Hi everyone, I’d like to take up this very important topic again. It looks like the consensus is that BLF is definitely not dying, and this lively discussion has proved that. I also gather that the majority recognize the negative aspects of Reddit and would like BLF to continue as a more traditional forum, while also mentioning some important features the BLF currently lacks.
So BLF is not dying. However, what is dying is Drupal, the engine that runs this forum. A bit of background: When I launched BLF in 2010 I already had some experience with installing and administering a non-flashlight-related forum with a different forum software. Frankly, it was awful, and most of the other alternatives at the time were also really bad, even the proprietary paid ones. Drupal version 6 at the time had a much more clean and logical architecture by design, and it got a lot of the fundamentals right that other platforms just didn’t have at the time. For example, I believe that one of the reasons that BLF took off so quickly in the early days was thanks to Drupal’s search engine friendly URL scheme, which was really unique at the time. Drupal also had a much better security record than most other content management systems and forum platforms from that period. In terms of forum functionality, Drupal has never offered much beyond its barebones core forum engine. But at the time Drupal 6 offered a huge array of community-developed add-ons that made it look and act like a fairly traditional and featureful forum for its time. From there I eventually upgraded to Drupal 7, which was an extremely painful and problematic migration, but there continued to be good support from the Drupal community for necessary forum functionality. Drupal also is a general-purpose site engine (“Content Management System” / “CMS”), and it offered some extremely flexible tools that allowed me to create our unique-to-BLF email notification system and community spam control and offensive post reporting system. This allowed for a very unique and flexible paradigm that contributed to making BLF such a unique community. Of course I would remiss if I didn’t mention the BLF users; this site would be nothing without all of you, so a huge thank you to all of you.
Fast forward to roughly 2018 and the Drupal project started posting notices about the end-of-life of Drupal version 7 within a few years. They of course encouraged site owners to migrate to Drupal 8, which had been available concurrently with Drupal 7 for some years before that. But Drupal 8 was such a radical paradigm shift that not even the official drupal.org website was migrated to version 8 for a long time after its release. All of the community-maintained add-ons that almost all site administrators depended on for basic functionality had to be re-written, and since Drupal always required a lot of tweaking to make it work as a forum, most of those add-ons fell by the wayside. Drupal has always been very barebones out-of-the-box, and site administrators have always depended on community add-ons, but with the advent of Drupal 8 they started treating it as more of a development framework for experienced web programmers, which personally I am not. And I’m not the only one. There were lots of upset Drupal 7 site administrators, and their complaints together with the COVID-19 pandemic made the Drupal developers decide to extend the Drupal 7 support lifecycle for another 2 years. This year it was supposed to end definitively, but they announced that it would be extended again for another year and possibly extended again after that. So I want to be clear that I have always taken the security of the BLF website very seriously, and the site continues to receive regular security patches for Drupal 7 directly from the Drupal project team. However, there will likely be no new features added to the Drupal 7 ecosystem, and Drupal 8 will soon become Drupal 9, which is no longer a general-purpose extensible site engine that normal web administrators like myself can install and mold into something useful without coding custom modules. The migration from Drupal 7 to version 8 or 9 would also be an absolute nightmare, with basic forum functionality still lacking and/or provided via limited third-party add-ons.
So, that wall of text is a long way of saying that we need to start seriously looking at alternatives for migrating BLF to another forum platform. I will only consider free and open-source forum platforms that I can host and administer myself, and that’s non-negotiable. There are a bunch of options in that field, some better than others, but only a few come close to providing what we need. It’s important to keep in mind that the unique BLF culture has largely grown around the hands-off community moderation systems for flagging and banning spam users and also for reporting and automatically removing posts that violate the BLF rules. This was implemented via a flexible framework that Drupal offered, and it’s really hard to reproduce it in other platforms. There are also some potentially useful new features such as pinging @usernames, thanking users for posts (an up/down Reddit-esque voting system is not and will not be considered), uploading images, and improved threading and searching functionality. And finally, as much as I hate to admit it (I personally detest using smartphones if I have access to a real computer), we need a “responsive” (i.e. adaptable) website layout that is easier for mobile device users. It also needs to be a secure platform with manageable security and feature updates and a solid development community behind it with high probability of remaining viable for the long term.
Of the very, very few options that come close to meeting the above requirements, the one that most stands out is called Discourse. In your web searches you’ve probably run across posts in Discourse forums, and there’s a good chance you’re already using it in forums about other topics. I have some experience installing and administering a Discourse forum from a few years back, and it has continued to improve since then. It has tons of useful features and is generally user-friendly, mobile-friendly, and administrator-friendly. It also is unique in providing a community moderation system with automatic tiered levels of user privileges based on the user’s behavior and level of trust, similar to what I’ve put together piecemeal over the years for BLF via the Drupal framework. Discourse has been very much developed from the ground-up to cope with the glut of spammers and trolls on the modern internet.
It will still be a gargantuan undertaking to migrate BLF to Discourse, and it will take some time to plan a path forward. There are still some major technical challenges to migrate the massive amount of users and threads and posts with different post formats (Simple Post Editor vs. Advanced Post Editor) from Drupal to Discourse. I’d be happy to discuss the details of those issues with our technically minded users that are interested in helping to find a solution. I’m still not entirely sure if it would be more feasible to migrate all of the posts and private messages, or if it would be better to just migrate the current user accounts to the new forum while maintaining the current Drupal site in read-only mode or even converting it to a static HTML archive. But please be assured that I deeply respect the wealth of information that has been created by our incredible user base, and it won’t be disappearing.
There is no rush to commit to this migration. Let’s take it slowly as a community, and I’ll do my best to listen to your useful feedback. The Discourse forum is highly configurable and extensible. Both you and I need to kick the tires and see how it works in practice, and as time permits I’ll slowly start working through the checklist of how to configure it and customize it for our needs. So to begin this journey, I’m please to present you all with this temporary testbed Discourse server :
The test site is open for registration. At the moment none of your current BLF accounts have been migrated, although account migration is probably the most easy and straightforward aspect of this process. So at some point I might try to migrate all of the current BLF user accounts (username and email addresses, possibly the location if filled out in the user profile, possibly the avatar, and probably not the current password) to the testbed server, and I’ll let you all know if that happens. Meanwhile I’d encourage everyone to create an account and play around, keeping in mind that the testbed server will eventually be deleted with all of its posts, and the final migrated version of BLF will continue under the budgetlightforum.com domain name.
I just got the new testbed up and running today. I’m also evaluating some different server OS options behind the scenes, but that’s not very relevant in terms of functionality to BLF users. So I was mainly focused today on the core aspects of the new server OS and setting up the email notification mechanism (the entire internet email system is a dreadfully complex and arcane maze of gotchas and little-known rules and protocols that make my head spin). The new site is still in its stock form and has almost no customizations to its behavior, so don’t judge it as the final product by any means. Since it’s using a brand new temporary domain name, outlook.com and probably some other email providers will not accept the new account registration emails from the test server, but Gmail and Yahoo do work in my tests. Please let me know here if you can’t register a test account there and I can probably set it up manually for you. Again, please remember that it’s just a testbed server that we will gradually customize and test before installing a different instance for the final BLF migration, and all of the testbed data will eventually be deleted.
Hope to see you over on the test server! But please don’t stop posting here either because we’re not done yet. Have fun!