Checks and balances

Destry Wion 6 September 2012

We bring you an uncharacteristic issue this time; a few articles now, and a couple more later. Most articles this issue are written by the Editor-in-Chief. But don’t let that scare you off, Issue 3 is a solid issue. The Editor’s letter introduces some interesting ideas for the magazine that could involve you. We’ll have two Spotlight articles when all said and done; one on a girl and one a boy. Checks and balances. The Bloke delivers a two-fisted tutorial for getting you started with building Textpattern plugins. Destry muses over theoretical designs following the recent Wikipedia Redesigned hoo-ha. We have our first guest author contribution this issue. (You should think about contributing your own sometime.) And a mystery author will review our newest Exhibit entry, which will wrap Issue 3 when posted.

It was a year ago this month that a handful of gents (sadly no lasses) sat around a table in central London and first talked about the revival of TXP Magazine, among other things. Six months later the magazine relaunched, and now six months more we’ve arrived at the release of Issue 3.

TXP is moving forward, sure and certain, but there are still many hurdles to hop before the magazine finds its publishing rhythm. As a publication effort that has relied on volunteer time, so far, I’m not always sure milestones from issue to issue will be met, and with this issue they haven’t been. For the first time, not all of the articles are ready. But rather than prolong what is already a stretch between issues, we’ll publish some of the articles now, and the others later when the last one is done. We’ll slip them into the table of contents and give a shout when they are there.

I might as well point out one more thing that will be obvious later anyway. All articles this issue, with the exception of Stef Dawson’s signature piece on How to create a Textpattern plugin, are written by yours truly, the Editor-in-Chief (unless a miracle happens). That should tell you a lot right there. We’re working with a skeleton crew, but combined with summer vacations, the code crunching efforts leading up to the awesome release of Textpattern 4.5, and the fact we can’t yet seem to woo guest contributions, I must fill in the blanks to the best of my ability when nobody else can.

The irony is that I should be writing fewer articles, not more. In fact, I should just be writing this column alone, so I can put more time to magazine business. No doubt if TXP was generating a little revenue to pay for guest contributions, we’d probably see more of them, and maybe our valuable editors would find more time too. And when I say valuable, I mean it. I’m extremely grateful for the time they can give. But there is a minimum level of input needed from people to keep the magazine on track, if not speeding along.

On that note, I’d like to open up a couple of ideas for your feedback, because until the magazine does find a comfortable publishing rhythm, we need to be light on our toes and ready to try new tactics.

Publishing schedule

One factor in all of this is the regularity of our publishing schedule. The periodicity started out on a firm quarterly agenda. In Issue 2 I introduced the idea of pushing that up to bimonthly, which was no doubt premature. The motive being to narrow the delta between the peaks and troughs observed in site analytics. In other words, when each issue is published, the magazine gets a lot of hits and the graph spikes (the peaks). Then, because of the duration between issues, it gradually drops off to a low number for a spell (the troughs) until spiking again on the next issue. The behaviour is much like a cardiograph, and we don’t want to see it flatline.

An open issue model

The analytics pattern alone is not a reason to change our schedule. But if 2-3 months is still not enough time for all team editors to draft one article while guest articles are still being solicited, then maybe quarterly, bi-monthly, or any other regular interval between issues isn’t the right call. Maybe an organic publishing pattern would be better; when the content is actually ready around other demands, and not when it’s supposed to be ready.

We’ll keep the issues concept no matter what; it provides a logical means for organising content, and the architecture is already setup for it anyway. But instead of publishing a batch of articles at a set period of time, articles making up an issue could be published in a series, as they are finished, until an issue is full (i.e., every column in the magazine has a represented article). In web publishing parlance, this is called an ‘open issue’. A new issue begins when the full set of articles for the previous issue have been published. Articles requiring more time to prepare, like Site Watch, would probably be published last in the series while opinion pieces and so forth would appear first.

While an open issue approach would see articles published more frequently, albeit in fewer numbers, a given issue could take longer to complete because we’d no longer be working on a deadline, or at least not as much. Frankly, this scares me a little, because without some push from a deadline, people tend to procrastinate more.

Publishing patterns of other web magazines

Other web magazines follow different schedules too, and no two are really the same. As a reader of many web magazines, I like the varied publishing patterns and schedules different publishers use. As a publisher, I’m curious which patterns might be stronger than others. On the surface, web magazines seem to be of two models: issue-based and blog-style.

Blog-style (there’s probably a better classification that’s eluding me) is where articles just keep coming on no apparent schedule and are archived by key terms, etc. There are far greater numbers of blog-style web magazines, though I’d guess for many it was just easier to architect them that way when starting out, and now years later, with so much content already published, they can’t really change into an issue-based format. Nor would they want to if their models are working, and obviously they are. Examples of blog-style magazines include .net, Smashing Magazine, UX Mag, Johnny Holland and so forth. These magazines tend to be just one content offering of many provided by the publishers; shops, event calendars, job boards… all are typical accompaniments of blog-style magazines.

