Webmaster forum - Re: Site Announcement?

Author

Re: Site Announcement?

MGW

2006-11-19, 7:41 pm

On Mon, 06 Nov 2006 04:11:41 GMT, "Heidi" <blackcat2@XXXXXXXXXX>
scrawled:

> Heidi wrote:
> : David J. Hennessy wrote:
> :: Let me guess... you have JavaScript enabled, but not Flash?
>
> Views fine in Firefox and Opera. I just checked it in both and now see what
> I was missing.
> That said it is a bit of a scroll still to get to the content... =)
> IE6 though does not display the large toilet. =) I checked my settings and I
> have no clue why that is.

Perhaps because it likes the major parties and isn't amused by the
site? Or else it dislikes plumbing?

--
MGW
Statistics means never having to say you're certain.

(Note: my Hotmail address is seldom checked)

Dylan Parry

2006-11-19, 7:42 pm

Matt Probert wrote:

> I should reduce the length of the individual pages within my site,
> thereby making each page more tightly focused

This is something that I meant to mention to you a while back when you
were asking for opinions, but I don't think I ever got around to it!
It's a sensible idea (imho), which would probably also help with the SEO
of your site -- as it is, with potentially dozens of different topics on
the one page, they are pretty loosely focused, which probably doesn't
help with your rankings.

I for one would welcome you posting an example of how you intend to
accomplish your new site architecture ;)

--
Dylan Parry
http://electricfreedom.org | http://webpageworkshop.co.uk

Programming, n: A pastime similar to banging one's head
against a wall, but with fewer opportunities for reward.

Matt Probert

2006-11-19, 7:44 pm

On Mon, 06 Nov 2006 09:40:03 +0000, Dylan Parry
<usenet@dylanparry.com> wrote:

>Matt Probert wrote:
>
>
>This is something that I meant to mention to you a while back when you
>were asking for opinions, but I don't think I ever got around to it!
>It's a sensible idea (imho), which would probably also help with the SEO
>of your site -- as it is, with potentially dozens of different topics on
>the one page, they are pretty loosely focused, which probably doesn't
>help with your rankings.

All perfectly correct. All addressed. All in progress <g>

>
>I for one would welcome you posting an example of how you intend to
>accomplish your new site architecture ;)
>

This is a real problem.

Take a gazetteer for example. I work alphabetically, thus one gets:

Gazetteer
A
B
C
.....
Z

There are a great number of places starting with the letter S.
Breaking these down to a reasonable number per static page, leaves a
very real problem (challenge?) with the navigation, about 150
different pages, which in turn appear as a long list of navigation
links, or could be broken down into a deep level of nesting...

Gazetteer -> S -> Sa-Sb -> Sa-Sag

It's not a problem with dynamically searched for data, the user just
enters a term and has delivered the data. It's with the browse
approach that issues arise.

I have created a human site map grouping together spheres of related
data:

The Natural World (Animals and Plants, Rocks and Minerals)

Humanity (Actors, Actresses, Gazetteer, General Information, Money and
Finance, Other Notable People)

Science and Technology (Aircraft, Computer Viruses, Medicine, Ships,
General Science and Technology, Warfare)

Arts and Culture (Architecture, Costume, Food & Drink, Heraldry,
Movies, Music, Mythology --- Greek & Roman Egyptian Celtic Norse
Voodoo Aztec & Mayan Hindu Sumerian Japanese Native American Enochian
Other, Recreation, The Plays of Shakespeare)

Language (Abbreviations, Languages, Latin, Slang)

About The Probert Encyclopaedia (Advertise with us, Contact Us, eBooks
to Download, Frequently Asked Questions, Information for Web Masters
--- Add your URL Free Content, Privacy Policy, Quiz, The Name of
Probert)

And even this was a considerable challenge.

As for the solution, I haven't one. Analysis reveals comparatively few
causal browsers, the majority of data access is through search
facilities, so this leads me to reduce the priority for the browse
navigation, with higher prioroty being given to spreading data among
more pages which assists search speeds, transfer times and, helpfully
enough SEO also.

Creating and publishing an encyclopaedia is not as easy as some people
might imagine, indeed I have been greatly buoyed to notice some ideas
being imitated by larger organisations which have appeared on the
scene after my humble efforts first appeared.

Matt

--
Woe to him that willfully innovates, while ignorant of the constant.
http://www.probertencyclopaedia.com

Dylan Parry

2006-11-19, 7:44 pm

Matt Probert wrote:

[Breaking down a site into a new architecture]
> This is a real problem.

Yes. ISTR that you currently use a flat file system for generating your
articles at the moment. Have you considered the implications of using a
RDBMS instead?

I think that the best way you could separate items out is to have only
one article per page, rather than having several articles, even if they
are related - rather displaying a "related articles" section at the
bottom of the page. This would have the added benefit of reducing your
bandwidth ;)

Back to using a RDBMS; if would be altogether possible to write a search
function that dynamically searched the articles, and built/maintained
the indexes for you rather than having to run a script once a week and
uploading its output. Of course, you could still have the site 100%
static by generating all the pages offline and uploading them, but at
least you would have the benefits of a RDBMS to help you to classify
individual articles etc.

I'm probably rambling and making no sense now, so I'll stop.

--
Dylan Parry
http://electricfreedom.org | http://webpageworkshop.co.uk

Programming, n: A pastime similar to banging one's head
against a wall, but with fewer opportunities for reward.

