AV-Portal 3.23.3 (4dfb8a34932102951b25870966c61d06d6b97156)

Lightning Talks - FOSS4G SotM Oceania 2019

Video in TIB AV-Portal: Lightning Talks - FOSS4G SotM Oceania 2019

Formal Metadata

Title
Lightning Talks - FOSS4G SotM Oceania 2019
Title of Series
Author
License
CC Attribution 3.0 Unported:
You are free to use, adapt and copy, distribute and transmit the work or content in adapted or unchanged form for any legal purpose as long as the work is attributed to the author in the manner specified by the author or licensor.
Identifiers
Publisher
Release Date
2019
Language
English

Content Metadata

Subject Area
Abstract
This is the lightning talk session from FOSS4G SotM Oceania 2019, organised by OSGeo Oceania and held at The National Library in Wellington, New Zealand from November 12-15 2019. 0:00:17 · Alex Lum - The World of Wikidata 0:05:53 · Andrew Harvey - Practical Mapping Tips 0:11:18 · Brent Wood - A new paradigm in web visualiation of species data 0:14:58 · Caitlin Adams - Lessons from a first brush with user experience design 0:20:16 · Chris Morgan - Agile Delivery in the Geospatial Industry at FrontierSI 0:25:15 · David Garcia - How to Map an Island in the Pacific 0:30:55 · Ewen Hill - Mapping in the International Year of Indigenous Languages 0:36:24 · Hamish Campbell - QGIS Geometry Generators 0:41:50 · Jessica Leiria Schattschneider - An open-source geospatial framework for beach litter monitoring 0:47:13 · Joeli Varo - A great giant of isolation in Papua New Guinea! 0:52:01 · Kamsin Raju - Challenges of Mapping Informal Settlements 0:56:08 · Michael Speth - High Performance Map Cache Generation 1:01:32 · Nyall Dawson - (Not) Merry Christmas, QGIS! 1:07:26 · Richard Law - How I use Docker and Make for reproducible workflows 1:12:52 · Rose Phillips - Point Clouds, Pixels and Open Source Awesomeness – Investigating LiDAR Data 1:18:23 · Wing Ho - Offline Maps in the browser! FOSS4G SotM Oceania is the coming together of Oceania's geospatial open source and open data community - with four days of workshops, presentations, a community sprint and social events.

Related Material

Loading...
Aliasing Building Freeware Inheritance (object-oriented programming) Identifiability Open source Variety (linguistics) Capillary action Virtual machine 1 (number) Database Data storage device Open set Mereology Code Area Wiki Different (Kate Ryan album) Ontology Diagram Statement (computer science) Descriptive statistics Area Programming language Knowledge base Key (cryptography) File format Lemma (mathematics) Projective plane Data storage device Bit Database Instance (computer science) Set (mathematics) Similarity (geometry) Category of being Type theory Predicate (grammar) Network topology Statement (computer science) Website Convex hull Personal area network Quicksort Object (grammar)
Code Texture mapping System administrator Time zone Mereology Semantics (computer science) Area Query language Web 2.0 Wiki Order (biology) Query language Circle Diagram Multiplication Descriptive statistics Point cloud Programming language Service (economics) Link (knot theory) Geometry Texture mapping Open source Translation (relic) Virtual machine Linker (computing) Text editor Quicksort Point (geometry) Aliasing Web page Identifiability Service (economics) Programming language Open source Link (knot theory) Linked data Twitter Radius Maschinelle Übersetzung Tesselation Weight Expert system Code Set (mathematics) Visualization (computer graphics) Query language Mixed reality Point cloud
Point (geometry) Mobile app Type theory Group action Mobile app Texture mapping Information Texture mapping Field (computer science) Bit Complete metric space Field (computer science)
Area Mobile app Android (robot) Mobile app Group action Texture mapping Texture mapping Mobile Web Android (robot) Bit Open set Field (computer science) Digital photography Digital photography output Conditional-access module
Web page Default (computer science) Asynchronous Transfer Mode Satellite Texture mapping Open source Information Validity (statistics) Capillary action Open source Open set Tracing (software) Wave packet Neuroinformatik Wiki Type theory Wiki Software Different (Kate Ryan album) Term (mathematics) Text editor Error message Reading (process) Asynchronous Transfer Mode
Area Web 2.0 Mathematics Programming paradigm Email Information Visualization (computer graphics) Electronic mailing list Website Species Descriptive statistics
Web page Programming paradigm Web portal Set (mathematics) View (database) Plotter Direction (geometry) Interactive television Range (statistics) Archaeological field survey Maxima and minima Content (media) Plot (narrative) Number Data management Prototype Hierarchy Graph (mathematics) Visualization (computer graphics) Drill commands Species Website Pairwise comparison Address space Physical system Pairwise comparison Noise (electronics) Distribution (mathematics) Variety (linguistics) Information View (database) Archaeological field survey Web page Maxima and minima Hierarchy Maize Visualization (computer graphics) Series (mathematics) ECos Species Quicksort Prototype
Web page Programming paradigm Email Open source Distribution (mathematics) Dependent and independent variables View (database) Multiplication sign Range (statistics) Electronic program guide Perspective (visual) Emulation Twitter Locally compact space Different (Kate Ryan album) Visualization (computer graphics) Species Integrated development environment Information Pairwise comparison Design of experiments Address space Distribution (mathematics) Email Dependent and independent variables Graph (mathematics) Texture mapping View (database) Prisoner's dilemma Archaeological field survey Bit Sign (mathematics) Personal digital assistant Series (mathematics) Species Quicksort Arithmetic progression
Group action Code Variety (linguistics) Decision theory Archaeological field survey Open set Content (media) Cube Energy level Data conversion Curve Email Variety (linguistics) Inheritance (object-oriented programming) Digitizing Consistency Archaeological field survey Interactive television Mathematical analysis Electronic mailing list Bit Digital signal Open set Process (computing) Order (biology) RWE Dea Quicksort
Group action Building Transformation (genetics) Multiplication sign Disintegration Open set Prototype Iteration Different (Kate Ryan album) Software testing Information Website Design of experiments Curve Collaborationism Beta function Information View (database) Digitizing Software developer Point (geometry) Digital signal Group action Connected space Process (computing) Cube Website Software testing Iteration Quicksort
Suite (music) Building Multiplication sign Workstation <Musikinstrument> 1 (number) Design by contract Student's t-test Mereology Regular graph Staff (military) Computer programming Field (computer science) Data management Frequency Mathematics Different (Kate Ryan album) Software Program slicing Circle Software testing Software framework Extension (kinesiology) Form (programming) Predictability Multiplication Focus (optics) Scaling (geometry) Texture mapping Information Software developer Projective plane Feedback Mathematical analysis Planning Bit Staff (military) Extreme programming Continuous function Data management Personal digital assistant Self-organization Video game Conditional-access module Iteration Right angle Cycle (graph theory) Resultant
Point (geometry) Time zone Greatest element Digital photography Texture mapping Multiplication sign Expert system Mereology Family
Stylus (computing) Dependent and independent variables Texture mapping Multiplication sign Projective plane Bit Mereology Word Wave Video game Right angle Office suite Library (computing) Descriptive statistics
Heegaard splitting Context awareness Word Texture mapping Human migration Right angle Data conversion Mereology Family Navigation
Programming language Execution unit Texture mapping Texture mapping Multiplication sign Shared memory Water vapor Tracing (software) Numeral (linguistics) User profile Digital photography Inclusion map Numeral (linguistics) Pattern language Right angle Information Representation (politics) Local ring Family Metropolitan area network Family
Point (geometry) Geometry Metre Pattern recognition State of matter Set (mathematics) Multiplication sign Channel capacity 1 (number) Topological algebra Mereology Entire function Rule of inference Field (computer science) Area Geometry Goodness of fit Square number Physical law Office suite Local ring Area Rule of inference Programming language Dialect Metre Texture mapping Information Tesselation Building Software developer State of matter Bit Hecke operator Set (mathematics) Line (geometry) Symbol table Symbol table Type theory Category of being Factory (trading post) Right angle Family Electric generator Local ring Determinant Flag
Geometry Metre Dot product Code Multiplication sign Polygon Menu (computing) Topological algebra Function (mathematics) Flow separation Symbol table Number Root Core dump Buffer solution Problemorientierte Programmiersprache Quicksort
Point (geometry) Ocean current Geometry Randomization Texture mapping Polygon Bit Line (geometry) Topological algebra Cursor (computers) Symbol table Goodness of fit Loop (music) Circle Right angle Series (mathematics)
Geometry Programming language Curve Angle Polygon Arrow of time Damping Quicksort Line (geometry) Endliche Modelltheorie Set (mathematics) Inflection point
Point (geometry) Geometry Texture mapping Hoax Graph (mathematics) Equaliser (mathematics) Polygon Open source Shared memory Raster graphics Symbol table Query language Different (Kate Ryan album) Universe (mathematics) Software framework Iteration Damping Software framework
Quantum entanglement Standard deviation Integrated development environment Different (Kate Ryan album) Multiplication sign Open source Mathematical analysis Form (programming) Form (programming)
Point (geometry) Group action Computer file Code Computer-generated imagery 1 (number) Shape (magazine) Mereology Field (computer science) Attribute grammar Medical imaging Different (Kate Ryan album) Vector space Form (programming) Texture mapping File format GUI widget Projective plane Electronic mailing list Attribute grammar Open set Product (business) Digital signal processing Type theory Process (computing) Software Software framework Resultant
Time zone Distribution (mathematics) Mobile app Texture mapping Computer-generated imagery Time zone Shared memory Data analysis Database Product (business) Product (business) Medical imaging Latent heat Mathematics Lie group Pattern language Information Website Form (programming)
Group action Presentation of a group Integrated development environment Hypermedia State of matter Order (biology) Vulnerability (computing) Local Group
