Technology Proposal for MT

Technology Proposal for MT

featured-image

MT has become, without a doubt, one of the most influential websites of its kind on the Internet. Both beloved and vilified because of their reach and their fame, the editors of MT should be proud of what they have been able to accomplish.

However, the Internet moves fast and staying still often means falling behind. If MT wants to continue to inspire thought and research they need to move away from a paradigm of disseminating information outwards to a new paradigm where the editors are not just content producers but are more traditional editors of content provided by a wide variety of sources. Response time to research, both in terms of advancing new findings as well as removing incorrect facts, will become more important in the future as sources of information continue to widen.

Recently, MT underwent a much-needed design and layout change. It is very much appreciated, but we can and need to move further than just a redesign. However, there are two groups who have different needs that must be satisfied in any future planning.

The editors need a solution where creating and editing content is as easy as possible. However, developers and designers need a solution that removes as much of the display layout and logic from the content as possible. CMS's such as Wordpress usually provide a great middle road by providing simple content editing that is later combined with the site layout after editing. However, CMS's as a technology are complex and are targets for security flaws, and when combined with multiple users of various permission levels can quickly become immensely complex for an average user.

 --Update--

 I don't even discuss wikis, which has been pointed out by a few people.  It's true, I don't, and it's true that a wiki would take care of much of the needs of the editors and people who want to contribute.  However, I feel that adopting a style used by sites like FAIRMormon have a few shortcomings.  Using FAIRMormon as an example, here are the problems that I feel need to be addressed with a wiki-based system.  None of them are, of course, inevitable or impossible to overcome.  They just seem to be problems that FAIR, a wiki which has been operating for a number of years now with a large volunteer force, still has.

 FAIR has a single voice on every topic.  Those voices may contradict each other at different parts of the site, but within each article there is only one point of view.  The history and discussion pages where (presumably) discussion and debate occurs are blocked off.  Edits are thus private.  We don't know the editors, we don't know who contributed a particularly bad edit, and we don't know if there are other viewpoints that are not being expressed.  It produces an unfortunate image of an "official" page on a subject.  But history and sociology are messy subjects.  We as a public need to have access to these discussions.  We need to know what the biases are in the construction of these pages.  We need to know why or why not certain items of information were written in this or that way, or were entirely omitted.  This produces an air of anonymity, which can be a powerful tool *and* a powerful weapon.  Anonymity provides some of the weaker apologists with a cover: their apologetic theories are imbued with the strength associated with FAIRMormon as a whole because nobody knows whether Daniel Petersen or some lowly high schooler made the argument.  Anonymity also can be a potent weapon: often FAIR has made sport attempting to portray their critics (whether they are also critics of the LDS Church or not) as fearful if they hide behind anonymity.  In the inverse, with a site devoted to an accurate discussion of LDS history and sociology, anonymity can prevent editors who have difficulty in keeping bias out from being found out.

 The answer to this, of course, in a future wiki system is to ensure that the history and discussion pages are open to public view.  Not too hard.  I just think that a system where the entire site can be openly accessed with improvements suggested, even down to the very nuts and bolts of the Javascript and CSS, is more in line with open source ideals.  I know a lot of people disagree with me and can be much more eloquent about how wrong I am.  I look forward to it when an eventual wiki-based site arises that attempts to be the honest FAIRMormon with fuller context to the issues.

Static Site Generators

One halfway point between the needs of the content creators and the web developers is the new idea of static site generation. A static site generator does not store the content in a database, but rather stores content in text files (usually Markdown or ReStructuredText). Upon command, the content files can be "compiled" with a site layout to produce a collection of HTML files ready to be deployed to a simple web server. Most web developers are familiar with the development pattern of Model-View-Controller. In a very rough (read: wrong) summary, the idea of an MVC web app like Wordpress is that the Model is the content written by the users, which is interpreted by a Controller, to produce (usually in conjunction with a template) a View for the end user to see. A static site generator could be though of as a Model-View-Controller pattern where all possible views are pre-built and saved as files.

The benefits, from a web developer's viewpoint, of a Static Site Generator (SSG) are speed and security. When a page is requested from the server there is no interpreter to be run, no specialized compilations to perform. Instead the server just grabs the HTML file in question and serves it. Files can be browser and cache-able, as well as compressed, providing even more speed improvements. And from a security point of view when there are no actual moving parts there are no vectors for attack: no SQL injections to guard against, no Status 500 errors to exploit. Also, the hardware needs of the server are at their most minimal.

