 Beth sy'n deall, yn unig. Felly mae'n ddwyg yn gallu gweld eich cyfnodol yn deissir. Felly, mae'r cadwch Cymru dros ddod yn gyfawr i ni, fel wrth gweithio fe non ym Mhulgu Paelahann. Bydd hyn yn cael ei ddweud am hwnna i ddweud, maen nhw'n gweithio ffas Fanshawn DxW. Mae'n ddweud yn ddweud yn cael ei gwybod yn cyflwr. Mae'n dweud â'r cherdd, yn gwlad. Mae'r cadwch Cymru yn cyfnodol yn ymgylchedd. Felly rydw i wneud weithio bod yn gweithio. Roeddwn i yn mynd yn eistedd, oedd eistedd felly maen nhw'n dechrau i'r slyg o blynyddu yng Nghymru, drwpw Llywodraeth i'r hyn nhw dda'r hyn sy'n dweud. Roeddwn i'n sanctwm Feippur Llywodraeth. Mae'n gweithio ar gyfer twestewid â'r parw hwnnw i'r hyn yw. Roeddwn i'n gweithio chael hiad ynghylch ar ni. Roeddwn i'n ddechrau hon o gyfrifwyr, a'n gweithio ar arhoed, Felly, we work with NGOs, nonprofits and public sector organisations, predominantly.. ..bein gweithio ar 2011, and we're a worker co-op, co-owned about 18 of us now.. ..and we continue to grow slowly, organically. Feel free to get into it if you've got any questions.. ..or if there's any questions during the talk, feel free to put your hand up.. ..and we can take them as we go. So this project came about, our involvement in this project came about.. ymwneud o gweithio i ddiwrnoddau cymaintolfa i ddweithio ymwneud, neu oedd eich gweithio gweld fynd i ddweithio'r ddiogel, a dw i'r holl o'r ddweithio'n gweithio'n gweithio'r ddweithio'n gweithio. Felly, mae'n gweithio'r ddweithio'n gweithio i ddweithio'r ddweithio'n gweithio, a oedd yn Llyfrgell pethau Gwyrddon, Offryd, Ffyrdd, ac mae'n ddwylliant o fodwch ar gyfer FFW i ddefnyddio'r gweithio yn ein ddau'r ddau'r arfer o'r reisarch ac yn y bwysig o'r ddau'r cyffredinol cwmprwyddiadau sydd gennymau drwy ymgyrchol ymgyrchol, ac mae'n ddatblygu ei bod yn gweithio'r ddau'r FFW. Mae'n ddau'r ddau'r ddau'r cyffredinol sydd yn gweithio'r gweithio'r ddau ymgyrchol, fel ychwanegwch ar gweithio'r mig oedd yn gwneud o'r ffordd ydych chi'n cael ei ddweud o'r cymryd. Mae'r gweithbeth yn holl bwysigol o'r ddweud arall, ond oedd yn cael ei ddweud o'r cymryd yn y ddweud. A dyna'n golygu o'r cyfnodd y cyfnodd yw Andrew Katz, y cyfnodd ddweud o'r llai oeithr. Cymru i'w dda'r blaen i'r llai o'r cyfnodd, ar hyn yn y ddweud yng Ngharffwyr, ac y cyfnodd cyfnodd yma yn ei ddweud o'r llai o'r llai o'r llai o'r llai o'r cyfnodd. Ac yna'n ymddangos hynny'n hystafell yw gweld y gallai cyfnodd. The point will use to work with Brighton and Ho, city council, where Andy works, and now as a developer. I'm not sure were you there at the time where we're working there. A couple of years ago he worked there for a year to help Brighton and Ho put together a fantastic Drupal site or Drupal 8 mae foesu pwysig i'r ddadl wedi我lywodau gwneud sefydlu ILEF a MIGOL i'r dduffyn yn de hun o ddufyn, mae nhw'n buenas. Rydym yn cyflawni wefweyl y ddiddorol gyda'r reilins yn ei fod yn ddiddorol gyda'r ddufyn a'i fod yn beth ai'r newydd. Rydw i'r llai bwysig o'r ddodol, oedd ychydig yn dechrau i ddadl y ddysgueth gan gweithio gyda'r ddedaeth. Rydw i'n cum Mandawnau i'w ddod Diolch A bod bod i'n ymwneud eich Ffantyr gyda'r Cyflawn i dydag o'r Cholwyddiadol yr Oedysgol. Mae'n rhaid i'r Cholwyddiadol. Yn ymwneud eich llwyddiadol. Daer日wg i yôl ymwneud ar gyfer gyfer gyda Ffantyr Gyda, felly mae'n gwybod nhw'n cael ei gydag o Gymraeg Thosegol. Mae'n gwybod nhw'n gwybod nhw'n arlendod eich Mweol, a gywedd pan fydd yn Y Ffantyr Cyflawn, ond Gydangodd William, Being at head of GDS Government Digital Services is now heading up the digital sector of Coroling Council and says can you come and help us build a website. So, with the connection between Brighton, we said hey Brighton guys can we use your code, can we use you know let's not reinvent the wheel here. So the London Byer Coroling website was developed in a fraction of the time, around about a third of the time. gwneud y cydn upbeat o maes o ganddo a wneud. O beth o'i pointo yw rhoi'r cydnod yn ei cael cyfrifod y gallwn o'r rhaid. Yn oedd y cyfrifod yn rhoi rel..! Roedd ydych chi'n gofod yn unig y gwir. Roedd ydych chi? Roedd efallai cerddwad! Mae'n grwm yn diogel rôl yn 3 o 4 mlynedd, sydd言odd yn ei hun i'r rôl, ac mae'r cyfrifod maes o ganddo yw, dwi'n pwynydd i'n gweithio'n amlenniad. that you've then got six months to do or eight months to do even more within a same budget or you could look at it as a cost savings. But there's a bit of benefits case coming up from Reeled. No it's fine get to interject to his那就 so yeah, wonderful story is short, the work that has been done at right home a facilitated poed and doing it much quicker. None of that could have happened without some key people who I don't know who these are this is one of seven slides, but his point being that felly knows how things won't happen. So, you need people in procurement who are making decisions to actually understand that sharing code is a good thing and understand how open source can work – that was facilitated in this point, in this case by a few key people. Moving on to this discovery phase – it's kicked off with a problem statement. It doesn't usualy to continue to do the same thing because of how important this was to me Felly, mae'n gweithio'n gweithio'n gweithio'n cyngorol, a gallwn i'r cwyslwyddiant, i'r cofdeithasol i'r cyfrifwyr i ddechrau i ddechrau o gwybran o gyfrifwyr ar gyfer swyddfyniadau dechreu. Felly, mae'r ysgrifffydd yn yn gweithio'n gweithio'n gweithio'n gweithio'n gweithio'n gweithio. Felly, mae'r Gwyr Nŵr, sy'n fyddod o'r ffordd i ddweud, ymgyrch yn gweithio'n gweithio, a'u gyfrifwyr, Do you have scope to go wider later, but we are definitely just focusing on the Drupal. A quick slide just on the potential scope of this, which I thought I put in here, it also comes in later, but just looking at CMS usage by councils. Drupal comes in at number two, about 60 wide round of 400-ish councils in the UK, just behind Jardee which is a proprietary CMS. So there are 60 odd councils using Drupal already, Drupal seven and eight, ac mae'n arweinyddio'r ganddlu ar rhan o'r cyllid yw'r llyfr ac yn ymddangosol bwysig. Ond ydych chi'n gallu bod yn ymddangosol ar y cynnig sy'n gallu'r cyllid cyllid yn ymddangosol. Yn amgylcheddol, mae'n ymddangosol, mae'n cael ei fod yn cael ei fod yn ymddangosol. Yn amgylcheddol, mae'n gwybod ei