Posted by chrissearle@dru... on January 1, 2009 at 2:38pm
A new (since september - am a little late) scouting site is now up. Drupal 6 based with a fair few contrib modules.
For those of you who can read Norwegian it's up at http://www.speidermuseet.no
English speakers - sorry - no english text there at present.
Comments
Nice Site.Can you point out how you did some of the features?
Nice site.
Google Language Tools does a half-way decent job of translating it to English.
Can you point out how you did some of the features on the site?
I see you're using the Tagadelic module on the home page, so you must be categorizing content with Taxonomy.
How are you handling the Newsletter signup on the homepage? Are you using Simplenews, or adding the collected eMail addresses manually to another list?
How are you doing the Flickr integration? Are you using the Flickr module ?
Are you using one of the google map modules that exist, or are you throwing some custom code in to a block/page region ?
Site build
Sure - let's see.
We're running drupal 6. The drupal install is managed with git with sites/all/modules and sites/all/themes using git submodules (see my blog for some rough notes).
Theme
This is Marinelli with one change. We remove all the images from the marinelli/img/banners directory and replace with our own set. All need to be 970x200.
Modules
It's a relatively simple build - throw the modules in and a very little config ;)
Categorisation
Yes - taxonomy in the base. We have one main taxonomy called tags - this is configured as free tagging and is used by tagadelic. There are a couple of fixed taxonomies used internally too (which association a post is related to, simplenews' taxonomy etc).
Newsletters
Plain old simplenews with mimemail this
Content types
Page for static content, blog for news. Only real addition was markdown for the editors (had issues with both tinymce and fckeditor). Views 2 for some simple lists (like the list of films - taxonomy based view, or the related content list when viewing a blog - a nice taxonomy based view that - find content with similar tags then exclude the current post).
Images
Image handling for posts is at the moment image and img_assist although we're testing imce and imagefield. However - since the editors need to be able to place 0 or more pictures at places where it applies to the content rather than at a particular weight (as a CCK field) imagefield doesn't seem like a good fit right now.
Photo archive on flickr is using the flickr module for blocks - we've not done too much in posts with this.
Videos are currently on vimeo - we just use the embed code in the content.
Map images - that's the gmap module running - both as a block and as input filter.
Comments
Comments are open - with only mollom protecting (really love mollom - fantastic system). Comment avatars are being provided by gravatar.
Sharing
We're using service links - it's simpler than sharethis which was slowing the page load down.
Internal
Navigate to assist administrators
XML sitemap for google et al
Couple of custom modules for selling the yearly magazine and membership - I'd like to migrate to ubercart (have done some testing) but I'll have to write a custom payment module for that to handle norwegian giro system - and I've not had time.
I always use the PEAR Wiki
I always use the PEAR Wiki Filter. I find the MediaWiki format far easier, and more robust, than Markdown.
FCKeditor used to me my default WYSIWYG editor, but then it started having problems with IE6. I've had great success with TinyMCE, however recently I've encountered problems with earlier versions of Safari on MacOS 10.4.
Nice call on the Service Links module. I've been disturbed how much people are using external content providers like ShareThis and AddThis. I had a conversation with a content producer recently who had no problem that a flash based twitter block would slow down a web-page load time to a minute or more.
I haven't seen the Gravatar module before, but it looks interesting.
I implemented Mollum on one site that was getting hit by comment spam. I am a big fan of reCAPTCHA though. If you can stomach the hit from loading content from a third-party site.
Ubercart may be my new best friend. I've deployed three online stores with Ubercart, and each time I do I love it more. It could be easier to integrate with other modules, Event and Signup for example, however, I usually stick some data in to the $_SESSION and then throw that in to a product attribute. After the order is placed I use Ubercart's hook_order to manipulate other data. It's brute-force, but effective.
Markdown
Chose markdown as I also use that in textmate on my own machine. However the other editors may be more wiki based - will have to check (and also see how many articles will need to be updated).
wysiwyg - issues with safari 3 and 4 beta together with img_assist - kept forcing all images to thumbnail irrespective.
Gravatar I really like. I use it widely and it's nice to hit a new site, comment and there is your gravatar. Users can have several e-mails registered and different gravatars per email so I have a scouting address with a scouting gravatar etc.
Mollom - I like the integration - and got really really good support from Dries with it in the early days (felt a bit odd that Drupal's lead was suddenly chatting to me).
It also did very well for me - see http://www.chrissearle.org/system/files/images/Mollom%20%7C%20Chris%20Se... - it did that without me even knowing the system was being harrassed - and without any slowdown.
I'll have to look into hook_order. I'd need to generate a check-summed number based on the order id. And the user and seller would have to be able to get to it. Will need to dig into this when I get time.
Event/signup - use those on other sites.