Do splash pages make you invisible to search engines

Is that splash page related to his search engine invisibility, despite
his lofty stature, and despite the large number of links pointing to
his site?

Kill the splash, add articles about fly fishing, fly rods, etc.
A good article will automatically have lots of related terms, which will
help.

Nov 10 '05 #4

On 10 Nov 2005 09:37:24 -0800, "Sa***************@gmail.com"
<Sa***************@gmail.com> wrote:

Here's a question I don't know the answer to:
Neither do we - tell us the URL and we might have a clue
His site does have a graphical splash page, with only one link
on it, that says "enter"

Dump the splash page - for all sorts of good reasons.

Nov 10 '05 #5

Stephen Poley

On 10 Nov 2005 09:37:24 -0800, "Sa***************@gmail.com"
<Sa***************@gmail.com> wrote:

Is that splash page related to his search engine invisibility, despite
his lofty stature, and despite the large number of links pointing to
his site?

A splash page per se will not destroy visibility, though it will harm
it. But some of the silly things that people do on sites, especially
those with splash pages, will destroy visibility. Without a URL we can't
say much more.

--
Stephen Poley

http://www.xs4all.nl/~sbpoley/webmatters/

Nov 10 '05 #6

Sandy.Pittendrigh

RE>Without a URL we can't say much more

I didn't make the webpage. It is a nice looking site.
That it isn't showing up on the search engines is a systematic bug of
sorts.

http://troutrods.com

Nov 10 '05 #7

mbstevens

RE>Without a URL we can't say much more

I didn't make the webpage. It is a nice looking site.
That it isn't showing up on the search engines is a systematic bug of
sorts.

http://troutrods.com

Yeeek! Deprecated markup, images for text, no doctype, slow loading on
both the splash page and the opening page.
There's more to worry about than the number of hits from a search engine
right now.

Anybody click the link to the people who made it? I can't even get in.

Nov 10 '05 #8

kchayka

That it isn't showing up on the search engines is a systematic bug of
sorts.

http://troutrods.com

I get the impression whoever built the site thought that meta keywords
had merit and would be enough. They thought wrong.

--
Reply email address is a bottomless spam bucket.
Please reply to the group so everyone can share.

Nov 10 '05 #9

On 10 Nov 2005 13:49:43 -0800, "Sa***************@gmail.com"
<Sa***************@gmail.com> wrote:

http://troutrods.com That it isn't showing up on the search engines is a systematic bug of
sorts.
Bugs of several sorts. The fundamental one is that it was coded in the
late '90s, with the standards of the day (If this site was built any
time post 2000, then it's really a bit of a crock)

Search engines are text-based. They don't really do pictures. Give them
some text and they'll index it happily. Don't give them _TEXT_ and you
may as well not exist. A "splash" page always hurts search engines,
because they're pretty but light on text that a search engine can make
use of.

This site has effectively two splash pages. The home page is almost
content-free (as far as a search engine sees it). Now if your incoming
links are to this homepage, then Google will see them as "unimportant"
because there's no content and it will see the real content pages as
"poorly rated" because no-one is linking to them. This all hurts your
placement.

There are two things you could do with this site. One is to re-develop
it, which will cost you money. The other (recommended) is to fix the
most obvious problems and leave the rest be. This newsgroup will
probably recommend the first and insist that you fix the "doctype" first
of all (you don't know what a doctype is. No-one ever caught more fish
by knowing what a doctype is (except maybe Blinky)).

I would suggest the following. It requires a _knowledgeable_ web spod,
but it also needs very little of their time (and so would be cheap to
do). This list isn't exhaustive (you want comprehensive reports, you can
pay me for them!)

- Dump the splash page.

- Put alt attributes on the images, _especially_ those that are used to
emulate text and most importantly on the ones that are the text links

- Put title attributes on the images too (a good web developer will use
slightly different wording on title and alt, a sloppy one is unlikely
to)

- Put title attributes on the navigation links.

- Put some text onto the homepage - even just a paragraph or two.

- Put some text around the link to the Japanese version

- Avoid using images for text, especially for the contact details. In
particular, this kills any attempts at geographical searching. Maybe all
the business is mail-order, but you'll still be searched for as "That
rod guy in Montana"

- Ditch the imagemap at the page footer and replace it with a text-based
navigation menu.

- replace the JavaScript-only links to the "enlarge the image" links
with ones that work when JavaScript is turned off too.

- Don't use "Earthtalk" as designers. They're clearly clueless about
useful web development (their own site is an abomination).

I _wouldn't_ go for more re-work than this. The next steps are obvious
(fix the invalid markup, use an up-to-date HTML version, a doctype,
avoid <table> markup, avoid <font>) however they are more work to do and
less payback for doing them.

You could even do some automated switching for Japanese / English
versions, based on the http accept header. This is easy with decent
hosting and gives the Japanese users a better initial default page.
It is a nice looking site.

That's a contentious statement hereabouts. The site suffers badly from
"Looks fine on my computer" syndrome, which is _very_ common. On one of
my desktops it's unusably small, on another it's unusably big. This sort
of problem is symptomatic of web designers who don't really understand
the web and don't appreciate that not everyone has the same settings the
original designer did.

