What is the current end of life policy for major point releases (is there one?), i've had a hunt around the Moodle website, but I can't find anything. There has been some discussion on the MOODLE-TECHNICAL-UK mailing list recently about when official support for Moodle 1.9 will end. Quite a few people are saying that they are unlikely to migrate to 2.x before mid/late 2012 (or even beyond), but there is some concern that support for 1.9 may not last that long.
Given that Moodle 1.8 was discontinued around the time 2.0 was released, would it then follow that 1.9 will be discontinued when 2.1 comes out in June 2011 (if the roadmap is correct), or will 1.9 be considered as a special case and be given extended long term support?
I don't have access to HQ secrets, and I'm sure someone more informed than me will reply, but if you look at the version history you will seen that there isn't a rigid pattern but that, recently, 1.7 was supported for nearly 2 years after 1.8 came out, and 1.8 was supported for nearly 3 years after 1.9 came out so I wouldn't expect 1.9 not to be supported less than a year after 2.0 came out.
I also cannot give an official answer.
However, the fact that 1.8 support ended around the time that 2.0 was released is largely coincidence. The policy is more that releases will be supported for N years after the initial release. Or is it N years after the next version is released? I am failing to discern any clear pattern in Moodle_version_history.
I think it also depends a bit on what is actually going on with each version. Some examples:
- By the time 1.8 stopped being supported, it was already clear from the download stats that almost everyone had moved to 1.9.
- New new roles system in 1.7 was a bit upheaval, and that was a pretty rough release. 1.6, by comparison was quite good. You will note that the last 1.6.x and 1.7.x releases were on the same date.
- Moodle 2.0 was another big upheaval, and 1.9 the release before that, was a really solid release. Therefore, expect 1.9 to be supported for some time still. In fact, I seem to remember Martin saying somewhere that 1.9 was going to get support for some time to come, but I can't remember where that was.
Actually Tim Williams was correct.
The official policy we've been using for ages (which admittedly may not be written down anywhere!) is that Moodle HQ commits to supporting the two most recent stable versions. That is, at any point in time (eg now):
- We work on the next version (2.1) but don't provide production support
- We support the most recent stable version with constant bug fixes (2.0.x)
- We support the previous recent stable version, but mostly only serious bugs and security issues (1.9.x)
- We don't support versions before that (1.8.x, 1.7.x, 1.6.x ...)
This would normally mean that we would stop supporting 1.9 when 2.1 comes out, but given that it's such a big jump to 2.x, and the release timing was bad for the north hemisphere, I plan to keep putting resources into supporting 1.9.x until December 2011. I doubt we can stretch beyond that, because it's getting harder and harder for our developers to remember how to write code for 1.9
Of course anyone else can jump up and offer to maintain 1.9.x after December if they like.
I know for a fact that there are large Moodle sites will be staying with Moodle 1.9 after that date.
As Northern Hemisphere sites often have a Summer update cycle (i.e June-July-ish), Summer 2011 is perceived as too soon to implement Moodle 2.0 so that puts it out to Summer 2012.Potentially 6 months after 1.9 support has ceased.
Many of these sites have considerable amounts of customisation and links with other systems and there's no possibility that everything can be rebuilt and tested before Summer 2011.
I agree what Howard wrote above. Our insitution (60 000 user accounts, 3800 course areas) was planning to move to Moodle 2.0 in Summer 2012, since next summer is too near, and an upgrade during the academic year would be unrealistic.
I really hope that 1.9 continues to receive security updates at least until the end of 2012.
I'm in the middle of managing the roll out of Moodles across British International Schools and we have had to go with 1.9 because the third party modules needed to support the curriculum aren't yet ported to 2.x and I didn't have money in the budget for us to do that ourselves. The rollout will be finished Summer 2011 and migration to Moodle 2.x is timetabled to take place Summer 2012.
I'm confident that we have the capacity in our company to support 1.9 (because I could foresee, given the history of support that Tim and Martin have both outlined, that this might happen) but I can see that many schools in the northern hemisphere won't be so lucky.
