Monday, September 26, 2011

Are you living in a bubble?

I've been trying to extract a domain from Namespro, a registrar. After (I hope) springing it from them, I gave them a little honest feedback. I'll provide their answer here verbatim, enough said. Okay, almost enough: I want to note that, considering the spelling mistakes, the support person who wrote this probably believed it. And for reference: my favourite registrar (hover.com) provides users the ability to generate auth codes themselves, not to have to search badly worded help systems to find the right invocation. In this case, it required a request via a hidden form, and a simple email request was ignored. After the invocation, they came back with an "are you really, really sure" reply to the ticket, which it took me a few days to notice as well. They provide no phone number to actually talk to a support person. In other words, unless you really know what you're doing and have a lot of patience, the chance of getting your domain out of Namespro is pretty slim.

Me:

If you really want an honest answer to your question: it's because of your policy of making it so hard to transfer anything our of your system. You appear to deliberately obscure the process of moving services out of your clutches. I'm not interested in doing business with a company that I don't trust to be upfront and honest.

Them:

Thank you for your reply and comment. The process of getting an Auth Code is not made public is because very few Namespro.ca users transfer out from Namespro.ca to another registrar to being with.

That being said, we have forwarded our suggestion to our management team for further review.

CiviCRM multilingual and customization: nice progress!

Today we've launched the 2012 French for the Future National Essay Contest. The full site is someone else's, but the contest submission pages are built using CiviCRM.

The essay submission pages are actually implemented as event registration pages of CiviEvent, which isn't exactly what CiviEvent is built for, but was close enough and provides some nice functionality that other alternatives (e.g. a simple CiviCRM profile) wouldn't provide.

Before starting, my biggest fears were around the custom presentation bits. My first CiviCRM implementation was back in 2006, and ever since then, my standard wisdom for anyone using CiviCRM is that CiviCRM is great as a CRM, but less great in exposing itself on your website. Partly, that's because CiviCRM is CMS agnostic, so it can't use all the great tools that I'm accustomed to with Drupal (notably the forms api), but also because of the whole Smarty/QuickForm architecture, which was okay for the day, but now way behind the competition.

On the other hand, CiviCRM has put a lot of work into improving this aspect over the past few years. My personal favourite trick, ever since learning jQuery and having that available for CiviCRM pages, is to use it via the theme layer or a custom module for almost any cosmetic changes. It's cheap, easy, safe and powerful. But it won't do functionality changes - for that, the new hooks are a nice complement.

Another great example - the new ability to add multiple profiles to event signups. Breaking up the fields into groups (which have to correspond to separate profiles), is both sensible from a useability point of view, as well as backend organization.

What I hadn't experimented with so much was the multilingual option, which until recently (i.e. mysql 5.1) required very unsafe mysql permissions for the site. But after one earlier success, I jumped in and have been surprised all the way along - the functionality has been perfect, and the administrative tools more than adequate. Particular thanks to DaveD for pushing forward the new "Word Replacement" feature (but no points for changing the name from string replacement and hiding it in the options section, I almost didn't find it!).

So my report from this project is: awesome, and thanks. I had a highly detailed-oriented client, and could satisfy almost all her requests.

CiviCRM also was able to import previous year's data, and next we'll be setting up donation pages and testing CiviMail.

Thursday, September 08, 2011

Really Getting into Drupal 7 .. upgrade from Drupal 5

Drupal 7 came out in January of this year, but I've only been dipping my toes before now. I ordered the new version of Pro Drupal Development and scanned for the interesting bits, and have created four new relatively simple D7 sites so far, even including some quick themeing.

But as of yesterday, I've finally started to climb the D7 learning curve, with a long delayed project to upgrade The Big Guide to Living and Working Overseas.

Working Overseas is my longest running project - I started it with a custom CMS while working at a previous employer. After leaving to work on my own (that's another story) and turning entirely to Drupal, Jean-Marc found me and I've been working with him ever since. Early on, I insisted he let me convert the site to Drupal and though he wasn't convinced at the time, he's been continually delighted ever since. He appreciates all of Drupal's many fine qualities and particularly its flexibility for all the custom stuff he wants, but even more - using Drupal is a sales bonus for him since he sells to universities and they all use Drupal.

So his site has now been ticking along since October 2007 in Drupal 5, undergoing a new theme in 2009, and continual new features implemented with a grab-bag of contributed and custom modules. In other words, a perfect nightmare for an upgrade.

But being optimistic and hopeful like Jack Layton, I brashly suggested we skip D6 and upgrade straight to D7. My thinking was that about half the process of an upgrade is the fiddly bits that you have to chase down manually, so it would only be 50% more work to upgrade to D7 than D6, roughly. Ha ha, you're thinking....

But there's more -- I also put him in contact with Thomas Cheng who I'd worked with as a designer, and he was convinced to simplify a bunch of his site and features, and even to do a redesign. So the D7 upgrade is a good idea because it provides all kinds of front end goodness like useability and accessibility bits and HTML5 support and better themeing tools. And since I don't have to upgrade all the customizations, it's not so bad, I'm thinking.

So, how does the story end? Stay tuned.