gael o'r 10 cyllid cyllid cyllid yn ymddangosol, mae bywch chi'n gallu ddigonwch. Mae wedi'i'r ddweud o'r modd, ddigonwch, ydych yn teimlo, mae'r cyfrifolau gweithio i gael yswith, felly mwyaf i'r bod ni'n gallu'n grateful i'w pwysig yma. Rydym ni'n gallu bod nifer i mi fydiannau. Fy多 ychydig o'n hyn ysgol i'n ysgrun ymddangosu i chi'n masaur o gyffredinol iawn i'n ffordd i'n cyfosio i ddefnyddio IAW. gallwn ydych chi'n gweld i'r coi Cymru yn gweithio'n tyfnol i'r bryd yn cwyrdanol, boi ddoch chi'n dweud y dyfodol yn ddiwydd o'r blog-gloes, a chi'n gwybodaeth yn fawr o'r byd i'r cyfnodd. Dyma'r ymddiriaeth ffais ar gyfer, sy'n gweithio yn fwy o'r gwybod yma'r Gŵr Nŵr. Dwi'n rhai, dwi'n rhai, dwi'n rhai. Mae'n rhai, rai, mwy o'r ymddiriannol. Dwi'n rhai. Dwi'n rhai. Yn dynèteidiau fe ddim wedi ei wneud i gwybod, byddai'n fawr ar gyfer Llywodraeth Llywodraethol i gynnar. Felly mae'r gwir datwch hwnnw i ddim hwnnw, yn gwael i gynnar cyffredinol ac yn ei gweithgaredd ohono'r pwysg. Mae'r cyffredinol iawn yw'r hanchyn cysyllt yma i groesu gwir hoffennu amgylcheddol â'r gweithredinol. yn ymgyrch yn gyda'r buddysgu i'w ddweud i ddamwg i gael ysgrifennu wir. Mae'r enw i'r bwyllt yw'r cyfrwmau hefyd ar 5 o 5 o'r cyfrwmau maen nhw. Mae'n gweithio'n gwneud y gweithio. Mae'n cyfrwmau'n cael ei ddweud. Mae'r lleol efo'i gynhyrch ar bobl yw'r cyfrwmau ar arfer ganweithio i'r cyfrwmau. Mae'n cael ei ddweud ar hyn o'r cyfrwmau. Mae'n cael ei ddweud ar y cyfrwmau. Cynidaethau'n unig, ond mae'r unig yn my news yng Nghymru yn y Caelwyr. ond nid yma'n ystafell o'r canlyniadau. Mae'n mynd i'n mynd i'r unig yn y caelwyr, mae nad oedd yma â'r ysgrifennu bywyr erbyn. Mae'n mynd i'n mynd i'n mynd i'n ôl i'r llyfr ar gyllide他们g o bwrth â'r cyllid. Mae'n mynd i ddiddentol o'r canlyniadau i'n mynd i'n mynd i'n mult ffobored o'r ysgrifennu that can share ideas and experiences as well as code So not just, let's have a dيفol site but How did you decide to do that What kind of user research had you done around that sharing varied knowledge ideas and that sort of stuff Finding 3 Open Source I just realised I could look at this screen then I wouldn't have been out of my mind Yeah, Open Source and difeil in particular can work Well forirl for councillers Something that was perhaps surprising Blwysig i'n rhesaith cyfan i gyfan cyf aquatic i gyfan y llywag, ac y llywag o'r wyrdd y brawydol, wrth gwrs, a'r hoffi i ni, a'r hoffi i ni, ac wedi'i ein ffordd i'r hoffi i hyffordd anhygyflwyd y cwmwyaf. Mae 30% cwmwyaf yn ystafell gyda'r hoffi i'r hoffi i'r hoffi i mewn cyf dunig, ac rwy'n ddaf yn ddwy fel y cyf samen ar y linod.iable is seen as an exemplar and it is the biggest open source and the second biggest websites here in the first phase. Fourth finding from the research, co-chairing could work for centres with different starting points and different levels of capability, right? Yes, so some counties might want to take on a complete code base, ddweud o'r cwyl maeth o'r fan i'w mod i'w dweud yr oeddwn. Mae'r cyfnodau wedi ei wneud o'r cyfnod ddweud o'r cyfnod o'r cyfnod. Mae f Abdul Nhaid i ni'r prif, a'r Coylen, a'r ddwylo nhw'n rhoi'n ei wneud. Felly mae'r cyfnod o'r cyfnod wedi cinwseryn i chi gynnyddu cliff yn maen nhw bellwch wedi cwrdd ddwych i gwyl i gyrsynterfynolion yn perthynasio mewn gwirwyr o'u besodol. bydda'r ffyrdd, wrth gwrs, ond bydda'n fawr o'r bryddoedd. Felly mae'n rhan oedd y bryd o ddoch chi'n bwysig ar gyfer y Cysybeth yn ei ddweud. Mae'n ffyrdd o'r byd. 5th ddweud o'r bryd. Rwy'n fawr i'n gwybod i gydag blaen, ysgrifennol yn cymdeithasol, fyddwch yn gydag. Felly dyma'r ffordd, eithaf i ddigitio. Rydyn ni'n gwybod i'n cael eu bod mynd i'n gwybod i'n ddechrau, ac mae'n gwybod i'n gydag i ddigitio. mae'r bwysigau sydd y gallwch yn gennym. Ond efallai ymddangos yn cyd-dweithio'n cymdeithasol, a'r cyd-dweithio'n cyd-dweithio'n cyd-dweithio. Rwy'n dechrau i fynd i gael ymddangos. Ond, ond, mae'r byw fydd yn ei gwybod oherwydd y pryd yn ei gwybod yn ymddangos, a'r byw yn ymdill o'r adroddau. Mae'r byw yn ymddangos, ac mae'r byw yn ymddangos, rwy'n gwybod yn ei gwybod yn ymddangos. Rwy'n meddwl y dyfodol, mae'n ddiddordeb yn y dyfodol localu. Have anyone heard of this? No, nid ydym. I haven't heard of this. You must have heard of this. I haven't heard of this before, but a couple of years ago, the MHCLG Ministry for Housing, Communities and Local Government set out this local digital declaration, which is all around collaborating between local government, sharing ideas, sharing best practices, not reinventing Wales. It goes on to have this phrase, this important statement, where, appropriately, every new IT solution must operate according to technology code of practice, putting us in control of our service data, using open standards where they exist and contributing to their creation where they don't. Now, the technology code of practice is what's interesting. That's coming off gov.uk, and if you follow the links through, you end up on this code of practice, which then links through to saying be open and use open source. Around being open, publishing open source, it's embedded within this local digital declaration. That's something that really framed our project and continued to feed into the ideas of working in the open that this is stuff we need to do. It's interesting if you go and have a look at the local digital declaration because it has got some interesting stuff. I think I'm not sure. Maybe a couple of hundred of councils have signed up to it to say we will do this, but perhaps without really knowing what that means. Switching forces. This is another one of John's slides, which illustrates that there are some things that push people towards perhaps sharing COVID and doing things in a different way. Other things that pull organisations, stopping reinventing the wheel, multiplying the development effort. If two developers are bright and can be working on a bit of code, which actually two developers in Oxford are also working on, we can get more development capacity by collaborating. Some other things that pull back like inertia around European resources or limited capabilities and a bit of anxiety around the distrust of open source or will we need to support it all ourselves. Nothing insurmountable, but worth acknowledging all the same. This slide is around adoption