 Rhywun, dydd. Rhaid? Rhaid? Yn y gallu gyda'r clynyddus. Rhywbeth, mae'n cael ei wneud yma'r october 2023 yn Fcor Hachathon. Rhaid? Rhaid? Fynd yn ei wneud. Rhaid i'r gweithio ar y ddiweddol yw'r rhagwyr ar y tair hwnnw, ydw'i ar y ddechrau, ac yn eich i weld i'r gweithio. Mae'r gweithio'r gweithio, ac mae'n hyn o'r ffunt o'n gwych i'r gweithio gweithio, a'r ffawr a'r enthysgwysbeth fydd yn fawr yn ydym. Yn ddod yn y gweithio'r rhain a'n mynd i ddim yn llidiau, a'n ddiddordeb yn y gweithio'r rhain, felly rydyn ni'n golygu'r rhain a'n gweithio'r rhai ychydig o ffwrdd. Rwy'n fyddo'n gweithio, mae'n mynd i ddim yn gweithio'r rhain. Rydyn ni'n gweithio'r rhain o'r rhain, It's been a very smooth experience made it very easy for us. Thank you everyone. I'm really looking forward to pouring some of the details of this hack-a-thon. I feel like we've done a lot of stuff together. Just as we have with the daily summaries we're wide through some of the different groups to hear about things you have all been working on. Mae'r ddechrau i nhw i chi weithio i ddweud eich cwm yma, maen nhw'n rhan o'r ddweud o'r meddwl, gan ymddangos o hyd yn ystod. Mae'r ddweud yn rydych chi'n iawn. Mae'r dduch i chi? Rhaid i? A chyn nhw'n ddych chi'n cynnig beth angen rhaid, ychydig yn siŵr i'r wneud cael myfyrdd i chi ddim yn ni'n rhaid. Mae oedd yn cyflym gyda'r ennillodd, mae'n ddweud thai ni ddweud o'r ddweud, So, I think everybody looked like this cat this morning, just opening and opening poor requests. Maybe we'll just move over right to the numbers. I sort of gave up counting per day. It's just a summary for the whole summit now. So we had 51 PRs opened, 42 merged. Maybe not the same ones that were opened so we still need to do a lot of reviews. So if anybody is around to still do this, then it would be great if you could help out next week still or if you feel like your PR is getting stuck in limbo, just keep pinging us on request review. Somebody will hopefully take care of it. We have 28 new modules and also one new sub-work flow and a bunch of issues opened and only some of them closed. I think that's it. Hello, hello. Yes, that's better. I don't know what happened. So hello everyone. First of all I would like to thank Adam. He managed to achieve what I wanted to do when I joined Sekira, which is get rid of Python within all of the NFCOR pipeline. We removed the last 300 lines of Python within the NFCOR pipeline. I'm proud to say that. I think I'm one of the only by-information years that doesn't use Python at all. Yes, it works. So pipeline-wise, Metal Seek has been released, which is super good. Crema Catalana. I think some people were picking attention last night, so that's nice. We've been adding NF tests for a lot of sub-work flow, Louis and June when working on that. Molcart is fully functional and only missing some tests. Where are the... great job. Differential abundance is ready for a new release, and waiting some of the module like CI issue. Super good job as well. I'm waiting for the last test to finish into fetch and yes. I'm like... You've seen all of the memes that we've added and yes, that's really that. I have like 48 tests that finish. I'm just waiting for the last two. So I was really hoping to meet all that and release. I'm afraid I won't be able to do a live release. I don't know. Yes, say something. Great work, everyone. So I just wanted to say a few words, because the main theme of this particular hackathon has been NF tests, and how we're going to adopt it more broadly on NF core and put high-test workflow to bed a little bit and transition over. It's been immensely useful having some of the brilliant minds that we've got here in a single room where we can sort of figure out how we standardize on this on NF core in general, and I think we've made awesome progress. Fetch NGS, which Maxime potentially might release soon, is sort of our template POC example of that. But we've also had other discussions about how we can refactor workflows to make each of these components a bit more self-enclosed that will help with testing as well as configuration and various other things that I think are going to be awesome once we've adopted them. Again, great progress, and I think we've ticked all the boxes in terms of coming here and trying to figure out exactly what we want to do with NF tests. The authors of NF tests are standing at the back there, so if you have any questions, bugs, issues, problems in life generally that you want to go and speak to them about, then go and pester them. They're here in person, they'll be here for the summit. Great, thanks, great work guys, thank you. Yes, so just one last test for me. It's getting there, it's getting there, so yes. I think I'm going to wait for it and maybe join in on the fun once again to finish the release. Either after Julia or after Ben. Julia, your turn. Okay, hello, so big news for infrastructure in our last day. Nicholas could release a new version of NF validation, so huge celebration on that. And also