You can download my slides with notes here: http://dl.dropbox.com/u/6541807/fun-with-location.pdf

I bookmarked a number of resources related to my talk as well: http://pinboard.in/u:spinnerin/t:secondconf2011/

Thanks to SecondConf for inviting me. I hope you’ll all be inspired to go out and experiment with location in your work.

Christie, Sherri, and I gave a well-received talk on event planning at OSCON (“Event Planning for Geeks”). We’re still building this out, but an initial event planning handbook is live on the Stumptown Syndicate site. You can also view our slides. We realized in the course of writing the talk that we have way more material than we’re comfortable squeezing into a <1hr slot, so we'll be working on expanding this into a 2-3 hr. long workshop.

In response to several conversations we had over the week, we set up a Citizen Code of Conduct site to make it easy to share our open source citizenship-focused guidelines from Open Source Bridge. The document is CC-licensed, so you can easily re-use it for your own events and projects.

Another cool thing happened while we were busy at the conferences: “Collective Agency” announced its plans for the space formerly known as Souk. Stumptown Syndicate is happy to be one of the initial workgroup partners, and we’ll be working on how we can use this space to support the local user group community.

So—onward to August.

Open Source Bridge, June 21-24
The 3rd year of our conference for open source citizens. In addition to being on the planning committee (talk to me if you need a media pass or want to schedule a BoF), I’m moderating a panel on building and maintaining open source communities.

Indie Web Camp, June 25-26
Assuming I don’t fall over from exhaustion first, I’ll be at this unconference/hacking event for people working on distributed social software.

Community Leadership Summit, July 23-24
I’ve never been to CLS before, but this unconference on building tech communities has been recommended to me several times, so I signed up.

OSCON, July 25-29
Christie Koehler, Sherri Montgomery, and I will be presenting Event Planning for Geeks, a condensed guide to successful unconferences, code sprints, and beyond.

SecondConf, September 23-25
I’ll speaking at this Chicago conference for iOS and Mac developers, on some combination of mobile devices, game design, location, and not being creepy (unless, of course, that’s your goal). Creepius will probably make an appearance.

WhereCampPDX, October 7-9
Portland’s location and technology unconference is on year 4! I still love helping plan this event. We haven’t really started the planning process in earnest (first everyone has to get through OSBridge) but you can join the mailing list to find out what’s happening as soon as we know it ourselves.

I like:

• The art! It’s a cute game. Someone clearly put effort into dreaming up a number of fun robot designs.
• I know from WhereCamp that the developers thought about how to make this a game you could pick up and play anywhere in the world, and I think these considerations put the game on a solid base. It probably makes more sense in an urban area, but there’s something to do anywhere you can go outside and walk around.
• Doko Air: a feature that lets you send a robot off to explore the world.
• The photo tool. Most of the gameplay centers around taking little travel photos of the robot in different places, and being able to scale and reposition the bot in the frame makes this more fun.
• The magnet: this is another nice consideration for how the game works out in the world. If a battery or robot piece is someplace you can’t physically reach (like the middle of a highway) you can pull it closer with the magnet tool.

I was less happy with:

• The reliance on user-generated content. All of the activated bots are named and described by other players, and all of the storytelling comes from what people put into the journal entries. This makes the game easy to expand, but it also means that there’s a lot of uninteresting bots wandering around, with just one or two items in their journal, and unless you have a pack of eager players in your area, any bot you drop off will probably just sit there until you stumble across it again yourself.
• The lack of interaction between players who are in the same space, or between robots at all. I can’t hand a friend my dokobot, or batteries or parts, when we’re in the same room, even though this would be a completely natural activity. Instead you drop off the bot, and the other person picks it up. Also, dokobots are basically invisible to each other—I could have a dozen sitting around my apartment, and the gameplay would be exactly the same. This seems pretty contrived.
• There are way too many inactive bots available to pick up. In any area I scan, I find several, including ones that are “rare” west coast only or other sub-groups. I currently have 24 inactive bots in my workshop (holding area), and I have no idea what to do with them, since I can’t trade, discard, or combine them into other things. I don’t think it’s unreasonable to make people walk around a while between picking up parts. Having only one or two appear in each scan might be more balanced.
• The graphics and the game text indicates that dokobots have different structures, but all of them have the same abilities in the gameplay (recording their location and taking pictures). Why can’t the headphone-wearing music lovers record audio snippets? For that matter, why do all bots have cameras? Couldn’t some be text-only terminals? This is another area of the game that feels artificial, instead of innate to the story.
• This might be a bug: Doko Air currently allows you to put robots you don’t currently have in your possession on a flight. While that was useful for retrieving a bot I sent to Singapore, it doesn’t really make sense to allow it.
• The scanner is kind of annoying to work with when I’m walking for more than a block or two, because every time my phone goes to sleep, the view zooms back out and in again to find my location. Unless the scanner has been sleeping for a while, it makes sense to start from my previous location and slide the map over to the new spot.

