Weve been open source since late 2008. In that time weve had 40+ version releases, 6 major ones, and weve always maintained backward compatibility. You can take a 5.0 website and hit Update all the way through 188.8.131.52. Its always been important to us to keep everyone moving forward together.
For the first time in 6 years, weve decided to not put backward compatibility first and foremost. We really wanted to have a clean slate, so we could embrace ideas that involved deep changes to the system. The ecosystem of 3rd party tools and PHP itself has changed quite a bit in the last half a decade, and we dont want to be stuck in the past. Theres so much good new stuff in concrete5.7 that weve been able to include because we simply decided folks would have to migrate by hand instead of upgrading with a single click. We know that when you start playing with 5.7 youll see why we made this hard choice, but we want you to also know were not going to abandon upgrades again lightly. Wordpress NEVER does this, Drupal ALWAYS does this - were right in the middle: once every half a decade or so seems reasonable.
This does throw our typical release approach somewhat on its head, however. Weve made a commitment to continue supporting 5.6.x for security and critical bugs for at least a year. Were going to make some changes to our marketplace here to support 5.7 add-ons/themes as well as 5.6 ones too. In the past weve often talked about upgrading websites feeling like working on cars engine, while driving down the highway at 75/mph. In this case, were assembling a whole new car next to the one were in, driving both, and still at 75/mph.
So whats 5.7.0 for and how stable is it? Heres some absolute truths you can share with your bosses and clients:
Just a quick official word...
5.7.0 will release on September 12th.
This will be a stable production ready version of 5.7, with no marketplace integration. We're getting it out there for our add-on/theme developers to build towards knowing nothing major will change. We also want our larger clients who may be solving their web challenges with the core and custom code more than the add-ons/themes we sell to be able to start working on their big projects with 5.7 sooner rather than later.
184.108.40.206 will release on September 30th.
This will include marketplace integration, bug fixes, and whatnot. We will be working frantically on docs during this period as well (editors guide has already started coming together) so expect that and some more site improvements here to support the maturing ecosystem..
Happy labor day all!
We just wrapped up 3 days of showing off 5.7 beta at OsCon with a great response.Bluehost very generously donated some of their impressive booth space to our project. If it wasn't for the auspices of these guys we and many other open source projects wouldn't be where we are today.
As we had hoped, 5.7 wows everyone who looks at it. We had folks intrigued with concrete5 from well known companies thinking about their web presence to agencies starting to tire of training clients on using blogs to manage websites, and even a passionate brony needing to rebuild his fan site. Some stronger partnerships seem likely to mature, and we're extremely motivated to finalize the bug fixes and get a production ready version of 5.7 out pronto.
Check out some pics of the event hall below, and start playing with the beta now!
5.7 beta's release is just a week away!
We could really use some help in GitHub.
There's lots of dashboard pages that need to be reskinned and Evan's juggling quite a lot in there. His new personal hero is Job who is grabbing issues that we've assigned to Evan and getting them done first! What a hero! We owe Job a beer or two, but he just got some karma points and free add-ons.
As I've often heard and frequently said: "You can have it fast, cheap and well done - now pick one, maybe two." We all demand 5.7 to be well done, we haven't been throwing wads of cash at it, and so yeah - it's taken a while.
All that is OVER my friends! We couldn't be more excited to tell you we have reached a point where we can actually commit to a launch date for 5.7 beta...
Yes just 19 days away... We will be at OsCon through the gracious auspices of Bluehost. As we launched concrete5.0 back at oscon 2008, it seemed only fitting to do the same with 5.7 this year.
You can see what's on the milestone list at github right here.
We could really use some development help to get over the finish line though. Our team is banging away at the 65 remaining issues at github, but precious few of them require the secret magic of being at PortlandLabs to complete. There's a lot of dashboard reskinning, there's a lot of testing, there's plenty of opportunity to throw an hour or two of development help in here and make a real difference. Get a pull request accepted and of course your name will be in the release notes, but we'll also give you a special badge and karma points for helping get 5.7 over the finish line.
It's been quite a marathon and we're just about there, we'd really appreciate your help!
Just a face lift, but a much needed one. Here's a discussion thread to chat about it.
We started many moons ago with a little marketplace of just our own add-ons. It was exciting to have found a revenue model for our free software.
As we looked for ways to differentiate ourselves from the sea of other CMSs, the idea of a clean, supported marketplace like Apples App Store made a lot of sense. We started accepting submissions from 3rd parties. With the excitement of Christmas morning, Andy and Ryan would personally review each submission as we had time, of which there was plenty of to go around.