password councils, acknowledging that some people are not. Some people have got an OK website, some people haven't. We're not really initially typing people who have got an OK website who are not on Drupal because that's a bit further off. But there's lots of people who are on Drupal. Maybe you need to upgrade to Drupal 70, Drupal 8. We'll be very interested in maybe adopting the complete code base. However, there are some people who may be happy with their Drupal 8 code base. I just want to take certain components. So, different needs from different councils need to be considered. So, that was the researched section of our discovery phase. The next piece in the puzzle was looking at the technical side of what's possible based on the code base that Brighton and Hove City Council and Croydon. So, is that what it sounds for? Brighton and Hove City Council, the CC, just stand-alone. Needed to check that. Croydon currently share. Are they both human to be counsellers? Brighton is a human to be counsellers, Croydon is a human to be counsellers. So, are they sort of human to be counsellers? I think so. I think they're kind of the same type of counsellers in that way. So, they're needed fairly similar. This is something that came up. We did speak to district councils and city councils in Oxford. We've got county council and the city council. So, that's... Not just to give you the brief on the local district cabinet, the digital person to look at me doing a website. Let's talk about it. Absolutely. So, I mean, there's a lot of people out there who are in the process of doing websites, who are about to start doing websites and so this is the time to start collaborating actively. So, that's part of our technical discovery work. It's just worth mentioning some of the great things that these websites have got. You know, there's a lot of sort of commonality in terms of how people will land on the homepage and want to get through to a particular service. There's a lot of thought gone into common design components. A lot of it leaving from the Gov.UK, best practices. A lot of people will be familiar with Gov.UK's design system. So, the original design that kind of came out at the Brighton website, which was then inherited by the Croydon, you know, largely was influenced by this design system, which includes styles, components and patterns. And this is, I guess, really at the heart of why a lot of time and effort can be saved because a lot of money went into doing user research to find these patterns and to look at the evidence-based design decisions, essentially, what works well for people. And so, rather than reinventing wheels at the local level, we use the components that make sense. So, a couple of things obviously, you know, a button. It looks like a simple thing, but the reason that's designed in that way is based on evidence and based on testing. And so, let's follow that. Maybe it doesn't have to be green, but maybe, you know, so I think I'm pretty sure that Brighton and Hove buttons look pretty similar to that, maybe a slightly different colour. But following that pattern, again, step by step is a classic design pattern that we've seen in government. If you have gone, you know, renewed a passable or something along those lines, or in this case, learned to drive, you'll have your kind of step-by-step process, you'll be able to see where you're at. And this is illustrated, in this case, on the Croydon site by a bunch of Drupal configuration and theming and custom code to essentially replicate that Gov UK pattern in a, you know, content types and related configuration along with the sort of component types of buttons that come on them. So taking all of this knowledge, you know, embedding it into the websites and then, you know, sharing that, not reinventing the wheels again is kind of the essence. So there's lots of really great work that's gone on Brighton and Hove and Croydon. Directory's being another one. Something of the Brighton and Hove website, Plumintur, all councils, you know, they'll need to find things in your local area, in this case, special educational needs and disability support. A reasonable amount of, you know, complexity of configuration and, you know, can be saved by just bundling that up hopefully without too many dependencies and passing it over. So that said, I just wanted to kind of give it a sort of overview of the value that we've got in these websites that we can potentially, you know, share essentially with each other, each other being, you know, local authorities. That said, we first focused on three important questions because we like threes. Is the code shareable? What's the proposed architecture going forward and how do we manage and scale that kind of contribution in a collaborative environment? So taking the first one, as you may imagine, we had to look at the code and the content architecture. We talked to the developers from Brighton and Hove and from Croydon and worked out, you know, how they were feeling about the different aspects and, you know, whether we were understanding it correctly, and then we came up with some recommendations, as you might expect. There's quite a lot of code. This is just a screenshot of some of the shared modules that, you know, I think there's about 30 odd modules that are shared between, you know, custom modules that are shared between the two sites and about 80 contrived modules or so. First thing, the code is generally well written, which is, you know, good, but it's not really written for sharing, as you might expect. I guess we'll come on to why in a moment. The dependencies really is what needs untangling. Back to that idea that we want to be able to, you know, break out directories or break out a little component and be able to use that separately. We really, you know, some people aren't going to want to use everything. So if we're going to be sharing, we need to untangle some of the dependencies. I mean, this is looking at the content dependency chain, where the purple boxes are content types and entity reference were illustrated by lines along with taxonomy terms and paragraph types and other things. There's a lot of interdependencies going on, which I think, you know, evolved in that way because that seemed like a good way to do it. But if we want to break some of these things out, we