 Where to start? The preamble for this presentation comes from the utter frustration that I have with rolling out open source and open standards based software and technology in a large public sector organization. So at the sort of when I started in 2014 as the CIO with the public health service I was at with the private health service in Townsville before that and there was a number of projects I wanted to do and I thought you know once I finally got to the top tier of IT management I thought I would have enough prowess, political capital to get my own way. Much of my dismay that's not the case especially looking for around the public sector because obviously we have agencies and governance layers. So at the beginning of the year I actually started writing all of these excuses down in the back of my diary just because I thought it would be a cute little exercise in sort of trying to keep me sane. Towards the end of the year they actually started grouping together so not with any sort of scientific method at all I put together a list of the ten most common barriers is the polite way of saying excuses some might say or you know basically crap that people say why you can't do open source technology and if you want to know number one had about 86 hits and number and number ten had about four hits so that's sort of the range. I'm looking to tell you what the end is and all that sort of stuff because it's not actually proper statistics. So who are we? We're a public health service that covers 23,000 times the size of Singapore that's the health service area. We're also a tertiary referring hospital for that sort of green triangle rhombus shape thing there so we get tertiary referrals from Papua New Guinea, Solomon Islands, bits of Micronesia and that sort of stuff. So we're big and it's a fun hospital to work for and health service because we get a lot of really different cases. The static population of our catchment area for our secondary and tertiary is about 300,000 so obviously probably about one one-fifth of the population sorry one one-twentieth of the population of Singapore. We employ about 6,000 staff it's about two-thirds clinical one-third administrative in rough terms. We provide 65,000 inpatient occasions so that's people who actually obviously get admitted to hospital and we have more than 62,000 emergency department attendances each year. It gives you an idea of how big the the organization is. I have a long sordid history and open source. I was part of the formation of a software startup in 1996. We did quite a lot of open source technology, quite a lot of open source development. It went incredibly well for many years and went incredibly badly for a number of years. Those years were 2000, 2001. So I moved on with the Australian overseas aid service. I worked in Vanuatu and as a side project I set up the first free mail exchange server in Vanuatu which was used to provide basically email services, just a dial-up email service. We will send mail server kicking in the background. It certainly changed a lot of things. Really cute application of a small technology that made a big difference because finally NGOs and non-government organizations in Vanuatu were able to actually use email and join the rest of say the donor organizations and things sending and receiving email. I used to be part of, I used to lead a political party in Australia called the Australian Democrats in the ACT. It's the president of the branch there and we were actually successful in 2004 in getting open source legislation through the Legislative Assembly basically by changing the Procurement Act that basically the government needed to actually look at open source software first before procuring proprietary software. Didn't really do much good but we had a good tilt at it. In 2006 I was with the World Bank sort of back to funding system with Australian overseas volunteers again and we put together a PNG coffee corporation logistics system. We were able to raise the farm gate price of coffee by about five to ten percent which was quite good for them. Obviously that's the main thing that they involved. I was in Cambodia in 2009 helped with the development of MoonOS. It's not really going as well as it used to but it was first Cambodian language operating system around which did quite a lot of good things and obviously currently I'm the CAO of Townsville Hospital and Health Service. I'm not going to look at me but I just wanted to say I am actually deeply committed to open source technologies and the use of open source methodology to try and get outcomes for organisations. I'm not a hardcore coder or programmer in any way but I guess I sold my soul to being management now so I can't really go back. Number one reason. Difficulty. Like I said it scored 86 or 84 times in the back of the diary. Three main reasons. It's tricky for the uninitiated user to use. User interfaces are far inferior those of other operating systems. Not that anyone really uses them in my workspace. It's just not what I'm used to. So the first barrier that we're up against is basically an uninformed user base that doesn't really understand the opportunities that they are with open source mitigating the challenges of actually using it. I'm going to go through the slides reasonably quickly because at LinuxConf I ran a similar scenario and we had a lot of questions and I didn't have enough time so I'm going to try and get through this fairly quickly and then there's looks like a lot of informed people here so it'd be great to have a bit of discussion about what your thoughts are on that. The second, it's the