I would suggest that the support "lifecycle" (if that's the right way of putting it) needs to be long enough to accomodate both the southern and northern academic years. Given the popularity of Moodle these days and that this is a major change of functionality, I'm sure expiring Moodle 1.9 at the end of this calendar year is too early.
Hope this helps,
Please don't panic. Firstly, just because Moodle HQ doesn't put resources into support for 1.9 at some point in the future it doesn't mean others can't or won't. And secondly, if there are still significant quantities of 1.9 installations out there by December then I'll of course direct more of our efforts into supporting 1.9 for longer. I am probably being overly optimistic about the number of sites that will migrate by then, but we'll see. Perhaps our policy should be something like "we support any old branch that has more than 20% of the total registered sites".
On the other hand, people do need to start preparing for 2.0 and kicking the tires at least. We need to break the chicken/egg cycle where everyone is waiting for everyone else to do the work. If there are deficits or problems for your implementation case, please get involved and report them, vote for issues and write code for them. If the threat of a deadline helps this then I'm all for it.
For some context, at the moment we have over a thousand Moodle 2 sites registered on MOOCH, and there are 20 or so big active ones (including Universities) with thousands of users. And these are just the ones who've bothered to register (which in my experience at Moots is a low percentage). And let me remind you that you are using Moodle 2 on moodle.org right now too
Due to popular demand, I've put together an official page that lists how long we'll update Moodle releases with new versions, and have committed Moodle HQ to at least 31 June 2012 for Moodle 1.9.x.
I hope this assuages any fears anyone had (possibly having the misleading "End of Life" in the title of this thread wasn't helping much!)
PS: this page lists Moodle 2.1 for release on 31 June 2011, I think a correction may be needed.
Martin always has a boundless optimism when it comes to release dates. Always take his predictions with a pinch of salt.
(Or, are you confusing 2.0.1, a bugfix release of 2.0 released before December, with 2.1. the next major release?)
(Or is their 31 June = our 31 December, with all that season switching.)
Since we've switched to a date-focussed release plan, you can be more confident of the prediction dates.
We're only working on one or two new features at a time, and only integrating things that are already tested, so it's easier now to just push anything that we don't get to forward to future releases.
This chart is helpful. Thank you.
July 1, 2011 for 2.1?
Recently, I lead a group of over a dozen faculty testers at a health sciences university representing 7 schools through an open-source pilot evaluation of Moodle 1.99 for 10 weeks and then Sakai 2.7 for 10 weeks, based on 300 pedagogical variables. Moodle 1.99 hands down beat out Sakai 2.7 in nearly every category (quizzes, gradebook, content organization, etc.) vis-a-vis usability, stability, and functionality. We are currently on ANGEL but are now poised and eager to transition to Moodle starting this fall (2011). However, our faculty buy-in has been for 1.99 and its plug-ins with particularly affinity to its simple file structure, flanking block layout, and the version's overall stability. Our faculty have been astonished at how stable Moodle 1.99 was compared to ANGEL and WebCT Vista. Stability has been a major issue with vendor LMS’s in the past, especially when we were thrusted onto the latest versions due to discontinuation of support by Blackboard. One faculty literally commented: “I’m surprised at how stable Moodle [1.99] is compared to ANGEL and yet it still meets all of my online teaching/learning needs. In this day and age, a stable application is what our online instructors need.”
Our tentative implementation plan calls for a 1.99 implementation that would be enterprise-wide in May 2012, when community support will be discontinued. This will open up a pandora's box. Our faculty will be averse to transitioning from ANGEL to Moodle 1.99 and then to Moodle 2.0. In fact, this news will deplete faculty enthusiasm and buy-in. Furthermore, we would have to retrain faculty on use of new tools such as conditional release, quizzes, and wikis, soon after they have are just feeling comfortable mastering 1.99.
While I appreciate the augmented reporting and functionality in quizzes (as a health sciences university that needs question stats similar to USMLE boards) and the hubs, etc., our faculty might assert that before moving to Moodle, we should compare Sakai (OLE) 3.0 with Moodle 2.0 for to be fair. We are not interested in stanching our current momentum and conducting another open-source pilot evaluation but with a summer 2012 deadline looming over Moodle 1.99, we might not have another choice. Howevewe if Moodle 1.99 support was extended we could be on Moodle 1.99 until 2013 giving our faculty time to acclimate to the LMS platform change before ushering in 2.0.