So, how might I fix the things I didn’t like about the gameplay?

The first one is that I would hire a couple of writers to craft interesting adventure stories for bots that players could participate in. There’s an underlying problem called participation inequality: the proportion of users who are deeply engaged in creating content is much smaller than the set who are interested enough to browse around a little. Many people might install Dokobots and scan around a couple of times, but far fewer will add several journal entries, actively seek out other people’s bots, and so on. You need a pretty big player community for that to generate enough good content to keep the more casual players coming back, but providing your own story content for people to latch onto helps bootstrap this.

What might those stories look like? Maybe my DrinkyBot needs to get to Toronto to meet the Cocktail Mixing Master, and while it’s there, the two of them won’t just take pictures of the drinks, but collect recipes, then go back out into the world to share the details of that fabulous Negroni. Or there’s a little bot who lost its puppy (also another bot) and players need to use their scanners to help find out where the puppy is, then get the bot to the same place. These don’t have to be terribly complex, just detailed enough to give people something to play with. It could even require that players have to go online and try to recruit other people in the right city to help out. I’ve seen techniques like that work well as part of ARGs. And a locative game like this is a kind of alternate reality experience: there aren’t really robot parts all over the street, but we’re going to pretend there are and see what happens.

The last thing I would do is spend some time thinking about how the game story fits into the player’s world. If my neighborhood’s streets are scattered with electronic rubble, isn’t someone going to want to clean that up? What does it look like when there are several dokobots in my local pub, or at the soccer game? Is everyone okay with helping these tourists along? Do the bots ever get bored and want to go home (but they can’t, because their ship exploded)? The game as it stands now is basically a collecting activity: you get more batteries, and different types of bots to activate, and sets of pictures and paths on a map. Once you’ve collected all the things you can, what happens next?

The meetings have a similar structure to most user groups: there’s a technical talk or workshop given by one of the members, and afterward the group moves on to a local pub for the “splode” part of the evening, a general discussion of whatever we want to talk about from our professional and personal lives. The guidelines for men’s participation have changed over time; the current rule is that all women are invited to attend and participate, and men are welcome as the guest of a female participant. The goal is to create a safe space for women to talk about our technical work in a friendly and open environment.

Why safe space? It’s no secret that the technology industry can be antagonistic, heavily competitive, and hostile toward outsiders who don’t immediately prove their technical competence. While not all women have experienced problems with this, many women (and men) have, and being a member of a visible gender minority often only increases that sense of otherness and hostility. Creating a group that is officially women-focused and has clear guidelines for inclusive behavior provides support in that environment. Men in the industry are part of a majority group; there are no defined “men’s user groups” or “men’s software conferences” because in many cases, that’s already what happens by default. In the open source world in particular, women make up a very small percentage of participants, and I find that having space where I feel visible and normal and not weird for being female is extremely valuable.

Besides that, having a cross-technology group where we can discuss our work with our peers is exciting. I’ve used this to present topics that don’t fit into any single-technology user group, like text game programming or privacy issues in software design, and I’ve attended other topics that don’t relate to my own work at all, but are interesting because of the enthusiasm of the women who are presenting it. It’s also a great environment for women who have never given a technical presentation before, and might not feel like they have the experience to present at a conference yet. This helps us learn the skills to move on to bigger events, meaning that over time we can encourage better representation of the women who are already present in these fields.

The best thing, the part that really demonstrates why this group is important, is seeing the effect it has on the women who participate. We’re learning to negotiate for better salaries, to find jobs we enjoy, to present our work to a larger audience, and we’re building friendships we can rely on when the problems of our industry are hitting us personally. It’s not a “fix” for the imbalances of the technology world, but a support system that helps us continue to be a part of this industry and still have fun.