default and this is mostly your technical people that give you this. The exchange of documents with suppliers or customers or internal people. Proprietary software vendors have already gone into things like universities and training colleges and trained the students to use their software. This was a large problem in the engineering sector in Australia because there's a particular AutoCAD developer and vendor that will actually give the licenses for free and the software for free to the students so that by the time they actually get into the engineering firms they don't want to use anything else because it costs them too much productivity even if there is a mandated company default software. Not identical in terms of functionality, user interface, performance, plugins, yada yada yada. That's a little bit of a bleed from the last slide but the reason it's in this one is that basically people get used to using things in a particular way and don't really want to commit any extra training. In an organisation like a health service you quite often have people coming to you saying well you're putting in this new software, what sort of training are you going to provide, what are you going to do and those sorts of things. It's interesting that computer literacy and literacy within information systems isn't actually considered to be a core part of people's jobs or a skill that they need to maintain. For example a nurse needs to continue to have basic life support. She needs to know, well they need to know how to do CPR and give mouth to mouth those sorts of things and that's just part of their job. They need to keep that training up to date to keep their job. However IT skills and literacy is not one of those things. It's an interesting dichotomy in the zeitgeist. Number three, support. You would have thought it was number one but it actually comes in at number three. Paying for guaranteed vendor response. Admittingly the open source sector isn't very developed in Australia and often we lack large or what would be considered to be large organisations to actually back software to provide solutions for the organisation. Free Libre open source community. So obviously the acronyms change. Foss, floss, Libre or open source are interchangeable in our community so I'm just going to stick with floss at the moment. Community responds quickly to queries posted on forums and pages but yeah sorry and the dot dot dot dot is but yet you can't pay them to do so. Well that's the idea that was put forward at that time. Externalising risk by giving projects to separate third parties. So it's almost like a catch-22 because there aren't any large vendors of open source software. There's a catch-22 we can't externalise our risk. Now within the public sector as opposed to say the private sector or community sector, it's not so much financial reasons why organisations make decisions or the financial impact or those sorts of things. It often comes down to political capital and the willingness to actually risk one's own what would you say profile within your organisation. So risk externalisation is a major underlying motivator for many of the public sector leaders that you're trying to actually commit to making open source purchases and if they can't externalise the risk it's a no go basically. Longevity, there's no guarantee that proprietary software vendor will stick with a product and on the other way small developers may not be around is sort of the coefficient of that one. Yeah that one's loaded and I think I'll go on a side tangent if I go through and explain that too much but that was an interesting one from one of my staff members. If an open source project is small there's also a danger that the person behind it may lose interest and any quick search of source for which will find many orphan projects as I don't know who quoted is probably an anonymous quote but they always say that success is many fathers but father is the orphan. Large vendors are likely to be around in a few years and to honour their commitments they give you. Well I don't exactly agree with it but it's a reason that's often put forward by many public sector leaders why they don't want to adopt open source. This is actually not so much a reason why but a reason sometimes actually question whether or not something's actually open source. The cloud which unfortunately I mean words like the cloud, digital disruption and innovative excellence get overused in the Australian public sector in its verbiage and I'll probably vomit the next time I hear it used by an uninformed user but basically even if the hosted software is built entirely on open source software you don't normally get access to the software anyway so to the code I should say the code base either in escrow or in some other mechanism to make sure that if they were to go out of business or if they were to be attacked or any of those scenarios why you want to look at their source code or even quality control you don't often get access to it because they just say look you're buying it as a service you don't have rights to do that those sorts of things. The benefits of using the cloud solution as a service model often outweigh the disadvantage of not having access to the source code that's the counter argument for that is basically if it ain't broke who cares I don't have to fix it why would I be why would I be worried about the source code so yeah let's let's use the cloud but it doesn't really matter to me if it's open source cloud or it's not open source cloud or how much the vendor wants to put the code into say open forums no sorts things it's yeah