 Hello, so I'm Kyle and today I'm going to talk about KPM Core, explain about what this library is and I think that most of you know what KDE partition manager is so I will not explain that and I will talk about the release of KDE partition manager 4.0 and the goals for the future of the library. So for those who don't know me yet, I'm Kyle and I'm a Brazilian KDE contributor and I work with KDE partition manager Calamaris which is maintained by Adrian and now I'm starting a new KDE application which is called Mark and it's a dataset builder for AI and supervised machine learning and also I'm a student at the Federal Institute of Bahia. So these are my presentation topics, I will start explaining about KPM Core, about my work on Season of KDE 2018, about my work on Google Summer of Code 2018 and I've been working as a mentor in Google Summer of Code 2019, I will explain some new features, released in the KDE partition manager 4.0 and the goals for the future. So for those who don't know what KPM Core is, it's basically the core of the KDE partition manager as the name suggests and it's basically a library for managing partitions, doing some disk operations and it includes now RAID support, LVM support and it can be used by not only KDE partition manager but other applications. It is used by Calamaris, the partitioning module of Calamaris and it has been in development for some amount of a good number of years and first I will explain my work on Season of KDE 2018 which was focused on the replacement of a maintained Libyata smart library to the parsing of the smart control command in KDE partition manager. So Libyata smart is a library to manage the smart support, to see some information about your disks, about the status of your disks and some other data but it's not maintained since 2010, 2011, so we decided to use the smart control instead which was released in this year for smart mount tools and one of my works involved the parsing of the smart control JSON output. I could remove the Libyata smart library dependency and I also work on something more than the smart support which was the improvement of the authentication support with Kout so users can now do only one step of authentication instead of many and during Google Summer of Code 2018 I could finish LVM group support on KDE partition manager and I've also implemented some graphic interfaces for LVM support in Colomaris as well. I've also implemented the MD rate support so full support actually not only the creation of rate devices but the edit, the manipulation and all the other procedures. It's not full stable now that's why we couldn't release it with KDE partition manager 4.0 because there are some issues with the devising mapping but one of our goals that I will explain later is to merge the MD rate work with KPM core and there is also a plan to integrate not only MD rate but DM rate which are two different types of rates for Linux. During Google Summer of Code 2019 we worked with our students Shubhan on the support the authentication process from Kout to Polkit Qt1 because we needed to remove the extra layers from Kout and use Polkit Qt directly and our students also needed to improve the key Dbus communication because now the authentication process is done by the helper we start the Dbus helper which we can use to do the process that needs authentication. In KDE partition manager 4.0 we could port the KPM core backhand from Libby ported to SF disk. This is part of one of our plans to make KDE partition manager more portable including the portability to free BSD and my work with the smart support was merged and we included the better support for Lux2 which was implemented in KDE partition manager 3.3 by Andrius and now we have the support for some new file systems such as APFS and Microsoft BitLocker. We also included modern C++ and all these ports both of smart and SF disks made KPM more portable. And goes for the future as I said before we want to bring full compatibility with free BSD and we also want to finish the Polkit port because now there are some issues with the current implementation especially on the scanning process and we want to evolve FS support include some specific features for file systems and merge MDE rates and finish the implementation of DM rate and another goal is to increase the number of automated tests because now we do not have a proper number of tests in KDE partition manager and that's it. I would like to say thanks to KDE for this opportunity and all the other opportunities that I have received. Thank you, Kyo. Are there questions about KPM core? It's not his fault, it's Calamaris' fault, okay? How did you know what I was going to ask? We want to use Calamaris in neon and one of the limitations is full disk encryption. Is that your fault or is that his fault? Where is the responsibility for that? Probably it's Andrew's fault, but I'm not like, say that it's his fault. I don't like other people. It's my fault. Other questions? Yep, we got one up there. Is it possible for you to go into some of the technical reasons why you kind of moved away from KOF and closer just to Polkit itself? Okay, so one of the decisions was because we do not use all the features from KLT so it's kind of... There are some dependencies that we are not completely using. So we just needed Polkit to authenticate the helper. It's basically that because we call the... We start the debug service with a super user and yeah, it's basically that. It was more Andrew's decision than mine. Partitioning is a bit of a weird creature because you need to run multiple commands. You need to DD zeros onto your MBR. You need to reinitialize the MBR. Then you have to make partitions. You have to run multiple file systems. You have to run a bunch of things in a row and all of it takes a long time. There's special authentication stuff happening anyway with the helper. That kind of makes KOF a weird match. Any other questions? Otherwise, I want to ask you about ZFS. Because with ZFS on Linux now being the source of truth for ZFS and we have unified FreeBSD, Illumos and Linux ZFS implementations, that's one of the most portable file systems around. Do you see KPM Core supporting ZFS soon? Yeah, it's one of our plans for the next version is to work directly on file systems, not only including new file systems that we do not have support now, but including the, as I said before, some features. Because now we only work with FS with the file system support. In a more, how can I say that? In a more abstract way. Cool. Yeah, I can do that. Cool. We've got one more question over here. Is supportive hardware crypto, so use B-Stick with something like a Nitro key or similar that holds the key to your encrypted devices. Is this something that you currently support or have on the roadmap to support? I'm not the best person to talk about that because I've not implemented the Linux support, so I'm not aware about encryption. It's better to send a message to Andrews talking about that. Yeah, I'm not the best. That's already right in order to be able to look for the key. It needs to know that. Yeah, it's a complete integration thing where KPM Core enters into the equation but also whatever your distro does. Or calamaris. It's not clever enough. Anything else? Otherwise, there is coffee and possibly cake. Hang on, Johan is keeping you away from coffee and cake. I want you to know that. Maybe I'm missing something, but as of this, do you understand, does KPM Core understand GPT tables? Yes. That was the best question ever. It was short and it had a one-word answer. Thank you. Let's give Kaijo a hand.