Does anyone plan on supporting 1.99 outside the community? If not, what would you advise us to do if our faculty are enamored of1.99? How will this change affect the support policies for Moodle hosting/support vendors such as Moodleroom?
You are in a slightly difficult position.
Starting now, or in the next few months, it makes so much more sense to start with Moodle 2.0, or even 2.1 when it is released at the end of June, to avoid all the re-training issues. Starting with 1.9 is a dead-end.
However, stability is a key selling point for you, and since a lot of big changes were made to the code in 2.0, it is not as stable as 1.9 was. With each successive point release (2.0.1, 2.0.2, ...) it is getting more stable, but 1.9 was one of the most stable Moodle releases ever, and 2.0 has a long way to go to catch up in that regard.
I don't think the change in functionality is too big a worry. Moodle is still fundamentally Moodle, even if a lot bits of the UI have changed. I think you end up with much worse re-training needs if you do a full roll-out of Moodle 1.9 to everyone, and then upgrade later.
So, I think you are better of taking the short-term pain and rolling-out Moodle 2.1 from November.
I'm afraid you are just in the right place (good buy-in from you staff to change to Moodle) but at the wrong time. A year ago it would have been obvious to use 1.9. In a year's time it will be obvious to use 2.x.
By the way, it sounds like you have done a really good job of running the decision-making process in your institution (but then I would say that, you chose Moodle )
Hi Mavan, do not despair, while I agree with what Tim is suggesting, I think reality will kick in - for both the Moodle community and Users like your staff. In one way it is unfortunate you have come to Moodle so late, but I suspect there will be a lot of the current Moodle sites remaining at v1.9 for quite a while yet, and some form of support may be developing for it. You know there are always people who will refuse to move even when the volcano is in the process of blowing up.
The question is always support. I expect that a small group will get together and form a common interest group to maintain v1.9 for a year or two, perhaps even to 2015, but sooner or later, they will drop off. In the meantime, v2.x will improve and v1.9 courses will be converted easily and restoration of backups will be sorted, and all those issues will be remedied. The UI will become simpler and while it will still be a jolt for people to switch, they will do so more willingly when they realize that the opportunities for more diverse approaches, more imaginative learning structures, simpler and more stable grading and recording procedures and so on, in v2.x are much greater than they are in v1.9. So I expect you will be fine for a couple of years, at least.
One way you can help yourself, is when you set up your new Moodle, do not do anything fancy. Use only proven stable plugins, like Book, and Lightboxgallery, and core. Nothing overly fancy, or flashy. Keep two or three exact copies of your production site live and the downloaded codebase. You can work out what you need to do with it if you have something on your production site not working properly that works in your supporting site. The forums will still be here and you will find there is a lot of documentation and other materials around that will keep you going.
The hard part is still going to begetting your staff to move when it is both obvious and necessary, but that is human nature.
It would be really useful for future releases to have some kind end of life roadmap. I think this mattered less for some of the earlier releases, since 1.8>1.9 was rather less painful than 1.9>2.0, but I think it would be a good thing if Moodle users can plan their upgrades with more certainty in the future. Redhat publish such a roadmap for RHEL releases on:
this is probably more than is needed for Moodle, but can serve as inspiration.
On a purely personal point of view I had hoped for 1.9 support running to the end of 2012. We are not tied to the academic year for our upgrades, the Moodle systems we host tend to be set up for outreach projects, with a lifespan of 3-5 years. A major system upgrade is generally not possible for budgetary reasons, so knowing the point at which we will have to start 'self supporting' our chosen Moodle version would be very useful. I fully expect to still be running some 'self supported' 1.9 based systems in 2013.
Well, in my opinion the support policy is completely understandable, and supporting 1.9 till December 2011 is surely a great effort. On the counter part, it is also true the situation of many institutions that needs a longer timeframe before upgrading to 2.0
Wondering if it could be possible to keep 1.9 support longer but limiting the effort only for security issues ?