that's that number five is one of the ones I get very passionate about in arguing and debate. Closed hardware well obviously our session beforehand obviously was a great leader into this and I think most of the points were covered in the previous session proprietary hardware has specialized drivers often closed source is is not available from the manufacturer. What's really interesting about that is that we have quite a lot of biomedical devices that are locked down like Fort Knox and quite often when integrating a system into a hospital scenario the biometric device or biomedical device may not actually work with even though it might be running on say some sort of light window system or light Linux system because the interface even though there's HL7 a number of standards if because it's proprietary often they're standalone devices so when trying to implement a electronic medical record and get say I don't know regular observations out of a particular bedside biometrical device sometimes the vendor won't actually allow you to actually plug your interface layer or your middleware layer in with theirs and there's a number of great examples of that on the internet. When I was working for an engineering firm and that point number three is important and our previous session was also you know basically the developers don't have access to the actual code and the hardware so therefore they find it very hard to build things for that hardware and that then becomes an ongoing loop. You don't really know what the black box is doing. That's a more of a thing why you should have open source but sort of the counter argument from you know there's that thing in the corner it doesn't break so we'll just trust it works and it's got a big you know stamp on it from a company that we trust we'll just leave it alone but literally my network so on a tangent I've about 140 staff I have about 7000 desktops and about I think it's close to 10,000 devices on my network as my slab of the network and then obviously we have a wide area network which has about 90,000 devices on it. The amount of devices that I don't even know about and it keeps me awake at night and thank God I'm in Singapore I can get some sleep but where the manufacturer just actually has a VPN tunnel straight through into its box and it's sitting in my network and sure I've got a quasi layer two layer three network sitting there and a bit of security but hey anyone in the room can tell you the security problems with that yeah and that leads into the last point that of course the vendor needs 24 by 7 remote access so A obviously from the previous points if they're a small open source vendor they're probably not going to be there 24 7 need them and B you need to give them access all the time anyway it's not like these devices it's some it's that sort of idea that you often get that people think that if a biometric device goes down then someone's going to die or something's going to happen but in the end people still make people better machines are only there as a tool you know I mean it's it's it's a hard argument to win liability I would have thought it came up higher but it didn't because people don't actually like to think of it this way what it is in liability indemnity matter a lot a lot in the risk adverse public sector as I mentioned before many open source projects aren't backed by commercial organizations I either had to sue most most organizations public sector organizations don't actually sue but it's sort of that same zeitgeist that basically you know there's nobody there to be responsible there's nobody we can go after if it really hits the fan you know and it's a major reason why people don't want to do want to go with it I would have thought that it would become number one but people don't you have to sort of tease it out of people so what is actually the reason why you don't want to go with an open source solution and in the end it is actually that sort of covering one's own reputation budget now this is a complex one that I've sort of jammed a couple of reasons together so it doesn't really it doesn't really fit one particular sort of idea but they all are financial and it comes down to the way within Australia and I think probably other other countries have a similar sort of public sector budgeting cycle that an organization has given a certain amount of money and it really needs to expend that money by the end of the financial year otherwise it won't get an increase in funding or a similar amount of funding it's a it's a common sort of issue so a few things there people think free when they see in front of open source they think of free beer and not free as a freedom so often it's very very hard to actually get them to understand that you really should pay for your open source software even though you're not required to do so I mean obviously if there's some sort of professional services agreement that you can enter into with an organization and that's great but even if you're using some cute little some cute little piece of software that someone's made and just chucked up on Sourceforge or on GitHub and you should probably donate to the to the project it's just good citizenship and in an organization especially public health where being a good corporate citizen I mean there are many say that there are many medical for an example there are many medical procedures that we do that we were that were never fully paid for whether by public funds or by private funds yet we do them anyway because someone's going to die someone's going to get hurt so you know there's there's generally a culture you would