Exciting times in the world of WordPress and WordPress MU. Last weekend’s announcement by Matt that WordPress MU would merge into WordPress caused a flurry of activity and questions on twitter and on blogs, most notably with speculation that WordPress.org would run on MU and by jeffr0 who asked me on IRC what was happening.
Basically, the thin layer of code that allows WordPress MU to host multiple WordPress blogs will be merged into WordPress. I expect the WordPress MU project itself will come to an end because it won’t be needed any more (which saddens me), but on the other hand many more people will be working on that very same MU code which means more features and more bugfixes and faster too. It also means no more marathon code merging sessions. I certainly won’t miss that.
Meanwhile in the real world, there’s more merging to be done. WordPress 2.8 is expected next Wednesday and it has introduced fancy new stuff I haven’t finished fixing yet in WordPress MU. Expect an MU 2.8 beta sometime next week I hope.
In what I first thought was fabulous news, James Farmer has announced that WPMU DEV Premium has been relaunched. The site offered premium support for WordPress MU for a very long time. It also sold proprietary plugins which I’ve never agreed with (because of the conflict with WordPress) but now all plugins are GPL licensed.
Then I found out that you need to signup and pay a subscription fee to download them. I’m conflicted about it, because if I’m honest, while they’re sticking to the letter of the GPL, the spirit may be lacking.
So, should you signup there for a month, download all their plugins and upload them to WordPress.org? It’s tempting isn’t it? But no, you shouldn’t. This is real income for James, Andrew and company. If their plugins are uploaded elsewhere will they be updated? Will you signup for another month and grab them all again and upload each and every one to separate Subversion repositories? Will you provide support when things go wrong? I didn’t think so.
If it really bothers you that GPLed plugins are not available “free as in beer” then write your own and support it. It’s not something to be done lightly.
When we speak of free software, we are referring to freedom, not price. Our General Public Licenses are designed to make sure that you have the freedom to distribute copies of free software (and charge for this service if you wish), that you receive source code or can get it if you want it, that you can change the software or use pieces of it in new free programs; and that you know you can do these things.
Of course, WPMU DEV aren’t the only MU support people in town. Check out Ron & Andrea’s musupport.net and of course I recommend the Automattic Support Network where you’ll find me and the rest of Automattic.
64 thoughts on “WordPress MU Catchup: big merge, wpmudev goes gpl and MU support”
This is good news! I sicerely hope you continue to maintain the MU-parts of new WP w/MU included. You write clean readable code, and that I really like. I gather this is a milestone for 2.9, or is it even further up the road?
Good luck porting 2.8 to wpmu! I will be testing and give feedback.
BjÃ¸rge – thank you. No milestones set yet. No idea when this will happen.
Hey, thanks so much for the mention. 🙂
On the merge, I think it’s a great idea and it will be interesting to see how it plays out.
#wordpress: Donncha: WordPress MU Catchup: big merge, wpmudev goes gpl and MU support: Exciting times.. http://tinyurl.com/q6ttsm
I had to read that twice to try to get your argument regarding GPL plugins. Still not 100% sure I get it. So I thought I’d pop over to Brian Gardners site, since he was endorsed by Matt and had a nice link over on the WordPress themes page, to see how he did it.
Oh look, you have to buy a support package before you can get access to the theme. Seems pretty similar to me.
Yeah, it’s exactly the same. As I inferred above, I don’t like it, but it’s not illegal.
Anyone could (and they do from what I’ve heard) distribute Brian Gardner’s themes without fear of being hounded by his lawyers. Still, people go back to Brian’s site and pay good money for the themes as they must think they’re getting value for money.
Thanks for clearing that up – makes things a lot simpler for me.
Thanks for the mention 🙂 If it does turn out that you won’t be working on the codebase I will miss that as well. I’ve done alot of work with the part of MU that is the MU. I like your programming style.