Moodle 1.9.x seems already to have enjoyed a long run, and I would suggest that further development work is unnecessary but would not the occasional bug need to be looked at? Considering the current position of Moodle 2.0.x and the projected release of Moodle 2.1 and further versions, it is unlikely that many organizations will have sufficient time to implement changeover policies. I specifically use "changeover" here, as it is easy to look at Moodle 2.0 as a new product, rather than an upgrade. I suspect many organizations will be reluctant to change, applying the older logic of "if it ain't broke.." rather than embracing a much better product.
I can understand the desire to move on, and wanting to avoid clashes with the complexities of running two very different code bases. In the end, it is not going to be Moodle.org that will dictate when people change over, they will do so as their own business needs force, or allow, or suit, them, to change. These organizations can be encouraged to change, but cannot be coerced. That should be a major consideration, should it not?
As long as it does not copromise Moodle.org's proposed release schedules, support for security at least could continue until the end of the northern summer. As long as everyone was very clear that this is not going to change - they will accept it. A lot of large and medium-sized organizations will not feel they have been left in the lurch, where otherwise, they may feel so. I know my own school Moodle provider is talking about a 2012 changeover, but probably January as would most Australian schools, I suspect. The northern hemisphere has summer at the wrong time of year so they are always catching up.
Well, there have been a few occasions in the past (sadly too few) when someone other than HQ took over responsibility for a branch. The example I am thinking of was that for a while Catalyst IT took responsibility for maintaining the 1.5 stable branch.
So, this is open source, it just needs one (or more) volunteers to step up. I am sure Martin / HQ would love to cooperate with any credible volunteer.
Let us consider what it means to maintain a branch:
- Triage all bugs coming into the tracker with an affects version of 1.9.x as they come into the tracker. As Colin says, 1.9 has been around for a long time. It really only needs a few fixes for critical bugs now.
- Work out the correct fixes for any bugs where no-one else offers a patch.
- Review all patches, whether from the maintenance team, or from the community, and integrate or reject them once-per-week.
- On occasion, declare that such-and-such a weekly is going to be the next 1.9.x release. Ensure that all the right packages are built, write the release note and security advisories, etc.
I think that is all the key points. The details of most of these processes are in the developer docs on the wiki.
That said, I think that continued official support till at least mid 2012 would actually be in the best interests of the future of Moodle.
I thought that's been the situation for some time now??
A few points:
1) I suspect if heavy 1.9 usage continues past the end of the year, they'll probably keep supporting it (for security updates) anyway.
2) Just my suggestion: Other than that, if some institutions need 1.9 support for longer, perhaps Moodle HQ should look into offering paid '1.9 security patch' deals. This is similar to Sun/Oracle's support for old Java versions: there are new security updates for old Java versions, but you have to pay to get them (while all updates for current versions are of course free). So companies that are slow to upgrade can themselves fund the continued implementation of security patches in old versions.
If GPL makes this an impossibility in the 'closed' kind of way mentioned, it would still be possible to fund this via a group approach; find a number of institutions that want 1.9 security support, and get them all to club together to pay for it.
3) Even if Martin prefers not to organise this, it would still be entirely possible for other people to work on security patches - and again, you might be able to club together to reduce costs - by contracting a Moodle Partner developer to monitor security threats [from the various security mailing lists etc, plus check any fixes applied to M2 to see if they're relevant] and code up fixes. You can do this right now, if you really really want 1.6 security fixes for instance...
PS We will still be running our 1.9 system until approx end 2014, alongside our 2.0 system which will launch this year. However, our system is behind an additional security layer which prevents the 'general public' from launching attacks (only authorised users can access it), and most parts of it will actually be set to read-only, so we can probably afford to be a bit more relaxed about most potential threats.
I personally think that 1.9.x will be still usable for several years. All the problems are well known, if people want to use it they can help with maintenance. The core team is working on the current and future versions, please note that majority of HQ developers did not work on the 1.9.0 release at all.
