So, Totara is forking from Moodle and going their own way. What does this mean for Moodle?
The most interesting part of that article is the line that says: "it is their (Moodle) intent to clone recent Totara LMS versions to offer the market ‘Moodle for Workplace.’"
What is ‘Moodle for Workplace’? Is it a version of Moodle for the corporate sector? I like the sound of that!
I think that perhaps after some lengthy discussions about the "open source-ness" of Totara (See: https://moodle.org/mod/forum/discuss.php?d=245614 and https://moodle.org/mod/forum/discuss.php?d=171235 ) Richard Wyles has decided to give up the pretence of Totara being open source and to distance it as far as possible from free and open source Moodle by making Totara incompatible with Moodle and therefore Totara's clients more dependent. Does that make sense?
I get the impression that Wyles didn't appreciate that someone shared the Totara source code on Github, which was done at least once before Moodle HQ did it but quickly withdrawn. Moode HQ were simply following the letter and spirit of the GPL licence by sharing Totara's code publicly.
That some corporate users of Moodle would like a more corporate-like LMS is a legitimate request. I sometimes get requests to develop tightly "locked-down" courses where every step of the way is managed and controlled, everything learners do is measured and reported on, and social interaction between learners is disabled. If there's enough demand for these features, I'm sure they'll be implemented.
This is not really big news ... Totara had already more or less forked a while ago. The way their extensions are written make them impossible to merge back to Moodle core without rewriting them. There are a lot of core changes, making it hard even for them to track us over time. They kept their current code private (so it's not really open source) - same with their mobile app which also forked from our older one a while back. Contributions back to core have been minimal.
I had really hoped for some closer open collaboration on this between our teams this year but it looks not to be.
Thanks Martin. Are you able to say what ‘Moodle for Workplace’ is?
Not at this stage, no.
Contributions back to core have been minimal
I would like to offer a counterpoint to that because I don’t think it’s fair on the hard work our developers have put in to upstreaming and contributing changes over the last 5 years.
Here is a list of the top 20 contributor’s domains, calculated from the number of lines of code in current moodle/master using the email domain of the author, over the last 5 years :
- moodle.com 529125
- skodak.org 335213
- open.ac.uk 97749
- nicols.co.uk 59915
- moodle.org 39338
- gmail.com 28804
- netspot.com.au 17577
- totaralms.com 17420
- moodlerooms.com 13471
- vedrine.org 10903
- mondragon.edu 10790
- jamiep.org 8648
- danmarsden.com 8341
- grabs-edv.de 6284
- spvickers.freeserve.co.uk 4774
- massey.ac.nz 3707
- luns.net.uk 3622
- oakland.edu 3422
- goffstown.k12.nh.us 3399
- mouneyrac.com 3212
Totara’s contributions to Moodle have been on par or greater than any Moodle partner. If Learning Plans, Competencies and Report Builder all get integrated into Moodle you can increase that figure by another 100,000 lines.
I don’t think it’s fair to call that work "minimal”.
 While not a perfect metric I would argue it is reasonably representative (enough to make the point valid). Code used to calculate the numbers above is attached.
Will your contributions to Moodle continue now that you are forking the Totara code?
Crumbs! Look at the contributions of Google with gmail at number 6!
Simon, just of curiosity, does your calculation separate the work that folks like Petr Skoda and Sam Hemelryk did when they were at HQ from the work they did when they were with you at Totara? I see Petr's domain there in the stats so I assume his is accounted for there. Not a developer myself so I don't know the in's and outs of how git blame works...
Yes, they both switched to using totaralms.com emails when they joined us so it only counts their work for Totara against the totaralms.com domain.
Let's put those numbers into perspective:
So Petr Skoda's productivity has gone waaay down since he joined Totara. How long ago did he switch? Was it more or less than 2 1/2 years? As the numbers stand, that seems more like a loss than a gain.
And as a bar chart:
"Contributions back to core have been minimal." - Seems like a reasonable claim to me. Individuals make substantially larger contributions than Totara's staff combined.
This is a ridiculous argument! People contribute to the project whatever they want to and it is no way anyone's obligation to contribute. Unless, of course, they are working in HQ, then I guess they are paid to do their job because that's what Petr and Sam were doing - their job.
Matt, how much exactly have you contributed to Moodle code?
@Yuliya - None to core Moodle, but I've contributed a few plugin modules. Not that it changes anything about the point I'm making! ;)
@Joshua - Both Petr Skoda and Andy Nicols have contributed more than Totara but yes, they probably figured that their 10% financial contributions were paying for core development while they focused on their own customisations that they don't contribute back to the community in any useful way.