I’m perfectly okay with for pay themes, plugins, etc. as long as the folks buying feels they are getting good value for their money.
I can see both sides, and while I personally prefer to just distribute everything for free and ask for donations (which we both know are few and far between), I appreciate that they at least have it GPLed. Makes everything a lot easier for the users, easier for the writers legally, and hopefully they’ll still get paid since it’s something of a walled garden (albeit a very low, easily-accessible wall).
What will your involvement be with MU after the merge? Will Automattic begin throwing vast sums of money your way for consultation and programming?
Hehe, I’m already employed by Automattic. I don’t know what’s going to happen after but I will be involved still.
Wow! Donncha, thank you so much for your hard work on WPMU all these years! It’s going to be interesting to see how this pans out.
Thanks for clearing that up indeed! One more thing: we’re looking at merging various WP.org blogs into a WPMU instance. Is this now pointless? Would you recommend waiting?
I can’t tell you how exiting news of the WordPress/Wordpress MU merge is for me. Just in the last couple of weeks I’ve realized how valuable MU would be for my project, but as I began looking at it, I could see another huge learning curve for me. Since I’m in no hurry to set this up, I think I’ll just sit tight and wait for WordPress to work it’s magic–and make my life a whole lot easier! Perfect timing (in my little world) for this announcement!
Let me understand this right, are we expecting wordpress to become wpmu but remain as wordpress that will be working with buddypress. Should I wait til they merge them together as a single version to start out with the buddypress. Just being concerned about current wp to wpmu then new merged version.
Will it impact the current path setup like blog.dir to current format with wp standalone?
Since there’s no date set for the merge you’re probably better off setting up and running your sites now rather than waiting. There are many MU sites out there so a migration/upgrade is something we’ll have to take very seriously. In all likely hood it’ll probably be just like doing an ordinary upgrade!
Personally I don’t like WPMU approach, because it created new set of tabled in DB for each blog. I would like to see extra column ‘blog_id’ (or something like this) added to tables. Therefore I wonder how WPMU will be merged – if current solution will be taken as-is, or everything will be refactored to use extra column to distinguish data belonging to different blogs.
In general it is a good idea to do this.
We have scaled the WPMU approach to 7 million blogs on WordPress.com. It is proven to work. Your idea about creating an extra column has been discussed before and has its benefits. I think, however, if we were to have monolithic tables on WordPress.com we would have much bigger problems than having lots of small tables. Since the multi-table approach is proven to work and the code is already written that way I doubt the schema is going to change anytime soon. Never say never though!
Thanks for reply. It is good to know that WPMU approach scales so good. However for smaller installations (I think of few tens of blogs) there is no big difference. This also have added value – code (both SQL and PHP) for working with data from all blogs at the same time is simpler. Of course this can be done with WPMU approach, but you have to issue query for each installation and aggregate results in PHP, or create one big query with subqueries for each installation.
What do you think of providing two installation options: separate tables or additional column?
Hell would freeze over before that would happen because most or maybe all SQL queries would have to be rewritten in WordPress to support that.
You should check out http://lyceum.ibiblio.org/ and consider helping there. That uses a monolithic table structure but it’s based on a very old version of WordPress.
I agree that this will require much work – quick grep through code showed over 200 queries which may need update.
BTW, another trick is to create updateable views for each blog instead of tables.
Can’t say I am overly thrilled…a big part of me loved the feeling of MU being a small niche platform, with a small but strong support network. Obviously I will not complain when I’m using new plugins generated from the WordPress.org bump, and usually news like this would excite me – but I feel like some of the WPMU character will die with this merge.
Thanks for clearing this up Donncha.
I think you misrepresent the spirit of the GPL slightly. I don’t think the spirit of the GPL ever was meant to prevent people from charging for their code. Indeed I strongly suspect that it was a goal of the GPL drafters to allow for it.
The point of the GPL is to make sure that those who receive code also get the right to distribute and modify that code.