Nullified

2006-11-19, 7:45 pm

On Sun, 05 Nov 2006 08:22:06 GMT, "David J. Hennessy"
<david@maidix.com> wrote:

>OK... so what are the current unspoken rules about posting the URL of a
>new site that has Google ads in it?

Anybody else noticed snakepiss' absense from this thread? Seems like
the rules only apply to those he thinks they should apply to - yet
another reason why the FAQ is worthless ;o)

William Tasso

2006-11-19, 7:45 pm

Fleeing from the madness of the NTL jungle
Nullified <null@null.null> stumbled into news:alt.www.webmaster
and said:

> ...
> Anybody else noticed snakepiss' absense from this thread?

No

> ...
> the FAQ is worthless ;o)

Please do tell what type of web sites you make? Are they for your beneift
or do you build sites for others?

/just curious

--
William Tasso

Nullified

2006-11-19, 7:45 pm

On Mon, 06 Nov 2006 14:34:36 -0000, "William Tasso"
<SpamBlocked@tbdata.com> wrote:

>Fleeing from the madness of the NTL jungle
>Nullified <null@null.null> stumbled into news:alt.www.webmaster
>and said:
>
>
>No
>
>
>Please do tell what type of web sites you make? Are they for your beneift
>or do you build sites for others?
>
>/just curious

Oh God no, just for me (though I have occasionally built for local
pals).

Matt Probert

2006-11-19, 7:45 pm

On Mon, 06 Nov 2006 13:21:27 +0000, Dylan Parry
<usenet@dylanparry.com> wrote:

>Matt Probert wrote:
>
>[Breaking down a site into a new architecture]
>
>Yes. ISTR that you currently use a flat file system for generating your
>articles at the moment. Have you considered the implications of using a
>RDBMS instead?

No. I use a proprietory ascii based file storage system, which is then
accessed by a sophisticated suite of bespoke software which split the
source files into numerous smaller files and apply HTML markup, and
links between related topics, automatically determining relationships.

This means that the raw data can be stored in a natural format,
providing a means for very easy creation and amendment. articles are
simply typed into a text file, prefixed with an appropriate title and
on a new line the body of the article without the need to worry about
relating the body of data to other existing articles.

It is therefore a RDBMS, not a variant of SQL (many less technical
readers will assume that RDBMS is a synonym for SQL/Access)

>I think that the best way you could separate items out is to have only
>one article per page, rather than having several articles, even if they
>are related - rather displaying a "related articles" section at the
>bottom of the page. This would have the added benefit of reducing your
>bandwidth ;)

Indeed. However, there is a challenge with this. The encyclopaedia
receives the bulk of its income from advertising. Reducing each page
to a single article would result in a great number of very short
articles which would be swamped by advertising.

Another issue is the number of disk files. With over 200,000 articles
you can realise the issue. I don't know about XP, but earlier variants
of Windows could not cope with that number of disk files within a
single directory, perhaps I could migrate to Linux, but that would
require months of work which I can not justify the expense of.

>Back to using a RDBMS; if would be altogether possible to write a search
>function that dynamically searched the articles, and built/maintained
>the indexes for you rather than having to run a script once a week and
>uploading its output.

I don't understand. I currently use sophisticated bespoke search
scripts which dynamically search the articles. These are assisted by
an index, but that is a simple pointer linking words to files
containing articles and it is automatically updated when the "build"
takes place, several times a week.

>Of course, you could still have the site 100%
>static by generating all the pages offline and uploading them, but at
>least you would have the benefits of a RDBMS to help you to classify
>individual articles etc.

The site is not 100% static at the moment. There are static pages - to
allow casual browsing. But it is also dynamic, the "research"
facilities dynamically collate all related articles for a given search
term and produce a single output. So, just as an example, one might
research 'rocket' and receive masses of dynamically generated data
about various rockets and related systems.

>
>I'm probably rambling and making no sense now, so I'll stop.

That's quite alright. It's a very unusual project, with very unusual
problems. <g>

Matt

--
Woe to him that willfully innovates, while ignorant of the constant.
http://www.probertencyclopaedia.com

Dylan Parry

2006-11-19, 7:45 pm

Matt Probert wrote:

> Another issue is the number of disk files. With over 200,000 articles
> you can realise the issue. I don't know about XP, but earlier variants
> of Windows could not cope with that number of disk files within a
> single directory

You'll be okay with XP. The limit to the number of files in a single
directory[1] is 4,294,967,295 as long as you're using a NTFS-formatted
drive.

_______
[1] I don't believe there is as such a limit to the number of files in a
single directory, but there *is* a limit of 4,294,967,295 to the number
of files within a single /volume/. Ergo an imposed limit per folder.

--
Dylan Parry
http://electricfreedom.org | http://webpageworkshop.co.uk

Programming, n: A pastime similar to banging one's head
against a wall, but with fewer opportunities for reward.

Red E. Kilowatt

2006-11-19, 7:48 pm

David J. Hennessy <david@maidix.com> wrote in message:
LHp3h.3457$Tz.739@trndny01,
> Careful about those "you" statements, as they may be inaccurate. ;-)
> However, if that's how the site comes across, that's valuable info.
> Are you making a suggestion? Initially, all I really wanted to do was
> put the graphic up. Then, I started to put more info up. I see it as
> an ongoing project with some potential. Or am I better off with just
> the graphic (and nothing else)?

Regardless, how about a cleaner toilet? That one looks pretty nasty. I
might be willing to flush my vote, but I'm definitely not willing to
touch that handle.

--
Red