--
Cats have nine lives, which is why they rarely post to Usenet.

Nov 10 '05 #10

Beauregard T. Shagnasty

mbstevens wrote:

Anybody click the link to the people who made it? I can't even get in.

Yep. Sad, isn't it? I especially loved the black-on-black text on:
http://www.earthtalk.com/

--
-bts
-Warning: I brake for lawn deer

Nov 10 '05 #11

mbstevens

Beauregard T. Shagnasty wrote:

Anybody click the link to the people who made it? I can't even get in.

Yep. Sad, isn't it? I especially loved the black-on-black text on:

There is this whole class of web-design books written by friends-o-flash
and/or print-graphic designers on a lark. That's about all my local
library buys (the one notable exception being Zeldman's book).

I suspect there's more money/advertising behind this kind of stuff than
behind anything actually having to do with web design. I fear we'll
never be rid of it. Perhaps, if the general prevalence of humbug ever
breaks, instead of our sinking into a new Dark Age...

Nov 10 '05 #12

Jim Moe

RE>Without a URL we can't say much more

I didn't make the webpage. It is a nice looking site.
That it isn't showing up on the search engines is a systematic bug of
sorts.

I agree: it is a beautiful site; the graphics are excellent, the layout
appealing. Too bad the fonts are so small; there is no reason for such
small fonts.

The Description and Keywords meta's are a good start. They must,
however, be reflected in the content for search engines to grant them any
weight.
There is no content until a 3rd page is visited. There is the splash
page, the "Home" page, then (at last!) some content.
I suggest dropping the Splash and Home pages, and making "the company"
page the first page people see. This gets people right to the business of
the rods, and content for search engines to index.

--
jmm (hyphen) list (at) sohnen-moe (dot) com
(Remove .AXSPAMGN for email)

Nov 11 '05 #13

Jim Moe

mbstevens wrote:

Anybody click the link to the people who made it? I can't even get in.

It is truly amazing. The only thing there is a Flash link, and it does
not work since I *only* have Flashplayer v7. It is impossible to enter the
site!
Maybe it is a pre-qualification screening for clients. If I can't get
into the site, I am not sophisticated enough for them.

--
jmm (hyphen) list (at) sohnen-moe (dot) com
(Remove .AXSPAMGN for email)

Nov 11 '05 #14

kchayka

Andy Dingley wrote:

On 10 Nov 2005 13:49:43 -0800, "Sa***************@gmail.com"
<Sa***************@gmail.com> wrote:
http://troutrods.com
it isn't showing up on the search engines

- Put alt attributes on the images, _especially_ those that are used to
emulate text and most importantly on the ones that are the text links
- Put title attributes on...

FWIW, this may have little affect on the rankings, if any. AIUI, some
SEs only index alt text if it is a link, others not at all. I don't
think any SEs index the title attribute.

Both alt and title, however, can be beneficial to visitors other than
SEs. That's reason enough to make use of them where appropriate.
- Avoid using images for text

This is indeed the best advice. Plain text rules. :)

--
Reply email address is a bottomless spam bucket.
Please reply to the group so everyone can share.

Nov 11 '05 #15

Nick Theodorakis

On Fri, 11 Nov 2005 04:23:29 +1000, brucie <sh**@usenetshit.info>
wrote:

splash pages are useless, annoying and keep the visitor away from the
content, get rid of it.

<http://seattlepi.nwsource.com/fun/Bizarro.asp?date=20051108>

Nick

--
Nick Theodorakis
ni**************@hotmail.com
contact form:
http://theodorakis.net/contact.html

Nov 11 '05 #16

Manfred Kooistra

kchayka wrote:

Plain text rules. :)

And that's sad, isn't it, as human beings are visual oriented while
their speech faculties are only an evolutionary add on. Or do you
"read" the street on your way to the supermarket? It seems to me that
computer developers should learn at least the well-known basics about
human perception and psychology.

While all the tips to include more text are really helpful in dealing
with the shortcomings of search engines, it's still no fault of us web
designers that programmers are too stupid to come up with a search
engine that can process images the way human beings do.

Currently, search engines are limiting human potential. I won't adapt
to them.

And by the way, simply using the current version of Firefox on a Mac
gave me the Earthtalk website in its full functionality. I have never
downloaded any plugins in my life, and still everything everywhere
works fine. Are you still using that Microsoft crap?

Nov 23 '05 #17

Safalra

On 13 Nov 2005 06:27:46 -0800, Manfred Kooistra wrote:

kchayka wrote:
Plain text rules. :)
And that's sad, isn't it, as human beings are visual oriented while
their speech faculties are only an evolutionary add on.