Of course that describes the current site, which is also flat files. But the current site does nothing to separate the concerns of the content creators and the developers. Using a SSG means that the content itself is still stored separately from the site layout. Extensive modifications can be performed upon the look, feel, and behavior of a page without altering the content and vice-versa.

Markdown

Markdown, as a formatting language, is becoming more and more well-known. It is the formatting language used by many popular websites, such as GitHub and Reddit. Chances are that the editors and other content contributors are already quite familiar with it (it is the formatting language used by Reddit, for instance, and we know that many of the MT editors are active on Reddit). This proposal itself was written in Markdown (not that this is much of a big deal as it isn't really using much formatting). You can see the original file here, if you'd like.

--Update--

I've moved this blog from a Pelican backend to Ghost, which is also Markdown-based but with a much more efficient web interface.

GitHub

One other benefit of moving the content pages to their own files is that it becomes far easier to make and track changes to them. The penultimate piece of this new technology system for MormonThink would consist of hosting the entire source code for the site publicly at GitHub. I understand that this extreme transparency might make a lot of people nervous, but in the end MormonThink is devoted to facts, truth, and accuracy and transparency is one of the best ways to force these ideals in technology.

Also, having the source at GitHub introduces an entirely new aspect to the development of MormonThink for the future: additional content. As a public repository of code, anyone would have the freedom to download their own copy of the website to do with as they wish. In reality, there is nothing about this process that is now impossible and indeed I've had to copy every accessible file from MormonThink as I've pursued this project.

GitHub, and its underlying technology of git, are centered around this idea of widely distributed code because of how those widely separated copies of the code can be used to make the project better. Users can do more than just make a copy of the code: they can alter it (creating what is called a "fork" of the code). In the case of the future MormonThink site they can add new pages or even entire sections, they can clean up and alter existing pages, they can upload images, and much more. This is because they have full control over their own copy of the code. Again, there is nothing preventing this behavior even now so while this might sound quite scary it really isn't. Anyone can post their own altered version of MormonThink even now, but these unofficial versions have no hope of attaining anywhere near the audience they need to endanger the fame and authority of the current site. Opening up the code would not change this: MormonThink.com is still the official page.

But users can request that their changes be "merged" back into the official code (the technical term for this request is a "pull request"). And here is where the editors of MormonThink get to continue to be editors. GitHub has extensive tools set up around users and permissions involved with these merges. This process of third-party, independent coders and content creators contributing their own effort back to the main project is how open source projects like Linux are built. Allowing these requests is fully the responsibility of the owners of the original GitHub repository.

Example

Let's assume that it's now a year in the future. The existing pages of MormonThink have been migrated to a combination of Markdown content pages being compiled by a Static Site Generator like Pelican or Jekyll. The source code, including all of these content pages, are publicly accessible at MormonThink's GitHub repository. A reader, let's call her Janet Law, notices that there's an error on a page, say the page about the Expositor affair. She wants to know how to fix it and sees a link at MormonThink about how to contribute. There she finds a set of simple instructions that lead her to create an account a GitHub and to fork the code for MormonThink. Now she has her own personal copy of the code. She opens up the files that discuss the Expositor history and makes the changes she thinks need to be made and saves them to her personal copy. Now she initiates a "pull request" to MormonThink asking them to merge her changes back into the main website. A MT editor sees her pull request and takes a look at the changes. GitHub color-codes the changed files so that it is easy to see exactly where changes have been made. The editor, while impressed, is concerned about one or two of the changes. He comments on the pull request asking Janet Law for more information on one point and to rewrite another point. Janet makes those changes to her personal copy and saves it again. Now her pull request is asking to merge in all of those changes, too. The editor takes another look at the request and decides that it's a good change and he approves the request. With that approval Janet's changes are now incorporated into the official website. If there are any problems it's not very difficult to roll the site back to a previous version of the code as git, the technology underlying GitHub, keeps a full and complete record of every change made to the code and can roll back and forwards the active code easily.

In our example we still have an official MT editor, but now much of the content development and maintenance can be moved from their own responsibility to a wide army of volunteers who want to help. Of course, every editor can go through the same process and approve their own merges if they also want to add content, but they are now editors in a very real sense.

(Also, as a quick aside, the created content does not only need to be in English! Translations of MT content are sorely lacking, and opening up the code and the contribution process this way would do wonders for the ability of foreign language speakers and writers to quickly amass a large amount of translated copy.)