Joshua, I don't understand what "mark" you're talking of. I'm stating a simple fact and Martin's claim that Totara's contributions have been small is clear in my mind at this point. Totara seem to be more interested in making PR statements than in dealing openly and sincerely with the Moodle community. It maybe that more than anything that rubs people up the wrong way.
It's disappointing when anyone leaves the Moodle community though.
Looking at all these numbers and comments there are few things that strike me about this:
- It is clear that the vast majority of code in core Moodle is written by HQ employees (moodle.com + moodle.org + skodak.org + nichols.co.uk). Since Petr switched his email when he moved, I think its safe to count Petr's numbers in with HQ there.
- I am generally surprised by the low amount of code contributed back from all the partners. To a certain, extent this is perfectly understandable since the partner developers are spending much of their time working on partner specific code that doesn't get shared back. However, this coupled with the relatively small number of plugins in the database that seem to have a partner backing them, suggests to me a relatively small set of contributions back to the Moodle community as a whole. This is of course wholly separate from their significant monetary contributions back to core, which enables all that HQ code writing in the first place.
- It seems obvious to me that Totara is in line with the Partner numbers in this regard and certainly seem to stack up well in comparison to most of them... However, this is a bit of an odd comparison to make given how Totara is different from a Partner. Given the number of users that Totara claims to have on its website, it may be more apt to compare them to the OU. The OU is the only other end user facing contributor that isn't actually a Partner that is high on that list. That's not a really good comparison either though... Clearly Totara is a unique group in this list.
- Matt, I don't see how the data justifies your claim that individuals are more significant contributors to Moodle than Totara. We don't know how many people there are in the GMail line and that is likely to be a large number. The only individual that approaches the Totara team output in that list is Jean-Michel Vedrine, who has been going nuts lately working on the Lesson and the question import formats. Jamie Pratt is high on the list also but he frequently gets contracts from the OU for their projects. Dan Marsden is working for Catalyst.
- At the end of the day, this thread misses a crucial piece of the puzzle. Whether Totara has been an excellent model contributor to the Moodle project as a whole or not is somewhat irrelevant. They have been an contributor, and not an insignificant one either. An open source community's strongest point is always the collaboration and communication that it brings together. I realize that people aren't all just going to "get along." But when the community loses members, it weakens the community. I don't think anyone around here is saying that Moodle could do without those 17420 lines of code provided by Totara. Losing contributions from the team that wrote those lines at the end of the year is a loss. Regardless of where fault lies on this, having fewer developers working on Moodle hurts Moodle. And that is disappointing to me no matter how that happens. As a teacher and not a developer, I care about Moodle, the product I can use with my students. Those lines of code represent features and bug fixes. If we weaken the product then we're weakening the thing we all should be focused on.
Another thing to consider is that the contributions shown here (https://moodle.org/dev/contributions.php) are for core commits only. They don't take into account contributed plugins, or solutions submitted through comments in the tracker.
It is much more difficult now (but not impossible) for a non-HQ programmer to get their code into core. This is due to new processes introduced a few years back to help control quality for the Moodle codebase, and ensure that there were controls over what went into core. That made sense considering how Moodle had grown in importance to the industry.
If you look at some of the non-HQ people in that list, their last commits were several years ago, primarily due to the needed quality control processes put into place for Moodle.
But, core code contribution is only part of what is contributed to Moodle. For example, Simon Coggins from Totara, has been and is an active participant in the Moodle community, both in the forums and the tracker. If Totara's forking means we lose his participation, that is a loss.
@Mike I'm a non-HQ programmer who spends most of his time on contributed work. However, I do have recent core commits to my name and for the moment am still on the first page of contributors. I will admit that getting code into core is like scaling Everest in flip-flops because of the rigorous quality controls, but perhaps that is a good thing.
Thanks to MDLSITE-4154 contrib developers will shortly getting a badge .
And indeed code commits should not be the whole picture for contribution. Perhaps we should stop worrying about statistics and concentrate on improving and supporting the wonderful thing we have in our own unique ways.
Hi Gareth -
I will admit that getting code into core is like scaling Everest in flip-flops because of the rigorous quality controls, but perhaps that is a good thing.
It is absolutely a "good thing". I hope nobody thought I was saying otherwise. The processes that have been introduced have made Moodle way more reliable and secure. My point was that core commits are not a good measurement of contribution. So I think we are in agreement.
Perhaps we should stop worrying about statistics and concentrate on
improving and supporting the wonderful thing we have in our own unique
I don't think the quality control/process changes has only lost us former contributors.
I see many more cases of different people submitting patches. The process is now well documented and the same for everyone. That is in contrast to the old system where there were people with CVS commit access, who could easily get their changes in, and everyone else, for whom it was almost impossible to get your changes in.
Evey release, Dan Poltawski does a count of the number of different people who have contributed at least one fix, and that number is tending to go up.