Area Point (geometry) Collaborationism Programming language Dialect Group action State of matter Channel capacity Artificial neural network Archaeological field survey Attribute grammar Food energy Group action Attribute grammar Data management Strategy game Software Telecommunication Telecommunication Strategy game Remote procedure call Implementation
Building Vapor barrier Open source Texture mapping Artificial neural network Materialization (paranormal) Mathematical analysis X-ray computed tomography Raw image format Number Attribute grammar Goodness of fit Telecommunication Cuboid Relief Programming language Texture mapping Information Channel capacity Open source Mathematical analysis Attribute grammar Connected space Data mining Computer configuration Integrated development environment Telecommunication Order (biology) Connectivity (graph theory) Musical ensemble
Area Point (geometry) Random number Crash (computing) Information Software Open source Crash (computing) Multiplication sign
Point (geometry) Geometry Dot product Texture mapping Information Link (knot theory) Open source Open source Electronic mailing list Open set Field (computer science) Type theory Medical imaging Term (mathematics) Energy level Data logger Local ring Resultant Local ring
Area Point (geometry) Presentation of a group Texture mapping Software developer Projective plane Topological algebra Cache (computing) Process (computing) Cache (computing) Different (Kate Ryan album) Information Local ring Local ring
Slide rule Complex (psychology) Presentation of a group Server (computing) Service (economics) Connectivity (graph theory) Virtual machine Database Data storage device Data analysis Neuroinformatik Subset Web 2.0 Cache (computing) Single-precision floating-point format Physical system Texture mapping Server (computing) Projective plane Bit Cartesian coordinate system Flow separation Web browser Single-precision floating-point format Cache (computing) Integrated development environment Software Website
Statistics Texture mapping Tesselation Scaling (geometry) Transport Layer Security Multiplication sign Zoom lens Projective plane Database Instance (computer science) Subset Cache (computing) Type theory Web service Process (computing) Befehlsprozessor Cache (computing) Befehlsprozessor Computer hardware Energy level Process (computing) Local ring
Intel Building Greatest element Texture mapping Building Multiplication sign Projective plane Bit Staff (military) Perspective (visual) Information technology consulting Supercomputer Type theory Hochleistungsrechnen Virtual reality Process (computing) Computer configuration Software Telecommunication Mathematical singularity File system Physical system
Service (economics) Link (knot theory) Open source Scaling (geometry) Multiplication sign Real-time operating system Mass Web 2.0 Estimator Befehlsprozessor Information Data conversion Multiplication Installable File System Computer architecture Physical system Texture mapping Projective plane Information technology consulting Supercomputer Cache (computing) Process (computing) Befehlsprozessor Telecommunication Mathematical singularity Vertical direction Website Near-ring
Group action Email Computer file Software developer INTEGRAL Multiplication sign Software developer Projective plane Control flow Translation (relic) Group action Mereology Number Revision control Different (Kate Ryan album) Term (mathematics) Finite-state machine Quicksort Local ring
Web page Multiplication sign Projective plane Open source Line (geometry) Template (C++) Demoscene Template (C++) Medical imaging Latent heat Data management Message passing Event horizon Right angle Quicksort
Statistical hypothesis testing Execution unit Group action Server (computing) Server (computing) Software developer Multiplication sign Projective plane Memory management Stress (mechanics) Group action Area Number Revision control Resource allocation Queue (abstract data type) Speech synthesis Damping Block (periodic table) Freeware
Computer icon Programming language Easter egg <Programm> Feedback Electronic mailing list Maxima and minima Compiler Twitter Revision control Computer configuration Revision control Damping Process (computing) Task (computing) Window
Group action Building Context awareness Greatest element Pixel Code Confidence interval Multiplication sign View (database) Function (mathematics) Open set Mereology Information technology consulting Subset Neuroinformatik Software bug Avatar (2009 film) Medical imaging Repository (publishing) Process (computing) Endliche Modelltheorie Pixel Sanitary sewer Point cloud Scripting language Email Software developer Point (geometry) Moment (mathematics) Open source Shared memory Metadata Sound effect Parameter (computer programming) Lattice (order) Control flow Complete metric space Variable (mathematics) Entire function Chain IRIS-T Configuration space output Software testing Right angle Quicksort Resultant Writing Stability theory Geometry Service (economics) Observational study Link (knot theory) Open source Software developer Computer-generated imagery Repetition Machine vision Field (computer science) Twitter Revision control Graph (mathematics) Authorization Reduction of order MiniDisc Plug-in (computing) Matching (graph theory) Server (computing) Interface (computing) Projective plane Mathematical analysis Letterpress printing Set (mathematics) RAID Symbol table Mathematics Integrated development environment Software Personal digital assistant Boom (sailing) Point cloud Library (computing)
Point (geometry) Identifiability Computer file Open source Multiplication sign Correspondence (mathematics) Range (statistics) Archaeological field survey Design by contract Port scanner Device driver Metadata Number Element (mathematics) Heegaard splitting Latent heat Read-only memory Single-precision floating-point format Information Traffic reporting Social class Point cloud Programming language Noise (electronics) Email Information Validity (statistics) File format Point (geometry) Projective plane Metadata Plastikkarte Coordinate system Attribute grammar Maxima and minima Total S.A. Set (mathematics) Cartesian coordinate system Complete metric space Arithmetic mean Process (computing) Angle Software Point cloud Data compression
Web page Surface Noise (electronics) Service (economics) Server (computing) Cross section (physics) Pixel Open source Interior (topology) Digital signal Open set Open set Front and back ends Computer configuration Operator (mathematics) File viewer Point cloud Energy level Endliche Modelltheorie File viewer Traffic reporting Library (computing) Point cloud
Point (geometry) Slide rule Context awareness Presentation of a group Computer file Open source Link (knot theory) Software developer Multiplication sign View (database) Raster graphics Element (mathematics) Attribute grammar Wave packet Web 2.0 Differenz <Mathematik> Goodness of fit Googol Single-precision floating-point format Endliche Modelltheorie Extension (kinesiology) Point cloud Service (economics) Scaling (geometry) Texture mapping Information Demo (music) Tesselation File format Block (periodic table) Interface (computing) Digitizing Surface Software developer Point (geometry) Image warping Metadata Database Web browser Inflection point Uniform resource locator Process (computing) Raster graphics Physics Website Endliche Modelltheorie Right angle Procedural programming
Slide rule Link (knot theory) Texture mapping Structural load MIDI Virtual machine Library catalog Menu (computing) Bit Web browser Goodness of fit Error message Emulator Artistic rendering Error message Window Surjective function
Web page Game controller Server (computing) Freeware Service (economics) Computer file Structural load Multiplication sign MIDI Virtual machine Control flow Library catalog Web browser Client (computing) Event horizon Computer icon Revision control Cache (computing) Single-precision floating-point format Artistic rendering Error message Plug-in (computing) Surjective function Enterprise architecture Texture mapping Demo (music) Tesselation Control flow Cartesian coordinate system Cache (computing) Graphical user interface Googol Error message Event horizon Software Logic Quicksort Window
so I'd like to welcome first Alex Lum talking about the world of Wikipedia or sorry wiki data thank you so I'm sure I don't need to tell you what Wikipedia is but you might not know what a wicked data is you can probably tell from the name is it like Wikipedia but for data so yeah that's pretty much it [Applause] so wicked out await live in 2012 it's a free and open knowledge base that can be read and edited by both humans and machines so that's an important part of it it's machine readable it's one of the sister there's about twelve sister projects of the wikimedia foundation of which wikipedia is the best-known but there's also several others including wiki data wikivoyage Wiktionary wiki source and others wiki data contains about 67 million items and I'll show you what an item is sort Li and about 5.8 billion triples and I'll show you what they are
so wiki data is essentially an RDF resource description format database or a triple store a triple is a data entity composed of subject predicate and an object so like these these bar build diagrams so in wiki data that can be an item can be linked by a property to another item or an item could have a qualifier which links it to another item or a property which links it to a value here's a more complicated diagram of that as of the wicked out of schema but I'll just show you through a basic if you go to the website which I haven't put anywhere on here but it's wicked art org and you can get to it from any of the wikimedia sites i brought up wellington here so there's three parts i want to point out the first thing you'll see at the top there is a label there's also a description and a bunch of aliases and you can also see these languages i went oh sorry the label description end aliases in a variety of different languages wicked adder has two kinds of statements or two kinds of each item is will be one to one or one of these things or some both but generally not generally is one or the other of an instance so that is something a an item a specific object that exists or it can be a subclass which is kind of like an ontology tree and it can be so you can have a city is they is a subclass of a human settlement and so on or a capital city is a subclass of a city and so on and you'll also see a wicked out of properties and items so I mentioned those in the triple definition so as I said this is an item for the city of Wellington it's an instance of a capital city and an urban area and it's linked to other items so capital in urban area also items in wiki data and their subclasses of other things and so on and so that's how it's all built out and the last bits you want to see are the two types of of statements and statements so are those ones where there's a property an item a property and another right above or a value and identifier so this is a very important part of wiki data identifiers which are the the codes or I did or primary keys or identifies that you'll find in in other data sets and a lot of these elite there's thousands of them linked to from from wiki data which