Many would argue that it's speech that separates us from the animals.
Communication based on speech is far more reliable that that based on
images, sounds, or any other medium, especially once a degree of formality
is invoked. Most of our communication takes the form of speech or writing.
It seems to me that
computer developers should learn at least the well-known basics about
human perception and psychology.
We do. That's where Human-Computer Interaction and Usability Theory come
from.
While all the tips to include more text are really helpful in dealing
with the shortcomings of search engines, it's still no fault of us web
designers that programmers are too stupid to come up with a search
engine that can process images the way human beings do.
<childish taunt>Well if you're so clever, you do it!</childish taunt>
Seriously though, the problem of understanding images isn't even well
defined - think how often humans misinterpret artwork (especially of an
slightly surreal or abstract nature).
Currently, search engines are limiting human potential. I won't adapt
to them.

That's a rather perverse view - like saying that vehicles limit human
potential because they can't yet go at light speed, and then walking
everywhere instead.

--
Safalra (Stephen Morley)
http://www.safalra.com/hypertext/

Nov 23 '05 #18

kchayka

Manfred Kooistra wrote:

kchayka wrote:
Plain text rules. :)
And that's sad, isn't it

I think not, considering that most sites using graphics instead of text
(Flash included) tend to use font sizes that are way too small for me to
read. That's the sad part, methinks.
Currently, search engines are limiting human potential. I won't adapt
to them.
And I won't adapt my browsing environment to cater to designers that
insist on treating web media the same as paper. So there :-P
Are you still using that Microsoft crap?

If you're talking about IE, I have never used it by choice.

If you're talking about the OS... while I don't like MS, I see Windows
as the least bad of the available choices.

I personally can't stand using a Mac for more than 5 minutes at a time.
Linux doesn't have the apps I want, among which is a decent graphics
editor (I *loathe* the gimp) and an accounting package. The stuff I run
on Windows won't run under wine or cross-over office. IMO apps should
run natively, anyway. Good or bad, that leaves Windows. :-\

--
Reply email address is a bottomless spam bucket.
Please reply to the group so everyone can share.

Nov 23 '05 #19

On 13 Nov 2005 06:27:46 -0800, "Manfred Kooistra"
<ma**************@gmx.de> wrote:

Currently, search engines are limiting human potential. I won't adapt
to them.

Good. Saves us all a lot of time and trouble.

Reconsider if you ever have any content worth indexing.

Nov 23 '05 #20

mbstevens

Manfred Kooistra wrote:

kchayka wrote:
Plain text rules. :)

And that's sad, isn't it, as human beings are visual oriented while
their speech faculties are only an evolutionary add on.

Living creature are collections of evolutionary add ons.

Or do you
"read" the street on your way to the supermarket?
Is this very epistolary conversation more like speech or a movie?
I haven't seen any pictures so far.
It seems to me that
computer developers should learn at least the well-known basics about
human perception and psychology.
Sorry, but you don't seem to understand it very well at all.

While all the tips to include more text are really helpful in dealing
with the shortcomings of search engines, it's still no fault of us web
designers that programmers are too stupid to come up with a search
engine that can process images the way human beings do.
A true line, finally.
Currently, search engines are limiting human potential.
Only if you feel you must depend on them exclusively. I find them quite
helpful.
I won't adapt
to them.

Or can't? Being a visual person doesn't mean that your creativity or
skills will deteriorate if you learn new things that you're not used to.
Inflexibility is nothing to brag about.
--
mbstevens
http://www.mbstevens.com/

Nov 23 '05 #21

Manfred Kooistra

Safalra wrote:

Most of our communication takes the form of speech or writing.
Well, that's not true. It is a standard misconception, while the truth
is that the larger part of our communication is non-verbal. Think about
how you orientate yourself in a city (by seeing the post office and
knowing that the next street is yours), how you understand people (by
looking at their mimics and gestures, their clothing and their general
behaviour). Think about why telephones and especially SMS, E-Mail and
Chat are such a good way to misrepresent yourself and lie.
think how often humans misinterpret artwork

We are not talking about art. We are talking about visual orientation
(navigation, menues and such). On airports these pict-systems work very
well and even across languages. We are also talking about the
communication of emotions - and human beings (and even art, e.g.
movies) are quite successful at that.

Currently, search engines are limiting human potential.

That's a rather perverse view - like saying that vehicles limit human
potential because they can't yet go at light speed, and then walking
everywhere instead.

No, search engines are like hiking in high heel shoes. A car can do
things that no human can, while all a search engine can do is handle
more information - while at the same time limiting this information to
a certain type. A car does not limit me, I can always get out, but the
current structure of the internet is based on verbal information, so I
cannot get out - except click myself from page to page, and that is
what most people do, if they search for images (especially, as I said
earlier, as it is so easy to lie with words - try any term on google
image search and see what you get: mostly unrelated stuff).

But to clear things up: I'm not against the internet or search engines
as they are, I'm only against those people that tell everyone that
alt-tags, text-links and stuff are the only right way to do things,
when someone asks a clear and friendly question that leads in a
different direction. It is a reaction that speaks of lack of respect,
of a need to dominate, and I hate that. I only meant to say that the
internet is large enough for everyone and that here we should answer
the questions that are put to us and not bully the person asking for
his or her preferences and taste.