Some known 1.9.x problems:
* htmleditor - it is relatively tricky to switch to tinymce or ckeditor, core developers are not going to work on that, because we need to fix 2.0 issues and improve the integration in 2.1
* KSES text cleaning has potential security issues - blacklisting is notoriously problematic when browsers include new functionality such as HTML 5, it would be great if somebody could help improve/test experimental htmlpurifier integration
* course category sorting - it runs out of integers on 32 bit platforms on large installs, the code there needs some rewrite, unfortunately the changes in 2.0 are PHP5 only so it can not be reused
Hopefully git will help a lot with both maintenance and contributor cooperation. Let's not forget Moodle is opensource, if you really want 1.9.x you can maintain it forever yourself!
A parable: A mayor wants to stick a facility on public ground so he can show something for his tenure. The facility really needs parking and new parking would require trees to be cut down, which the public won't stand for. So the Mayor builds the facility without parking, knowing that once the public starts using the facility they will demand that the trees be cut down.....
Providing a relatively simple way to move to tinyMCE from htmlarea in Moodle 1.9.x would be hailed by hundreds as a HUGE deal. But by pushing out Moodle 2, development on software that is stable and will be in place for years to come will be eschewed. Yes, because devs want to work on Moodle 2, but also becaue devs don't want anyone working on Moodle 1.9? Does it really make a difference if the software one develops, which will arguably have a very short life anyway, is used from 2011 to 2013 as opposed to being used from 2013 to 2015?
Petr's remarks remind us that Moodle is NOT like LIbreOffice.....
I should not better get involved in these flamewars again - I need to get back to bug fixing because that is my preferred way to get Moodle better.
We've used TinyMCE 3.x in production for the Moodle 1.9 branch for a year-and-and-half and it hasn't been that difficult. Its available in the CLAMP LAE version of Moodle and while the LAE has numerous other changes, they're documented and easily removed, if not desired by the institution. If I had to go with 1.9 using HTMLarea, I'd be considering 2.0.
Sorry to veer off topic, but since it came up here -- how does one encourage a reluctant Moodle host to employ TinyMCE 3.x?
Your thoughts are much appreciated.
Why would a "reluctant Moodle host" employ anything that is the least bit outside of core, let alone do much in the way of supporting Moodle 1.9? I think the answer you have found, though one that you perhaps weren't expecting, amounts to a simple redistribution of wealth Specie seems your best bet. Keep us posted ;=}
Said Moodle host was reluctant to support TinyMCE 3.x before 2.0 came out. Said doing so was problematic, but others seem to be doing so successfully.
Not understanding your meaning here: "a simple redistribution of wealth Specie seems your best bet"
Don't worry about Marc - he may have read too many books at some point (sorry Marc )
Thanks for the explanation Howard.
Howard pegged it, I read too much
But even you Howard have to acknowledge that the response by the MP here should have been, "We will do what you ask on the following conditions... because the work you are asking us to do will have the following consequences....", since it is a manageable hack, as opposed to, "NO!"
btw, don't you know that the book I just finished reading is entitled, "Heresy"?
TinyMCE is not your basic core html editor in 1.9. Why should any webbhost go out of their way to support something in Moodle that is not core when it will invariably bite them in the end? Redsitribuition of wealth? You give them bags of cash to deal with the hassles. Specie? SIlver and Gold... just being cute....
It can be done... but what on earth would pursuade your web host to take this on? As mentioned, CLAMP may be a better solution for you. They seem to be a sober group that are not easily stampeded into drinking any koalade....
We pay a Moodle Partner to host our Moodle site. They have rules about which modules we may add and so forth. We want to be able to use Safari with Moodle. The basic core HTML editor in 1.9 doesn't allow that. So, after reading that many have had success with TinyMCE, I asked our Moodle host about TinyMCE, as I've described and they refused. If others have had success with TinyMCE, why do you feel that it will, as you say, "invariably bite them in the end"? Initially our host refused to install NanoGong, but then sometime later came round. No biting there that I'm aware of.
I can certainly ask about CLAMP.
Ahhhhhh....... you pay a Moodle Partner and it is the Moodle Partner that is not willing to provide the modifications you wish. Such a strange response from an organzaition that should have no isues whatsoever ith making the changes your request - lol.....
