I’m working on a Moodle project right now and I’m fighting it all the way. I have no control over the mark-up except the very top and bottom of the page. There are tables all over the place that just don’t need to be there but I can’t do anything about it. Can’t hack the core because of the security issues.

It is natural that the trade-offs made by the system you are most familiar with are the ones you understand best and are most comfortable with. It seems to be dangerously easy for the human brain to generalise from ‘the way I am familiar with’ to ‘the best way’. Humans seem to be much better at rationalising than being rational.

Another factor that drives systems to be designed in weird ways is that the platforms we develop on are not perfect. Moodle started off on PHP 3/4 and MySQL 3 and there are some weird design decisions that can be traced to that.

Similarly, Drupal started in the PHP 4 days and came up with its completely custom hooks system that basically re-implements OOP in PHP to avoid the problems with PHP 4 objects. If you were starting today, you would probably do more with plain PHP classes/interfaces rather than a custom hook system.

What’s my workflow for upgrading the security hole? Download Drupal, replace files, reupload Drupal. Run update.php — done. My code will still override at the proper place and it’s a much cleaner solution.

I admit Drupal has a much higher learning curve then anything else I’ve worked on but I’m not also able to enable those with no idea what Drupal or the traditional web languages powering it are. I have lowered the cost of entry for all of our employees much as Moodle has for those in the LMS world.

You’re post is somewhat valid — these are coder-snob issues. Both are valid solutions to the problems put forth. But which is more sustainable? The one where individual files have to be modified or where almost anything imaginable can be overridden. This is more in fitting with how cascading style sheets work as well as the functions and hooks cascade down in a logical order, one overriding / performing an operation before/after the previous one.

If Moodle had the planning put forth to do this, you don’t think there’d be more killer-app like functionality to it? Again, while I’m sure Moodle meets the needs of it’s end-users in teachers and instructors; gorgious, flexible, sustainable code could also produce the same usable end product. I just think Moodle could learn a lot from the way WordPress and Drupal have their stuff organized / managed is all.

Believe me… I want Moodle to succeed; I REALLY do, more then this initial rant / posting could probably indicate. I just can’t get behind it from a developers perspective.

Good follow up / discussion though! I love to see this topic keep going Disagreement / conflict leads to growth and reshaping of ideas.

On the subject of Drupal doing everything with hooks. Yes, it is an elegant architecture that allows everything to be overridden, but only once you understand what is going on.

I am a complete Drupal novice. I once wanted to tweak something simple in Drupal (I can’t remember what now), but I had to give up after a couple of hours when I could not even begin to work out which bit of code was doing the thing I wanted to change. (Sure, I could learn to hack Drupal by reading all the right docs, the point is that I did not have the time or inclination to follow a full tutorial.)

By comparison, suppose you want to tweak what happens in Moodle when viewing a forum. Well, go an look at the forum in your web browser, and look the URL. …/mod/forum/discuss.php?… That makes it totally transparent where to start reading the code.

Of course, once you have found the bit of code you want to change in Moodle, then you have to edit it in-place. You can’t override it from a separate module. That is a pity, but then git/svn/cvs can help with that.

i was also aware that there is an in process development of Drupal Moodle integration.
i really hope that will come along successfully for the benefit of both frameworks. after all, we should not waste energy in duplication code when it is already implemented elsewhere

Drupal can become a web platform or course platform or LMS platform. I would argue that Moodle can only become a LMS platform because that’s exactly what it was designed to be.

Part of the issue w/ throwing money at the issue is that open source kills proprietary solutions so I think it’s an unfair comparison . I hope Moodle matures more, I know lots of effort is in integrating Drupal with Moodle and I think the two could make a great relationship. Moodle for the LMS components, fed by Drupal with the CMS components. One does learning management really well, the other does content management (which MOST courses have a lot of content hopefully ).

I don’t want Drupal to become a full blown LMS; just stating it could

thanks

I was also shown an interface that referenced the files directly through a drop down and made it easy to modify parts of those files messages via a web interface. Again, modding files via the web, not such a good idea IMHO. Make them settings or a module to override things. Again, I’m biased now as there are fundamental aspects to the way Drupal organizes things that give it so much more power over other projects that I don’t know I could work on a project that doesn’t organize itself that way. If I could, I’d still be grumbling about conventions (let’s be serious, none of these are going to last “forever”).