JavaScript or not JavaScript

[John Timney \(MVP\)]

Damien,

I think thats a fascinating figure and a huge number to consider, would you
care to speculate as to why it could be your target market?

 

[Mark Rae]

But having looked at our revised figures, we're now running
at 27% of sessions not having script available.

As John said, 27% is an enormous proportion, so much so that if I were you I
would want to reassure myself of its accuracy.

How are you collecting this information? Is it something you're written
yourself, or is it something like WebTrends...

 

[Damien]

Damien,

I think thats a fascinating figure and a huge number to consider, would you
care to speculate as to why it could be your target market?

Not really sure why it'd be our market especially, just the bare fact
that our experience is saying 20%, whereas global stats are showing 6%.

One speculation might be that our clients will tend to access our site
from their workplaces (not paying for internet at home), and that it's
more likely to be something locked down in a corporate environment.

Or maybe we just don't have a large enough sample size yet (we have
1500 clients having had, between them, 2300 sessions on the site. The
20% is based on the number of sessions in which JS is disabled).

Damien

 

[Mark Rae]

One thing this thread has not touched on is the need to cater for disabled
users across a web site, and its an area where use of Javascript will
always score poorly and is a pressing design issue for web designers
worldwide with new legislation having the potential to affect design
choices.

Hmm - you're right - that's something I hadn't even remotely considered...

 

[Laurent Bugnion]

Hi,

That's my aim also when I design a web application (note: I develop web
applications, that's different than a website). However, this is not
always possible. Simple example: Our last web application (building
automation management station running in a web browser) is licensed. The
licensing scheme is implemented using a lease mechanism: The licenses
are leased for 1 minute, and the lease is renewed every 30 seconds as
long as the web application is active. However, to avoid that the user
has to send postbacks to the server every 30 seconds (!), we use web
services to "trigger" the server and renew the lease. This feature is of
course non-operant if JS is off, and in that case we do not allow the
user to use our web application.

Fr a website, I try to follow your "design ethos" as much as I can.

True, our site is still under 1.1, so we didn't get to use any of these
features. But having looked at our revised figures, we're now running
at 27% of sessions not having script available. I wonder if it's some
quirk of our target market, but I wouldn't describe 1/4 of users as a
few. (And no, we're not in a "security concious" market - our clients
are people with debt problems)

Damien

Maybe they can't afford a JavaScript enabled browser?

Laurent

 

[Damien]

As John said, 27% is an enormous proportion, so much so that if I were you I
would want to reassure myself of its accuracy.

How are you collecting this information? Is it something you're written
yourself, or is it something like WebTrends...

Because our site deploys different behaviours based on whether we have
javascript available or not, we have to collect the information
directly in our site (and record it in our database).

We send the user through a page which has a javascript block that
redirects using window.location and a meta refresh set for five seconds
(and if that path is taken, we decide that script isn't available).
That's all that this page does. I suppose we might get a false negative
under the following circumstances:
1) Part of the page loads, enough so that the meta refresh is seen, but
not enough that the script block is encountered
2) The remainder of the page does not arrive within five seconds
3) The browser respects the refresh tag before the page has loaded

Does anyone know if (3) is true or not?

Damien

 

[Damien]

I'm sure it must be. The reason I use theCounter.com's stats is because
their counters are one of if not the most popular counters out there, and
their stats are based on hundreds of millions of hits - 118,800,730 for the
month of August 2006 which, by any test of reasonableness is a
representative sample...

Interesting site. Can anyone explain the following two observations (or
am I reading too much into the statistics):
1) At the end of last year, no javascript was running at ~10%. In
January, it shot down to ~3%. What happened?
2) From January onwards, it's been slowly trending upwards from 3% to
the current 6%

Damien

 

[Mark Rae]

We send the user through a page which has a javascript block that
redirects using window.location and a meta refresh set for five seconds
(and if that path is taken, we decide that script isn't available).

Hmm - that certainly does sound pretty reliable...

That's all that this page does. I suppose we might get a false negative
under the following circumstances:
1) Part of the page loads, enough so that the meta refresh is seen, but
not enough that the script block is encountered
2) The remainder of the page does not arrive within five seconds
3) The browser respects the refresh tag before the page has loaded

Does anyone know if (3) is true or not?

I think it almost certainly does.

Have you not considered the <noscript></noscript> tag instead...?

 

[Damien]

Hmm - that certainly does sound pretty reliable...


I think it almost certainly does.

Have you not considered the <noscript></noscript> tag instead...?

Well, we're planning to swap the order in the <head> such that the
script appears above the refresh, but I still think it unlikely (with
such a small page) that it'd break in exactly the right place so
reliably to give lots of false negatives. Maybe (as I put in another
post), we just haven't had enough activity yet.

Damien

 

[Mark Rae]

Well, we're planning to swap the order in the <head> such that the
script appears above the refresh,

I'm not convinced that would make much difference...

 

[Scott M.]

But how many of the 20% of the sessions are distinct and unique visitors and
how many of the 20% are the same visitors returning?

 

[Coleman]

I would go to management and explain it is hard to implement "AJAX"
without the "J". The ability to disable script is primarily for
security purposes. If your site is a valid commercial site, there
would be no reason for the user not to trust it. However, depending on
your business... that may be another story.

Check if it is enabled and redirect to a page explaininig how to take
care of it. If you support multiple browsers, do it for each based on
the type of browser they are accessing from. Again, if you are on the
up and up, explain you're not collecting information, installing bad
stuff etc., (with a disclaimer that you can't be held responsible for
things they download, and all the other legal blah, blah...)