makes those identifies make wicked out a very important part of the linked data web so this is a diagram from this website LOD - cloud net and where we could add a tweet out as one of those little circles but it joins to hundreds of other data sources or linked data sources via those identifiers so that makes it a very powerful and important part of those because you can link it to all sorts of other data sets language
wiki data is very very flexible when it comes to languages so those labels and descriptions I showed you before and the aliases can be added for any language within iso 639 - 3 code I looked up how many there were there's a think about 7,800 so it's quite a lot we're not saying every item has all those languages filled out of course but the possibility is there and you can see Wikipedia uses those for these inter wiki links or if you look at an article about something the article about the same thing in a different language on the on the left-hand side of the page that is generated from wiki data multilingual wiki data labels can be used for many other things such a simple machine translation and even localized labeling for map tiles geospatially is we've got latitude longitude points administrative territories and many others including capitals but the main advantages are linking to everything in these internal datasets query wiki data it's the delightfully named query language Sparkle which is an RDF query language for semantic data the wicked add a query service where you can put in these sparkle queries or use the visual editor builder and there's also one for OpenStreetMap [Applause] okay so next up we have Andrew Harvey of Sydney and he's going to be talking about practical mapping tips in OpenStreetMap hi everyone so this is very much aimed kind of at beginners but even if you're an expert in OpenStreetMap you don't know everything and hopefully you'll pick up a few tips so first off how do you actually map in OpenStreetMap you can go out and do some
field mapping so you know if you like to actually get outside this is a great way to do field mapping you can do it individually or with team group of people organized mapping parties a great way to do mapping and to help each other
out so you know you don't really need much equipment you can go out there with pen and paper this is exactly how I started just printing off a map and just making notes on the pen and paper but
you know since then things have evolved and we have a few mobile apps which there's there's a lot of others out there but here are just kind of a few that I've mentioned that make it easier for you to contribute to OpenStreetMap so street complete certainly go ahead and try that one out because it really makes it very easy to add in little bits of missing information someone else might have added a sporting field but they never set what type of sport is played there so it Street complete when you're walking past it it will pop up and say what kind of sport is played here it's very easy to contribute really great great for for kids as well and a few others Maps not made great for adding points of interests osm and also great for adding a whole bunch of places
to have in story map and then you spit Bruce Best XI that's Buchi sorry I practice that like an hour last night but I still wouldn't get it that's also a great app for Android that lets you go and add more detail to OpenStreetMap so it's a bit harder to use but you can add in a lot more detail and then go map is also on iOS which is a bit quite similar and then also still
field mapping you can do photo mapping so with either map Hillary or open Street cam really great tools you can setup either just with a phone or action cam to go out and collect imagery both it's useful if you want to go back later and and map features from that imagery but also it's a great reference so that other people in OpenStreetMap community it's maybe looking at a particular area maybe someone has reported an issue can go and ground truth that by looking at some recent imagery and then of course
there's armchair mapping so you know if you're really lazy and don't like to go outside you can also contribute just sitting at your desk you can use the ID
editor that's the default editor when you go edit on the top of the OpenStreetMap web page so you don't need to install any software and you can map a lot of detail and a lot of different types of features with ID and then if
you if you'll like me and then you want to use a more advanced editor I'd certainly recommend Jocelyn you can just install it on your desktop computer it works really fast and you can go into a lot of detail and so in terms of data
sources you can use aerial imagery open data GPS traces street-level imagery all these kinds of things are useful for Open Street Map mapping so what the map certainly head to the wiki of Open Street Map so just wicked OSM's org and just search for the kinds of features that you're interested in mapping now you'll be surprised probably anything that you can probably see outside there's probably a wiki page or some kind of reference to it here about how to actually map it so it's really great documentation about how to map different kinds of things so certainly just go have a read because you'll be surprised so it doesn't matter what kind of what mode you are you can map features and information not just features but information relevant for that mode and then lastly validation so a lot of people think about okay let's let's map in open street map but validation is also quite an important step it helps to catch errors maybe people make and you it generally helps improve the quality of the data so osm char is a really great tool for that you can actually set an
area or a region that you're interested in and then you'll get alerts or kind of changes that happen they'll pop up here and you can actually go and look at that change and you can go and see what they changed you can see the description of what what they changed all the tag information and then if you do notice an
issue first I'd recommend to just add a change to set comment and if you go to this website here it's a great tool to see all the change set comments that are happening in any particular country so you can also look at the Pacific Islands
there go to mailing lists or slack there's a really great community resource there so thank you very much [Applause] thanks Andrew so next up we have brentwood and he's going to be talking about a new paradigm in the visualisation of species data this was not done as a lightning talk but I have done it in five minutes before so the goal a web portal allowing
users to interactively visualize data pertaining to modern fishery surveys these are very eco system approach many species many methods a whole range of information available and it's quite complicated and the relationships between all these things is quite complicated so we wanted a tool that would allow this to happen normally if someone builds something like this you have a complicated hierarchy of web pages from as you drill down through to
find the information you want about the species you want and a sort of information about their species in its own page etc and you finished up getting
lost in the maze so we wanted this to do
as much data as possible and as few pages as possible without the data turning into noise we want to see visualize comparisons between species distributions and directive plots species centric views of data and data centric views of species and the minimum number of pages you can have in a website that's the goal it does have tabs and pop-ups so technically you can get picky about it very mouse over driven so the user is just moving the mouse to change the content etc there's the address for the prototype it's actually working and has been quite successful as you move the
mouse over the graph the species is highlighted and you can see one species you can click on a species to select it and it will appear on the map you can
click on one of the columns of data for a data driven view of all the species and on that pop-up you can click on the species to change the species again to a new current species you can look at
things like biomass trends with ranges the map is showing both the species where as is found and where as absences were somewhere it's not found which is quite critical from a bio perspective you can see the prisons ever since there and a bit more detail on that map you can see multiple species and compare the distribution of different species on the map and we still haven't changed the webpage so I'm going to do this in well
less than five minutes but the response from users so far has been essentially almost word-for-word in several cases where that's awesome it's new it's different and it's very much a work in progress it's being Fadel developed for us by Kendall Astarte so thanks Toby I hope so I'm paying for it it's plotly d3 base for it mostly any technical details see Toby not me any other questions about it and sort of how we're doing it it's my email address is there [Applause] so come into the sage now we have Caitlyn Adams and she's going to be talking about her first experiences with user experience design okay so this is not a guide to how to do UX design it's the story of how we did a little bit of UX design I'm learning more and more about this all the time and I think it's amazing and given we're here talking about open source stuff if you're not talking to you users you should be so
the thing that I work on that frontier is I it's with Geoscience Australia and it's called the digital earth Australia's sandbox the idea is that you can log on interact with code that you know does all these really cool spatial things and you can use that to learn about the open data coop and that's awesome and when I came on it didn't have a lot in it and so my job was to fill it full of really amazing remote sensing analysis but I came on and I said well what actually needs to be in there I don't know I'm super new to all of this stuff so the best thing I could think of to do was to start asking people so we started with a really