SO, if you have asked a Moodle Partner to make such changes, and your hosting agreement and SLA provide that they will make such changes upon your order (and of course per the charges previously agreed upon) then you seem to have confirmed that it is arguably an inexcusable waste of time to host with a Moodle Partner.
If your agreement with them does not provide for the kinds of adjustments you are seeking, and I would surmise that based upon your very cursory comment about the limitations they impose, then you have purchased services that are specifically engineered to get money from you with very little in return. In that case, I suggest you consider renegotiating your arrangements as I rather doubt that your Moodle Partner provides much of anything for free.
I rather suspect that your Moodle Partner will dress up in a grass skirt and dance in the street if you offer them anough money, so I am guessing that you are on some budget package expecting your MP to provide you premium service at an economy price point. Probably not a reasonable expectation.
Anything beyond core makes Moodle upgrades problematic. Hence, if you hve some plan where someone else is supposedly keeping your Moodle current, it makes no sense for them to provide anything that will make more work for them. Seting up tinyMCE for Moodle 1.9 is just such an exercise. Nanogong is much simpler to manage, though your webhost would probably balk if you asked for asciimathml as well, as their installation often overwrites the others modifications ;=} Additionally, it is one thing to add an option for everyone getting served, quite another to set something up for just one client, which has to be addressed separately for that client eveytime nd minor upgrade may be uploaded.
And of course you can use safari with moodle, just not htmlarea ;=} But why use safari? Even the most hard-headed folk at my local district which is often as backwards as they get have finally decided to use Firefox on Mac... Not to say TinyMCE isn't great (though it takes a bit longer to display than htmlarea)-I can't wait to roll out the math tols that we will have available with TinyMCE.
As you can see, though, from the discussion, though folk will b e using 1.9 for years to come, HQ has decided not to move 1.9 to tinyMCE; a real pity too as that would be so welcome.
Image of Moodle host in grass skirt will surely prove useful at some point. Thank you.
We do use Firefox on the Mac, and of course it is fine. But our iPad initiative poses real problems--Safari is the only option there. I write pleading notes to our Apple rep (who assures me Safari developers are aware of the issue) and sometimes I post queries in these Moodle forums wondering why we can't all be friends.
In the meantime, a weird but happy development since our recent (free) upgrade to 1.9.10+ is that there is now a not quite correctly displaying, but still seemingly functional, WYSIWYG editor in Safari. More testing needed before I do happy dance.
They may be playingwith this hack:
http://www.lewiscarr.co.uk/content/enable-moodle-wysiwyg-editor-chrome-and-safari (quick patch could even be kludged for themes) which has been floating about for a while http://moodle.org/mod/data/view.php?d=13&rid=3945 (you didn't say initially why you wanted tinyMCE - lol.....
Or they may have gone to FCKEditor. I would be surprised if they went to tinyMCE, although who knows, maybe they were embarrased by your post ;=}
Of course an iPad initiative sound rather elitist - lol - bet you and your girls read books too
CLAMP schools (North American) are in exactly the place mentioned by so many in this discussion -- cannot reasonably consider getting 2.x out the door during Summer, 2011 so must wait until Summer, 2012.
IMO, 1.9 is only getting nominal bug fixes already. I may have missed some actual "development" in this branch and remain open for someone to correct my above assertion. I would also respectfully submit that 2.0 isn't actually a "stable" branch -- though admittedly I define stable as "suitable for production use for an entire school using 90% of its functionality".
Another chime in that we expect to be on Moodle 1.9.x until summer 2012, and sure would appreciate at least the guarantee of security updates.
One logistical issue which might help speed up that transition (for us poor Northern hemisphere-ers) would be shifting the release date for Moodle 2.1 to June 1, 2011, instead of June 30. With the current release date, I have no confidence we'll be able to install/upgrade, test, fix, document, and train sufficiently before faculty members need to be building their courses in August. With an extra month - even if it means sacrificing some bug patches - I can at least contemplate a Fall 2011 pilot of Moodle 2, which gives us more options for Spring 2012.
