (false)
dangers. They then have the option of accessing the site anyway, with or
without JS. If a user has JS on, they are automaitcally redirected into
the site, and never see the interim page.

That all said, I've never -- knock on wood -- gotten a complaint from a
user on any of my navigation schemes or the alternatives I've provided,
or on my use of Javascript in general. Opinions and mileage may vary,
though, as this is base don my personal experience and observations.

--

*** Remove the DELETE from my address to reply ***

======================================================
Kevin Scholl http://www.ksscholl.com/
(e-mail address removed)

 

[Stuart Palmer]

Rule of thumb, always try and provide a non JS dependant user functional
site. Yes, have JS in place, but try to offer the non JS option like putting
<a href="blah.html" target="_blank"
onMouseOver="window.open('blah.html');">Link</a> this will work for both JS
and non JS users.

Stu

 

[Richard Cornford]

If I use javascript on my page, how likely is it that the
viewer will not have javascript? Anyone have data?

If you have only one viewer then it would be quickest to ask him/her. If
you have a reasonable number of viewers then the chances of the site
being visited by a javascript incapable/disabled browser are 100%.

If you are looking for someone to pat you an the back and say that it is
OK to produce a javascript dependent web site because only a tiny
percentage of people don't use javascript and they are not really your
problem, then you will find such people. They don't have any more reason
to believe that that is true than you have for believing them when they
tell you it.

You will find "web statistics" everywhere. What you will not find is any
information about how those statistics are gathered, form whom and how
they were analysed. Certainly not enough information to make any
assessment of the validity of those statistics.

It is known that there are significant restrictions on what is
achievable when attempting to gather statistics about web usage. One of
the biggest restrictions being web cache systems that mean any HTTP
request may not even get to the server it is directed towards if an
intervening cache has a non-expired copy of the requested resource
available, so where is the log entry for that request? But there are
plenty of other significant restrictions in what information can be
gathered.

Web statistics produce a viscous circle. Only the people who believe
that the statistics are meaningful/useful contribute to those
statistics, the people who don't believe that meaningful general web
statistics can be gathered don't bother. But the people who believe the
statistics are meaningful allow those statistics to influence their
behaviour and create web sites that pander to whatever majority those
statistics indicate, creating IE specific and JS dependent sites.
Visitors to those sites using other browsers and JS disabled/incapable
browsers do not hang around on those sites clocking up hits because it
instantly becomes clear that they are wasting their time, and they do
not make repeat visits. The resulting logs reflect this in reporting a
very low percentage of visitors with non-IE and JS disabled browsers and
those logs are used to contribute to the gathering of the reported
statistics. Showing the resulting bias towards JS capable recent IE
versions.

So do the reported web statistics do any more than reflect the
consequences of the belief in the validity of those statistics? That
would make them nothing more than a chimera.

There are people who believe the statistics because they "sound right",
and we probably can be confident that the majority of the world's
desktop computers are indeed running a MS Windows OS and do have IE
installed and that IE is the web browser being used to access the
Internet. But statistics are not meaningful just because they "sound
right". Just because a majority of IE is expected does not mean that 95%
is the actual number.

I am often reminded of Hans Christian Anderson's "The Emperor's New
Clothes" when people start talking web statistics; stop believing and
what remains? But so many web development decisions seem naked without
them.

Two javascript statistics that I have noticed over the past months
probably best describe the situation: The first from the logs of someone
who freely admitted that because of browser usage and JS statistics they
had created a site that was both IE specific and JS dependent. They were
apparently getting 2% of visitors without JS (and 4% without IE).

The organiser of the second site, dedicated to HTML authoring and
directly promoting universal accessibility, but without a doubt a
specialist interest site, reported 80% JS incapable/disabled visitors.

(Neither went into any details on how those figures had been derived.)

The bottom line truth is that you just cannot tell how many visitors
will (or would have, all else being equal) visited a site with JS
disabled or incapable browsers. You can be certain that it will be more
than none and probably less than all. But is doesn't matter as there is
a considerable amount that can be done with javascript in a way that
enhances a web site without imposing any dependency upon javascript. It
is just a matter of designing the HTML, CSS and javascript from the
outset with an appreciation of the need for clean degradation in the
optional technologies. Its not an easy design task, and many will seek
any excuse to avoid it, but once the challenge has been risen to the
results can be very rewarding.

Richard.

 

[Mick White]

Richard Cornford wrote:

....

The bottom line truth is that you just cannot tell how many visitors
will (or would have, all else being equal) visited a site with JS
disabled or incapable browsers. You can be certain that it will be more
than none and probably less than all. But is doesn't matter as there is
a considerable amount that can be done with javascript in a way that
enhances a web site without imposing any dependency upon javascript. It
is just a matter of designing the HTML, CSS and javascript from the
outset with an appreciation of the need for clean degradation in the
optional technologies. Its not an easy design task, and many will seek
any excuse to avoid it, but once the challenge has been risen to the
results can be very rewarding.

Richard.

Of course it is possible to determine what percentage of browsers have
javascript disabled. It's just not practical.

One way is to sample [large number here] computers, and inspect every
one of them.

Mick

 

[Richard Cornford]

Of course it is possible to determine what percentage of
browsers have javascript disabled. It's just not practical.

When practicality makes a task impossible it becomes an impossible task.

One way is to sample [large number here] computers, and
inspect every one of them.

Some points (off the top of my head):-

1. Given a global Internet with hundreds of millions of users that
number should be better described as "very large" if the results
are going to be statistically valid. There would also have to be
a global distribution of sampling and that sampling would have to
be done within a limited time frame.