have thought a culture of basically of good corporate citizenship or good community citizenship but as far as software goes we're happy to rip off other people's IP and not pay for it budget structure is anti-open source it's it's funny in an in a it's something that I really tackle as a as a senior executive in a health service if you can capitalize the money IE it's a project expense it happens it's like purchasing some hardware or purchasing if you say you're doing project software is a little bit hard with this sometimes then it's far easier to get the money than if it's an operational cost because the underlying budget in Australia we've got this what's called an efficiency dividend which is basically that each public sector organization needs to try and run more efficiently by 2% each year and it's it's it's it's it's quite a anyway that's that's another debate so when I'm putting projects forward if I can actually capitalize the project I can make it a project a proper project sorry I'm using big the low and small p project then I'm able to get it through the budget and if I can keep it within a year then it's just like purchasing a new wheel for the car done you know where we go but a because open-source software especially the smaller and more of innovative stuff tends to be tends to be done by small projects and they they really want you to pay them a monthly fee which is what you want to do it's very hard to capitalise so there's this sort of counter intuitiveness actually being able to support some of this innovation and last one if the open-source solution is used to support critical applications quickly and cheaply it often goes unrecognized on and under resourced I can't tell you the amount of open-source software and some well how mostly open source software that just gets put in you know like people go oh I found this online and you know just run it and the way we go whether that is true open source you know built you know some sort of lampstack web server app or something right through to someone's pinch some excel code from someone else on the internet it very rarely goes recognize that it's actually open source and that it's actually that it's actually used anyway I'm slowing down I'll keep going segways very nicely into intellectual property the need to protect the organization from any intellectual property challenges challenges from open-source software basically if something goes wrong someone wants to sue us and it was this piece of software who's actually responsible chain of command liability once again replace replace any code found to be in violation of intellectual property rights so it's funny most of these actually come from our procurement people and they're pretty acetic but trying to actually purchase something you've obviously got to go through the procurement people because they pay for things and what it often happens is they're very very scared about the fact that if a another company suddenly buys out another company so like I know Oracle buying out Sun micro systems was a big one for us in a few years ago then some of that code then becomes proprietary because it wasn't strictly under the GPL or some other mechanism at the time therefore we have to start paying for it in the future I don't have one single case where that's actually happened but our procurement people are actually quite worried about that maybe they studied at university during their degree or something I don't know if our own staff contribute to an open-source project who actually who actually owns that intellectual property we're a public sector organization I would say the public does so you know and there are many mechanisms to do that but it's a we don't want to give out IP away just willy-nilly to anybody even though in the organization we often you know take intellectual property from other people and don't give them credit for it so it's quite a bizarre you know zeitgeist the last one who's intellectual property is it anyway it's often very very hard to actually nail down who created and who didn't and those sorts of things this isn't so much a reason but it was two long conversations I had and my stance is you know as a public sector organization or not public money go to supporting you know open-source solutions and open systems and providing code to the public other than then took company that's going to basically make profit and unfortunately the majority of software in in health care is actually for profit proprietary organizations that develop software on the public purse and then make a lot of money for it somewhere else and so yeah anyway I could go on about that one but I won't so that's my observations over the first would you say two and a half years of being a CIO and a public sector organization a lot of those I was hoping would start a discussion and ask some questions and so I'm happy to take any questions I thought some of them would be provocative okay first of all I work in a government related agency so this is pretty interesting to me and we do research so IP is obviously very important and I'm surprised that it came number nine indeed but one of the things that I have to deal with also is I mean not just open source adoption we we often see open source as kind of a shortcut or if it's already there it's working well we could use that but there's a bit of a fear also about copy lab licenses because then you can't sell that as your own IP because you have to you have to give all all the this product that you just developed using with the aid of the GPL or with other similar licenses you have to give that out for free