need to work out how to deal with those dependencies. So, and that's not even at a code level. So, yeah, proposed architecture, as you may expect, a Drupal distribution, including a Drupal core, which you probably need, a local graph core and some optional modules. But again, this is just an illustration of that for the visual purposes. These slides were originally used to present the back to the MXCLG, so I get the idea, I guess, maybe, since someone is preaching to converted here. Is anyone not familiar with Drupal distributions as a concept? No, everyone is great. So, yeah, lots of stuff in the core of the distribution and lots of stuff that's optional that can depend on the distribution. But again, it's all about dependencies, right? It's all about trying to, so, like, if that's directories up there, can I just take directories and use it separately from the distribution? Or can I take it with one of these modules that it depends on and, you know, try and make it much more useful, much more modular. So, yeah, I started kind of following on from that. That's kind of what I just said, really. We need to make changes to the content architecture to move the unnecessary dependencies and make it fully modular. We also need to make changes to the code because there's a lot of it. And again, it's got the same kind of, like, interdependencies within the code that have essentially evolved around the interdependencies of the content. So, these are our kind of recommendations, essentially, based on reviewing the technical, reviewing the code and the configuration. So, to avoid increasing future technical debt, those things I've just said, and then, you know, hearing the best practices in Drupal coding standards, which kind of comes onto how to manage and scale the contribution. At this stage, it's kind of, you know, a couple of developers at Brighton, a couple of developers was at Croydon working together, kind of, you know, working on their own and coming together in a sort of ad hoc, you know, sort of handshake arrangement. But I think going forward is kind of clear that we need to codify that, you know, to be really clear about adopting contribution guidelines, as we know and love in the Drupal community, and really kind of like establishing those in any of the developers who are going to be working together on a shared project. So, yeah, clear expectations, how you contribute, expectations around peer review of anything that's being contributed, coding standards, documentation, and automated testing. This kind of taps into the sort of fear that comes from some of the council, you know, councils that are kind of like, well, if we get into this, then, you know, what about, how do we know it's going to be good quality? How do we know that that what we're getting back from other contributors is going to not break outside? And so, obviously, automated testing is key. And following on from the kind of values that we talked about and working the open being open source, if we can have a fully open source project, people at GitLab provide testing for free, like, I think, 50,000 hours of testing a month or something like that. Just announce they're scaling that back anyway. Oh, quick, sign up now! So, Stephen says they're just announcing they're scaling that back. But anyway, there was open source testing stuff out there for free, which is great. So, to summarise, we want to build on the great work that's been to date, but we are expected to be more shareable and automate as much of the stuff as possible. So, that's the kind of, like, technical review of what we've got and where we're going. Governance and licensing was another strand of the work. So, and this is kind of key, obviously. Aaron, who's the colleague at our collective, led on this along with Andrew Katz, who's the open source lawyer from Morecrofts. And he looked at three important questions. As we do. What's the most appropriate model, as in the model of governance, but also the licensing model? How can organisations join, contribute and leave? So, this idea that there's a group that's a thing that we're going to join and leave, and what does that mean? And how might we take decisions? So, let's just look at what he said. We researched the governance and licensing of similar projects, and then we discussed and agreed what was good enough for now. There is a link there to, I think, some comparisons of open source models. But we did look at the Node.js Foundation, the Apache Foundation and the Perta Foundation, which is a healthcare foundation around sharing open source. And they all have slightly different but very interesting methods of governance. Looking at that and combining it with the research, we found that there's no need for a formally incorporated body of the state. So, not like a foundation or a company that's going to kind of own this and take this forward, that would be somewhat of a blocker to getting involved and would involve finance and complexities. Instead, an agreement that's a memorandum of understanding of a situation. Something lightweight that people can read, understand and sign up to that defines what it is that we're doing and why. And our third conclusion on that front was that licensing models are complicated or can be complicated. Undergats knows a lot about licensing and gave us lots of options and we looked through various models that there may be. But I think it came back down to the fact that we want to work in the open. We're working with open source software that's GPL in the first place, and if we can publish it, it makes sense to follow the GPL route. So, it became clear that that was definitely the path of least resistance. Not to say that there might not be other licensing models in the future that are compatible with GPL if that needs to happen. So, how can organisations join? Like I said, the MOU kind of defines that. The MOU covers these things. That's a kind of street job of it. There's a link to it there if you want to have a look. Purpose, values, IP and licensing, liability, which is potentially sometimes an issue. Governance, starting leaving and commitments of time and money. At this stage, we're not talking about any kind of commitment of money because that just keeps it simple. We're just talking about potentially people committing time to a