simple Google Forms survey and this was just Li mailed out point-blank to everybody on our mailing list which I think was 600 people and what we kind of discovered out of that was that people have a really wide variety of backgrounds they were government like people who heard of it who are in government or in industry from the level of decision makers who wanted actionable insights to the people who were doing the technical work and that all of those different groups had a really wide variety of needs and sort of the other thing we learnt was that people had had a pretty small exposure to the sandbox and that kind of tells you something but in order to understand it a bit more we decided to kind of contact some of those people directly and say well would you be happy to talk to us about it so we go to this user interview stage we were sort of having short you know 30-minute conversations over Skype with these people and we got to learn about what kinds of problems they're trying to solve but also we tried to ask them well how do you like to learn we were trying to one understand whether the sandbox as this sort of demonstrative tool was the right thing and from that learning about that we were able to talk directly and understand their sort of problems and needs and really this sort of consistent problem emerged the reason people weren't using the sandbox the learning curve was too steep there was no gentle introduction it was too hard to get on to nobody really knew what it did so nobody used it and that was an amazing thing to learn because
what it highlighted to us was we actually needed to build something else so what we did was we sat down this is me and Chris Morgan who's talking after me and we said well what if we could make a really simple website that was user focused that provided the information people were actually looking for and specifically could we do that so that it addressed the needs of those decision-makers who wanted to know whether they should pursue using digital earth Australia in the open data cube through that the technical users who were going to have to install the thing and then directly work with it and so once we sort of built a gentle sort of prototype we went into user testing so
we build a very rough website and then sat down with those people we'd interviewed originally and said we'll go away for ten minutes just look at the website do what you like and that was great because we'd leave them alone for ten minutes and we'd come back and we'd say well what did you look at and they kept you know telling us all these different stories that gave us so much insight into who they were as users and whether the information they wanted was accessible and then we could follow up and ask them well you know if you didn't find what you were looking for why not we clearly didn't make it obvious enough so that helped us to keep developing this and we did this as a sort of iterative process which gave us a sort of gentle you know this gentle introduction website there was a really great prototype for us and allowed us to come back to Geoscience Australia and say you know we want to work more on this education than this outreach stuff because it's what's really important so we learned a lot from engaging with our users you know I was super new to experienced user experience design and it's now pretty much all I want to read about all the time and doing that direct connection actually helped us identify that the reason people weren't using the thing that we were building is because it was too hard and that we needed to address the gap in the learning curve first before they would really do that and that was the most valuable use of our time we were able to define those different user groups and certain different information and through that iterative process we've started to really build something that's genuinely useful thank you next also from frontier si we have Chris Morgan and he's going to be talking about how frontier si approaches agile development for geospatial everybody so I wanted to really just talk about our journey going down an agile transformation at frontier sire so we started as a collaborative
research center for spatial information which were in about 14 years and so that came towards the end of that program but our partners wanted to continue doing that kind of work trying to bring government industry and research together to solve spatial problems so not frontier sized form over the last two years and we're non-for-profit or about 30 employees our project management base of the way that I was doing it was mostly waterfall delivery so it had about 1 to 2 people per project often much multiple projects per person running about 6 to 18 months per project there's a lot of changing priorities to these projects so depending on how important something wants things were moving around a bit as well as milestone and delivery based contracts and so this was mostly because we've moved from having government funding guaranteed to now having to have part in the base projects so we're looking to lock in resources and have enough of that sustainability for our business but there was a lot of interest from management to move some more toward an agile delivery for the different benefits that have been shown in some companies about moving to agile so that's around in increase quality customer focus change predictability and things like this which I have seen in larger organizations and volunteer programs been part of have seen these benefits but ok and to explain agile a little bit this is one way I like to talk about it is you go from a traditional or waterfall method is where you have you do your analysis upfront then you do all your design get that signed off you build it and then you test it and that can be fine when it's a simple project but when it's a little bit more complex you tend to find issues in testing or even when you're building it and it requires a lot of rework so it adds a little bit more about working out what are your most important features breaking them down into slices and then delivering on those ones which means have a lot of benefits so the question here was how do we do agile at frontier si there's all these different methods and frameworks there's cam and the scrum there's extreme programming this scale is safe there's all these different things but we're trying to through in a in this kind of case what we have researchers we have developers how do we find something that can work for everybody so we want to do is go back to the agile principles so you can google these that come up under the agile manifesto which I think are really useful to look kind of build not just really based on a framework based on what you're actually trying to achieve behind these frameworks so what we did is I spoke to the executive about let's try an experiment with a couple of projects implementing some of these principles and see how we go so the key things that we implemented was prioritization with stakeholders at the front of the project and then either of lis throughout the project regular cycles depending on the project can be one week can be four weeks it depends on the project then each of those cycles have reviews and showcases with those stakeholders and see how it's going get feedback plan for the next iteration based on that feedback and then having retrospective is really important to go have what's going well what isn't what can we do about it and actually doing that as a cycle not at the end of the project or right at the start so and then also company-wide Showcase is really great for when you have multiple projects running people be across what everyone else is doing and highlighting that and getting that learning journey from there the results were from those projects we had a really higher customer satisfaction and the big thing was a visibility of the work and transparency there and be able to prioritize based on the results of those projects are really good there's a had really high staff satisfaction because they had a lot more ownership over the work that they're doing and they have a really clear priority about what they're working on so we saw a lot of substant contracts we had removing a lot of milestones and making them a lot more vague because they were finding that it's actually quite restrictive to them when they can't change because you've set a milestone and they need to go to their legal department and change the milestone and you're going around in circles when you build that trust they want to just have you work for a period of time because I know you can deliver for them and so now we're implementing this program these principles across all the next projects that we have so it's only the start of the journey and we still want to try and move away from people having multiple projects at once and we want to try and keep building this up and trying to build more teams around it and it's a never-ending thing but I just really wanted to highlight that the principles behind agile I think you're really good and there's something worth looking at and not always having to go down a framework for agile because not always gonna suit your needs at your company thank you David come to the stage please so next we have David Garcia and he's gonna be talking about how to map an island in the Pacific thank you okay I'm super nervous okay hello everyone Mabuhay at Korra I'm David I make maps and it's nice to meet you there will be some difficult topics ears so every every field station oh just shake her hand and imagine the sea and feel the ocean and be the ocean I'm a peachy student writing a story by living a life with OpenStreetMap and always do an extension anyway so they're going make
nice maps I should be writing the PhD but I do this every day it's qgs obviously to blend I really love it so addictive I come from this place and
island called Philippines and my
experience with the OSM was started with typhoon here and basically you're using it to to do urban planning and you're
doing my Python and it was good because
the world was helping us and if you were there helping us thank you very much from the bottom of my heart this is the
club and the city that was destroyed and this is bored before and after and this is very useful until today when I go back there the maps are still there a
part of my career was to go to war zone next Isis was in the Philippines it's still there and the government they're still fighting and was a battle and we did the same thing this is a photo of the war zone in and I felt sad I said okay I know nothing about GIS so what to
do this time I kind of stepped aside step aside and let the local communities lead especially our Muslim experts and sisters and brothers who know the city more than I do that was OpenStreetMap on the base we're doing participatory planning the point here is is it's not just important to democratize it's important to diversify and sometimes it's not me who has to make the map sometimes I just have to listen and then we're doing some side
projects this this whole story but basically we're mapping human rights violations in the Philippines I was using QJ a sad story going I love the Philippines because of this because it's very dangerous to be a researcher there right now so I think
Zealand and so I think thinking out there for the scholarship and but but
the sad was well as doing the beach you know nothing happened in Christchurch and that made me think a lot about things in life and it happened close to my home and I said okay what what what do I do now