2. Not all web browsers are running on desktop computers and the ones
running on mobile and portable devices might not be that easy to
track down.

3. What assumptions would be made about computers with multiple
browsers installed. Mine, for example, has 24 browsers on the
partition that it is currently booted from and another 30 odd on
the other two bootable partitions. Some have never even seen the
Internet, others get used for browsing all the time. Not
necessarily a common case but with IE so heavily integrated into
the operating system it is unlikely not to be present on a Windows
box, so would the presence of another browser be an indicator of a
user's preference for that browser? And then, which user, in a
family or business context, and what proportion of the internet
access from that box is accounted for by each and any user of any
preferred browsers?

4. Does a browser that is JS enabled at the time of sampling always
get operated with JS enabled, when it can be as easy as hitting a
key on the keyboard to toggle JS support on and off?

It may be theoretically possible (if impractical) to sample a
sufficiently large quantity of Internet accessing hardware across the
globe within a suitable period but the results of such a survey would
not tell you how the internet was used. It would be answering the wrong
question.

Richard.

 

[Randy Webb]

Richard Cornford wrote:

....

The bottom line truth is that you just cannot tell how many visitors
will (or would have, all else being equal) visited a site with JS
disabled or incapable browsers. You can be certain that it will be more
than none and probably less than all. But is doesn't matter as there is
a considerable amount that can be done with javascript in a way that
enhances a web site without imposing any dependency upon javascript. It
is just a matter of designing the HTML, CSS and javascript from the
outset with an appreciation of the need for clean degradation in the
optional technologies. Its not an easy design task, and many will seek
any excuse to avoid it, but once the challenge has been risen to the
results can be very rewarding.

Richard.

Of course it is possible to determine what percentage of browsers have
javascript disabled. It's just not practical.

One way is to sample [large number here] computers, and inspect every
one of them.

I have access to "inspect" over 300 computers, none of which are running
IE as the default browser, and none of which are setup with javascript
enabled by default. Does that mean I can assume that 0% of the web uses
IE or Javascript enabled browsers?

 

[Mick White]

Richard Cornford wrote:

It may be theoretically possible (if impractical) to sample a
sufficiently large quantity of Internet accessing hardware across the
globe within a suitable period but the results of such a survey would
not tell you how the internet was used. It would be answering the wrong
question.

Richard.

Agreed, my comment was somewhat tongue-in-cheek. But my shackles rise at
the utterance of the word "impossible".
Capisce?
Mick

 

[Mick White]

Randy Webb wrote:

....

I have access to "inspect" over 300 computers, none of which are running
IE as the default browser, and none of which are setup with javascript
enabled by default. Does that mean I can assume that 0% of the web uses
IE or Javascript enabled browsers?

You could assume that, but you'd be wrong.
Mick

 

[Richard Cornford]

... . But my shackles rise
at the utterance of the word "impossible".
Capisce?

A quick text search of my post did not find the word "impossible", or
the word "possible" (which may, if present, have been preceded by
"not").

Richard.

 

[kaeli]

A quick text search of my post did not find the word "impossible", or
the word "possible" (which may, if present, have been preceded by
"not").

Richard.

Of course it is possible to determine what percentage of
browsers have javascript disabled. It's just not practical.

When practicality makes a task impossible it becomes an impossible task.
--------

I myself happen to agree with that sentiment. It's just so large a task
that it is, for all practical purposes, impossible.

--

 

[Mick White]

When practicality makes a task impossible it becomes an impossible task.

Well, it sounds good, but it's circuitous reasoning. But the tenor of
the argument is that we can never know the true percentage. To this I
offer a tried and true cliché: Never say never.
"What is impossible now may be possible tomorrow"
Mick 2004

 

[Richard Cornford]

(e-mail address removed) enlightened us with...

When practicality makes a task impossible it becomes an impossible
task.

But the message that inspired Mick White to propose his computer
sampling strategy was:-

< >

- which apparently did not contain the word "impossible" and so should
not have raised shackles (hackles? - <self>assumes typo).

OTOH it is, for all practical purposes, an impossible task.

Richard.

 

[Randy Webb]

Randy Webb wrote:

....



You could assume that, but you'd be wrong.

Of course I would be. And thats the fallacy in any attempt to determine
what percentage of browsers are in fact IE and how many of them have
javascript enabled or present.

 

[Thomas 'PointedEars' Lahn]

There really is no such thing as reliable statistics of any kind for the
Web, at least not on any far-reaching scale. The best you can do is
monitor the client's logs, and adjust accordingly on a project by
project basis.

And what would that accomplish? If you create JavaScript-only sites,
you will only log users with enabled JavaScript support. Think twice.


PointedEars

 

[Thomas 'PointedEars' Lahn]

Rule of thumb, always try and provide a non JS dependant user functional
site. Yes, have JS in place, but try to offer the non JS option

Full ACK

like putting
<a href="blah.html" target="_blank"
onMouseOver="window.open('blah.html');">Link</a> this will work for both JS
and non JS users.

It will, but it is suboptimal anyway. Rules of thumb:

1. Never force new windows upon users if not required.
2. Always cancel events if they are triggered by common behavior.
3. Never use meaningless captions for links, like "here".

So use

<a href="blah.html">Meaningful caption</a>

or

<a
href="blah.html"
target="_blank"
onclick="window.open(
this.href, this.target, '...,resizable,scrollbars,...')"

Meaningful caption</a>

[Appended fullquote]

Please do not do that, it is nonsense in the
context of an easily followable discussion.


PointedEars