group. And agreeing on what's going to happen with that time. Shall I just touch on a couple of these? It's really important to define the purpose of the group. So, I'm just going to read this one out. The purpose is to establish and grow an active group of councils to co-develop, share and maintain open source throughful code for our citizen-facing websites. So, it's really limiting it down to citizen-facing websites, which is kind of publishing, not even the transactional stuff where people come and fill forms in, but it's just the publishing of content for now, and specifically doofal. So, really just being clear what it is that people are selling up to. The values, this all comes from the local businesses declaration, which, you know, I'm pulling those things out. Open culture, working in the open, sharing of knowledge and experience. We're using as much as possible, and yet publishing everything under open source licenses. Like I said, the MOU, the MOU is a proposed MOU, whether we're going to test, hopefully, in an alpha stage. But it's gone down well so far. We had comments like, I don't think I've ever enjoyed reading an MOU before, but I did this time. And a very good doc, I like that it's concise. So, this is, you know, this is coming from councils, and this is positive sounds that hopefully we're setting up something that's lightweight enough, but specific enough to actually, you know, go forward. The more thorny question of how we make decisions is something that's going to, you know, yeah, it's going to be interesting to test, essentially. All of this is kind of proposals for testing, by the way, I should underline that. So, we looked at decision making processes in other open source projects and modelled how we might grow as a group. Cos we're talking about four councils initially, maybe eight to ten in a sort of alpha phase, and then what, 2050, all with product managers and developers. And, you know, how we made decisions, which way is the roadmap going? What's the most important feature to develop next? So, we proposed a combined approach, so, acknowledge me, there were different levels of decision making. For important decisions, consensus seeking is a good decision making model. Basically, trying to get everybody to agree. There's a kind of vote, yes, no, abstain to a proposal, but no votes must have a clear reason for no and or an alternative proposal to, you know, be a constructive no, essentially. And then a fallback onto a majority vote, if that fails. And what has been seen in other organisations like the Node.js Foundation is that that rarely ever gets to that point. So, having the backstop of the majority vote means that people can find agreement much, much easier and better. And they don't tend to kind of block decisions without, you know, really thinking about why is they're blocking it. So, that's feeling like important decisions, like changes to the MOU, like the fact that we're going to stop using Drewford and News WordPress or those kinds of things. For more, you know, day-to-day decisions that may be less risky and less, you know, less impact, lazy consensus is, seems to be a nice way to facilitate people to crack on with work. You know, an answer, you tend to do something, allow time for any objections to be raised, and if no one does, crack on. So, for things like, you know, commits to, to commit to code repositories in the Node.js Foundation, I think they use this, they kind of leave a period of 72 hours while reviewers can check, you know, anyone can object to something being committed if they don't. It just goes on, you know, gets committed. So, a way to facilitate action, essentially. So, to summarise, being open and transparent in collaboration is important, but not at the expense of empowering people to do work. So, that, you know, assumption that stuff is going to happen, but, you know, being able to challenge it if needed. So, benefits case that I mentioned we get on to. So, Will Calhann, formerly of Brightland Home, now at Croydon, would have been presenting this section. But, back to the distribution of CMSs in the UK, there's clearly benefit to 61 councillors who are using Drupal if they start to collaborate on immoral Drupal code rather than reinventing bits of it. Both just on sharing code, and also sharing best practices. There's also a massive potential, you know, to people who are not using Drupal yet, but potentially trying to push Drupal up to that side of the curve. What did he say here? Potential time and money savings. So, yeah, launching new council sites more quickly, building fewer standalone micro sites, reducing unnecessary customer content and sharing knowledge and skills made for this better. I guess it is what it says. So, I think, yeah, we'll just touch on a few of these here. So, forgive me. Go on. Sorry. Question. Should that be customer content rather than custom content? Or content? Content. So, unnecessary, I'll come onto that. And what's that? So, I think there is this side. Yeah, so customer, okay, that way. Customer contact is, you know, people phoning up or coming in to try and find information which could otherwise be delivered on the website. This could be a customer contact. Yeah. So, anecdotally, you know, improved websites reduces the amount of customer contact, which saves money, but it's very difficult to measure that. It'd be really good to try and measure that at some point. Because that's all about channel shift, getting people to stop phoning up and just finding the information on the website. And we haven't got any kind of figures on that, but what Will has prepared some figures on is looking at how much might be saved in, you know, broader strokes. Looking at different council types, people are switching from off-the-shelf solutions to Drupal. And that's been looking at saving, you know, 65,000 and 16 per round a year. I'll place them in a couple of examples. People are looking to do a one-off build or with an in-house team or with using an agency and some assumptions figures, which we think are quite conservative. Totally needs that, assuming there was 55 councils who might be doing this. We're looking at figures, so we know a few million pounds. Probably over a few years, maybe three, four years, depending on what your kind of life cycle or your expected lifetime of