so I travelled a bit I went office word for a seminar and then it was nice because it was an exhibit about about maps this was in last July and I was
very happy because I saw this this is a map of our Pacific ancestors in the Pacific stick chart stylus and the waves and it's very nice because at least I saw it for the first time and I love it it's a map it's our ancient knowledge but the sad part was that the creation
was bad so that was good it was there everyone can see it but it said there at the description pacific mapping traditions decline in response to colonization and as well i don't know it was because of not just in response of so it's not just the data the map being or the participation that matters how we tell the story about the mapping matters to and then kind of okay what's the
context of my research well you know it's sad history of the pacific about about the nuclear attacks and the colonization and all that and still
happening right now with the island split and all that sad history very difficult conversations anything okay how can we move forward as you know as one community and despite this we have a future together because everyone is very
interested in the pacific right now apparently but we have a shared future
because we have a common past especially for my pacific brothers and sisters i didn't know that the navigators came from southeast asia that part wasn't even taught to us in the philippines
because when I came to Ottawa the words are the same about the land the sky is amazing it's kind of us knowing who you are again and that's what we're doing
here in one boat and trying to remember things so if you look back
OS M it's about three things at least for me missing maps missing mappers and the missing map stories and I think that I'd like to contribute to is for example the the missing maps please go on Friday by the you're gonna
have a mapathon I'm gonna make my maps one map every day I hope of an island I
got to be to osm by tracing reefs I really love this i trace roofs and reefs almost every day and apparently there's so many reefs in the pacific that haven't been traced yet even back home in the Philippines there's some there's
some a pattern and Friday please be there we're gonna talk about these things how to collaborate and also the missing map makers someone create that
this I found after the Heidelberg conference the missing mapper so what do we do with this and then I'd like to I
also like telling stories about cartographers who were forgotten like the legend Marie Tharp seemed at the
seafloor and I just made inspiration by the way okay my time is almost done and
the missing map stories because mapping
is a very social technical thing is social and technical is saying that I'd
like to help with the podcast so I devote one day every week be my guest and let's share stories about osm thank
you very much and please remember [Applause] these guys are too good there at a time I can't yell at them so next up we have you and Hill and he's talking about mapping in the year of indigenous languages thank you everyone here today and we're on the lands of a whole lot of tribes who've come here and I won't mention them all my mountain is Mount and and on my River is Yarra River my boat is normally an a380 my family from Victoria but my family also come from Scotland and Auckland so apologies for that my town is Richmond and my name's Ewan Brits Pascoe who's read written this book and it's a book that I think every Australian should should read says to learn your country start by learning the Aboriginal names they show how deeply and intimately how old people knew the land and it's really really important that I feel that we need to capture all of this on the right hand side there's a map which was on the back of a spear chucker showing all of the water holes it's a really complex because some of them permanent water holds some of them are intermittent and it showed the paths through there this is and this map we don't know how old it is but it it certainly predates white man so - Lee numerals on there because it was taken from a book we had 250 languages in
Australia about 800 dialects sadly most of them are gone the endangered ones are only spoken by very few families we lost a lot of the First Nation through a whole lot of diseases and through massacres so we don't actually have a whole lot of information about a lot of the languages so we've got 13 which are actually active in the community nowadays this year is the International Year of indigenous languages and I'd like to point out points three and five so integrating indigenous languages as a standard setting so OpenStreetMap you can do that and elaborating new knowledge to foster growth and development so it's something that we really need to capture but wait there's always a but the first rule is and the first nation custodians you need to talk to them straight up and really understand where they want to be some some communities may want you to be part of it some communities may not each one's different and you've got to spend a bit of time and respect and make sure that they've got a solid understanding of what you want to do this is a fire which took four to six weeks I think to put out its in a peat field and basically we used several of those helicopters per day at about 200 thousand dollars a day the local milk factory was closed the aged care facility was closed we had to move people out of the community it was a really really difficult fire because it was seven meters deep and 70s a 76 hectares give or take and that was one of about four fires that were burning in peat really interesting normally when you're fighting a fire you have a safety officer and a safety officer can just say stop everything and you have to adhere to that we actually had a senior elder who could actually say stop everything we don't you can't go on beyond this point so I what we did was we had we had fifty square meter grid squares and we actually made a mark the the areas that we shouldn't go into in yen we built a four kilometres hype line so what can you do understanding your locals OSM feel proud that you're actually trying to keep something going gotta make sure a little legislation there's a whole lot of legislation in Australia federal state and local and this is all we've got of Aboriginal names in oh s M just a quick one this is from wellington in 1869 and if you had a look at the first tile we had six names of communities we've only got one there so not everything is correct thank you thanks Ewan next we have Hamish and he's going to be talking about qgs and geometry generators hello we all know our cutie Isis good okay right doesn't know what a geometry generator is okay okay it's a symbol
type that generates new symbols as sub symbols which is which is pretty cool saram so my name is Hamish and I like to make maps but unfortunately I don't have the category okay that David or or Chris do to make really pretty maps but I do have the heck is I for looking at a feature and wondering how it can be manipulated and maybe unexpected ways so might talk about that briefly today mmm
so if you go to QGIS and you load up in your lair and you go to the symbology tab you find the menu thing and go down you'll find the Roses geometry generator and what it does is take the root the the data that you've got in that layer and then you can run a bunch of code there's sort of a nice little DSL there like a domain-specific language that you can use and you can use that to create a new symbolize that uses the modified geometry or a completely new geometry which is cool because it means that you can keep your core data your base data separate from the rendered data that you want to display now what is a really neat feature I think more people should use it so just a really simple example
hey it's some dots so we can go into a
geometry generator and have said that a few times this but um you can go I want a polygon I want it to be the buffer of the geometry for some number of meters and the output is that so we're
rendering polygons instead of points hey I could have just you know buffered that I could have just used a really big symbol for that right but that gets you know the idea across
what if we're a little bit what are we thinking a bit more well we could make a geometry narator from those points we could create a line from the geometry which is the point to the current canvas cursor point and will differ to the circle around the canvas point and now
we're generating this and it changes as you move around the map I don't know if this is a good idea but it's fun all right um here's some lines right that's pretty straightforward
what if we create a geometry generator that does this little thing that some this array for each place generates series especially a for loop it interpolates the lines into a bunch of points and then let's add some randomization to each point and then render there's a line again hey we've
got a wobbly line that's pretty cool so now we can do like some more sort of human styles we could then go and like maybe stack a bunch and it looks like you know something that you've drawn or some Lightning or something that looks pretty cool as well so there's some really cool things you can do it there and the end the the language can be
pushed pretty far so what are we take those same lines again what are we doing here we're creating we're breaking it up to reach basically Bend in the line we're making a new line that's offset from the line we're tuning into a curve where I've sitting left to right depending on which angle that liners and then we'll put an arrow on it hey cool we can put little directional
arrows around our lines we haven't created a new data or still just allows data set we've got a bunch of geometries on top so that's pretty cool this is my favorite hack I love this hack let's create an empty polygon scratch layer there is no data here we'll use them base there so let's take a rusted-out a set of New Zealand elevation use your own elevation model
alright so inverted polygons on an empty polygon layer will create one feature let's iterate X's and Y's over our map and query the Russell layer for each point and then use that to create point symbols now we can create a victor layer
from an empty scratch layer using the data from the raster layers underneath to do this kind of like fake elevation layer that's pretty cool you can start doing more creative things like hey let's make it look like a an equaliser graph a whole of New Zealand again there's this that's an empty geometry there's no geometries there yeah so cool
that's it thank you very much thanks a Miss Jessica so jessica is going to be talking about an open-source framework to map litter on the beach hi guys so I'm Jessica so in the last couple days I felt like a little spy coming on my way from Brazil to be my first foss4g conference and I'm really happy to be here and share with you what my colleagues and I and from different universities in South America how we've been applying foss4g for bitch litter monitoring so marine litter it's growing