Regards
Coleman

 

[Damien]

I would go to management and explain it is hard to implement "AJAX"
without the "J". The ability to disable script is primarily for
security purposes. If your site is a valid commercial site, there
would be no reason for the user not to trust it. However, depending on
your business... that may be another story.

Check if it is enabled and redirect to a page explaininig how to take
care of it. If you support multiple browsers, do it for each based on
the type of browser they are accessing from. Again, if you are on the
up and up, explain you're not collecting information, installing bad
stuff etc., (with a disclaimer that you can't be held responsible for
things they download, and all the other legal blah, blah...)

Regards
Coleman

We're a charity, and we seek to help people with debt problems, so
fairly legit. However, we've always been keen to do no harm, so we get
by without javascript (but enhance the experience with it), and we're
not even allowed to set cookies :-(

Damien

 

[Paul Sture]

Excuse me for jumping into the thread at this late stage, but as the
"guilty party" who triggered Mark's question, I thought I'd put my point
of view.


I personally think that allowing non-JS users into a site, albeit
without the "smooth experience" is the way to go (and if I choose to
turn JS off or use an ancient browser, it's my decision, and I am quite
happy to accept the consequences of that decision, in terms of
potentially ugly screens, trickier navigation etc).

How did our conversation start?

Mark posed a question over on a Mac newsgroup asking for comments about
a particular bit of behaviour in Safari, and here was my initial
response:

---- start quote ----
Ah. I have JavaScript turned off in Safari, and see nothing but a blank
page.

Please consider making the website friendly for folks like myself who
prefer to run with JavaScript turned off.
---- end quote ----

(I should have said *friendlier* there, as I was not asking for the full
shooting match, just something better than a blank page.)

My intention was to persuade Mark to display the logo and a bit of text
about the site for non-JS users, explaining why they couldn't get any
further.

Instead he came back with a much better solution. Not only could I see
the logo, but clear instructions on how to enable JS in a variety of
browsers. Better yet, when I persevered, I could get into enough of the
site to see what it was about (approx. 80% of it), and make a decision
as to whether I'd come in with full JS enabled.

True, our site is still under 1.1, so we didn't get to use any of these
features. But having looked at our revised figures, we're now running
at 27% of sessions not having script available. I wonder if it's some
quirk of our target market, but I wouldn't describe 1/4 of users as a
few. (And no, we're not in a "security concious" market - our clients
are people with debt problems)

Interesting. In that environnment I'd tend to put it down to running
oldish kit, whether it be stuff at home or cheap/free public access
points. Do you have any stats on how many are using dial up access? Cost
savings there may be a motivating factor. I know that when I was on dial
up access I watched my phone bill like a hawk.

Looking on the grim side, there could be a lot of out of work IT folks
using your services ... :-(

 

[Paul Sture]

Hmm - you're right - that's something I hadn't even remotely considered...

That is something I am acutely aware of. I live in a spa town, where
there are quite a few associated health clinics.

The official town website is bad enough to navigate with a mouse, let
alone a laptop trackpad. I would have thought that with the potential of
using the website for promoting this aspect of the town, they would have
thought of accessibility issues.

 

[Paul Sture]

Not really sure why it'd be our market especially, just the bare fact
that our experience is saying 20%, whereas global stats are showing 6%.

One speculation might be that our clients will tend to access our site
from their workplaces (not paying for internet at home), and that it's
more likely to be something locked down in a corporate environment.

Since you mentioned that it is a charity, do the counsellors themselves
recommend access points (public libraries etc) for those with out
internet access at home? If so, that may give you a pointer.

 

[Paul Sture]

Me too.


Me too.


They were actually using a Mac...

Yep, although as I have mentioned to you already, I'm having a look at
Opera's functionality in this area.

Now a few comments on why I routinely run with JavaScript disabled.

o - With older browsers it was an easy way to get rid of unwanted pop-ups
and redirections, especially unwelcome when I was on dial up and
it could cost me.

o - I had a potentially embarrassing episode at work when reading a
newsgroup item which redirected via JS (eventually) to a porn site.
Fortunately I caught it in time, but turning JS off was a no brainer.
OK, that was an ancient version of Netscape, but the lesson stuck.

o - I do have a security conscious background, and keep up to date with
what potential security holes are out there. For example, it is only
recently that potential exploits were found in the FireFox/Mozilla
family of products, and the recommended precaution was to disable
Javascript. US-CERT reports for April, May and June 2006 can be found
for these at

http://tinyurl.com/fmtyo

o - A couple of years ago I came across a new job website, and the
authors were asking for comments. Fine, but they refused point blank
to support Netscape, and wouldn't let you into their site with it.

*But*... My employer was at that time still running on Netscape 4.7.
They had a huge amount of internal applications which all needed to
be validated on newer browsers, and we were expressly forbidden from
downloading or installing anything more modern meanwhile (the price
of adopting web enabled applications early on in the game).

The website authors couldn't get their heads around the fact that
I wasn't free to simply install the alternative browsers they
suggested.

Going slightly off topic now, but I must mention that from approximately
2001 on, one of the (formerly major) job agencies here steadfastly
refused to support any browser except MSIE for accessing their website,
using the argument that "95% of our visitors use MSIE, we don't care
about the rest".

Big whoopsie there, as the employer that I am talking about above had
over 30.000 well locked down PCs still using Netscape.

The 5% that the agency "didn't care about" included the hiring managers
for those 30,000 plus people, so in this case at least, looking at the
raw numbers isn't quite so straightforward.

Just some food for thought from "the other side of the fence"