I know a month of work doesn't appear out of whole cloth like that... I keep asking the Physics department how that time machine project is coming, and they keep saying they're not quite there yet. But it would be a major help to the Northern academic schedule to have a stable release at the start of summer instead of midway through it.
Yes, I think Joe has hit on the key point here.
No-one expects (or should expect) any new functionality to be introduced to 1.9.x
However, I agree with others here that important security updates should be a policy until end of 2012.
We're using 2.0 on two live sites now - and it's not been without challenges, and some continue. I think for the many 'self hosting' Moodle users around the world it's not a great idea to hurry them along to a version that is not settled (and as Shane mentions the re-training aspect is a major consideration that hasn't really been such a huge factor in previous releases).
I don't think the whole Northern Hemisphere / Southern Hemisphere release date is important - it's going to affect someone somewhere equally wherever the release date is made. In fact, I think the release dates have previously favoured NH (partly due to the UK OU involvement and contributions). Nothing wrong with that, I just think it's unhelpful to try and use this as a main criteria for release dates. A clear roadmap help everyone planning, regardless of location
Sorry to butt in again, but the main questions have to be obvious by now. How can support for v1.9.x be dropped and maintain Moodle's credibility? However, can Moodle.org maintain the expense of doing so? Should it be expected to? Marc has pointed out that Moodle 2.0 seems more like a beta than a production release and the forums seem to support that claim. As it stands, it appears that there are very few production sites using v2.0, once people realise that upgrading is fraught with difficulties as their custom modules are just not sliding in nicely or their favourite plugins are wrecking their upgrades. New installs are not allowing them to restore their v1.9 courses. How many questions relating to these two issues alone are the forums fielding?
Marc has also reminded us about the gap between release and implementation of Moodle 2.x and misses what would have to be the critical point. People, administrators, management, will not change until they are comforable with the new product. 2 years may be a bit extreme, but it is certainly foreseeable. While the technology is changing fairly rapidly, the learning curves of new products seem to be getting steeper, (but that just might be me, I am not sure). I am not sure if there is a viable answer to this, but even if it is a minimal team and updating monthly rather that weekly, that would, at least, be something - irrespective of Moodle.org's usual practice of weekly updates. Unless something happens which requires an urgent response, regular monthly security updates are better then we get from most big companies.
I would expect many Moodle Partners (us included) will still have a few sites running 1.9 after Dec 2011 and it will be in our interest to maintain a stable 1.9 version (and fix any Security issues that might arise)
Even if HQ drops "official" support - we will still be able to submit patches for any security issues and have them included in the normal weekly integration cycle.
So - the push will be to move to Moodle 2.0 (which will be even more awesome by Dec 2011!) but if people are still running 1.9 you can expect the community(especially some of the Moodle Partners) to continue maintaing 1.9 (especially with Security patches) for some time.
If by dropping official support it helps encourage others to upgrade - I think this is a good thing! - especially by Dec when we would have ironed out a lot of different things!
I wonder if the shift in approach to more frequent releases calls for an adjustment of the support policy overall, but most especially with regard to the move to 2.x. We like many others who have chimed in are planning to switch in (northern hemisphere) summer 2012.
Isn't the problem here (assuming for the sake of argument that Moodle 2 represents improvements across the board) that its design precludes easy transition? Frankly, the "crisis" rests solely upon the suggestion that Moodle 2.x is "the most recent stable version" a proposition that may be arguable if one could consider Moodle 2 code in a vacuum, but is dubious in practice because it is virtually impossible to move a complex production Moodle from 1.9.x to 2.x.
Personally, I can;t wait to see what Moodle 2 looks like when it gets out of beta
Martin's confirmed that there will already be extended long term support (otherwise 1.9 support would be dropped as soon as 2.1 comes out). It just might not be quite as extended or long term as you want yet.
But, like I said above - I think the power here is probably in users' hands. Look at the graph on the bottom of the stats page. The registration data has a tendency to get out of date and include test sites that were never used and stuff like that, so a certain degree of older versions can be ignored, but if by the end of the year it continues to look like 1.9 Pac-Man is eating all the other versions (with 2.0 not even being as big as the older versions)... is Martin really going to drop security fixes for 1.9? Well, it's up to him, but I doubt it...
