Looking for:
- ideas
- articles
- war stories from the trenches
- methods you have used
- wisdom derived from dealing with this before
- anything!
. . . to help convert this site to XOOPS.

The situation:

I have a client with a large site which was created as a bunch of individual pages.
400 to 500 pages, or more.
The site is very successful - well over a million page views per month.
Bandwidth allocation purchased is 600 GB/month.

The site has apparently just grown organically since 1999.
The owners know little about the web.
The web person knows how to make small web sites (a few pages).

The site is like hundreds of tiny web sites.
Each newsletter, article, advertiser, trade show photo collection, etc.
is in it's own directory, with it's own images, and it's own HTML.
This includes the navigation images.
So the same navigation images are on the site hundreds of times. (bandwidth - oink, oink)
All of the navigation images have generic and inconsistent names (image001, image009, etc.).

The page layout, nav images, even the site logo image varies all over the place on the various pages.
There are no templates used anywhere.
There is no consistancy anywhere.

They initially contacted me to help them reconfigure their phpAdsNew setup to do run-of-site ads and section ads.
Since there are no templates, no headers, no footers this is a challenge.

I just looked at the phpAdsNew setup last night.
No zones at all.
~180 advertisers x ~4 ads each = ~720 ads.
All of these ads are placed in the site with phpAdsNew direct selection code.
Kept finding my mouth hanging open.
And kept thinking omigod, omigod.

So, they need to convert to a template based system and use ad zones.

I am trying to visualize how to convert this content to XOOPS.
No consistent header or footer HTML to strip to get to just the content.
Looks like a ton of cut and paste.

Anyone out there who has dealt with a similar situation please share your experiences.
How did you approach the project?
What recommendatios can you make based on your experiences?
Anyone know of any online articles about this issue?

Thanks to all for your ideas and help.

No templates, no apparent directory structure, no standard naming conventions . I think that it is going to be very unlikely that you'll be able to create a script to migrate content, and I'd agree that this one is going to be just one massive cut 'n past job. I'm glad it is you and not me , and I hope this is not a quoted job as it has serious potential for blowing out into a nightmare job. I find it almost unbeleiveable that such a high traffic site was able to develop this far without something being done about it. Have you given any thought to what modules you'll be using in Xoops?

Well I haven't done any large site conversions, small ones are a snap. It sounds like you have a real monster on your hands. I don't envy the task except that with my monetary situation I'd try to tackle it anyway.

Humble advice I could give:
Map out the structure and content flow you want before analyzing how big of a mess it is already. You don't want to mimic the mess there now. In fact I'd suggest not trying to mimic the site too much at all since you won't be appreciated if it looks or feels the same. The much improved maintenance will be under appreciated because the current web monkeys have a long standing process they are comfortable with even if it makes us cringe to watch it. Be prepared for that letdown.

Since you will be dealing with high traffic carefully consider the server set-up. Separate and optimized web and database servers might be preferred. Perhaps even a caching proxy server in front of the web server is called for so you don't have to compromise the back end caching decisions.

It does sound like a lot of copy and paste but you won't have to deal with the headers, footers, and common graphics more than once. The CMS will make that part a breeze as long as consistency is a desired goal.

You might visualize the various content management needs and decide on suitable modules that will serve them up for you. The ads functions might need some custom work depending on what is required. News, graphics, and the usual content can be done with available modules I think.

Sounds like a lot of work. Let me know if you need any help.

Hi.
In your place, I'd have no doubt at all. No, this isn't a case for a sane template system and overall optimization: finding a way to get dinamic/section based phpAdsNew selection codes would do.

If you want to be a tad more helpful, I'd suggest getting those .html files parsed by PHP and trying to implement simple custom header/footer/nav stuff from that instead of trying to put this hairy monster into a DB

Thanks all for the advice.

So far I am only engaged for the phpAdsNew (PAN) work.
I have only been talking to the wife of this husband and wife team as she handles the ads and PAN. Looks like they have 2 employees and the "web guy" is local help.
They live in the boonies.

I made it clear today that changes must be made to the site in order to do the ads the way they want. We have discussed it before so they are aware changes will be required.

The good thing is they have already published and sold the new ad programs to their top advertisers. In fact, the new programs were requested by the top advertisers. They now must deliver.

I think in the short-run putting in some SSI headers and footers will allow them serve the run-of-site ads. This will also allow pages to be updated one-by-one without breaking anything. Then eventually all the pages would have common header and footer code allowing much easier conversion to a database driven site using a conversion script.

Just getting the image duplication issue under control should be a huge bandwidth savings.
But again, this will have to be done page-by-page. UGH!

The home page is 933K! 933 freakin' K !!
They do have great content and the site users are avid enthusiasts of the subject so that is why the users appear to put up with the huge pages and horrible site structure.
I also looked at some of the page HTML - spagetti.

Modules?
They have articles, news, classified ads, vendor directory, picture galleries, etc. Have not really thought about it in depth. I think they are ideal for a forum, but they are wary of the "bad guys" in forums criticizing the vendors.

No doubt they would benefit from a CMS.
I cannot even imagine how they maintain this site.

I just looked at the ad reports I downloaded last night.
Adviews are running around 15 million per month.
All of the advertisers are on a flat monthly fee (a freakin' bargain!).
They like the flat fee concept - so I think that is what I am going to propose.
Per their hosting services fee quote to me, it is probably costing them $600-$1,000
per month for the bandwidth.
What I can save them in bandwidth alone could pay for a monthly fee.

The dedicated server is up to the job.
3.06 GHz, 2 GB , 36 GB SCSI, FreeBSD.
I did ask if they were using a cache accelerator, she did not know.

I am supposed to get full access to the server tomorrow so I can investigate more.

Thanks for the input everyone!