because that's the third of the license indeed indeed and hence why so many new licenses have been made your Apache licenses in the area my tea licenses and stuff but then if you have dominant pieces of software like word press for example yeah you can't really change your license on that no so we've had to find fairly obscure that this for a basic CMS function well that's kind of a it's a very good product hmm how many of you work in large organizations save more than say a hundred people one two so what's how much open source do you work with yeah indeed yeah I mean it's good to hear that Google actually does quite a lot of open source stuff I guess my my often my frustration coming to these sorts of events is that we as we in open source tend to be in our own little bubble sometimes and I'm right at that crossroads every single day fighting the trenches to try and actually get people to use open source and I just find it incredibly difficult to try and you know a cute story on the side is that I'm Libri office okay so Queensland Queensland health is was predominantly on an XP environment earlier this earlier late last year projects now going through to actually clean out and make go make go to Windows 7 which is a fantastic option that's sarcasm and we through various leadership shortcomings we we only ended up with for the 90 odd thousand desktops we only ended up with about 60,000 license or 6,000 licenses for an office upgrade so from 2003 to 2010 which is of course then we haven't have and have not sort of environment and so my automatic answer was well there's Libri office you know let's let's just roll out Libri office and so I had a package by my own team I had it rolled out and I just basically rolled it out to everybody they're now in the process even though they don't have a replacement for everybody for Windows 2010 to even go to 365 or something like that they're now trying to actually claw back and discourage people from using Libri office because of their perceived the perceived threat that it causes with document version control and those sorts of things not that the jump from 2003 to 2010 will cause the same problem so it's a it's a it's an interesting mix and another sorry go ahead oh no I was going to start another story which I can say well I guess related to this maybe it's the other topics so you rolled out Libri office and relatively successfully despite maybe some pushback but did you have to overcome any of the challenges you mentioned here related to you know having a service providers having someone to provide support how did you rationalize that if you even have to people just kind of accept it in that case well it's mostly basically people were so desperate for something that was better they just lapped it up but I'm currently in a lot of in a lot of conversations people that are because so within Queensland there are 16 health services so we're just one of 16 we're the third largest so the two large hospitals are in Brisbane our capital city and my hospital is the largest but actually it's the largest hospital outside a capital city in Australia so it's about a 650 bed facility plus eight other hundred odd bed hospitals and so when I take leadership on something people kind of go oh because you know we're big we're not the biggest but we're big and no I'm currently in in discussions with them about but it like I said it comes back to that literacy argument it's like well do you really need a vendor for an office product shouldn't you just use it and get on with it I'm obviously we put through through group policies and that sort of stuff we made sure that it saved them to 2003 first and that sort of stuff but yeah it's mitigation another great example though is we were using for our IT asset management HP open view product and we're transitioning at the moment away from that and the new the new system that they're going to go to I won't even discuss but we we've sort of fell short of actually our asset control and those sorts and licensing control so I obviously quick market scan snipe IT is not a bad product I instructed my server team there's four of them to load snipe IT onto a onto a lamp stack that I've already built for them basically because they're like oh we don't do Linux so I built the lamp stack and got it all running but I needed them to do the LDAP and stuff to actually you know hook it into the network and that sort of stuff literally while I was away on leave in not to in November last year they took snipe IT they ditched my lamp stack obviously CIOs can't be too technical I don't know where the problem starts but and they actually engineered it they got online clever little sods that they are they got online and they worked out how to shoehorn the thing into an iOS environment so they could actually run it and so even with direct instruction to stuff that are directly reportable to me they went out of their way like like a couple of days worth of serious effort to engineer it in a way to give me the result I wanted to but have the back end that they wanted to that's the sort of sort of cognitive dissidents that you actually up against a lot of the time it's an interesting story. You mentioned something about monthly subscriptions from difficult to get to what you know in fact it's it's it's actually a real problem because often when the federal government which it doesn't have a responsibility in Australia for health care it's it's a for secondary health care which hospitals are I should say often when we get projects funded from the federal government to do something it often is and this is where the