I thought I would throw in my 2 cents. Not long ago I post a simple question, "How many of you have upgraded to 2.0x for your production site?" I received no responses to this question.
I upgraded one of my production sites to 2.0 and later that afternoon I went back to 1.9.10, which is a very stable version. The bugs I ran into I did not expect when upgradeding to 2.0. My personal feelings are that for the number of normal users using 1.9x you would think that there would have been alot more thought given to migrating all of that data from 1.9 to 2.0.
Also I feel the way or the lack of ftp ability into courses need some work, as well as the media componets.
I think I will approach 2.0 like I approach Microsoft releases, waiting until all the beta bugs are worked out then upgrade. I figure I will upgrade my sites in a year or so, maybe by then the kinks will be worked out.
Indeed, William, but folk do not want to talk about whether Moodle 2 should have been released (hence the circumlocuton, Howard, regarding the great ape.)
Perhaps one should simply stick with calling 1.9 "mature"? I don't think anyone is disagreeing here about core points: a) 1.9 IS stable, b) 1.9 hosts an extensive set of modules that are not yet supported by 2.x, c) until there are stable tools with which to migrate courses to moodle 2.0, a move to Moodle 2 must be site-wide or tediously address upgrading one course at a gtime, d) there are still upgrade issues with 2.x as well as bugs, and we have seen some bugs introduced post 2.0 (as well as bugs removed), e) any move to 2.x would require very extensive planning, testing and training of support staff, intructors and students.
Given the above and the assumption that IT departments NEVER have the resources they would wish to have to engage in such a migration, the process is going to require a number of parallel installations potentially requiring significant additional tangible assets as well as substantial manpower from departments already arguing that their staff is maxed out.
I remember the Win2K deployment planners promoted by Microsoft, lol, though I Do NOT remember any win2k migration that went smoothly - though there was a good deal of black humor about forests and trees ;=} I don't think that Moodle is served well by emulating the Redmond model, either in the GUI (where making things more Windows-like is NOT an improvement) nor in the life/development cycle.
I believe I predicted that we wold see Moodle 2 on majorproduction sites two years from its release..... I am sure Martin scheduled things to make me look prescient - lol.
I think that your assertion that it is "virtually impossible" to move a complex site to 1.9.x to 2.x is neither true nor particularly helpful. It is perfectly possible given sufficient time. And that's the crux of the matter. Long term and larger users will be relying on customisations, integrations, themes and existing plugins all of which will (unless they are lucky and they have been converted) require rework and testing. This is likely to be a non-trivial task requiring significant amounts of that.
What nobody has yet mentioned is that, particularly in larger institutions, an amount of re-training will be required for staff users. For large universities this could literally run into thousands of users - many of which will display a certain cussedness (with which many of us will be all too familiar) at the thought of any change.
I don't want to overplay this but for many the transition from 1.9 to 2.0 is a significant task requiring medium to long term planning and execution. I would not like to be "under pressure" because of premature termination of 1.9 support.
I agree with your assessment - there is more to upgrading than just code/database changes. We have upgraded many sites to moodle 2.0, most of which have been relatively straight forward. Usually the problems come from customisations or third-party plugins.
However, from our experience so far, training or retraining of staff (technical support, course/instructional design etc) is usually the major factor in organisational planning for an upgrade.
As Tim mentioned, there is an opportunity here for someone to step up and be a maintainer of the stable branch and he has outlined the responsibilities very well. Perhaps something the Moodle Partners could fund/support?
+1 for Shane's point about staff training for 2.0 software being a big consideration in organisation planning before upgrade.
Though i completely agree with moodle should not stop supporting 1.9 support having a large user base for that version,
I would be very happy to maintain 1.9.x branches just incase.
There seems to be a typo on this page:
Which says "Support for 2.0.x will end June 2012 "
That should read "Support for 1.9.x will end June 2012", shouldn't it?
Yes it should. I changed it. (I'm sure in general it's frowned upon for people other than Martin etc to edit this page but in this case, at least the 2.0.x was obviously wrong and should be 1.9.x.)