environmental concern sanic pollution and entanglement of marine animals and the accumulation of toxic substances are just some of the problems related to it
and which litter monitoring is one of the first steps to better understand the problem so and help to map from which polluting activities are most contributing to this problem different efforts around the world are coming to to monitor it and they are basically they use basically paper-based forms and no data standards yeah and so you all
know where I'm going lack of interoperability between them and more time between fieldwork and the final analysis so we thought maybe
foss4g can be helpful to make this process more efficient and still flexible so yeah this is what we came from with what we came we basically use three different software's common ones to do all the process so QGIS to design the GIS project using common sampling designs for bitch little monitoring and having questions that most groups tend to answer and put these in a J's collector cute collect the data and then the data is already structured to be read in our environment and with the with our code that we build for it so and we apply it in three different beaches in the south part of Brazil and so going for the first step in QGIS the project is basically one base map and one point layer with 11 types of attributes which are common attributes use it for in paper-based forms different we'd get types allows us to come to use different data formats so the user have the they have access to the to the little list and you can also or she attach geotech images along the
monitoring so here what it looks like in the field the result of it's a shape file already structure and ready to be read in our environment and
yeah it's not complicated code and in the end we come with six different products which help us to understand better the litter patterns and answer common questions so the first broad products are related to general little items so which polluting activity contributed more in different bitches and the third product is related to temporal changes along these bitches and the fourth and fifth so as related to a spatial distribution are these items being accumulated in some specific zones along the beaches and the last one is a map which can be share it and help also to understand better the problem so in the
end the conclusion was that was pretty efficient much easier than use paper-based forms the data was already structured and we got really excited with it but we still have many problems we had problems with the geotag images and we think that a shiny app can help us to make it more make it prettier and most people more people can use it and yeah if you want to talk about it just
fine you're in the conference or yeah in the media in social media thank [Applause] thanks Jessica next we have Joe Lee and he's gonna be talking about Papua New Guinea and its isolation so take it away Joe Lee thank you good afternoon oh I would like to
talk about the Giants of isolation in Papua New Guinea not only in Papua New Guinea but for small island developing States small island developing States as defined by the United Nations Development Program says that a distinct group of developing countries facing a social economic and environmental vulnerabilities disasters and all these are some of the Giants that we have to overcome that we need to defeat in some way or the other in order for us to progress I will highlight some of them during the course of the presentation they did UN member countries we vetted Pacific Island countries PNG is one of them Fiji Tonga Samoa and you name 15 priority priority
areas as highlighted by the battle strategy of action in 1994 the red one that I highlight there simply says that regional institutions and technical cooperation that's why we are here we need more vibrant dynamic regional collaboration and also the dissemination of knowledge technical know-how to small island developing States that's what we need to do now looking at the Giants of
isolation one of them is languages in Fiji we have three languages and 300 plus dialects in Papua New Guinea we have 800 plus languages and dialects as well according to the survey the diverse cultures and traditions very beautiful cultures and traditions which we embrace and defines our way of living rugged geographic and topographic attributes very very rugged terrain we cannot move easily from one place to the other as I access canoe or you walk to reach one point or the other small domestic markets and heavy dependence on few external remote markets and infant communication network I would not say disconnected communication network we are connected but only it at its infancy stage we are still developing simply we are small island developing States looking at the languages as a giant in
order to to to come up with this as the risk-reduction languages can be a barrier for dissemination of information if we take it from a positive note looking at languages we cannot eliminate languages we just have to embrace it because it defines how fast it defines us it defines our values as well one way in which we can overcome that is using open source mapping so that people not only educated people but the local the uneducated can see the maps and know where they are they want abilities for what disaster that they are vulnerable
these are the Giants that I highlighted that diverse cultures and way in which we can address them looking at the number three rugged Geographic and topographic attributes we we need a remote sensing and knowledge in which we can do analysis of our environment especially we have raw materials like mining minerals and other materials that boost our economies [Music] we need communication and connectivity as well as capacity building so just to end with this don't think outside the box just think that just think like there is no box okay and now we have comes in Roger talking about some of the difficulties you encounter mapping informal settlements hi good evening I'm commissioned mmm so I work for you
inhabit art in Fiji for informal settlements informal settlements a lot of people get confused with informal settlement squatters and slums but informal settlements just means that they have some permission to live on their land but they don't have legal titles or leases or any of those things so un-habitat has identified 16 informal settlements in Fiji we have like 280 something but we've only been working with 16 I work with the Nandi settlements namely corrode the area now AG Kumar and what we plan to do is because these settlements are affected by climate change we plan to map out the communities and then decide if we can help them adapt or if we need to relocate them so some
issues that I've come across with using open source software or data is the first one QGIS crashes are the most random times like my screenshot here look at how many layers I have just like points of four cities and one base layer and it's still crashed next I have okay
GPS inaccuracies um so I tried to use gps's I try to use phone GPS s I try to pin points using Google Earth on my map but you you see all these points where the houses are the red dots and the blue dots they're all so varied so using sorry using cue field as well like Jessica had mentioned that gives you high levels of inaccuracy as well I've talked to her about it next access to
local open data sources okay um so in Fiji we have a few data sources such as speck geo Pat cress and Vanuatu I as bunwich s is not completely open we need to get permission from Land's to use it but if you see the top image that's peg geo and I searched the term Nandi which is my town and seven results came up upon which only one of them was relatively useful and that's it whereas when I search for Wellington has like Wellington City Council links you know all of them available I won't list it down but this is just one example of how many like types of information is available for one city and I looked at links and they had about two thousand plus data records for just Wellington which is amazing because I have seven um
next um poor base maps of unmapped areas I'm sure you're gonna tell me that my
job is to map the specific areas but all I'm trying to say is dude there's no reference points look at my map for an nd and look at the map for Wellington look at the difference of how mapped and unmapped they're both both the towns are and my final point is as you've noticed throughout my whole presentation they're very very ugly Maps so my final point
and this is my final favorite ugly map is because we have lack of local expertise all the maps they represented wasn't done by me it was done by some of my colleagues and this map particularly is done by one of my colleagues who has a master's in GIS and this is what she made to show the inaccuracies for the GPS points this
thank you [Applause] now we have Michael and he's gonna be talking about high-performance map caching yep the topic is high-performance map cache generation at mannakee phenoix Landcare research Michael Smith - presenter and DevOps Andrew Kalley is our cartographer and geospatial developer I want to talk about a recent project where Andrew and I were tasked with increasing the performance of caching our maps
mannakee phenoix informatics developed and operates several web map applications our mapping services rely heavily on WM TS we cache the maps and store the caches and Berkeley databases BD B's for short our map systems use map server and map cache Apache modules the first website above is ad a also known as the Antarctic data analysis Fraser Morgan is giving a presentation tomorrow at 11:15 at Thornton seminar summer please go and find out more the second site is our environment and that provides New Zealand environmental data the third is the s map online and that provides New Zealand soil data there are several other sites that use our mapping systems and finally the one that's underneath is our web map services and that provides api's for accessing a subset of our maps before starting the
high performance project we use a single virtual machine approach with all software installed with puppet on the host vm this slide just shows you the relevant components and complexity here
are some statistics regarding our Maps our Maps consists of 92 BD beasts and 280 layers about 3.4 terabytes of data and 920 million map tiles we generally cache down to 12 zoom levels using our existing processes updating the maps takes a long time for the high
performance project we picked one of our web services to focus our efforts on and that was s map online since 2011 caching was conducted on paren on a single VM so it has 12 CPUs and 32 gigs of RAM we update the data about twice per year and the compute time takes 8 days so that's a long time we thought an easy win would
