Organizing and Linking Pages in a Large Website

[Dirtwad]

If you have a 50+ page website and most of these pages are in their own folders i.e gallery/ contact/ etc. How do you link back to the home page in the root if all of your folders have pages titled 'index'?

I know you can just put in the complete address http://mywebsite.com
But is there a better way?

Thank you for helping!

[pauland]

../index.html (.htm) should do the trick

[Dirtwad]

Thank you Pauland! I should have tried that...for some reason I assumed it would be directed to the index of the sub folder. Which leads to the next question:

How does one add the 'rel = no follow' to a link in XWD?

[bjornkn]

And remember that '../' brings you up one level only, while '/' brings you back to the root, no matter how many levels down the hierarchy you are.

[Dirtwad]

Thanks Bjornkn! I did not know that. So you are saying that the two dots .. before the forward slash / brings it up one level only? So what do I put for a page to go to a dif page that resides in a dif subfolder?

I am trying to 'Silo' a 50+ page site so I am spliting the site into seperate folders. Never done this before with XWD. I will share the SEO results in this forum when it is complete.

[bjornkn]

2 dots goes one level up, and '../../ goes 2 levels and so on.
'/' goes directly to the root. It's just like command line DOS, where cd / brings your prompt to the root.
If you want to go to a page named dif.htm inside a dif folder that is located inside the folder where your main page (calling page, current folder level) you would write 'dif/dif.htm'.
But if it is one level up (parallel to your current folder) you should use '../dif/dif.htm'.

[Dirtwad]

So if I understand you correctly:

If I have a folder in the root named /services that contains 2 pages:
1. named index.htm
2. named /policies.htm

A link from 1 to 2 would not be: /services/policies.htm

But rather: ../policies.htm

And a link from 2 to 1 would not be: /services/

but rather: ../services/

Is that right?

[bjornkn]

No, that is not right.
With relative links it is always needed to know where you are calling those links from, because all paths are calculated in relation to the current path.
If the folder 'services' contains two pages then they are at the same level ( no '/' in front of page 2.), and the folder could be moved or renamed, and the cross-links would still work.
A link from 1 to 2 would simply be: 'policies.htm'
And from 2 to 1 would be: 'index.htm'

Because the 'services' folder is at the root for 1-> 2 it would also work with: '/services/policies.htm'
as well as '../services/policies.htm' , but then you can't rename your folder without breaking the link.

And 2 -> 1 could be: '../services/' (because pages named index are loaded by default when a folder name is used)
Or even '/services/'.
Usually it works fine without that ending slash too.
But there's no need to make it more complicated by adding such paths when there's no need to..

[Dirtwad]

Awesome clarification bjornkn! Thank you for your help!

Are you familiar with the 'Silo' structuring of large websites? To do this, I need to add 'no follow' to several links through out the site.
To prevent Google from crawling certain pages and passing page rank.

Which presents my final question:

Where do I add: rel="nofollow"

Should it be added like this? /about.htm"rel="nofollow"

Thank you for your help. I will contribute some SEO and Xara examples of this finished Silo site when it is finished.

[bjornkn]

I'm not familiar with that 'Silo' system. Actually I thought you just meant a slang for 'SEO' system, which apparently wasn't that far from the truth

The best way to test out such things is to try them out. Make a preview and look at the source code it generates.
I don't think that you can make such rel= parameters with normal links in Xara. I did a little test, and Xara translates those quotes and spaces to:
<a href="/about%20rel=%22nofollow%22" onclick="return(xr_nn());">
which I don't think will work.
But you could always build those <a> tags yourself using HTMLplaceholders. If there's a lot of them it will be quite tedious though...
You might consider using robots.txt instead?
http://www.robotstxt.org/orig.html