Then things grew, and grew, and grew - and we couldnt keep up. So we added the PRB as an open source inspired committee of reviewers who could collaboratively help alleviate some of the initial poking around at submissions, letting us just be the final approvers. For a while, it worked great. We knew which add-ons to look at first, and while everyone wanted things to move faster, the marketplace maintained a level of excellence far above our peers.
As the marketplace grew, there were struggles around serving both the developer and site owner side of the equation in one environment. (e.g. What impact should difficulty level have? Whats the best way to install themes on existing vs. new site?) Theres a disconnect between our original vision and the reality today. We hear from our 3rd party developers that they feel like theyre putting great effort into providing support, and we hear from customers that its not always delivering the experience they thought they were getting. For example, just this week I was told:
"...In my opinion you guys should rise bar with quality control. This is second time in three years that I'm seriously thinking walking away from concrete for exactly same reason. Poor quality of the apps and bad support. I think you should rate developers for support so before we by their item we know who we deal with..."
Its natural for something important that is growing to have adolescent pains as things change. We believe there are two major problems with the PRB and marketplace today:
- We are trying to serve a diverse customer base with widely different expectations for what a finished product and good support means with a single experience. Is our marketplace a code repository for concrete5 developers who want to share with each other and maybe make a buck? Or is it more like Shopifys app store, with add-ons that are all safe, tested and running against a consistent environment in the will that always works no matter your technical ability? While we still believe we can serve everyone, we need to make some changes to speak more directly to the extreme ends of the spectrum.
- Due to demands on our time, we are unable to keep up with approval of new work in anything close to a timely fashion.
Its time for some well considered changes to improve the processes of the PRB and give different types of marketplace customers the experience they need. Heres whats happening
Submission to the PRB and getting a listing live quickly
1) The automatic checking system (called linter) has been improved. It now runs both on new submissions and new versions of existing listings.
2) If your new submission passes all the tests, you are given a launch date of 1 week in the future. Your listing will go live in 1-week even if no one on the PRB approves it. It will have a Has not been approved by PRB member! message on the listing, you'll have to flip a check box to get them to show up in the marketplace, and it will not be shown in concrete5s in-dashboard shopping UI.
3) The PRB will now have a small administrator group. They can approve add-ons and themes without anyone from the core team signing off. This adds their personal endorsement to a live listing, and/or launches the listing right then with their endorsement.
4) If your new submission did NOT pass all the tests, your listing will not go live until one of the PRB administrators or a core team member launches it.
Submissions will be rejected with limited cause or explanation. (e.g. wouldnt install, code looked sloppy in first file I opened spend some time working on this and resubmit plz. ) The PRB is not a free school, thats for the forums.
5) A new public Submitting to the marketplace forum area will be created for the entire community to help people get their code right.
6) Issues with version updates that fail linter tests will be worked out in that public forum, and not in the existing PRB interface.
1) PRB administrators and regular members have term lengths of 6 months. At the end of that period they automatically lose their PRB membership and have to request it again.
2) It takes 3 PRB administrators agreeing to accept a new PRB member.
3) It takes unanimous PRB administrators and the core team agreeing to accept a new PRB administrator.
4) To start we appoint: JohnTheFish, Mnkras, GoutNet, Tallacman as the first PRB administrators. For this special situation, their terms will be staggered (9, 8, 7, and 6 months respectively. ) Everyone else has had there term end and should re-apply now if they're interested in being part of the new PRB.
5) To request to join as a PRB member, submit your concrete5 user name to this google spreadsheet where PRB admins will review submissions on a monthly basis: (http://bit.ly/1vSIRZ3)
1) We will have a linter server you can just throw stuff at and get results from sans PRB. Dont abuse it or it will go away.
2) If the community wants to build additional checks for the linter to run, weve open sourced the framework and an example test. The PRB administrator group will decide if your check will be added.
3) The PRB interface has changed to be more usable. Were collapsing information that was getting in the way and were adding some filters/status to help keep things organized.
New style types, presets, one page vs. all pages, less framework, etc.
(Play on YouTube in HD for better readability on code)
We've been pretty quiet about what type of changes are going on under the hood in 5.7 until now. Let's just say, quite a bit:
New jQuery 1.x
JS & CSS precedence loading and minifying
New authentication types
New model view controller approach
Namespaced core (PSR-4)
Core & Override directory moving to the same level
Doctrine and steps towards total ORM
We're actually putting aside the time to redesign concrete5.org again in May. Here's our priority list and some out of date comps to get your mind running. Let us know if there's something we should be thinking about that we're missing.