Nov 23 '05 #22

Sandy.Pittendrigh

I got a lot of good feedback on this thread.
I'd like to ask another recursive question, about
the original subject:: splash pages.

I'm not an HTML developer, so my grasp of the
terminogy is limited (I'm more of a Java/database guy).
Would it be possible to maintain
the currently spurious splash page, while also including hidden
navigation somehow, perhaps in a NOSCRIPT tag
or a hidden DIV, so the site's owner could
still have his splash page, while still directing search engine
robots into the interior portions of the website?

Nov 23 '05 #23

"Sa***************@gmail.com" wrote:

Would it be possible to maintain
the currently spurious splash page, while also including hidden
navigation somehow, perhaps in a NOSCRIPT tag
or a hidden DIV, so the site's owner could
still have his splash page, while still directing search engine
robots into the interior portions of the website?

Possible: yes. A good idea: no.

Serving different content to search engines is called "cloaking". It can
result in your site disappearing from Google altogether. Just get rid of the
splash page. It's a useless appendage.

http://www.google.com/webmasters/gui...s.html#quality
--
phil [dot] ronan @ virgin [dot] net
http://vzone.virgin.net/phil.ronan/

Nov 23 '05 #24

Jan Steffen

Beauregard T. Shagnasty schrieb:

mbstevens wrote:
Anybody click the link to the people who made it? I can't even get in.

Yep. Sad, isn't it? I especially loved the black-on-black text on:
http://www.earthtalk.com/

"The newest version of the Earthtalk website requires the latest
version of Macromedia Flash Player."
(thanks god for custom colors)

Funny, I have Flash. It's just JS I have disabled.
They really have no clue at all.

Jan
--
Rediscover the web:
http://www.spreadfirefox.com/?q=affiliates&id=18179

Nov 23 '05 #25

Shawn K. Quinn

begin quotation
from Sa***************@gmail.com <Sa***************@gmail.com>
in message <11**********************@g44g2000cwa.googlegroups .com>
posted at 2005-11-10T17:37
[about site that sells custom made bamboo flyrods]

But if you search on Google for 'bamboo flyrod' or 'split cane flyrod'
he doesn't even show up in the first ten pages of links.
Typing in a colon, followed by his domain name show 534 sites
link to his site. This doesn't make sense. It isn't supposed to
work that way. His site does have a graphical splash page, with only one link
on it, that says "enter"
This doesn't make any sense. Once I've typed in the URL of a site, I've
already entered. Why should I have to "enter" again?

Are you sure that's really an HTML link? What do you get from running
something like "lynx -dump -source http://www.example.com" where the URL
is replaced with your friend's URL? Can you get into the site running
Lynx interactively?

If it's a Flash movie or Java applet that has an "enter" link, that's
why. If instead of <a href="realhomepage.html"> it's <a
onclick="document.location = 'realhomepage.html'"> (or however you would
do this in Javascript) that's why.
Is that splash page related to his search engine invisibility, despite
his lofty stature, and despite the large number of links pointing to
his site?

Quite possibly. If you want to know for sure, ask someone who works for
Google[*], as most common people have no idea how Google decides what to
index and what not to index.

If Google[*] determines it's in their best interest for your site not to
appear in their index, or for whatever other reason doesn't want your
site to appear in their index, your site doesn't appear in their index,
and there's not really much you can do about it.
[*] Or any other Web search engine or Web directory.

--
___ _ _____ |*|
/ __| |/ / _ \ |*| Shawn K. Quinn
\__ \ ' < (_) | |*| sk*****@speakeasy.net
|___/_|\_\__\_\ |*| Houston, TX, USA

Nov 23 '05 #26

Shawn K. Quinn

begin quotation
from Sa***************@gmail.com <Sa***************@gmail.com>
in message <11**********************@g14g2000cwa.googlegroups .com>
posted at 2005-11-10T21:49

I didn't make the webpage. It is a nice looking site.
That it isn't showing up on the search engines is a systematic bug of
sorts. http://troutrods.com

* Complete lack of text on the home page.

* No idea that the graphic is the "enter" link until I wave the mouse
around and get a hint that that's an enter link. It's entirely
possible even if this page were indexed, users would see this graphic,
think that's all there was, and immediately hit the back button.

* Complete lack of text on index2.html, save the copyright notice.

Advice to everyone: Do not hire Earthtalk Studios, who made this site,
as it is flagrantly invalid HTML and poorly done all around. It's right
down there with "sorry, this site won't work on your browser, please try
again from a PC running the latest Windows and Internet Exploder" or a
blank page.

--
___ _ _____ |*|
/ __| |/ / _ \ |*| Shawn K. Quinn
\__ \ ' < (_) | |*| sk*****@speakeasy.net
|___/_|\_\__\_\ |*| Houston, TX, USA

Nov 23 '05 #27

Harrie

Philip Ronan said the following on 11/14/2005 15:51 +0200:

"Sa***************@gmail.com" wrote:

Would it be possible to maintain
the currently spurious splash page, while also including hidden
navigation somehow, perhaps in a NOSCRIPT tag
or a hidden DIV, so the site's owner could
still have his splash page, while still directing search engine
robots into the interior portions of the website?