be to just scale up our Hardware on a single VM we used AWS and tried a few instance types so c59 Excel and t5d 18 Excel this is certainly more expensive than running on Prem so we needed to exercise caution when running the VMS the c59 excel took 40 hours and about 90 US dollars while the c5 d-18 excel took 30 hours and 136 US dollars processing
went down from 8 days to 30 hours but now we're incurring a cost we thought we could do better for both processing and hoe end time for hosting our costs mannakee finola is a partner with Nessie so from our project's perspective Nessie is a free resource one note on Maho Iike is equivalent to the sea 518 excel from AWS our mapping system is a bit
complicated to get set up puppet manages the software installation and puppet isn't really going to work on the nodes and either as running VMs on compute nodes we initially thought about compiling the software and manually configuring that but it's a really a nightmare scenario we pose this to Nessie consulting and they came up with much more elegant solution containers Nessie staff helped us to containerize our mapping system but the build pipeline shows how the singularity container is created that's the top left and also the bottom is showing how we run it on Nessie caching can easily be divided up by layer and type this means that all the S map layers can be built at the same time there is no requirement for communication between jobs and all jobs right to the file system
we ran into a few issues both in the conversion and running of the singularity jobs NASA consulting with was instrumental in helping us resolving these issues they were great to work with so I highly recommend their service if anyone needs to paralyse their jobs we built our estimate caches using 72 CPUs per job with 128 gigs of ram that's one full node on ma who Iike the process required 44 nodes and the compute time was only three hours that it's about 64 times faster than our original system and 10 times faster than AWS this now opens up the possibility of near real-time mapping updates and we're now looking at the same pipeline and architecture to create weekly updates of map caches from sentinel data and
finally here are some links for our web map sites so please check them out from when you have time thank you yeah we have Nile and I'm very curious to see what he's talking about it's titled merry christmas' qgs actually it's titled no you Merry Christmas for kuja's oh actually or you could call it cross-cultural communication issues in a multinational open source project so Kuja is a massive project like there's a
community members from basically all around the globe if you kind of break
down a numbers there's like about 150 developers over time who have contributed but that's any like a small small small part of the community because we've also got the documentation team translators local user groups and then I like a whole raft of other people who are sort of integral members of that cutest community and they're all from everywhere so if we just look at the v?lib it's like taking a really narrow
snapshot of that community again the kind of most active developers on github and you look at where they're from we've
got people from you know Australia from Oceania from Asia Europeans us a lot of different paces and there's something in the way there but as well as like
geographically diverse it's also sort of diverse in terms of people's backgrounds so if we take one of their like oldest files in kuja's it actually starts with this header that's got a whole bunch of Bible verses there but in the same
project we've also got people who've added like a commit like this and added this sort of Pastafarian religion logo which is kind like a joke religion right so lots of different backgrounds lots of different cultures and it's it's raised some kind of interesting discussions over the time conflict and issues and that sort of stuff so today I'm actually gonna be cheating
and I'm gonna air some of this dirty laundry about what happens behind the scenes in the the cutest project we've some examples first example all right so
this is from the cutest hackfest that occurred recently in bucharest actually it's not a cutest hackfest anymore because we changed the name to cutest contributors meeting because hackfest was it was actually turning people away and it was really difficult for people to get management approval to send their employees to our access anyway there you know lots of fun midwifery this pull request came in to
update the pull request template on github to add that line you can imagine kind of the background of what was happening over there at the time got
merged right about a couple days later this poor request came in to say hey hang on a sec let's think about this you know some cultures totally reject alcohol consumption and we should be respectful for that what are we actually what message are we portraying by adding that to our sort of public image and that one
got merged so you know resolved there was discussion almost pages lots of comments example two I want to show
you this was in a massive period of
transition for Kuja so going from queue just to to queue just free heaps of work like really kind of stressful time for the developers and the project in general born out of this kind of
frustration of like one of the Vella purrs hit these issues with a coup de server so queue to serve 8 years like the server-based face of cutis speaking honestly in queue just to it was really quite shaky it was kind of like this basically and you know
this developer kind of ran into an issue and his work got a little more difficult because of q just server and he posted that thing to say actually let's just get rid of it as a thing but this kind
of this this caused a lot of conflict because a lot of people were really invested in that emotionally in financially and they're like what are you what are you saying here and your humors not working the good news is out
of that that big kind of conflict really positive result came because people stepped up and invested a lot of money in the cutest project itself also invested money and server was ripped apart 4q just free rewritten from scratch and now it's actually quite modern and you should revisit it if you've got bad experiences number three
again cute is two to three coming out of
this like really difficult week of discussion I fought or somebody for it was me there that I could be funny on the cutest list and just post his stupid joke comment about let's go to surplus plus 17s cutest hurry its consequence we lose the windows build who cares OS X users will have to build their own version of the compiler from scratch yeah they can do that silly joke some
people got it some problems some cultures and they kind of ran with it other people just took it literally and it's like you know might my users will never change their OS too because of this fortunately Nathan stepped in and save
the day and actually explained that's Australian humor doesn't work for everybody
right example for at one stage qutuz had this little easter egg we're going into Christmas somebody somebody made it so it has sent
a hat on the cutest logo was really well
received people on Twitter loved it you know they
said it made that day it was really
happy then feedback started coming in today actually how do we get rid of this
you know it doesn't it doesn't match my culture I want to get rid of it we don't
want it there let's get rid of it and yeah the pull request came in got merged
it got removed got turned into a plugin so you can still get your hat but you just if you're part of a culture that wants that lastly my story right for a
long time my experience with cutis was these mailing lists people's little
avatar was all in you about them their github picture you know their Twitter face that was my whole knowledge of these developers it wasn't until I met
them face to face at a cutest meeting and they became real people that that issues like that that my vision of them changed and I actually learned it there you know emails terrible that's terrible for actually communicating who people is thank you and now we have Richard law he's going to talk about how to create rep reducible workflows with dagger I'm fish to last by the way so we're nearly there up to the hours already all right so we have a reproducibility crisis and I really mean a crisis in November 2011 the Center for open science launched the reproducibility project involved about 270 authors aiming to reproduce the studies of 300 studies and I think it was in the field of psychology but it's not super relevant originally 97 of these studies claimed to have significant results the group of authors Twitter and seven even went to massive links to reproduce these studies Consulting the original authors and their datasets but only 35 of those studies were able to be replicated in a way that reproduce the original result of those 35 replicable studies the average effect was weaker than the original studies and in some cases the replicated study found results with the opposite effects the one they were trying to recreate so reproducibility is a foundational principle of a scientific method it involves recording your method in detail so other people can do the same thing therefore how we use computers to do science must consider how how to reproduce our results on other people's computers so I'm gonna share a way that I work there hopefully my analyses are reproducible by others and also by future versions of myself so this is the high-level view was there men in a moment the general premises stats the top and comes out at the bottom the bottom is things like images datasets perhaps even an entire Journal article if you're writing in lytic or something like that making sure that your inputs are persistent is also important for reproducibility but that's a different topic so starting from the beginning we create this thing docker image what's a docker image well think of it like a detailed reproduction of a chemists lab right down to replicas of the the beakers that they use to do their experiments the symbols or the tools you need to do your experiment in in our case that tends to be things like GTL you know this example has got rusty Rio and in its CDF as Python libraries and it's important to be able to clear the exact version of these libraries in case later on you discover maybe there was a bug in something that you used cool so in this case we're actually using a docker image that someone else has made and extending it slightly next we is a raid a related tool called docker compose and that essentially allows us to declare the interface with that computer so where our data is sitting where the output should go environmental environment variables and things and on the Left we've got sort of a development configuration that you can use with you know subsets of your data and on the