now NF validation releases have names and they are named after ramen types. Then we also catch up in all the PR reviews that we didn't do yesterday. So Sofia worked on adding the admonition syntax and she was also adding a warning for DSL to pipelines and last we asked the gods of AWS to get GPUs to run some tests on GPUs. And that's our last numbers for today, so one issue closed, one PR open and four PRs merged and also five reviews and today we didn't win any quiz. That's all. Howdy folks. Wait, this is wrong. Is this outdated? I can't, this is it. Is this like a different version of the slides? Let's see if I can use a Mac. I don't know what just happened. Okay, I had to refresh, that's funny. I found another meme for you guys. This is how I feel whenever I submit a next flow PR and Paulo rejects it because it would break existing pipelines, things like that that nobody cares about. Come on. But you know what you don't need Paulo's approval for is a next flow plugin. So we've got a lot of those today. First of all, huge shout out to Jordi. He knocked out another big item. This custom process directive for dumping the version, like the tool versions. And as a bonus he added another directive that allows you to create like custom trace fields, which is something I could have used in my PhD research. Hopefully it will be useful to you guys as well. So that PR is sitting there. So thank you to Jordi. He's really like our secret weapon at Sakura. So go and tell him thank you when you see him. Mahesh is still working on this joint operator. He's having some issues writing next flow plugins. I guess you don't need Paulo's approval, but the plugins are kind of a Wild West. There's sometimes things don't work. But I've been working with him and we're trying to get those kinks worked out. That's a mess. We'll figure it out Mahesh, I promise. We'll get there. Francesco looks like he's just about done with this training material he made for variant calling. So I guess that's somewhere and you can go look it up if you want to see it and go talk to him. It looks pretty cool. He's got some cool diagrams. I think it was a scaladraw. I like the little hand writing diagrams. It's a good time. Okay. We got for the NFCO2 footprint. It looks like they're just about done and they're testing it with 23.10 that just came out. I'm just about done with the RO crate format. I think there's just a few things left to add and then we'll get some feedback from those RO crate people and what they think. Also if anybody in here, if you're interested in this RO crate format, would love to get your feedback once I've got a PR that you guys can look at. Rob Newman, well he finished this yesterday but I didn't merge it until today. This is the only thing I actually merged because it's really simple. I can't merge these things up here because it's a next flow PR and the carbon footprint people are doing their own thing so they don't need me for that. So that's good. On this channel factory to load basically this sample sheet data from a provenance manifest but an interesting turn of events we might actually just use quilt to do that. So the idea is that you could do a pipeline run, publish your quilt package and in your quilt package you can include all the metadata you would normally put in a sample sheet and then on a downstream pipeline where you would normally load a sample sheet you would have some kind of channel factory like this from something like from quilt where you would give it that URL to your quilt package and then you could parse out whatever metadata and files you need and go from there and just do that as much as you want. I think that's everything I've got. This was my first hackathon leading a group and I got to say it was a lot of fun. I really enjoyed working with all of you and spoke with. Looking forward to the next one. Final test has passed. Can I plug in my computer or am I going to break stuff? Just pressing buttons. Yes. I usually don't do that much as you can see. Merge pull request. Merge. Now let's do the new release. Tough new release. We are good. I'm lazy so I just use this kind of stuff. Then I think we are good for everything and publish release. Boom. That's quite a finale to finish on. I think that sums things up very nicely. We came into this hackathon saying we wanted to focus to be adoption of NF test and to wrap it up with the first release of a pipeline which purely uses NF test as a testament to everyone's hard work. It's passed three days so well done everyone. Great stuff. With that, thank you everyone. We'll see you maybe in Boston one or two of you for the next hackathon. That one's in person only. Otherwise we'll see you next year. If anyone left the apron last night you can come and get it. Might be able to identify the group by the food splatter. That's up here. Just come and grab it if you're missing yours. Otherwise, thanks again. Take all your stuff with you and we will see you at the next flow summit in a few hours time. Thank you.