Possible: yes. A good idea: no.

Serving different content to search engines is called "cloaking". It can
result in your site disappearing from Google altogether. Just get rid of the
splash page. It's a useless appendage.

http://www.google.com/webmasters/gui...s.html#quality

Interesting read.

It also says "Don't create multiple pages, subdomains, or domains with
substantially duplicate content.". A friend of mine has two domains (one
with the cipher "2" and one with the word "two") resolving to the same
virtual webserver (so it has exactly the same content), will this hurt him?

He didn't do this for his ranking, just because the name could be
written in two ways.

--
Regards
Harrie

Nov 23 '05 #28

Harrie

Harrie said the following on 11/15/2005 16:05 +0200:

It also says "Don't create multiple pages, subdomains, or domains with
substantially duplicate content.". A friend of mine has two domains (one
with the cipher "2" and one with the word "two") resolving to the same
virtual webserver (so it has exactly the same content), will this hurt him?

Just a thought:

Would it make sense to make a permanent redirect (HTTP code 301) from
one site to the other?

--
Regards
Harrie

Nov 23 '05 #29

"Harrie" wrote:

Philip Ronan said the following on 11/14/2005 15:51 +0200:

http://www.google.com/webmasters/gui...s.html#quality

Interesting read.

It also says "Don't create multiple pages, subdomains, or domains with
substantially duplicate content.". A friend of mine has two domains (one
with the cipher "2" and one with the word "two") resolving to the same
virtual webserver (so it has exactly the same content), will this hurt him?

Quite possibly, yes.

What he should do is host the site at just one domain, and redirect all the
requests from the other domain to this domain. Apache servers can do this
sort of thing quite easily with a simple rule in the .htaccess file.

--
phil [dot] ronan @ virgin [dot] net
http://vzone.virgin.net/phil.ronan/

Nov 23 '05 #30

Harrie

Philip Ronan said the following on 11/15/2005 17:19 +0200:

What he should do is host the site at just one domain, and redirect all the
requests from the other domain to this domain. Apache servers can do this
sort of thing quite easily with a simple rule in the .htaccess file.

This was what I was thinking about when I posted a second message one
minute before your reply. Thanks for confirming it.

--
Regards
Harrie

Nov 23 '05 #31

Andrey Tarasevich

I got a lot of good feedback on this thread.
I'd like to ask another recursive question, about
the original subject:: splash pages.

I'm not an HTML developer, so my grasp of the
terminogy is limited (I'm more of a Java/database guy).
Would it be possible to maintain
the currently spurious splash page, while also including hidden
navigation somehow, perhaps in a NOSCRIPT tag
or a hidden DIV, so the site's owner could
still have his splash page, while still directing search engine
robots into the interior portions of the website?

As you probably understand already, this depends heavily on the particular
heuristics used by a particular search engine when it is trying to determine
what to index and what not to index. It might be that there should be certain
absolute amount of text present on the page in order for the search engine to
accept it. The easiest way to tell whether the methods you mention above will
actually work is to try them.

If you find the method to solve the problem without removing the splash page -
don't remove it. Most of the nonsense posted in this thread about splash pages
being useless and wrong is nothing more than... well, plain nonsense, probably
made up after it was discovered that splash pages might create problems with
certian underdesigned search engines. (Of course, the heuristics employed by
those search engines are not directed specifically against splash pages, but
rather splash pages became "collateral victims" here.) A splash page
implementted in good taste greatly impoves the look and feel of the web site,
and also provides an easily recognizable and relatively static facade to it. The
splash page on the site in question does indeed look good. Don't remove it,
unless it is absolutely necessary.

--
Best regards,
Andrey Tarasevich

Nov 23 '05 #32

On Tue, 15 Nov 2005 16:01:35 -0800, Andrey Tarasevich
<an**************@hotmail.com> wrote:

A splash page
implementted in good taste greatly impoves the look and feel of the web site,
and also provides an easily recognizable and relatively static facade to it.

What concrete advantage does a splash page offer ?

Books are relatively linear. We open them at the splash page. The web
is anything _but_ linear - you can build a site around an "ideal"
navigation path, but your users will have other ideas. They will enter
the site through the most relevant page that the search engine offers
them, or through the page that they personally chose to bookmark. You
have minimal control over either of these (without using server-side
redirection).

Nov 23 '05 #33

Stephen Poley

On Tue, 15 Nov 2005 16:01:35 -0800, Andrey Tarasevich
<an**************@hotmail.com> wrote:

If you find the method to solve the problem without removing the splash page -
don't remove it. Most of the nonsense posted in this thread about splash pages
being useless and wrong is nothing more than... well, plain nonsense, probably
made up after it was discovered that splash pages might create problems with
certian underdesigned search engines. (Of course, the heuristics employed by
those search engines are not directed specifically against splash pages, but
rather splash pages became "collateral victims" here.) A splash page
implementted in good taste greatly impoves the look and feel of the web site,
and also provides an easily recognizable and relatively static facade to it. The
splash page on the site in question does indeed look good. Don't remove it,
unless it is absolutely necessary.