right we can override certain parts of the configuration to when we want to run data on the full set or and or write to a particular place with output should be written officially next once we've defined docker compose context who build the service so we've got here a docker compose build model and once it's been built well we can run model cool once it's running we've got access to all our data in this reproducible environment and this is where a new tool comes in called make it makes been around for a while this is not new it's you know new in this context and well how make kind of works again this is a top to bottom kind of thing if you start at the bottom we define a target so your target might be you know the data set you want to produce and each target can have one or many two things that depends on so in those dependencies are themselves targets which have dependencies and so on and so forth it's essentially a big chain of dependencies starting from your raw data and processing down to some kind of final output cool so now we're actually already at the end that was surprisingly quick introduction to dr. and make so the big picture is that if we think back to the beginning I talked about the they're reproducing those studies if you go to that link in the bottom there you can actually download an a script to the authors of that reproducibility study published where you can reproduce the study that reproduced the studies and which is quite cool but more importantly it gives you so much more confidence about the output of that write because you can actually go and run this analysis yourself it's there and complete detail warts and all these probably bugs in their code for all I know and isn't that just great you know it's it's this this philosophy of open reproducible science built on the transparent foundation of open-source software and I think that's just awesome so thank you second to last we have ROS Phillips and she's going to be talking about pixels point clouds and geo sauce awesomeness yes do open source I'll let her take it away kyaaa everyone today I'm going to be talking about point cloud pixels and open source awesomeness specifically how we use open source tools at land information New Zealand to process and publish our Creative Commons elevation data firstly diving into the classified
point class that we publish from less
tools less info is a great way to summarize information about a point cloud data from finding the total number of grandpa's to find in the maximum scan angle found in that data that's a really useful tool you can also use last validate to check if your point clouds have any corrupt elements in them and also figure out if your point cards follow the file specification that they were delivered in we use these commands that language information is Eiland to cross-reference our point cloud data with the corresponding survey report and contracts to make sure what we've been delivered is what they said what is what they told us basically you can also use and the last step tool to compress the point cloud files into up to eight times small smaller file so that's really useful if you want to transfer the data to someone else or just store it for your own means peddle is another great
piece of open source slider processing software it has a really broad range of tools and applications I will just skim over these an example is you can translate from one point cloud format to another this software has the most complete set of point cloud file format drivers that are available through open source software and you can also remove noise you can identify ground points as well as splitting and merging point cloud data the P Dow P Dow pipeline tool is a really safe way to combine all the tools that are available under the software into a single Jason reference we use P Dow pipelines to reclassify our points if needed changing our metadata headers and to also add some information specifically coordinate projection references and at the same time doing this we can also compress our point clouds finally we publish our am
cast by point cloud data through open topography one of the options when you're downloading and this data is to view them through the 3d point cloud viewer you can do some really cool operations with this viewer you can generate cross-sections and we can do volumetric analyses and also filter classifications all for free this has been built on the front into the report viewer and in the backend a Greyhound server with an int wine library all open source to check out our point clouds just Google search open topography and just click on on the top of the web page bite the find data tab now diving into
our our elevation wrestlers which are derived from our unclassified point cloud data we publish digital elevation models representing and the ground level elevation as you can see here but and we also publish this digital surface models representing the first terrain feature found at their pixel and that is the first return - noise we use G down
Pacific liji down info do that edit and did I walk to inspect a raster data that is a single bedded gridded sq file or geotech file we also use G Delta trains file formats to change our noted no data value if needed and to mosaic Erastus together we also use a Postgres database with the post es extension and procedural language extension to verify that the Resta tiles that we given correspond to 1 to 1000 scale and elevation reference as we given thousands and thousands of tiles and our references have reference has Fox been million attributes processing them in the Postgres database is very time efficient way to make sure our data is what we expect it to be we also use python with with all these open source tools to combine elements together to create a nice GUI interface and in their specific context to make sure that the correct naming engines have been used that said to download a raster lighter elevation data a digital elevation or surface models you can go to data on CVT in Zed to
learn how to use some of these really cool open-source tools and to do some pointy 3d exploration you can check out these medium blocks that I've made to find these you can just Google search point craft medium aligns it and one of the first URLs is a link to all the blocks that are available and the land information New Zealand site thank you at our last presentation we have wing Howe and he's got the most technically difficult one to show so hopefully works he's going to be showing us how you can view web maps offline cool all right hi last talk of the day good job for making it through so I had a bunch of slides I was gonna go through ah my name is wing I'm a front-end developer I don't really do much work with your special Oliver I do work with Terra that we do do a special things there and yeah so I had a bunch of slides but I think what I'll do was I'd start with my demo actually because otherwise you're gonna get bored to
death with all the slides before so a
couple of things to note I don't use Windows this is a Windows machine so let's just see if this works because it works on my machine it's probably reproducible well that's uh uh thanks Richard um so I have a map here and I'm just gonna refresh that okay
good error first thing of that's great um so
it's it's a map it works we're online and I'm just gonna try and what we're doing is we're catching a couple of things not all of us have access to high performance compute that we can use for free so so why not just load it up on to
the users browser so all right so I've just kind of used the up a little bit just because we need to cash a little bit of things we're using the browser um and hopefully I'm gonna try and do the big thing and we gonna try and figure
out his windows okay here we go sir hopefully you can all see that I'm gonna go to work offline we're gonna go fine
and as a extra precaution I'm just going to turn off the Wi-Fi here I hope I didn't disconnect anything and if we
refresh this map it should just work
that's probably still find Eretz same error but let's let's see
okay so maps all kind of works the tiles are still loading in and I can still go in and add whatever
data said that I just loaded in and yes
that is a paper on some of these icons missing so yeah that's so working offline and we were able to do that because we leveraged the users browser and to kind of stuff in all the tiles stuffing all the data and that's done with JavaScript so he's I guess we can
flick back to the stuff the boring stuff
so the I'll find maps in the browser so this is done with service workers they made available now in all the major browsers including Safari and including IE so that's props to them why do you want it so things you can do is you can go offline progressively and safely so what I mean by progressively is that what I'm showing you here if it doesn't work and or if your users browsers doesn't support service workers for whatever reason it's just older version or they're working on an enterprise machine that has like a lock down really old version of Chrome or something it's not gonna work it's you're not gonna break anything by introducing what we're adding here how do I get started we're using service workers to kind of control the cache on the client side so you can kind of do this sort of thing on the browser sorry on the server side but we're kind of just pull a logic into the browser instead of the server there's some technomages over here we're using I'm gonna show you how to install a serviceworker I'm just going to show you what how we leverage that to do the caching basically we add a whole bunch of files like all the HTML CSS JavaScript that you use to load your particularly single page application and then you kind of hijacked the fetch event and you pull out stuff in the cache and feed it to the fish so the browser thinks it's online it's like oh I'm gonna request things as usual and if it's not there it's not made available then it will just pull it out of the cache instead of going to the network but this way you can kind of go to the network in the background while returning a cache too tall or cache Dowsett or cache something immediately and then the user doesn't have to wait and they'll still get the most recent data on the next request cuz you're just kind of doing this storing the cash first yes it's serving the cash and then you get the network request for the next one where peck helps us through things if you're doing single page apps you're probably using webpack there's a plugin that kind of wraps up all this what I'm explaining in a nice little package and my time is almost up so I'm just going to tell you about some other things some other resources Google and Mozilla have cool things so that you can do with service workers and really showing the demo Thanks I'm also speaking here tomorrow morning so I'm gonna be talking about Google yeah the evil company [Applause]
Loading...
Feedback
hidden