capitalization comes into it it's often a certain amount of money to put in a certain capability as a project and so the money comes in it's then put against our budget but it's but whenever we're very rarely given money to continue that project on onwards so often what I have to do is think how long is the software going to be used for maybe four to five years and I sometimes actually have to capitalize the whole project and I put the whole lot in and so I will purchase a forward facing agreement for maybe five years so that I don't have to pay a monthly fee and raise the operating raise the operating costs of my department that's that's the problem and the problem is when you're mucking around with with innovative small companies and you're trying something out that might be huge in the future you can't do that you well you can but you're gonna fall short because there's no way to accurately you know sort of anticipate where you're going to go with that technology and so yeah it's it's a real bind I mean we've got a considerable amount of money for um there's quite a gap between indigenous health outcomes and well First Nations health outcomes would have turned me like to use and say the standard populace we get considerable amount of money to do certain things and it gets really hard to put technology in for this group because federal funding you know it's what it's a yearly based thing something maybe three years but the life of the technology I mean our patient administration system is 40 years old at the moment it's an old green screen piece of software from CSC and it still takes a long I mean there are banking bits of software that are probably over 60 years old sorry so you have so much problem how your service provider can you help well I sort of well because I used to work for a service provider and we and a lot of this kind of conflict with customer indeed indeed and I sort of ran this session it's a modified session from LinuxConf in Australia to try and it was a sort of a how you as a startup can get money from us the big government and it got very depressing actually but um but there are there are ways service provision can be done but you're looking to make sure that and not a lot of innovative not a lot of open source companies do this they don't talk about exit points so if you're going to go to a public sector organization that needs to capitalise the majority of its budget especially if you're being used for a particular project might be funded from a different party you need to say look we're going to provide this software say you're a SAS provider we're going to provide this solution to you for five years and the cost it is a fixed cost of this and you'll need to leave it that and what that means is that the actual innovation risk is then put upon the small provider which is very dangerous having been a been a startup myself not where you want to be you want people actually pay for what they use but quite a lot of time that's what you need to do but it doesn't mean that you have to be static there you can call that one project the biggest the biggest thing that kills what I've seen kills small innovation companies and stops large public sector organisations from taking on that technology is scope creep if you can get your scope sorted and you can get your requirements sorted and it's for a project then that's fine and if that particular if same our health service wanted to use it for something else you say well look that's actually a different project why don't you find some more money we'll give you another slab of the the pie I guess that's sort of answer well you know what you say you use a liberal office right yeah a liberal office have a version every couple of years every couple of months have different tasks so even that they want to provide them want to fix the scope well yeah I mean Libra is not a great it's not a really great example because we're using it for free I'm still trying to work out a way of actually somehow remunerating the open document foundation for Libra I still haven't worked that out yet I've only got a couple of I've probably only got maybe hundred and I think it's 120 users of my six thousand so far so it's not a big problem but I'd like to do something about it but the majority of the funding I have is for the office 365 project so I'm in a bind there as well but yeah it's sort of shifting sands but you know you're right you never know where technology is going to go it seems to be a lot of talk at least on the federal level where DTO or no sort of example I can't explain it so I don't know if I can cover that for this but do you think that is going to help in terms of pushing it down towards the very Australian so very Australian centric argument so the DTO is our digital transition office it's part of our department of communications at a federal level and what my colleague here is talking about is there's a change in our well there seems to be a change in our government finally to a more innovative approach and we're all very excited in Australia because we're a little bit behind everyone else when it comes to innovation at the moment and they set up a special department to actually look at well not look at but be sort of a clearing house for innovation is it's sort of how I see it I don't know the DTO is too young at the moment to actually see how much effect it's going to be the majority of their effort seems to be with other federal departments so as a state organisation I'm not so sure our main federal sort of interface is through the primary health care networks which is a federally funded GP