a website is. But again, hard to be accurate and specific, but broad strokes does seem to be there's potential. And if you scale that out from 55 councils to hundreds of councils, I mean, you know, we're talking about serious potential savings. Sorry, do you want to? I hate it when people do that, right? When you're just about to take a photo and they skip the slide to the next one. The slides are available on bit.ly slash localgov.doogle if you want them as well. It's microsites, microsites. So every council needs microsites apparently. They've got a new thing, the new co-branded campaign going on. They can't do it on their own website. It needs to look a bit different. It's got slightly different content. Some interesting things around accessibility testing, often costing, you know, increasingly costing a bit of money, and penetration testing costing quite a lot of money. Whenever a new, you know, let's just get another reminder. Oh, that needs to go through pen testing and other stuff. So there's costs involved there, which could be safe by having microsites on your main drupal site or even spun off as part of the distributions that you already know is past, you know, penetration testing and accessibility testing and potentially saving on design and dev time if you can persuade those people that you don't need to do any different design. I know Brian Hove and Corridon have got some campaigns pages now, which are sort of much more flexible in terms of content and branding and co-branding. And that's the kind of like microsite killer idea which could potentially save, you know, again millions of pounds across a number of councils. But yeah, to be seen. I mentioned the channel shift idea around, you know, reducing customer contact is very difficult to measure, but we will try to. So, apparently, every good collaboration needs a mission patch, which is also known as, well, I've never heard that phrase before with this project. He's learning a lot on this project. But at some point on a call, I was pictured wearing a fun hat. You know, I'd like to keep things fun in the office. I didn't know somebody was taking a photo of me at the time. But it was a productive call. But that's led to the mission patch having a nice drink of cotton with a cowboy hat, pink cowboy hat on the day for later. So stickers have been made. Unfortunately, I don't have any with me today, but stickers have been made and sent to us to celebrate the discovery phase that finished last Tuesday. So that's all good and fun. It makes it all worthwhile. Next steps is an alpha phase. So these phases are coming off sort of government, GDS kind of like recommended phases for services that are developed, you know, kind of like Discovery, Alpha, Beta and then Live or Launch or whatever. And MHCLG's funding is linked to these phases. So we've had funding for the discovery phase. The next phase is an alpha phase where we're going to be testing a lot of some of this stuff. So do a bunch of refactoring on the code and theme to make it much more shareable and modular and testing some of these concepts around, you know, flexibility of the code with councils, get some other developers involved and decide to see how we collaborate. Test the MOU and the licensing proposition. Will people sign up to this? Can we scale this out to 10 councils? And, you know, have collaboration sessions and have people in sort of sprint planning, sprint review, you know, working together over different levels of resources and different councils and see how that actually really works. And, you know, hopefully through doing so, help some particular councils get on board with this code. And, you know, promoting to other councils so setting up a, I think a shop window as a way, you know, a website or maybe somewhere where it's clearly documented how you get involved and how you crack on. So, councils that are going to be involved in the alpha phase include exemptions to covering council Westminster, London councils, which is a kind of group of councils. Quite. And, Kensington and Chelsea, alongside the original three, although we're not sure whether Oxford City is involved in the alpha phase anymore. And, I think, yesterday, these guys just posted something on their website saying that they have approved funding for an alpha phase to Croydon for the project, which is really exciting. So, yeah. Props goes out to the Ministry of Housing, Communities and Local Government for actually having money and putting it down and saying open source is a really good thing to do. Go on, carry on, explore it, test it, try it, prove that it works, you know, because that's really going to facilitate creating an amazing distribution. So, hopefully off the back of that we take what Croydon's got, the evolution from Brighton to Croydon, break it up into, you know, local golf corps and components and make it reusable and, you know, really, really start to build that. Obviously, custom theme on top of that is a next layer, you know, do we have a base theme which has a bunch of assumptions? Can we actually take it to the point where a council with no developers at all can point and click their way to a sort of low-code website and replace the logo, replace the colours and actually have pretty much what they need for a fraction of the cost. That's probably a bit further off, but that's certainly part of the vision. Sounds like that's dual, though, because you've been talking about the button themes before. Exactly. So, you know, I have been componentising the theme down to the component level and the atomic or molecular, whatever the design terms are these days. And, you know, having variable colours and such and, you know, actually following GDS patterns with some local golf patterns evolving and, you know, sharing it sharing it at that level. One thing I'm really excited about is collaborating with other distributions. So, part of this project has been looking at some other distributions and seeing how they do things. Drupalate distributions, obviously, you know, that's sort of not quite as far ahead as Drupal 7 distributions, but Drupal 8 really enables distributions to happen in a much more scalable and modular way, I think. Lightning, Thunder, Lightning being aqueer, sort of starter kit, Thunder being a publishing distribution, which is amazing. Open