I echo Andrea
I’m also curious to see how quickly people will move over, which will depend on how quickly the mu-plugins are moved over. Hopefully the new arrangement will require minimal changes there, or perhaps none at all i.e. backwards compatibility.
I can also appreciate how much easier your life might become, Donncha. Thanks again for all your efforts!
I’m very happy that the two versions will be merged. I only hope there won’t be any problems upgrading to the “merged” version as I asked in Twitter yesterday.
A merged version would be nice because 1. no more lag in upgrading. 2. no more plugins that won’t work because of using MU. Since I have both MU and regular sites, I won’t have to worry about which plugins I’m adding.
I hope you’ll still be active in WordPress’s development.
If you have problems with some plugins now, even after the recent upgrades to the plugins page you may have problems in the future too. MU a part of WP will have the same restrictions as it does now. Things like embeds will still be stripped by default.
most of the broken plugin issues I’ve had are due to the separate database tables for the different blogs and the plugin not looking in the right place for info or saving in the wrong place.
For example there is a sitemap plugin that required major hacking for MU and even then I couldn’t get it to work right. It would only sitemap the main blog (after the hack).
I would think that when MU is merged most plugin authors will at least consider the possibility that other blogs exist on a site.
Donncha, please excuse a layman’s question re MU merge to WP 2.x
If at my host I have a main acct and within that 5-6 other addon domains each with its own current WP install
How would one “merge” or introduce the MU function to those existing WP installs?
Or does one just enable MU within each of the existing installs?
One user to another, as an MU user currently, I would figure that you would perform database backups for each then export from the individual blogs then import each of those into the new version and Domain Map the separate blogs to the existing domain names then whack the old individual installs and databases associated with them. As always, do more backups than yo uthink you’ll ever need prior to starting. You can always happily throw them away later.
I agree with Chris. I’m a relatively newbie with WP, and have been looking at possibly using Drupal for centrally managing multiple projects for the long term.
Over the past couple of weeks I’ve been having to learn the hard way of how to harden WP for security reasons, and having a centralized way to do that for multiple domains would be a great advantage.
The WP/WP MU merger sounds perfect! I’d like to stick with one solution, and with this merger WP may be the ticket for my long term needs.
Regarding paying for plugins: I wouldn’t mind paying for ‘some’ premium plugins, but they’d need to be really worth it (and definately proven to be secure). I think having the bulk of plugins available for free, along with some tried-and-true premium ones available for purchase would be the way to go.
How will the WPMU merge affect the BuddyPress project. I don’t see anyone raising this but it must have some impact. Hopefully it will make it easier to impolement BuddyPress?
Has the WPMU/WP merge got anything to do with the leaked info regarding “BuddyPress for WP”? Would be logical, now that I think about it.
When in Paris last February, as I referred to the then freshly leaked info, Matt said to me he’d try it on his own blog first, when available (WP 2.8 seems too early still… WP 2.9 ? – I’d call it WP 3.0, as it would be a major change, and I would bet it might be available this Fall).
What I’m eagerly waiting for is for registered WP users – be it on WordPress.com or on any WP install – to be able to “interconnect” with each other wherever they are, making “friends”, carrying their profile and friends as they surf from one WP blog to another, just like you do with your own gravatar today, de facto creating THE ubiquitous social network of reference, 100% decentralized. (Well, you’d have to centralize some data, but from a user’s point of view, you would not feel stuck in a walled garden,and Oh my, what a beautiful business model for Automattic!) Definitely something you can bring to the bank.
And way sexier than any proprietary FB…
Pure Sci-Fi? Not so sure… long live open source :-)!
It would be particularly good if the merged WordPress makes it relatively easy to run multiple websites, each with its own domain, from one instance of WordPress. Imagine how incredible it would be, too, if you could install and upgrade plugins and themes in one place but available to all your sites on that server. Donncha, is that more or less the direction in which things are moving?