doctor sort of push at the moment sort of a primary care sort of situation and the my health record the national electronic health record and I haven't and what I find interesting when your question is I haven't seen the DTO actually involved in any of that discussion yet maybe because I don't have any visibility at the moment but and if they do get involved with that that might really be interesting because obviously with the failure of the last more failure with the non-success of the last private health public health record it might work out can I ask where you where you work what space you're in I'm actually here in Singapore so I just had a passing interest and well I can hear the Australian accent so yeah seeing you come in and taking a lot of learnings I guess from the UK and the US yeah and I think when Turnbull set it up when he was minister for communications I think he really was looking at the what was it was 2010 when they started there was it GOV I forget what the name was it GOV.UK or what anyway their digital transition and obviously the CEO of the DTO was actually from the UK who was working in that process so obviously we've tried to scout some talent there so no I'm really excited to be in Australia I was away overseas in 2010, 11 and 12 running my own open source business intelligence organisation in Scotland and when I came back to Australia I got a little bit depressed about where we were and it seems that we're moving forward if we've run out of questions yep I think we've run out of questions thanks everybody I'm around for the rest of the conference I don't need business cards on me I've already been completely pillaged for business cards but I'll have some more tomorrow and I'm around to talk to and all that sort of stuff thanks very much for turning out yep and now it's at Foster Asia yes so basically you're reaching to the choir right is there a way to make the public sector people more aware of these limitations and how their excuses or their reasons of their barriers are not as valid as they seem my hope was actually the opposite was that to actually bring the open source community in from the cold and say well look these are actually the expectations that public sector organisations have not that I feel I'm holier than they are or anything like that but it's really that you know we have a considerable amount of revenue that we need to use for innovation we only work with proprietary organisations and this comes down to a really great debate we could have maybe with three aside is you know is it our responsibility to seek out open source solutions because we're a public sector organisation and somehow coax coax the open source innovation into the organisation or is it the open source community's responsibility to aggressively go after our revenue and our issues and those sorts of things I think it's probably a hybrid between the both but the reason why I come to these sort of conferences to explain things is that I'm a die-hard open source person and every single open source organisation not every one sort of clad in sorry I mean the majority of open source organisations that I work with are just not on the page they're just not consumer friendly they're not trying to solve they want they they basically want you to meet them 80 down the road and they don't want to meet you halfway even if you are an open source person I mean I can tell you a hundred examples of where I've tried to use an open source technology from a service provider get them get their funny little hat and feathers whatever you need to do to be a government supplier I still think it's a dark art and bring them in and actually you know it's like dragging them in from the cold another process I've tried to use is we have a an education database that we use it's not it's it's got a learning management system there's a moodle in there as well but it's basically tracks courses that people have been on and what sort of qualifications they have it's a learning management system but a qualifications management system and that was developed in-house and we actually moved that into GitHub and we get in from we're co-located with James Cook University and we actually get normally we start off with a cohort of like 12 or 14 that whittles down to maybe about three or four hard you know hardcore people and they work on that project for us and so we have an open source and not that that whole project has been implemented somewhere else but parts of that project have been implemented somewhere else within Townsville and I'm glad that that's happened so even though I even though I'm not remunerating them they're really just getting course points and experience it's it's one way of sort of getting more open source to happen but you know I'd love I mean what it comes down to was back I mean another one was back when I was working for the private organization in Townsville we didn't have an electronic medical record and I literally tried to put a cohort of about five or six Catholic hospitals together and to either implement open Vista or one of the other open source ERMs that are out there electronic medical records I should say because the medical record when you're implementing it I mean you're looking at about $100,000 per bed so if you're a 650-bed hospital just do the sums on that it's phenomenal and most of that is in vendor costs and vendor professional costs and change costs which I think open source could be better at and that's actually the theme of my thesis so maybe next year I might come back with some learnings on that thanks very much for your time thanks guys