Publishers of issue-based magazines tend to focus on just the magazine alone; no stores, no job boards, or what have you … just articles. This group includes A List Apart, Contents, and TXP. (Our editorial process supports the idea of a store, currently in limbo, but it’s not part of the magazine itself.)

A List Apart issues are composed of two articles, always published on a Tuesday, and usually every two weeks. One excellent article per week is a solid schedule, and impossible for us.

Contents operates on an open issue basis like I’m describing, where articles are published as they are ready. Their article counts (versus ALA’s) are greater like ours too, ranging from 7-12 articles per issue. Interestingly, they launched in November 2011 and are currently on Issue 4, which is somewhat parallel with our periodicity if you consider our first issue was February 2012. They say issues are about every two months, but if that were strict they’d be at Issue 6 by this point. I can certainly understand how timelines can shift in an open schedule. Contents doesn’t have regular columns like we do, rather each of their issues is a separate theme, so the two models wouldn’t be exactly the same.

Like most popular magazines online, Contents has an amazing team of industry talent behind it, which no doubt makes it easier for them to keep plenty of articles rolling in. Our topic scopes are different, and without as much industry connection, but it should still be possible to keep article numbers high in a reasonable time frame. We just need to find the right fit under the conditions.

What do you think about the open issue schedule as described? Would articles published more frequently appeal to you, are do you like the full magazine issue at the end of every set period? This isn’t meant as a vote, per se, to decisively change our pattern; we’re simply getting a feel for what might be reasonable options to pursue. Any feedback does help.

Collaborative evaluations

Our skeleton crew is another factor influencing our publishing rhythm, which is complicated by not having guest article contributions, which in turn are likely rare because we can’t yet give authors a small payment for their efforts. A vicious circle.

Until (I’m still optimistic) we get the money thing figured out, we need to break the vicious circle in other ways. One is to compromise on the breadth of topics we provide by dropping some columns. But that’s something we’ve been doing already; e.g., we’ve never started the Showdown! column simply because we don’t have the people to support benchmarking other CMSs. Some of our other columns are rotational anyway, not always appearing every issue. Further, some columns are essential because of their relation with other vital features that can’t be dropped; notably Site Watch, inherently tied with the Exhibit. So the question, then (besides how to get more guest contributions), is how can we leverage support on some of the work that editors must do before the actual writing?

Of all the magazine columns, Site Watch is one that could be conducted more collaboratively with the community, but in a structured way. I talked about this in my first editor’s letter, but here’s a fresh spin on the idea.

The current process for preparing Site Watch articles

Site Watch articles rely on evaluating Textpattern websites, maybe even one of yours. In order to prepare each Site Watch article, we must fairly consider the website in question against a thorough set of criteria. There will always be a degree of subjectivity in site evaluations, but the criteria help to keep that subjectivity to a controlled level and focused on real-world concerns, like semantics, accessibility, mobile compatibility and so forth. Criteria that considers these aspects of site creation will be important down the road when we start limiting sites displayed in the Exhibit.

We have developed a criteria matrix (a spreadsheet with multiple tabs for different criteria types) that is composed of criteria from various and respected sources. For example, the base of our matrix is from 247 Web Usability Guidelines by David Travis of User Focus, with some adjustments that better suit our specific needs. Our Accessibility criteria is based on WebAIM’s WCAG 2.0 Checklist. And I’m augmenting other criteria categories for content and mobile, for example, with quality sources too. Nothing is reinvented; we’re using existing industry conventions and honing them to what’s useful for Textpattern site evaluations.

One thing we’re seeing is that picking and evaluating sites prior to writing up the Site Watch article is proving difficult to get done in time. Currently, the column editor first needs to determine a site to evaluate (not an easy task by itself), do the evaluation (a little time intensive, if not a bit monotonous), and draft the review based on the scoring and note-taking. That’s quite a bit of work for someone who’s already busy and not getting compensation to do it, even over a three-month period.

We are streamlining this process as much as possible. For the site selection process, we’ve provided the community with a site submission form, which helps us get candidate ideas in one location for easier filtering. We’ve received many good candidates so far too, so that much is working. The evaluation part is facilitated by the criteria matrix I mentioned. It begins as a Google spreadsheet template with all the criteria types ready, and the scoring boxes prepared with functions for handling the math. An evaluator just copies the template for each new site and runs through the checklists. As mentioned, however, this can be tedious as there are quite a few criteria across tabs. Lastly, the drafting of the site review is aided with a draft template that corresponds with the structure of the criteria matrix. The author simply summarizes her findings in each section for each criteria set, and can inject some of her own insights and feelings about the site too in relation to the measurement. All-in-all it’s a pretty tight process, but it’s still a fair amount of work for one person.

A possible collaborative model

The Site Watch editor will have to be involved no matter what in order to be familiar enough with the evaluation to draft the review, but maybe she doesn’t have to do everything, or exactly as the process defines. Maybe we let the community pick which sites to evaluate, taking that burden out of it for the editor, and also allow a rotating list of people to do the actual evaluations against criteria. To help prevent evaluator burnout, each person scores one criteria type (one tab of criteria in the worksheet) and makes relevant notes in the comments accordingly. This would also help keep them focused and not get distracted into other criteria concerns. The worksheet itself will collate scores across tabs, and the editor can use each evaluators’ notes against her own cursory analysis to draft the review.