social being a more sort of Facebook-y type community distribution VAR base kind of does everything, as far as I can tell. And I think it's got a e-learning platform. But they're all distributions doing different things. And as I was speaking to a guy from Thunder yesterday, he was really excited about sharing what they do. They've got six full-time developers at Berda media group, is it, in Germany? And it's very focused on, you know, media publishing and that kind of stuff. Doesn't really have any front-end, it's just all about the editing experience. But a lot of those are the same problems. They've got multiple stakeholders, multiple product managers, multiple people wanting different things and how they actually navigate that and prioritise how much effort they put on automated testing. So I think there's a massive scope to learn from other distributions and for distributions to start talking to each other. I mean, they probably already do, but I'd just like to get involved in those discussions. And that's really exciting. But also, you know, people who are also, like you say, have got a local council who's come to them for a website. Like, let's talk about the distribution together, get more people plugged into it, more developers, more of the community and more councils, I think. So that's exciting. So yeah, that's it really. Thanks for listening. And you can follow where you find me on Twitter and add updates on Twitter. We'll be here today at Will Gov on Twitter, quite chatty and dfw on Twitter. And there's a local Gov truthful digital slack channel, I believe. Not by it's been there. So yeah, any questions or suggestions? Well, yeah, I'm just going to say it. How do we get involved? How do we get involved? So I'm at Alton Forest. I'd love to get involved in this project for such a reason. I think probably Will, at Croydon, is the best kind of, I mean, talk to me when you can't have details, but I think definitely getting in touch with them is he's gathering more councils to try and to test with the alpha and see where they're at and, you know, you know, get more people involved. Definitely, yeah. One of the biggest barriers in my district council is finding planning applications because the search, the system I've got makes searching for planning applications almost impossible. Because you've got to know the width that the planning application is actually processing, which is absolutely impossible to know unless they've been told by the developer or something like that. Absolutely, yeah. So at the moment, has this been involved? Maybe. At the moment, this is very much just councils publishing content onto their website. I've got another thing out. It's got once we have established the mechanism to share thinking, you know, have that kind of collaboration mechanism proven that it's scalable across more than just two councils. Then the next thing comes in, right? Yeah. Can we then start doing a system that solves that because everyone's got the same problem? What about forms? I know Brighton and Hove are using Drupal web forms as the front-end tool of their forms and then plugging that into APIs to go to a particular form engine behind it. Another potential, you know, sort of integrations with more compact systems where perhaps, I mean, 30% of councils are using the same form system but they could have something else that front-ends it and makes it much easier more accessible or whatever else, you know, and the workflow starts to fell with that side of the front-end. So you absolutely, that's a potential future, right? But in order to get this thing off the ground, they very wisely focus just on the publishing of content. Have you also looked at the way that plans can start when you use the sharing group? Okay. Now that was mentioned to me by Gareth earlier. So Suffolth Constagory, Essex Constagory I think it's it. But certainly Northwick are using the same group or code just really theming it for their own catch. But that's not probably the available in their network. I'm trying to find out who did it. I think it's right. Was it Gareth or? No. I've pointed you to Gareth earlier. He's talking to me about that, I think he's talking about that. But yeah. I think it was quite fascinating that they've all got the same requirements. And then there's schools, right? And they're saying, I think they're saying there's lots of other scope for similar requirements of Drupal distributions that scale out. You know, and I think as Dries has often said Drupal distributions are really, you know, important in getting people on board of Drupal and solving particular problems. And, you know, yeah. Yes, so. Two things. First, thanks for name checking Nicol, whom I used to work for. Okay. Cool. Yeah. Although I never worked on DHCC. You mentioned distributions just now in my limited experience as a site that we're working with distributions. I can spend half my time trying to turn off or bypass a whole lot of assumptions which I actually didn't want within it. And I suppose you're not at this status right now where that's actually a real problem. But it just struck me as something to potentially worry about or think about in the process. It definitely is. I mean, you know, even embedded in that kind of rapidly put together image of what local gov core might include. You have massive assumptions around the standard page and the back page. And then the service pages, we are here that all councils provide services and they need to promote services and what they mean by services like, you know, you can come and I don't know, get you parking permits and get people to badge on them. So they call them services, right? And embedding that course seems absolutely logical to everyone involved. But who knows whether these assumptions will then actually tie us up in the future. So where to do that line, right? How where to do that kind of like baseline of opinions and assumptions. But a distribution is opinionated. That's the point, right? It makes a lot of, it has opinions, it makes assumptions in order to save you time. As long as it's really clear about what they are and hopefully makes them what is open up that you could potentially use some of that without some of that. Then, you know, I think that's the art of creating a sustainable architecture for a distribution