We recently had the brilliant opportunity to hear all about the development and launch of Craft 3 direct from the makers of Craft CMS, Pixel & Tonic.
Our Q&A with Brandon, Leslie, and Leah covers hot topics like the new features in Craft 3, plugins, the launch, the impact of GDPR, how the development process went in general, and what kept them on track.
The transcript is pages and pages long so we've split it into sections. This first instalment focusses on the challenges and hurdles faced, how they overcame them and the reaction Craft 3 has received since it launched.
Kicking off the questions, we asked:
What are you most proud of about Craft 3?
Brandon: For me it’s the fact that we made the launch date!
It was a crazy project that kept growing and growing and there’s so many different moving parts it’s hard to pick any one thing beyond the fact that it all came together in the end.
We had the whole team in for about 2 weeks leading up to the launch all huddled around this table hammering through this massive task list of things that needed to come together toward the end and the fact that it all got finished in time was pretty remarkable.
I’m really proud of all the work that everyone put in and it’s actually pretty crazy; I don’t think that we could have made it on time has we not have been here. The speed in which we were able to work together and collaborate on all these features and functionality is way easier to do when everyone is in the same room and time zone - usually everyone is working on their own hours.
Leslie: For me it just felt really good because it was also the ‘how’ we got it done; the spirit in the room, maybe this is too much of a hippy-dippy American term, but the positive energy in the room. I mean, there was a high level of stress but it deepened friendships, we had each others backs, we felt like a really solid team, it was fun to come into work everyday, despite all the high stress and pressure.
There was a lot of excitement and energy. For the late nights that everyone was putting in, it was a good time and a great feeling to get to the end of the countdown and go live.
Leslie: ...and nothing terrible went wrong.
Leah: We had a really great playlist that motivated us.
Leslie: From a feature perspective, when I talk about Craft in sales and clients, the thing I really love to be able to talk about are the performance improvements and the plugin store. Those two things, especially when we try to out the small business charm into those discussions, just work really really well and have been really well received. Performance gains are easily understood by end clients and the plugin store is also easily understood.
But probably the whole damn thing!
What were the biggest challenges?
Brandon: I don’t understand any of the math that went into the image editor. It basically started like let’s do something like the photos app on the phone but it turns out the math and the theories that go into replicating that kind of functionality was pretty intense, so that was definitely the single most challenging specific feature - was rotation plus crop in the image editor.
The plugin store was the most challenging bundle feature. That was the one for me that was looking forward at it and no idea where to start; how should we prioritise this? how should we tackle this thing? Th plugin store wraps around 4 or 5 different projects that come together. Between Craft ID and replicating Packagist on our end, being able to become a new Composer repository, getting the Composer updating stuff working, getting all the billing stuff working, learning Vue and building that out, and all of these different things that come together in order to get that working.
For me, personally, that was the biggest challenge - orchestrating that whole thing - that was the part of the project that got me scratching my head the most.
Did the launch go better or worse than you anticipated? If you had to do it all over again tomorrow, what would you do differently?
Brandon: I can answer the second part. The launch itself went fine. It was extremely intense leading up to April 4th and leading up to the minute everything went live. There was obviously various fires that we needed to put out but I think that was all expected especially considering how big the launch was.
Given the scope of this launch, we did everything we could have to make it easier on ourselves. From that perspective I’m pretty happy with it.
It was actually a 3 phase launch; we first launched Craft 3 RC1 and that introduced composer updating for plugins - the non commercial version of the plugin store, and then we released Craft ID and that was the second phase.
We moved all purchasing history we ever had over to a new Craft 3 install running Craft Commerce 2, migrate all the history. We launched big updates to our old web service so that it started interacting with the new web service (CraftNet) and then update CraftCommerce.com and basically that was the initial launch of the plugin store.
Ever since Craft ID launched, if you bought Craft Commerce or Craft CMS, you were actually buying it over the same APIs that the plugin store would ultimately start using as well. So that was the launch of the backend and then finally on April 4th, we launched Craft 3 GA version with the new plugin store that added the commercial functionality.
I think we did a pretty good job of trying to ease the challenge and certainly after each one of those launches we were thanking ourselves for doing it then and not waiting until April 4th because there were inevitably bugs that came up that were easier to tackle in an isolated state than alongside all the other stuff.
If we were to do anything differently, I would say we should have phased Craft 3 as a 'thing' into multiple releases.
Maybe Craft 2.5 could have been Craft 3. We could have done a version of Craft 3 that was just the Yii 2 port and maybe by this point it could have been Craft 4 or Craft 5. I think we started to feel a little guilty that we’d been sitting on all these performance improvements for years. The Yii 2 port, most of the performance improvements came within the first few months of the project. Maybe we could have launched those and then saved some of the refactoring for later, but who knows.
There definitely was a less stressful way to get to here but it is what it is.
We’ve learned our lesson. We’re not going to do anything quite so crazy for Craft 4 or Craft 5 - we’ll do smaller chunks going forward. We made it here, Craft 3 is out and it’s stable, people are using it, so at the end it all worked out.
Have you seen anything new/exciting/interesting done with Craft 3 that would have been difficult or even unachievable with Craft 2?
Brandon: Yeah! One obviously example would be our own CratfNet service. We took advantage of all of the applications development-type features that came into Craft 3; between migrations and the fact that Craft lets you modify the actual application config and do all these things that really position Craft as a decent application framework rather than just a CMS.
Multi-site, from a sales perspective, has generated a lot of interest from people coming from other CMSs.
