HTML Beginner

[Sam Hughes]

Nice strategic trim job of my post.

Strategic? I didn't know I was playing a game.

Can you name a browser that doesn't support linking to an ID? A simple
"yes" or "no" answer will do.

This is irrelevant, hence the trimming. Memorizing which specific browsers
don't support linking to an id is not useful; it is simpler to merely note
that there are some which don't. I don't expect Web Browsers to be a
category on Jeopardy anytime soon.

 

[Karl Core]

Strategic? I didn't know I was playing a game.

Apparently you think you are, because you think you can get away with making
a claim and not supporting it.

This is irrelevant, hence the trimming. Memorizing which specific
browsers
don't support linking to an id is not useful;

It is very useful when the burden of proof rests on your shoulders to
substantiate what you're claiming.
Either you can prove what you've stated or you cannot. Pretty simple issue.

it is simpler to merely note
that there are some which don't.

There are some which don't support tables. There are some which don't
support images. There are some which don't support CSS. There are some which
don't support scripting.
And in the real world, the bigger issue is "How many don't?"
I don't dispute your claim that some browsers don't support linking to an
ID.
You made an additional (implied) claim that the market share of these
browsers is substantial enough to be a concern for developers.

 

[Kris]

Many browsers that don't support linking to an ID have substantial, non-
zero market share.

I asked you to name one.
Just because you say so doesn't mean it is true.

I'd say that any browser that doesn't support linking to an ID will have
less than 1% market share.
I don't make business decisions based on 1% of my customers, and neither
should anyone else.[/QUOTE]

And it is not like the lack of support causes the information to be
accessible to those users. They just have to scoll down a bit and hunt
for it.

 

[Toby Inkster]

Can you name a browser that doesn't support linking to an ID? A simple "yes"
or "no" answer will do.

Netscape prior to version 6. Note: Netscape 4.8 was released in mid August
2002, compared with IE 6.0 which was released October 2001.

 

[Karl Core]

Netscape prior to version 6. Note: Netscape 4.8 was released in mid August
2002, compared with IE 6.0 which was released October 2001.

Netscape 4.x has a less-than-1% market share. I rest my case.

 

[Sam Hughes]

Netscape 4.x has a less-than-1% market share. I rest my case.

You rest it on what?

From the statistics I've read, Netscape 4.x and below have about a 0.2%
market share. Let's suppose they are the only browsers that don't
understand id-pointing hyperlinks. It takes 9 extra characters to use a
named anchor instead of an id attribute. Let's examine the costs of
those 9 extra characters.

Development time: nine extra characters takes about three or so seconds
to type, even when inserting the tags afterwords. Let's suppose your
time is worth, say, 120 USD an hour, to make the math easy and to pick
our numbers conservatively. In that case, 120 USD * 3 / 3600 = .10 USD.
With that math, a named anchor costs ten cents of development time (and
less if the named anchor is created blindly by a WYSIWYG editor which
many developers use).

Hard Drive Space: hard drive space is cheap and negligable.

Bandwidth: Suppose, for the sake of example, we were to order the Single
B package at <http://affordablehost.com/cpanel.shtml>, at $5.95 a month
with 5 gigabytes of bandwidth. To be conservative, let's suppose the
entire cost of the hosting goes towards bandwidth. In that case, we have
a rate of 1.19 USD / GB. Multiply this by (9/(2^30)) (the number of
gigabytes equal to 9 bytes) and we arrive at the price of 9 bytes of
bandwidth, 0.00000000997 USD.

With a 0.2% market share of browsers needing the named anchor, we have
only one in five hundred users who benefit from it. With the cost of
each named anchor being .00000000997 USD per transfer, with only one in
every five hundred named anchors ever needed, it costs .00000000997 * 500
= .00000498 USD in bandwidth for each use of the named anchor where id
would not suffice.

Now suppose that in all the NS 4- hits that download a certain named
anchor, only one in say, two thousand of them are actually linked to that
named anchor (just a guess, and a pretty conservative one). .00000498 *
2000 = 0.00997 USD. Which means it costs about one extra cent for each
user that ends up using the named anchor.

Now let's suppose that penny wasn't spent. In that case, the user would
get linked to the top of the page. As Kris wrote in


"And it is not like the lack of support causes the information to be
accessible to those users. They just have to scoll down a bit and
hunt for it."


There are generally two possibilities when the user gets linked to the
top of the page: either he knows he was meant to be linked to the middle
of the page, or he doesn't. Let's consider the latter case first: If he
doesn't know what happened (and this is the far more likely scenario), he
will be confused, wondering why the information he was looking for cannot
be found, and his time will be wasted (certainly more time than the three
seconds it would have taken to write that named anchor). He won't know
where he is, and a sale may likely be lost.

Now, suppose the user does know to scroll down and "hunt for it." (If he
is so savvy, why isn't he using FireFox?) How much time does this take?
Certainly more than a penny's worth. Certainly more than three seconds.
If the site is trying to sell something, it will lose sales, and if it is
the type of site such as www.html-faq.com or All My FAQs, it will be
doing badly at its job of being a good resource (seeing that the three
seconds of development time is worth less than the k * n seconds of
user's time wasted, k being the amount of time wasting "hunting" for each
link that fails to work, n being the number of users for which it fails
to work).