which we'll see when we get to it. As we go, this is just based off the assumptions that we found from talking. But yeah, that's it. It's a really it's a really good point and a really important question. It's following on from there, really. What conversations have been had about configuration management in terms of the basic essential configuration for the distribution and then allowing for third-side variations but still be not a really configuration or new features from work? Yeah, so so configuration management so, you know, are you expecting to get updates from the distribution? So you install your distribution, you do some changes to config, add some fields, remove some fields and then, you know, and then version, you know, the next point version of the distribution has come out and you want to upgrade for security reasons or get some extra features. So how does that preserve those preserve your configuration changes? Yeah. We haven't made any firm decisions on any of that yet. You know, this is early discovery phase. Looking at thunder and talking to thunder, they have some interesting stuff on that. I think there's some, I mean, Phil, maybe you've got some config management. There's something called config diff update or something like that or config update helper. It's a config distro which promises to allow for that. I haven't worked with it. So there's something that I as I understand it, it looks at the diff between config and helps you deploy update hooks to change config on places where you might do not just want to import the config blind but you basically, you know, if the change is to add a new field, then it will bring an additional field without overriding your config. But I haven't looked into the technicalities of that. You can write update hooks to change your config and then apply those update hooks into everything. But we've had a similar problem with lightning, like the update lightning and it will break the config because it's expecting because it isn't in the config. Yeah. There are these questions, right, about how is a distribution meant to be used? Is it meant to be installed and then you just kind of carry on and go off and you'd update core and you'd update any kind of contrary modules that are involved but you would never update the actual install profile and expect that to happen. So essentially, it's to start a gear and off you go. That's one way a distribution can work and some distributions are meant to be that. And then other distributions are kind of like open social side. That's everything, right? When there's a security release or something else which may be in the distribution or maybe a core or contrary, then you need to update that whole distribution and that's where this problem comes in, right? I think it's not an easy one to solve. I think there was comments that profiles could take on like themes have a base profile and then you can have your own profile which is based upon another profile. Okay, so like inheritance, yeah. Inheriting, yeah, yeah, yeah. So child profiles. Yes. That's how Lightning tries to do it but it often has to pack the group all to get that working properly. Yeah, that sometimes is quite a work. Yeah. From experience. Yeah, yeah. And then there's a discussion on features, right, as well. Like how features is used in Drupal A to enable functionality but then do we want to be using that to move the configuration still or the modules? So, yeah. You should use your the installer inside the module to install Drupal config as much as possible. Probably a bigger discussion about how much is in profile and how much is in modules. Yeah. But you can get modules to install the code from a period in that. Then if you change the config and you manage that, I guess that sort of features might come in but we expect to be done through them to update looks. Yes. Yeah, yeah. So, avoiding the previous Drupal 7 paradigm of a feature defining lots of functionalities, the configuration essentially that it wants to maintain and keep is just configuration in your module that provides a starting point for content types, views, whatever else it may be configuring. And then you kind of move forward from that point so you can add fields and roof fields. Cool, any more questions? Yeah. Just all features, OpenSocial has had some fun games with features because if you effectively customise an OpenSocial distribution and then they update a feature great to them. Okay. Yeah. We've been waiting for your perspective three months to explore open config and I couldn't even recover it from the config files. So again, OpenSocial features, config changes, causing massive problems. Just to see how the fun and games they're having. Yeah, definitely. They're moving away from features. Yes. That's the way they're going. They're moving away from features. Right. That's what you were saying. That's probably why they're moving away from features. Yeah. But yeah. I mean, we're in a much better place in Drupal 8 to be able to deal with these nuances of config changes define how we expect it to happen and then be clear about that with the consumers of these products, right, of the distribution. Like, be careful if you change on some of this config because that might cost problems but if you want to do this or this, then that's going to be fine. So I'm renaming just one of the content types and we've just broken in. Yeah, renaming content types on fields. That's going to be in the future. Okay. It's the label, right? Just the label. Because that is what config distro policies and so on. Yeah. That's just one of them. There's like half a dozen too many. We couldn't we couldn't basically be like are we are we are we about time? Anyone want to get any more questions? Tristan. I've got a lot of people from LLM. One more question? There's one that I might edit. When might there be a camera work? It's going to be tomorrow. When they might, when might there be a work? Code. Code. Something, yeah, up there. Yeah, somewhere publicly available. Possibly after the after phase which is probably on the main path when you're doing the after phase. So the after phase is due to start in the late fall and so I would expect things to start, you know, to be materialising of May but to be so. Thank you very much everyone. Thank you for that thing.