Does this sound like something you want to participate in too? We meet every 4th Tuesday of the month, and you can find details on the group’s website or Calagator.

I’ve been in San Francisco (and nearby) for most of the last week, primarily to attend this year’s WhereCamp, which was held at Stanford. I had a ton of great conversations about data, privacy, transit systems, community organizing, and how Portland is or isn’t like other places.

I also led a session on geo-games (games with a locative element, digital or otherwise) and design strategies, mechanics, and reward systems. We started off with a list of these items that I had gathered from earlier reading, and added more items and other details on the whiteboard through our discussion. Amber Case made a spreadsheet from the whiteboard notes: https://spreadsheets.google.com/spreadsheet/pub?hl=en&hl=en&key=0Asy5qINtkUStdEZPeE5mODVIa1hCV2tlOUFPZ3JxTkE&output=html

We couldn’t talk about games without playing a few, of course. Amber and Aaron ran a couple of demos of MapAttack, which is built on their GeoLoqi platform, and lets two teams compete for points by capturing dots while running around outside. At another session I was introduced to Dokobots, an iPhone game involving lost robots who crashlanded on Earth and have to be found and re-activated and sent off to explore things.

WhereCampPDX will be returning again this year as well. We’re tentatively scheduled for October 7-8-9 and ought to have more details on the site soon.

Matt asked in comments: “I’d love to see how this compares to the bay area.”

For the San Francisco-Oakland-Fremont metro area, the mean annual wage for Computer and Mathematical Science Occupations is $91,440. (The Portland metro area number for that was $74,890).

San Jose-Sunnyvale-Santa Clara is a separate metro for this report, and the mean annual wage there is a little higher: $109,130.

For more west coast comparison, the Seattle metro comes in at $87,620.

In comments on Hacker News, a couple of people expressed surprise that developers wouldn’t know how their wage compared to the local average. It may seem obvious if you’re already paying attention to this yourself, or work for a company that uses cost of living and prevailing wage to determine compensation, but many developers really aren’t aware of this information. I’ve seen that to be especially true for younger workers, or those who haven’t had a lot of mentoring in their career. It’s very common to assume that what a company offers you is an average amount to be paid for that job, and that other people around you are earning a similar amount. This happens even within individual companies.

Hacker News also had some interesting discussion of whether wage variance from city to city was related to how companies in those places were valued by investors, i.e. if investors penalize a company from being in Portland, that should affect wages, but if companies are getting an equal level of investment, then wages should match other cities. One of the things I’ve been wondering is whether wage variance in Portland is partly related to company revenues: I’ve seen a fair amount of anecdotal evidence that employers that are paying developers less than other local competitors are also doing less well as a business. I don’t really know where to find the data to study either of these issues, though.

Another thing on HN was a discussion of whether a company paying developers in different cities the same amount is fair. Alex from BankSimple says that they decided it was, but other people thought it penalized developers who lived in more expensive places. There’s several issues wrapped up in this: is the market for developer talent local, national, or international? Do developers have sufficient freedom to move or choose where they live that the trade-offs between bigger and smaller (more expensive and more affordable) cities is a fair choice? Much of this depends on the specific company, but given that programming can often be done from anywhere, without requiring the same built resources as other industries, this is worth discussing.

One last thing: I think all of this highlights how useful an annual wage survey for the software industry would be. I’ve been looking at the AIGA survey of design salaries and what they report. They break their data by categories like type of company, location, and whether the company’s client base is local or national or international. It makes it really easy to find out what someone at your job level, working for a similar company, could expect to earn.

QCEW data is split by industry, so programmers could potentially be covered under any number of areas, depending on where they work, but we can pick a likely area to start, like NAICS code 54151 (computer systems design and related services). In 2009, the last full year with data, there were 9,457 people working in this industry category in Oregon [2]. The total covered wages for the year added up to $722,755,908. So we can divide the two and conclude that the average annual wage was about $75.5k.

What if we want to know more about income by job title? For that we need the BLS Occupational Employment Survey. Here’s the May 2009 data for “Computer and Mathematical Science Occupations”: http://bls.gov/oes/current/oes_38900.htm#15-0000

The national median annual wage for this category is $72,900. In the Portland metro area, it’s $74,890. You can see that broken down further. For example, a research programmer averages $102,930 here, and the most popular job area by employment count, “Computer Software Engineers, Applications”, earns $91,300 a year.