So how many of your bookmarks point to splash pages?

--
Stephen Poley

http://www.xs4all.nl/~sbpoley/webmatters/

Nov 23 '05 #34

Pierre Goiffon

Harrie wrote:

http://www.google.com/webmasters/gui...s.html#quality

Interesting read.

It also says "Don't create multiple pages, subdomains, or domains with
substantially duplicate content.".

I've got a particular case in mind : what about a multilingual site ?

Let's say you've got a language selection page at the root of the .com
domain - this page only contains the enterprise logo with links to each
version of the site (fr, en, ...). This page is somehow a splash screen,
right ?
Should each dedicated language site be located on a subdomain
(<language>.<enterprise>.com) or a subfolder
(www.<enterprise>.com/<language>/) ?

What is the best way to do in such a case ?

Nov 23 '05 #35

"Pierre Goiffon" wrote:

Harrie wrote:
http://www.google.com/webmasters/gui...s.html#quality

Interesting read.

It also says "Don't create multiple pages, subdomains, or domains with
substantially duplicate content.".

I've got a particular case in mind : what about a multilingual site ?

Let's say you've got a language selection page at the root of the .com
domain - this page only contains the enterprise logo with links to each
version of the site (fr, en, ...). This page is somehow a splash screen,
right ?
Should each dedicated language site be located on a subdomain
(<language>.<enterprise>.com) or a subfolder
(www.<enterprise>.com/<language>/) ?

What is the best way to do in such a case ?

Content negotiation

--
phil [dot] ronan @ virgin [dot] net
http://vzone.virgin.net/phil.ronan/

Nov 23 '05 #36

Alan J. Flavell

On Wed, 16 Nov 2005, Pierre Goiffon wrote:

I've got a particular case in mind : what about a multilingual site ?

Let's say you've got a language selection page at the root of the
.com domain - this page only contains the enterprise logo with links
to each version of the site (fr, en, ...). This page is somehow a
splash screen, right ?
I wouldn't do it that way.
Should each dedicated language site be located on a subdomain
(<language>.<enterprise>.com)
You're likely to give yourself difficulties with handling
Accept-language negotiation if you structure it that way.

I'm not saying it can't be done, but it'd involve redirections and
other unpleasantness.
or a subfolder (www.<enterprise>.com/<language>/) ?

I wouldn't do that either. Sure: language negotiation shouldn't be
the -only- way to get to the content - lots of browsers haven't been
properly set up to show their user's language preferences, so it's
good to also offer explicit links. (NO FLAGS, PLEASE). But it would
be a mistake to structure the site in such a way as to cause
difficulties for the negotiated solution.

Nov 23 '05 #37

Pierre Goiffon

Alan J. Flavell wrote:

I've got a particular case in mind : what about a multilingual site ?
First of all, I didn't mention it in my first post, but there is no
content negociation in this site, and there won't be any - because
there's no good reason to add it, simply speaking.
Should each dedicated language site be located on a subdomain
(<language>.<enterprise>.com)

You're likely to give yourself difficulties with handling
Accept-language negotiation if you structure it that way.

(...)

or a subfolder (www.<enterprise>.com/<language>/) ?

I wouldn't do that either. Sure: language negotiation shouldn't be
the -only- way to get to the content - lots of browsers haven't been
properly set up to show their user's language preferences, so it's
good to also offer explicit links. (NO FLAGS, PLEASE). But it would
be a mistake to structure the site in such a way as to cause
difficulties for the negotiated solution.

I doubt you're thinking of a solution using <filename>.<language>.html
as an easier way to get content negociation ?

Actually, without any consideration about content negociation using the
accept-laguage http header, I was submitting the question having in mind
results in search engine indexing. So, if there are any differences,
what will give the better results : subdomain or subfolder ?

Nov 23 '05 #38

Nick Kew

Pierre Goiffon wrote:

I've got a particular case in mind : what about a multilingual site ?

IMHO httpd.apache.org does a pretty good job of that, within the
limitations of the availability of translations.

Why would you want to make more trouble for yourself and your users?

--
Nick Kew

Nov 23 '05 #39

Alan J. Flavell

On Wed, 16 Nov 2005, Pierre Goiffon wrote:

I've got a particular case in mind : what about a multilingual site ?
First of all, I didn't mention it in my first post, but there is no
content negociation in this site, and there won't be any
I don't see any reason to block that possibility right from the start,
but it's your web site in the end. If you don't want advice, there's
no need to ask for it.
Alan J. Flavell wrote: You're likely to give yourself difficulties with handling
Accept-language negotiation if you structure it that way.

I doubt you're thinking of a solution using
<filename>.<language>.html as an easier way to get content
negociation ?
For the filenames? It's an effective option, certainly. Irrespective
of what you aim to do with them (explicit links, or multiviews, or
typemaps).
Actually, without any consideration about content negociation using
the accept-laguage http header, I was submitting the question having
in mind results in search engine indexing.
But one doesn't put pages on the web solely for the indexers, any
more than another puts pages on the web solely for their visual
impact. By following a balanced strategy, the pages are available for
many different purposes.