We’re seeing new applications of Craft and that does also build on the fact that it’s more of an application framework now. There’s more use-cases for Craft these days beyond just building a website.
Brandon: Looking into the future, that’s where we’re trying to push Craft 3; to become this non-traditional CMS. We’re trying to move beyond just being a website CMS, we’re trying to make it more of an application platform, make it more of a solid headless option, and even further down the road we want to make it a solid serverless option.
Are you going to update the Happy Lager site to Craft 3?
Brandon, Leslie, & Leah: Yes!
Leslie: It would have been nice to have had the demo up and ready but we want to make some minor but important updates to the demo based on our own usage of it. We just ran out of time to get that done.
Brandon: It’s been on our list. It’s going to happen for sure. We want to improve it a little bit. We want to get SEOmatic involved so that people can start seeing that as well. It’s kind of a ‘must-have’ plugin if you’re building a website. Down the road further we want to get Craft Commerce involved as well in some form, potentially even redesign it.
We built the Happy Lager site at a time when Craft was still getting on its feet and at that point we were just trying to make the point that agencies should check it out. We knew that most agencies, if they were going to start using Craft, the first thing they were going to try it with was usually their own internal website. Just as a way to play with it before they were going to pitch it to a client. Happy Lager became like an agency’s site - a demo of an agency site.
We’re definitely not getting rid of it and we want to improve it. Initially, we just need to get it right in Craft 3, that the first step so that’s on the list. Down the road we want to start reinvesting in it and probably rebuild the site. At this point it’s probably better that instead of doing a fake agency site, we make something else that’s more interesting. It is more of a sales tool for agencies to sell to clients rather than our own sales tool to try to sell Craft to the agency. It needs to evolve a little bit I think to better meet that need.
There were a ton of things that we would have gotten in before April 4th if we could have, a lot of them just didn’t make the cut. The documentation was a big project that took longer than we expected. That’s mostly done now. We’re moving on to the next steps now and that’s in there.
Now that Craft 3 has been around for almost 2 months, what sort of reactions are you seeing from non Craft developers? Where do you see yourselves when compared to other content management systems and has this changed when comparing Craft 2 and Craft 3? If this has improved then what is it in particular about Craft 3 that is hitting the sweet spot?
Leslie: I think that Craft is still hitting the same sweetspot. Most of the time were being asked to compare ourselves to either Wordpress or Drupal. In both cases, Craft 3 just makes the argument stronger because it’s faster, it has the image editor, it has multi-site. The other thing in those comparisons, that hasn’t really got a lot of community scrutiny yet, but the Craft licenses have been a huge deal in those conversations with Drupal and Wordpress advocates.
The Craft license is a really dramatic change that we haven’t talked about too much publicly - it really is as close to open source as we can get without putting ourselves out of business. When we talk about it on those direct calls, the only difference between Solo and Pro is users. Users and some branding stuff. So when we go into those sales calls asking to be compared against Wordpress or Drupal, with Drupal we’re making it towards simplicity because if you’re considering Drupal you typically have a very complex build in mind so Craft is just a lot more elegant from a developer perspective so it’s going to be more cost-effective and the content author experience is considerably better.
You can make all those same arguments with Craft 2 but in Craft 3 every single one of those arguments is better.
In Wordpress you kind of go the other direction, if we’re talking to someone who has a $10k - $15k budget or lower, they ask why they should go with Craft and then it all centres around the content experience that they’re trying to get and do they value the external presentation of their brand or do they just need a rebranded theme. That makes the decision of, “do I go low-cost Wordpress or invest in a unique site with Craft?”, it makes that decision point for the end client easier - they’re not as concerned about Craft 2 versus Craft 3.
Wordpress and Drupal are the 2 CMSs that we get asked to compare ourselves to the most. Sitecore surprisingly comes into the mix a lot. With Sitecore we talk about that exclusively in terms of pricing. It’s almost never about the product. If you’re considering Sitecore it means you have $80k a year for 1 admin seat and you’re concerned about the cost of Craft, why is that? So the pitch gets turned around to taking that admin licensing fee, give it to an agency to build something of value quickly, and then we’re back into the same argument that we use for Drupal in that case.
From a sales perspective, it’s so much easier selling Craft 3 because the narrative is stronger but it’s the same narrative.
I think people switching over from other CMSs, they figure that out really quickly. That this is meant for unique, bespoke sites, it’s not a theme engine, we want you to be able to build things really effectively and uniquely to make the clients happy, and you need to stick around to really develop that with the client in mind. So it makes the decision, “is Craft right, or do I need Wordpress, or do I need an enterprise system because I have all these security audit trail needs and things that Craft will never be?” All of those arguments have become easier with Craft 3. It makes my job a whole lot easier than it was, and hopefully yours too!
The launch of Craft 3 is an exciting step forward for a platform that, in our opinion, was already great. The improvements and future plans make us even more confident that Craft CMS is a wise choice for anyone looking for a flexible, reliable, and enjoyable to use Content Management System.
Be sure to check out the other instalments in this series to learn more too:
- Ep 1: The Launch of Craft 3 (this one)
- Ep 2: Craft Roadmap & Competitors
- Ep 3: Craft Plugins & Support
- Ep 4: Craft & GDPR
- Ep 5: Craft Commerce 2
If you have any questions about Craft CMS, or the web in general, please get in touch firstname.lastname@example.org
Or feel free to come along to the next Craft CMS Manchester Meetup that we organise to hear what's new and meet some of the Craft community.