The breakout pages that show you more about the data for each job category are pretty interesting, too: http://bls.gov/oes/current/oes151031.htm
You can use that to find out things like which industries hire the most people with this job role, and which metro areas pay the best.

What other ways can we find out about developer wages? Well, last year Eva and I did a survey of the local tech industry that included questions about this. The median annual income [5] for our survey sample (667 people, but not everyone answered all questions) was $75k. The median for people who described their job role as a software developer of some sort [3] was $80k. The range, though, was from about $25k on the low end to $200k at the top, which seems pretty significant. 46 respondents were earning less than $65k/year, and 43 earned $100k/year or more.

There’s other kinds of surveys, too. This week a Hacker News discussion [4] led to someone coding up a little webapp that lets you ask a group of people how much they earn without revealing who entered what. It only displays the results if at least four people submit their info. Christie asked people on IRC and Twitter to fill one out for Portland software developers. So far the responses range from $0-90k, with 15 responses (the lowest non-zero amount is $38.5k). It’s still too small a sample to know how it compares to our other data, but if you’re a local developer you can add your own salary and see the responses here: http://salaryshare.me/5b9f602f99f87ea1492cc056292000ae

You can also get some employee-reported data on wages through a site called Glassdoor. Here’s a search for software workers in Portland: http://www.glassdoor.com/Salaries/portland-software-salary-SRCH_IL.0,8_IM700_KO9,17.htm
It lets you see things like the average salary for a Software Engineer at Intel is $85,427, while a Software Developer at Rentrak earns $61,500. Might be useful if you’re trying to decide which job postings to respond to, assuming they have data for the company.

Back to that variation issue: what sorts of questions can we ask about why some developers in Portland earn way more (over twice the annual wages) than others? It seems likely that there are factors including Portland’s local economy and business markets, individual characteristics like skills and education, and the specific technical markets a person is working in.

Possible reasons for higher wages:

• More education or experience
• Working for larger or more profitable companies (including companies with a national or international customer base)
• Good negotiation skills and knowledge of market wage rates
• Working in technology sub-fields where developers are in higher demand (such as iPhone app development)
• Having skills or experience areas that are particularly desirable in the current market

Possible reasons for lower wages:

• Less education or experience
• Working for smaller or less profitable companies
• Poorer negotiation skills or knowledge of market wage rates
• Working in technology areas where compensation tends to be lower (PHP web development, for example)
• Working independently or for a company that is struggling with funding or profitability
• Having skills or experience areas that are fairly common in the current market

So that’s what I have so far. It’s kind of a big data dump. I’m not an expert on this by any means, but if you have any questions about other data, who’s included in which numbers, and so on, let me know and I’ll try to find out and report back.

[1] The most notable exception is self-employed workers, who aren’t in this report because they’re exempt from unemployment insurance, which is the tax data used.

[2] Yearly average, based on monthly numbers.

[3] Software developer, software engineer, programmer, web developer, etc. The count of responses for these job titles with usable income entries was 189.

[4] If you’ve curious about the potential benefits and disadvantages for workers and companies to disclose salary info, this discussion pretty much has all the points covered.

[5] Total personal income, rather than wages, so some people may have reported a higher income due to other sources (like investments or a side business) than they would in the government wage data.

Note: Code ‘n’ Splode is a women-focused technology user group. If you need more details about the group and our guidelines for participation, see here.

It’s interesting, as someone who hasn’t paid more than casual attention to this area of games since childhood, to see where it’s ended up. There’s a very lively online community of people working in text games, but it’s primarily amateur, unlike most other areas of game creation [1]. It also seems pretty insular, in the same way that leads Elizabeth Bear to describe science fiction short story writing as a club scene [2].

One of the general things that caught my attention is that there’s a ton of informed, well-considered discussion within this community about storytelling in games, and about game structure and technique in general, but the scope of games in practice seems much narrower. If, like me, you’re not likely to touch anything involving medieval fantasy, fairy tales, or steampunk, that can really limit your options. I’m finding that the text nature of the game means I’m much more likely to judge the game on whether I find the story compelling, not just whether I find the gameplay enjoyable.