In the medium to long term, it's been my experience that following
common good-practice gives rather good results with search engines.

Doing something unexpected could produce spectacularly good results,
for a short time before they notice and take measures against it; for
the most part, though, it's been my impression that doing something
unexpected is rated to just make things somewhat worse than they could
otherwise be.

And I say this as someone who, when he searches for new information on
topics of interest to him, gets a bit tired of being shown his own
pages on the topics by the search services. So although I don't make
a close study of search engines, I guess I must be doing something
right.
So, if there are any differences, what will
give the better results : subdomain or subfolder ?

By which you're saying you've already decided it's not going to be
anything else. Well, in the end it's your choice. Maybe someone else
has a more-specific answer to that question - I haven't one.

good luck

Nov 23 '05 #40

Pierre Goiffon

Alan J. Flavell wrote:

First of all, I didn't mention it in my first post, but there is no
content negociation in this site, and there won't be any

I don't see any reason to block that possibility right from the start,
but it's your web site in the end. If you don't want advice, there's
no need to ask for it.

As you already know english is not my mother tongue and I think what you
interpret reading my posts could be different of what I wanted to write
and how I wanted to express it.

To get back to the content negociation point, it's indeed right to
mention this technique but it's not applicable to the website I was
thinking of, for lots of reasons :
- I don't think accept-language reflects at all what the user wants -
considering more of that that no one can assure all the translations of
a given document are up to date
- content negociation functionnalities aren't available on the httpd
- there are just a few pages in this website
- the pages are static and to do things well implementing content
negociation would be really time consumming, for a so small gain for the
website users
- indeed only new users are selecting their language, the others are
using the direct language URL
- ...

Anyway, content negociation or not, I wrote my previous post in this
thread with questions in mind about webpages indexing (that was what
that thread was dealing with, if I understood things correctly), and the
particular case of a multilingual website.

I don't know if search engines robots are making use of the
accept-language header ?
If content negociation is in place, it sounds to me a good practice (in
fact, I think this is absolutely mandatory) to include on each page a
link to all the languages available, at best to the corresponding
document, at least to the home page for each languages. How all these
links between each languages version will be considered by search engines ?

You're likely to give yourself difficulties with handling
Accept-language negotiation if you structure it that way.
I doubt you're thinking of a solution using
<filename>.<language>.html as an easier way to get content
negociation ?

For the filenames? It's an effective option, certainly. Irrespective
of what you aim to do with them (explicit links, or multiviews, or
typemaps).

Sorry, I think I didn't understand what you meant from the start ? If I
understand it right, your advice is to keep a structure (file naming,
folders, sites) that could easyly support any content negociation
technique ? Did I get it right ?
So, if there are any differences, what will
give the better results : subdomain or subfolder ?

By which you're saying you've already decided it's not going to be
anything else.

Err, no in fact, I've not decided anything, otherwise I won't have asked
here what is the best way to do :)

So, we can start from the beginning :
- a multilingual site
- the "international home page" : a page that is only aimed to give
access to one language or another
- each page contains links to the other translations
- content negociation on the accept-language header, or not (if yes,
users would directly see their language site version, without using the
"international home page")

Is there are anything to take care of with search engines ? Are links to
translations available on each page will be a problem ? How each site
language version could be organized ? Etc...

Nov 23 '05 #41

Harrie

Beauregard T. Shagnasty said the following on 11/11/2005 00:24 +0200:

Yep. Sad, isn't it? I especially loved the black-on-black text on:
http://www.earthtalk.com/

Also sad:

A 404 when visiting:

http://www.toriamos.com/

(with JavaScript disabled, for those who didn't disable it)

--
Regards
Harrie

Nov 23 '05 #42

Big Bill

On Wed, 16 Nov 2005 10:41:30 +0100, Pierre Goiffon
<pg******@free.fr.invalid> wrote:

Harrie wrote:
http://www.google.com/webmasters/gui...s.html#quality

Interesting read.

It also says "Don't create multiple pages, subdomains, or domains with
substantially duplicate content.".

I've got a particular case in mind : what about a multilingual site ?

Let's say you've got a language selection page at the root of the .com
domain - this page only contains the enterprise logo with links to each
version of the site (fr, en, ...). This page is somehow a splash screen,
right ?
Should each dedicated language site be located on a subdomain
(<language>.<enterprise>.com) or a subfolder
(www.<enterprise>.com/<language>/) ?

What is the best way to do in such a case ?

I'd forget about Flash altogether. Have one multilingual home page and
say, an English page which would be optimised as the entry page for
the English speakers, a French page optimised for the French engines,
etc.
You need really to get away from the concept of the two-dimensional
brochure. A brochure can only have one cover because it's a physical
entity, whereas a virtual entity can have, effectively, as many covers
as you want.
The concept of the home page, the one the domain points to, endures
though, and to a degree that still needs to be treated as if it were
the front page of a physical brochure.

