One thing I did in addition to all this was to make a bookmarklet that our editors can use to quickly toggle links on and off. It’s very simple (wrapping added by me):

javascript:document.location.href=’http://example.com/toggle_cookie.php
?whereto=’+document.location.href;

toggle_cookie.php simply reads the cookie and toggles it (“on”, or “off”), then redirects the browser back to the page you were browsing (via the whereto GET variable). Makes it handy for editors to turn links on or off with one click to their bookmarks toolbar.

I hope you don’t mind. but your beginner (dogfood) and a few others want to know where to get started. So did I back in October’02 when it dawned on me that you could use MT as a CMS. I’ve created a short list of links, a summer reading list if you will, on various topics that helped get me over the hump when deploying MT as a CMS for a church web site I maintain.

On another note. I both agree and disagree with your point made on Brad’s site regarding categories as subdirectories versus coded fields in the Entry More field. While I do use categories as subdirs, I would suggest using the Keywords field instead. I have an abbreviated example of how I did it using PHP on my site, though using Brad’s MT-Perl will get it done just as nice.

That said, glad to see this topic covered as thoroughly as you have.

1/ what’s the reason to split the AP-site into different blogs? the use of different layouts for each blog/web-site-section?

2/ i’m overwhelmed by the vast amount of infos in the web about MT. is there any good site/link/source which delivers not MT-hacks, but conceptual approaches for MT? (when to split a site into different blogs, when to start using template-modules for editing etc…)

While Brad’s solution is perfect for personal or techie use, it does nothing to help a client easily update website copy…

Unless of course, anyone can think of a way around that. There is Adam Kalsey’s Process Tags plugin with which one could put the text blurbs into template modules and include them into entries, but then you will still be excluding them from search which is the major benefit of Brad’s method. If you exclude all of the text on a page, from the search, then what good is it?

Thinking…..

Installing/hacking MT is about the extent of my development abilities so any guidance would be appreciated.

Also, I should mention that there is an excellent desktop tool called Zempt that is under Development. I’ve suggested that it would be a great feature to allow custom entry profiles so that you can specify the fields you want and have all the Key Values stuff be transparent to the user. For more info, or better yet to chime in and let the developers know this would be a killer feature, check out the forum thread.

I think the greatest reason behind everyone’s blog linking to stopdesign or adaptive path is because of Doug Bowman (stopdesign himself) and exactly what kind of a company adaptive path is.

Anyone who considers themself a web designer thinks that Doug’s designs rock (check out his CSS Zen Garden entry if you don’t agree).

But the main point is that a hero in the web community redesigned the website for a consulting firm who is a hero in the usability and accessibility industry. Many of the world’s top usability experts work at adaptive path, and because they were the ones who redesigned its guaranteed to generate a lot of fanfair.

The same type of press would have gone to Cooper if it was all done in XHTML and CSS, and if someone like Doug or Jeffrey Zeldman had done it. But because it wasn’t redesigned to be ultra standards-compliant, its redesign wasn’t as well publicized and talked about as adaptive path’s was.

All in all, the combination of stopdesign and adaptive path definitely gets web designers turning their heads, and when another company of this magnitude and stature undergoes a standards-compliant redesign, it’ll get noticed as well.

Michael, I can’t tell you how many times I’ve run into this problem. In the beginning, I made things a bit better by changing the welcome message on the blog to list a key to entries. In other words:

Entry body: text 1
Entry additional text: text 2
Entry excerpt: text 3

However, that’s not perfect because when one is entering an entry, the welcome message can’t be seen.

I thought about hacking the templates to change the names, but two things stopped me. 1) Hacked templates get overwritten on upgrades and 2) I figured that displaying help/title text differentially depending on which blog page is being assembled was either beyond the templates capabilities or a HUGE hack….

What I’m doing currently for a client is using Brad Choate’s Key/Values plugin to overload the proper fields and then making a separate add/edit entry page for the complicated blog which uses the MT XML-RPC API to work with MT.

In this way, you can easily create a “catalog blog” with many many fields without strectching MT so far as to break it, metaphorically at least… So far, so good.

However, I don’t understand what all of the hype over your new site is about. Sure, it doesn’t use tables, which is great, but lots of designs don’t use tables. The actual design is nice, but lots of other sites also have nice designs.

What’s the deal? It seems like every blog I read has twenty links to your new design. I designed a pretty site without tables too. Why is it not all over blogdex?

I’m using MT in a similar fashion and it’s almost perfect. My only problem is that my clients usually don’t post “blog entries” but events, reviews, announcements etc. and I often have to creatively misuse the entry fields for a different purpose. It’s always hard to tell them that the “Location” has to go in the Excerpt field… I’m praying that there will be the possibility to rename/redefine the content elements for a blog in MTPro. ;)

I’m interessted to hear how you approach that problem.

For static pages (like perhaps an about, or a contact page), I design the page, remove footers and headers and common elements that I call from a main site configuration file (which I refer to in my original post). Then I save what is left as an index template.

After that, I extract all pieces of text from the index template and replace them with MTInclude tags to the template modules which will hold the text.

At the end, between the header/footer/repeating elements that are handled by the site config file and the template module includes for the text, the index template is pretty bare.

By doing this, the user can edit the text through the template module or change the whole look or layout of the site through the site configuration file!

Including static text on a dynamic page is a bit less heady. I link the template modules to a file on the file system, and then include the file in the appropriate place via PHP include().

Just to keep things neat, I usually put all of these fragments hierarchically below /includes/pub. For instance, if there is a piece of text on the essays page (at /essays/*) I’ll usually write the file to /includes/pub/essays/ and give it a good descriptive name.

So anyhow, that’s how I do it…

Thank you, Jay (and Doug), for all the info about this project!

What makes it secure is that if you are not logged into Movable Type, or have the auto-login cookie (by clicking Remember Me upon login), all you would get by clicking any edit link on the site or the gateway page is a Movable Type login screen.

These links are simply shortcuts. The security is still handled by MT.

I don’t understand what allows HIM to see the links, while others can not. Could you explain this?

Thanks, and I must say, you’ve both done a great job on Adaptive!

I’ll look into it. Thanks.

Here’s a screenshot, again of Jad Fair’s site, with the administrator links (all blue) turned on. The edit links below each entry go directly to the entry editing screen for that entry. The edit link on the sidebar links to the static content template module for that piece of content. The post new entry does precisely what you think it might. The control panel link goes to the gateway page and shows up on every single page of the site when the links are turned on.

Like the edit links I put into MT-Search (either adapted from an idea David Raynes had or vice versa), the links are cookie-based and can be turned on an off at will via the gateway page (see the toggle at the very top of the gateway page screen shot mentioned earlier).

Yeah, okay, so I probably forgot the coolest thing. Isn’t this always the way with top-of-the-head writing?

Jesper, indeed. I don’t know why I didn’t think of it when MT first came out, but it’s indispensible to me now.

By the way, in case it was unclear, I make a separate static content blog only to make matters simple. One could certainly set up static content template modules in each blog, but then you start to create problems.

1) If your text is somewhere above the Local Site Path, then you have to use absolute paths or ../ relative paths in the output filename. My static content blog has the document root as it’s local site path.

2) The static content blog is just that: STATIC. It doesn’t need to be rebuilt every time you add an entry. Keeping static content inside dynamic blogs (i.e. ones with entries) slows down the rebuilding process.

3) Dynamic blogs already have enough templates to deal with usually. There is no need for ten others with static content. That’s highly confusing for the client.