The combined effort works to lower the preparatory overhead for the editor, and increases the odds of getting the review done in timely fashion. All contributing evaluators would be named in each published review they help with in a special section at the bottom of the article, and in the meta-notes of each Exhibit entry, for example a page like this one for 24 Ways.

In theory this could work, but we’d need a structured and fair process for picking which sites get done. You wouldn’t be able to pick one of your own sites. Websites would be selected for each issue by group voting, and if a site was yours, your vote wouldn’t count in that case. Something simple like that.

Once the site (or maybe even two) for a given magazine issue was determined, the Site Watch editor would ready the criteria matrices, and give the evaluators access to them, according to the roster, to conduct their piece of the assay. They could browse around the matrix, of course, but they should only be scoring and commenting on their own tabs. When all evaluator scoring is done, the editor takes over; prepares her draft for publication, making sure all evaluators are credited.

Selecting and assigning evaluators is the final piece. To make it as easy and systematic as possible, we’d basically take volunteer names and create a public list in alphabetical order. We’ll move down the list and assign evaluation tasks in that order, and repeat from the top when we reach the end of the list. Names are added and removed along the way, as volunteers desire, the list order adjusts, and that’s just the way it is. Volunteers could even indicate if they’d like to handle more than one criteria type, or perhaps a specific one if they have a knack for a given interest (e.g., accessibility).

The specifics of all this could be refined later, of course. We’d make some concise instructions available, and evaluators would get a look at the matrix ahead of time, etc. But before I spend time on any of that, I’d need to know if you would be interested in this kind of collaboration.

Would you be?

Ebook versions of TXP

We return to the subject of revenue, which, let’s be honest, is what could make or break this effort over time. In my letter last issue I introduced the idea of adding advertorials to our offer for advertisers and sponsors. We’re still working on this. We’ve also been kicking around the idea of ebooks of each TXP issue as another way to potentially garner a few bones.

Doing this would require some changes to the current layout and architecture, but that’s our problem. Without getting into trivial discussions of function and presentation, the idea would be to refactor the front-end design so content and advertising were more integrated, and so tools like Adblock wouldn’t work because they wouldn’t know article images from ads (we don’t use third-party ad servers anyway, so that could work for us). Website visitors would get the advertising, widgets, and so forth along with the articles. Still lovely and responsive, mind you, but the whole kit and caboodle. On the other hand, if people just wanted articles—and nothing but articles—without printing them out, and when no web connection is available, then the ebook versions would be available.

Regular readers might scoff and think nobody would bother with ebooks when the website is free, which may or may not be true. But keep in mind that few people even know about the magazine at this point, and as more writers get involved, the articles will get more interesting. Mobile reading is increasingly common, on- and offline and in varying devices with many more to hit the market. It’s perfectly feasible that full magazine issues with enhanced readability could sell for the price of a cappuccino.

However, some important points need consideration with regard to ebooks…

First, creating ebooks isn’t going to help our current skeleton crew problems, on the contrary. If ebook planning compounds regular publishing priorities, then they’re probably not worthwhile until our crew is more equipped.

Second, full magazine ebooks are somewhat incompatible with an open issue publishing model as described earlier. We’d probably want to stick with the current batch-publishing process, and the ebooks would be made available at the time each issue is published.

Lastly, TXP articles will need to keep improving in value if we hope people will pay for them, so this goes back to making headway with advertising success, and thus more guest author contributions. The vicious circle.

Any ebook readers out there?

Textpattern world happenings

A few updates on past conversations to close this issue’s letter out…

If you don’t already know, Textpattern CMS 4.5 has been released. Designer types should start poking around on the front- and admin-side markup to see the changes, and start thinking about the custom theme designs you’re going to create for upcoming theme competitions. What?

Butcher brings the beef

I’m happy to report that Stuart Butcher, the person who has been steadfastly managing Textgarden (expect that identity to change to Textpattern Themes) since the passing of its original creator in 2005, has accepted invitation into group thinking about branding and content strategy for the entire family of Textpattern sites. With Textpattern 4.5 out of the door, the editorial team, including its two core developers, should begin making some headway on harmonising the family ring, from the content up!

Another word on theme competitions

Now that we have Stuart and the themes site in the loop, we have our platform for theme competitions. On the road to Issue 4, I’ll be helping with content and architecture planning in some of the family sites, and particularly themes to get foundation in place for handling theme competition development.

There should be opportunities for people to help as plans map out, so keep your eyes peeled or your ears piqued. No doubt such opportunities will be made clear in the Textpattern Forum, among other places.

A parting glance

Listen, I know you readers are smart, and from all corners of the web industry. I know you have something worthwhile to share. Think about it, and if you’re the slightest bit interested in contributing to TXP, even if you’re not sure on the details, don’t hesitate to contact me and we’ll kick it around. I’m here to help.