BB
--
www.kruse.co.uk/ se*@kruse.demon.co.uk
The buffalo have gone

Nov 23 '05 #43

Shawn K. Quinn

begin quotation
from Harrie <sp*****@linux.org.invalid>
in message <43***********************@news.xs4all.nl>
posted at 2005-11-15T15:05

Philip Ronan said the following on 11/14/2005 15:51 +0200:
http://www.google.com/webmasters/gui...s.html#quality
Interesting read. It also says "Don't create multiple pages, subdomains, or domains with
substantially duplicate content.". A friend of mine has two domains
(one with the cipher "2" and one with the word "two") resolving to the
same virtual webserver (so it has exactly the same content), will this
hurt him?
Quite possibly.
He didn't do this for his ranking, just because the name could be
written in two ways.

Then set up the other domains to issue an HTTP code 301 redirect to the
one "canonical" domain (that would be the one used on business cards,
as an e-mail domain, etc).

Same for e.g. example.com and www.example.com resolving to the Web site;
the former should be a redirect to the latter.

--
___ _ _____ |*|
/ __| |/ / _ \ |*| Shawn K. Quinn
\__ \ ' < (_) | |*| sk*****@speakeasy.net
|___/_|\_\__\_\ |*| Houston, TX, USA

Nov 23 '05 #44

Harrie

Shawn K. Quinn said the following on 11/17/2005 19:30 +0200:

begin quotation
from Harrie <sp*****@linux.org.invalid>
in message <43***********************@news.xs4all.nl>
posted at 2005-11-15T15:05
Philip Ronan said the following on 11/14/2005 15:51 +0200:
http://www.google.com/webmasters/gui...s.html#quality
It also says "Don't create multiple pages, subdomains, or domains with
substantially duplicate content.". A friend of mine has two domains
(one with the cipher "2" and one with the word "two") resolving to the
same virtual webserver (so it has exactly the same content), will this
hurt him?

Quite possibly.
He didn't do this for his ranking, just because the name could be
written in two ways.

Then set up the other domains to issue an HTTP code 301 redirect to the
one "canonical" domain (that would be the one used on business cards,
as an e-mail domain, etc).

You basically said the same thing as Philip Ronan two days ago with his
reply to the same post. Thanks anyway. It has been solved.
Same for e.g. example.com and www.example.com resolving to the Web site;
the former should be a redirect to the latter.

I didn't think about the www prefix, thanks for that (so he had four
domainnames pointing to the same thing).

I've looked at Apache's mod_alias for the redirect directive and was
able to figure out how to redirect one domain to another, but without
setting up a new virtualhost for a domainname with and without the www
prefix, I don't see how I can achief something like this, since the
first argument is a URL-path, not an URL itself.

This is of course off topic for html authoring, if nobody answers I'll
take this to an Apache mewsgroup.

--
Regards
Harrie

Nov 23 '05 #45

"Harrie" wrote:

Shawn K. Quinn said the following on 11/17/2005 19:30 +0200:
Same for e.g. example.com and www.example.com resolving to the Web site;
the former should be a redirect to the latter.

I didn't think about the www prefix, thanks for that (so he had four
domainnames pointing to the same thing).

I've looked at Apache's mod_alias for the redirect directive and was
able to figure out how to redirect one domain to another, but without
setting up a new virtualhost for a domainname with and without the www
prefix, I don't see how I can achief something like this, since the
first argument is a URL-path, not an URL itself.

This ought to do it:

RewriteCond %{HTTP_HOST} !^www\..*$ [NC]
RewriteCond %{HTTP_HOST} !^$
RewriteRule ^/(.*) http://www.your-site.example.com/$1 [L,R]

--
phil [dot] ronan @ virgin [dot] net
http://vzone.virgin.net/phil.ronan/

Nov 23 '05 #46

Robert Latest

On Thu, 17 Nov 2005 12:30:08 -0600,
Shawn K. Quinn <sk*****@speakeasy.net> wrote
in Msg. <sl********************@xevious.platypuslabs.org >

Then set up the other domains to issue an HTTP code 301 redirect to the
one "canonical" domain (that would be the one used on business cards,
as an e-mail domain, etc).

Same for e.g. example.com and www.example.com resolving to the Web site;
the former should be a redirect to the latter.

Interesting, because I'm in the same situation: The webserver is set up
do direct all of www.domain.com, domain.com, www.domain.de, domain.de to
the same root directory, but we always (on business cards etc.) go by
the www.domain.com address.

So do you say that, instead of the current setup, I should create a
separate file for the three .de resp. www-less domains which issues a
301 redirect? How can I accomplish this except through a PHP script? The
webspace is just "bought" somewhere; I don't have access to the server
config.

A related issue which has been discussed in this thread: Multiple
languages. This site is German-English, and I have the html content set
up in two subdirectories "de" and "en". There is an index.html in the
base directory containing some JavaScript that evaluates language
preference settings and loads the appropriate start page (de/index.html
or en/index.html). The non-JS alternative is just a "splash" page
requesting the user to select her preferred language. Additionally both
the German and English index page provide a link to one another.

How do you like this approach?

robert

Nov 23 '05 #47