Prose

There's one final benefit in this example layout: it can be done without opening any code tool such as Dreamweaver or Sublime Text. For Markdown files contained in a GitHub repository there is a tool, called Prose, that can edit them on-line and save the changes to the personal or official code repositories. Prose is a great tool because it can instantly translate the Markdown being written into a rough formatted version so that an author can quickly make sure that they're writing and formatting their content the way they want.

Healthcare.gov (Not the Bad Bits!)

This technology plan for MormonThink isn't just an unproven idea. While the actual application parts of the website have been bogged down in bugs and errors, the landing pages and informational pages for the US government site healthcare.gov are produced using a Static Site Generator, a private instance of GitHub, and Prose. These aspects of the website have been lauded for their simple design and devotion to user-friendliness. It does kinda stick that the rest of the site is so buggy because it certainly reflects badly on the parts of the site served with the SSG.

Conclusion

The future of MormonThink requires some drastic changes (we didn't even get into the fact that quite a bit of their current content is heavily dependent upon tables of data and while these don't need to be deleted they shouldn't be the main content for average users to read). I want to propose that these changes should include moving the website to an open source model that would allow to wide involvement from users across the world. In my opinion, the best way to adopt an open-source model would be the use of a Static Site Generator, such as Jekyll or Pelican, combined with Markdown content source files. This very blog is actually produced via the same process. The difficulty in setting it up would involve the initial transition. Once set up, maintenance and continuing content creation would be very easy and manageable with a shallow learning curve no more difficult than the current process of local file editing via Dreamweaver and upload via FTP.

I welcome any and all comments and questions. For anyone who is interested you can find my current explorations of this project at the following two repositories:

  • nocoolnametom/MT-Transfer - This is a scraper I'm spun up in some free time that rips the existing site and converts it (mostly) into Markdown files that can be used with a SSG.

  • nocoolnametom/MormonThink - This is my attempt at creating a SSG-served version of the site using the files from the scraper. I'm not very far along on this project, obviously.

An Update After Posting

In conversation with others and in asking for feedback (here, on Reddit, and through email and chat) it seems that there is a general consensus that this process may be too technical for an average user. I still disagree but then again I am a professional programmer and my perspective on "user-friendliness" is very skewed so I shouldn't be very surprised if it turns out I am wrong. The general response is that 1) MT probably needs at least a good re-organization of their materials and sections but a full change of approach might be going too far, 2) if (and it's still a big "if") the editors favor a system where outside edits can be accepted it seems a wiki approach is preferred, 3) there have been a few responses from more technical people who really like the idea even if they may or may not see it as feasible.

So I will admit that the energy behind a more wiki-based solution for disseminating full information on LDS history is impressive. I support it as it still helps approach the ideals of the newer generation of Internet users who don't always want to simply read information but want to participate with sites in a more back-and-forth method. I still have my worries about such a project, but I imagine that those worries may be misplaced.

One further avenue of this discussion has also been brought up by a friend: the Internet presence of the New Mormon History and the resulting communities of people impacted by it seems to be unavoidably fragmenting. The rise of other sites and approaches, including but not limited to The 95 Theses, The Letter to a CES Director, MormonCanon, ExploringMormonism, and a few already-existing wikis shows that there are many people who want to participate and have already made movements of their own in this regard. While I don't think MormonThink is in any more danger of being lost in the noise than McDonald's is currently in danger of being lost in the noise of other restaurant chains who knows what the future may bring? However it goes, I doubt that MormonThink will ever truly go away or be forgotten, any more than Dialogue and Sunstone have gone away. Instead those trailblazers established an untapped field of study and of community and later gave rise to many different on-line communities in the Bloggernacle, such as By Common Consent and Times and Seasons, and they're still alive and actively contributing to the discussions.

So I'm going to leave this technical proposal up on my blog for others to peruse. I still think that this approach will work much better in approximating the process of editing a scholastic journal than a wiki ever would, but if that is the way things move forward I will not fight against it. I don't think a wiki is a bad idea: I think it's a good idea (I think it's better than the pace of change able to be kept by the current site written by a small group of private editors in their free time), I just think this approach would work better with the current editors of MormonThink and their need to retain a sense of strong ownership over the text and content. I also think that releasing the content itself under an open source copy-left would do wonders to help the information get out to an even wider audience (foreign languages in particular do not have much access to this information which is a major failing of most existing sites).

If my friend is right and the fragmentation continues I may someday resurrect this idea and throw my own hat into the ring. Until then, perhaps the ideas here will spark some interest in how to move forward with presenting the New Mormon History more effectively.

Tom Doggett

Tom Doggett

I'm a programmer, Ancient Greek reader, feminist, spouse and partner, and a dad.

View Comments