@GilCatt – your interconnection dream is already in the works! Matt has stated that Automattic has plans to do precisely that, using the technology behind their recent purchase, Intense Debate. Between that, the millions of wp.com users, the million+ wp.org sites and BuddyPress (which now, with this MU merge, will be an option for the massive number of self-hosted WP sites out there), the future is definitely interconnected and, best of all, not chained to Facebook.
What Donncha secribes sounds ideal re one install running many domains serving WP.
Still I hope there will be a way to “migrate” existing domains with WP installs back to one umbrella.
It will be critical for WP-savvy host companies to help ease that pain.
Think of what the dir structure would look like
main acct domain running WPMU with multiple addondomains
if plugins for all are centralized does that mean there ought be one image archive for all? how do you set permissions for that? or allow choice of which domains can access certain dirs or not.
How will server farms cope with this
what would actually be in the home of addon1? some index file?
But retrofitting existing installs to be compatible with one umbrella seems like it would be the biggest hurdle simply because Automaticc cannot control what web farms do or don’t to remain compatible.
@BGR – Just to clarify, I am Donnacha, not Donncha the owner of this site.
Yes, your points about image archives is very pertinent, I think there could be one archive to which the server would direct requests regardless of which domain they were addressed to i.e. domain1.com/images/arrow.png would actually be the same as domain2.com/images/arrow.png but the outside world wouldn’t know it.
If it wasn’t for leeching, I would love to run all my sites using mostly the same image sets, possibly pulling them from Amazon S3 or a CDN to improve speeds. It is such a waste to have exactly the same images downloaded when someone visits another one of your sites using the same theme. In an ideal world, a financial and technical model would exist whereby theme author’s could centrally cache their images so that someone visiting one site using, say, a StudioPress theme, would not have to download all those images again when they visit another. Perhaps the ISPs should all get together and agree to cache a few TB of the most commonly used images, which all site owners could rely upon to be in their users’ local loops.
Donnacha.. Donncha.. this is rather confusing …
Now if Donncha himself could confirm what you said regarding my comment, that would be nice 🙂
Donnacha – At the moment it’s not likely that the merged WP/MU project will do domain mapping out of the box as there’s a plugin to do that job. You never know though, it may happen.
BGR – you’ll probably have to import each blog separately into the new one.
GilCatt – I can’t confirm anything you said. You’d be better off asking Andy Peatling.
So does this news also mean that your super cool WordPress MU Domain Mapping plugin will get the ability for remote customer logins so they can manage their own blogs? I have been patiently waiting but eagerly checking to see if new versions of the plugin include this… 😉
Donncha, great post. I agree with you to some point with regards to the way wpmudev works, but I do think their hard work should be rewarded somehow. A balance must be struck and would you agree that at the moment it’s not there?
Look forward to the merge (since I work on both!) and hope to see you take wp/wpmu even further!
Man, I take a vacation and all hell breaks loose! Looking forward to the merge. Hopefully the MU community manages to stay strong. The general wordpress.org forums are a bit of a nightmare at times.
“Big huge 2.8 merge. Some things are broken. Only for the brave” – (just found out it’s actually for the super brave!)
It’s been a couple months since this announcement. Any updates on when this merge will occur?
Hi Donncha – are you able to give an indication on when the merge will occur? Even if it’s just ‘when it’s ready’ would be a help. Thanks!
Is merging WordPress and MU still on the task list, to date?
Do you think progress towards this may be held up *as a result of* the infrastructure that has been/is being created (consulting, proprietary plugin downloads, buddypress, etc) surrounding the WPMU architecture?
The two are so close now, the merge sure would solve my pains with modifying plugins to accommodate both types of installs (I manage many sites, some with MU, some not) and re-installing to change from a single site to multi.
Thanks for sharing this info here.
Any update on the merge?
Nope. If there was, I’d blog about it 🙂