DDD is one of the few regular software developer conventions in Perth, so it was very well attended with tickets selling out well before the day of the event. The concept is that it is organised by developers, for developers to attend and talk, and the subjects are selected democratically by the developer community, hence the name “DDD” — “Developer Developer Developer”.  There were six of us from Gaia Resources in attendance, including myself as an elected presenter, and there were some common threads to our responses to the day.

Ben in full flight at DDD 2017

Highlights

Unsurprisingly, the keynote and locknote speakers were well-received.

Gojko Adzic discussed in his keynote the challenges of developing and specifically controlling quality of cloud service based applications.  In particular, reliance on external services gives us some unique challenges compared to previous generations of software that were a lot more self-contained.  This leads to ideas such as testing in production, something that was previously frowned upon, and challenging our assumptions about data accuracy, especially as a result of automation gone wrong.

Kris Howard‘s locknote was about the social, rather than the technical, aspects of development, a trend that was present throughout the day with many non-technical topics being well-received.  Her concept of “leaving the tech industry better than we found it” aka “the campsite rule” highlighted the advantages of being a mentee or mentoring as a way of “giving back” and improving the tech industry, included useful suggestions on how to find and ask someone if they’d be willing to be your mentor.

Overall the day was well organised, well catered, and well run.  The talks were fairly diverse and informative, the staff and attendees were really friendly and inclusive, and you could tell that a lot of personal effort was put into making it a successful community event.  Of course, meeting people in the industry and catching up with other developers was a highlight. There was also a lot of genuine excitement about DDD itself and its growth and ongoing success.

Some other highlights for the team included Ian Hughes’ talk, “what a real sprint looks like”, Patima Tantiprasut’s talk about “Web, Wellness and Getting Sh*t Done”, Rami Ruhayel’s gentle introduction to neural networks, and Rob Crowley’s introduction to GraphQL.

 Tony, Kehan and Shay captured “in the wild” at DDD

Serverless Architecture at the Small Scale

My talk was about “Serverless Architecture at the Small Scale”, and it turned out to be a popular topic, with the room overflowing with audience members – standing room only!  This was my first time talking at a conference, and it was quite an experience.  There is a lot more preparation work that goes into a presentation than you might expect; all up I spent about a week preparing, practising and polishing my talk!

The “small scale” aspect is also very relevant to us at Gaia Resources.  Our work with citizen science groups, small collections, and researchers, is very much at the small scale, especially when compared to “large scale” users like Netflix, Target, Amazon, etc.  In fact a lot of the projects we deliver fit comfortably within the “perpetual free tier” offered by many of the cloud services.

We have used serverless technology on quite a few projects so far, including multiple survey applications (of the question-and-answer kind, as opposed to the biological survey kind), with a serverless API behind single-page application (SPA) frontends.  These projects have been delivered for clients such as the UWA Centre for the Built Environment and Health, BirdLife Australia, and the Clean Air Urban Landscapes hub.  Several other projects utilise serverless technology, including the Tasmanian Collection and the WA New Music Archive.

Demonstration

The main part of my DDD talk consisted of a live demonstration, which despite some technical issues, was successful in the end.  In fact I think the audience appreciated my perseverance and eventually making it work, and I received a lot of positive comments afterwards, both in person and through the conference’s feedback system.  The demonstration used some very simple code and the Serverless framework to receive an image from an S3 bucket, normalise it and add a watermark, then save the image in a different S3 bucket.  An email was then generated, and any images were accessible via an API created using API Gateway.  There is a GitHub repository containing the code that I used, if you want to check it out.

The concepts were demonstrated using a photographic image, because it was visual and therefore interesting for the audience.  However this is a technique that we have used successfully with uploaded CSV data files, processing the data and storing it in a database or sending it to another API.  The code required is very minimal, delivering a great effort-to-reward ratio.  I hope to do more such presentations in the future.

Outcomes

The various talks presented during the day, and our own discussions with each other and developers outside of Gaia, have given us a lot of takeaway messages to consider.  Having the opportunity to hear some of Perth’s best developers talk about their experiences and recommendations is an excellent opportunity to get away from our day-to-day work and find some fresh perspective.

DDD Perth was a great experience and everyone from Gaia who participated, as well as many others that I spoke to on the day, are keen to participate in next year’s conference.  If you’d like to know more about the talk or what we do at Gaia, feel free to email me.

Ben

The post Serverless Architecture at the Small Scale appeared first on Gaia Resources.

This project was built using our SurveyEngine toolkit, similarly to the recent Black Cockatoo Community Wisdom Survey (see our recent blog on that project here).  The project is summarised really well on their landing page:

This survey will ask you questions about your use of parks and green spaces (i.e., no entry fee or membership required) in your local area and around Perth. We are interested in what areas you visit and for what activities.

We are currently testing our survey and require people over the age of 18 to complete the 10 minute survey at two different time points that records details of recent park use. 

If you agree to take part, we will ask you for your email address so that several days after first completing the survey we can contact you again and ask you to complete the survey for a second time. The purpose of this is to test our data collection methods and evaluate our survey design.

The survey starts like a normal survey; it asks you to register and then sends you a “please validate your email address” email to get started.  The system then assigns you to one of two groups of users – one who gets maps, and one who doesn’t.  This is to support some A/B Testing that tests if the use of web maps is beneficial (and there will be some research on this in the future from CBEH).

The address selection with a nice mapping interface and address (or junction!) geocoder

Once validated, you then put in your home address (or nearest street corner) which is then georeferenced on the first map.  From here, you indicate if you’ve been to any parks in recent times, and the system then either gives you a map or an input box to pick a park, centered on your address, and then you fill out some information on your use of that park.

The component of the survey about your park use

Once you’ve filled out your details, and some demographic information, then the survey ends.  And three days later, you get another email reminder to come back and fill out some more information.  The team at CBEH then can hit a specific secured URL and the system will generate a data dump of the responses for them, which nicely closes the loop on the data.

The system is another SurveyEngine implementation, and as part of our ongoing evolution, it’s actually been developed in serverless infrastructure, using a range of services in Amazon Web Services (AWS).  This is a pretty new set of technologies (you can read more about the AWS serverless offerings here), and we’re using this in a few different areas already, like these SurveyEngine projects.  We even had Ben run an internal “What is Serverless” session at our regular drinks on a recent Friday which has us all thinking about the various possibilities of this technology stack.

Anyway, CBEH has been up and running for some time at https://survey.perthgreenspace.com.au/ – if you live in Perth, go and record some information about your use of the parks and green spaces in your area – and go into the draw for a $250 Coles/Myer voucher!

If you want to know more about how we can help you with custom mapping based surveys, or about serverless technologies, then feel free to drop me a line (andrew@gaiaresources.com.au) or start a conversation with us on  Facebook, Twitter or LinkedIn.

Andrew

The post Parks and Green Space Surveys appeared first on Gaia Resources.