Most of the things that normally combine to make play compelling are absent here, as well. There’s rarely a timing or physical performance issue [3], moves can be made at any speed the player desires, and while many games do award points for solving puzzles, that seems to have more of an effect on the sense of pacing (if you have 4 out of 6 tokens, you’re about that close to finishing).

Anyhow, enough general babble. Here’s what I’ve been playing and what I’m using to play it.

Games:
Lost Pig. This is my favorite so far, by a long shot, so I’m listing it first. You’re an orc, and you have to get the missing pig back to the pig farm where you work before daybreak or you’ll lose your job. It’s funny, well-written, and IMO the perfect size and scope for this sort of game. It also succeeds at one of the things that cause many text games to fall flat for me: everything you do enhances the story, instead of only helping to solve the game puzzle.

The Dreamhold. I started here because it was rated newbie-friendly, and I’d agree with that. I didn’t find the story very interesting, but the gameplay was fine, and after getting through a couple of puzzles I was interested enough to finish.

9:05. Short and meant to be re-played to get the full tale. It’s not hard, and has a funny twist.

Aisle. Even more than 9:05, this is primarily an exercise in storytelling, rather than a full game. You get to do one thing, something will happen, then the game starts over. You build up the whole story by reading what happens as you try different things. It takes place in a grocery store.

Violet. This is a game where you’re trying to work on your dissertation, but distractions keep getting in the way. Violet is the girlfriend who is going to leave you if you don’t get the writing done. There were many things I liked about this one, but for me the pieces were better than the whole. I got frustrated several times when the game told me I couldn’t resolve a puzzle a certain way because Violet didn’t approve. In the end I wanted to give up on the PhD, dump Violet, and go play in the park with the zombies and mole men.

Anchorhead. I really wanted to like this, because I’m a Lovecraft fan and it was highly rated on IFDB. Alas, I had multiple “throw the book across the room” moments within the first hour. The game map is wide open at the start, so if you’re not clear on what it expects you to do, you can wander off and completely screw up your ability to continue [4]. On round 2, I found the PC’s husband in the university library, attempted to talk to him to figure out what I should be doing, and wound up so frustrated I wanted to punch him (the game insisted that was out of character). Eventually I consulted the walkthrough, which told me my first move was to do something that seemed entirely out of character for someone who is dressed in “a tasteful ensemble from Ann Taylor” [5]. So I yelled at the game for a few more minutes then quit.

Enough games. Here’s the other part, about software for playing and writing them.

The first thing you should know is that IF games are generally not distributed as a complete executable package. Instead there are game files in formats like z-machine and glulx, which you open in an interpreter. There’s a push to make the most common game formats easily playable in a web browser through Javascript or Flash-based interpreters, which lets authors link to a version of the game on the web ready to go, but this doesn’t appear to be the default approach currently.

The main interpreter I’m using is called Frotz. It’s available as an iOS app that comes preloaded with several games to try, and you can download more through IFDB. For me, playing on a mobile device instead of my laptop is ideal, because I like to get away from the computer at the end of the day. I’ve also tried a couple of games using Parchment, and I downloaded Spatterlight but don’t have anything loaded in there yet.

To create games, I downloaded a program called Inform. Inform is very neat. It’s a natural-language DSL for creating games (z-machine or glulx), and so far it hits that sweet spot between ease of use and power very well. I’m currently in the middle of working through the tutorial by writing a model of my apartment. Next I plan to do something with zombies. [7]

More results to come.

[1] The electronic ones, anyhow. Is there a good category term for games that includes console, internet, and computer ones? Software games? I know there’s a certain amount of non-commercial activity in tabletop and RPG gaming, so I want to distinguish between that and this.

[2] Meaning a community where work is created by and for the participants.

[3] Depending on how well you type and spell, I suppose.

[4] I got lost in a twisty maze of streets, all alike.

[5] Obvious clue the game was written by a guy: the description of the PC’s clothing doesn’t mention shoes at all, but everything else suggests she ought to be wearing very plain pumps (probably in beige). If I were climbing on trash cans in anything with a heel, you can bet I’d notice, but the game blows right past that detail without remark. [6]

[6] There were many other things I found annoying, like not being able to get drunk or glare at people, but if I keep going I’ll have to rename this post to “Why Anchorhead Sucks and Everyone Who Gave it 5 Stars is Wrong”.

[7] Zombies and unconferences and learning to kill zombies at the unconference. I suspect the audience for this game will be about three people, but what’s wrong with that?