And remember that the ten-cent development cost and the one cent per
named anchor use cost are based on keeping all the numbers conservative.

Therefore, even with NS 4's pitiful market share, writing named anchors
is worth the effort.

 

[Sam Hughes]

And remember that the ten-cent development cost and the one cent per
named anchor use cost are based on keeping all the numbers
conservative.

I suppose it might take some people more than three seconds to type the
named anchor, but then again, I don't think their time is worth 120 USD.
It is a fixed cost per anchor, anyway.

 

[rf]

Sam Hughes wrote:

Therefore, even with NS 4's pitiful market share, writing named anchors
is worth the effort.

You forgot the last part:

Lets assume the product being sold is cheap. A beer perhaps. These retail at
about $1 or so. Lets say the net profit is 30 cents.

If only one person in the entire world ever buys even a single beer that he
otherwise would not have because of the anchor problem then the development
investment of 10 cents has been repaid threefold. Pretty good odds to me.

Now, I buy my beer in job lots of 30 tins. Make that ninetyfold. Bloody good
odds. The marketing bods would be drooling.

 

[Toby Inkster]

Netscape 4.x has a less-than-1% market share. I rest my case.

Depends on the market. It is still routinely installed in many educational
institutions.

 

[brucie]

In alt.html Toby Inkster said:

Depends on the market. It is still routinely installed in many educational
institutions.

which is one reason to think carefully about trusting your education to
them. if they're so stuck in the past using ancient broken software
whats your course going to be like?

 

[rf]

trusting your education to
them.

Those that can, do.

Those that can't, teach.

 

[Big Bill]

Depends on the market. It is still routinely installed in many educational
institutions.

It educates them how, one wonders?

BB

 

[Big Bill]

Those that can, do.

Those that can't, teach.

Hey, I can't. Gotta job?

BB

 

[Karl Core]

You rest it on what?

Therefore, even with NS 4's pitiful market share, writing named anchors
is worth the effort.

No, it isn't. It is not even worth the "10 cents per named anchor" to make a
business decision on a market which, even by the most generous estimates, is
..75% of users.
ANY company that is willing to make their business decisions based on such a
small segment of their (potential) customer base is destined to failure. And
in this case, it is even sillier. It's not like we're talking about a markup
practice which will completely alienated visitors (i.e. missing content/
inability to navigate, etc.).

 

[Sid Ismail]

: Now, I buy my beer in job lots of 30 tins.


Knew there was something odd about you.... LOL

Sid

 

[Neal]

Those that can, do.

Those that can't, teach.

That's such bullshit. Most of the time at least. Though there are utterly
useless teachers who have no real understanding of their content beyond
the textbook.

 

[Sam Hughes]

No, it isn't. It is not even worth the "10 cents per named anchor" to
make a business decision on a market which, even by the most generous
estimates, is .75% of users.
ANY company that is willing to make their business decisions based on
such a small segment of their (potential) customer base is destined to
failure. And in this case, it is even sillier. It's not like we're
talking about a markup practice which will completely alienated
visitors (i.e. missing content/ inability to navigate, etc.).

It does constitute inability to navigate, when it comes to incapable users.
Most users will have no way of knowing that they should scroll down. Years
ago, I remember how amazed I was when I discovered that a whole bunch of
links pointed to different parts of the same page, having been oblivious to
the fact for months, thinking they pointed to different web pages. Had I
originally arrived at the top of the page, I would have wondered why the
information was not what I was looking for (I was looking for fees, and
instead, I would have gotten a hole-by-hole description of tho golf
course), and I would have used the back button, wondering why the Web site
was broken.

I would say that using ids isn't much of a problem if there is a short
table of contents at the top of the page.

 

[Karl Core]

It does constitute inability to navigate, when it comes to incapable
users.
Most users will have no way of knowing that they should scroll down.
Years
ago, I remember how amazed I was when I discovered that a whole bunch of
links pointed to different parts of the same page, having been oblivious
to
the fact for months, thinking they pointed to different web pages. Had I
originally arrived at the top of the page, I would have wondered why the
information was not what I was looking for (I was looking for fees, and
instead, I would have gotten a hole-by-hole description of tho golf
course), and I would have used the back button, wondering why the Web site
was broken.

I would say that using ids isn't much of a problem if there is a short
table of contents at the top of the page.

Again, you're dodging the real issue: It is bad business to make decisions
based on such a small audience.

 

[Sam Hughes]

Again, you're dodging the real issue: It is bad business to make
decisions based on such a small audience.

It is impossible not to make a decision: Either ids are used or named
anchors are used. If the decision between named anchors and ids is not
based on the audience which gets affected by the decision, then what _is_
the decision based on?

It is worse business to make decisions based on nothing at all.

 

[neredbojias]

Without quill or qualm, Neal quothed:

That's such bullshit. Most of the time at least. Though there are utterly
useless teachers who have no real understanding of their content beyond
the textbook.

Teachers may know their stuff, but they're too lazy to put it to use.
That's why there's no drinking teachers.