 And then there's four people on YouTube or no, yep, okay. I see myself It sounds like a notification to people right on the in the YouTube app is that how these things work it does right? Yeah, I turn off all my notifications. I don't know. Okay. I'm gonna start right When it hits the top of the share screen Oops looks crazy Right everybody can hear me right all right. So my name is Victor I'm a private manager here at good lab and you are watching the good lab 11.7 kickoff meeting So you should see on your screen a kickoff document And that should be There should be a link if you're watching in the youtube stream There should be a link in the description of the video And so I'd like to start off with what the plan team is doing And so in 11.7, we're going to be working on these three features Scrolling horizontally in the roadmap view. So if you look at the roadmap view right now in good lab It's great. You can select quarters months or weeks And if you select quarters, it gives you this many quarters And that's the longest in terms of time That you can go and so you can't go further in the past or further into the future And so that's exactly exactly what this issue will address allow you to scroll backward and time or forward and time We're actually considering maybe you know using buttons or But suffice it to say we'll figure out a way to to do that The next issue is searching or filtering in the boards navigation drop down searching and filtering So what I mean by that is in the boards drop down you see Something like this currently and you see a lot of boards. So this is you know, sort of silly With the it's not we usually don't have the same board names, but you have many many different boards And so this will allow you to Be able to quickly find the board you're looking for in a typical issue board navigation And then the last one I wanted to highlight is importing csv Into issues. So what I mean by that is that a lot of times you might be using An external issue tracker such as jira or some other System and you wanted to import those issues into good lab And so we're going to provide a way to do exactly that in the in the format the comma separated format Which is very typical because many tools allow you to export a csv file And we want you to allow we're going to allow you to take that file and import it into good lab itself and so how we are planning to do that is Very typically you have an issues list here at the project level And you have a button to to export as csv. So we're going to have another button here Right next to it that says import and when you click on that You're going to get a modal something like this where you can allow yourself to Upload the csv click import and this will happen behind the scenes asynchronously And then you'll get an email notification Once You know it has confirmed good lab has confirmed that it's complete And then you can go back to the issue page and see see those results. So that's what plan is working on james up to you Thanks, victor The create team's got the three particular improvements were focused on in the 11 dots Place uh, this is a blockage request Which we began working on last release in 11.6 um, apparently for merge request approvals you're limited to Specifying a list of users or groups and the count of approvals and what multiple approval rules does is allow that to be Increased so that you can have different rules. So instead of just saying I need one of jane mary or bob to approve my merge request You can now say I need one staff engineer One product manager and one qa and they can be separate rules that get validated individually And so what this will mean is when you're if you choose to enable multiple rules You'll see a structured breakdown of the different approval rules So here we've got an example of some different teams that might need to approve Some different approval count requirements for each of those rules and who has approved and this is um Building on and working towards the changes we've been making to introduce A more sophisticated code owners approval system In addition to the manually configured merge request approvals The next change that we're working on is um To do with self approval merge requests a few releases back. We made it possible to Turn self approval blocking on off. Sorry. So by default the merge request author could not approve their own merge request And really this behavior is to make sure that The the user who's introducing a code change isn't the person who's approving it but the merge request author may be different from the commit author or There might be multiple authors in a merge request So if a teammate jumps in and starts adding a couple of commits on a more complicated feature Say a front end a back end developer collaborating with each other um Currently we only block the author of the merge request, which is Not strictly the same person. And so what we'll be doing is we'll be upgrading the behavior to actually block The merge request author and the commit authors in the merge request and that should um help I guess deliver the more expected behavior around disabling self approvals and the next um change is I guess Uh less exciting from a the visuals. There's no change in behavior except gitlab Merge request should be faster. One of the most expensive things that we do in merge request is render the syntax highlighted diffs And I wanted to pull out some really exciting work happening to tune merge requests and make them even faster We're shipping improvements in 11.6 and continue to ship more 11.7 And there'll be many more to come because performance is a very important feature at gitlab Jason how about verify All right, uh, let me get my screen shared here All right, so for verifying this release we're going to be finishing up the upstream and downstream pipeline expansion This is a feature that we started in the 11.6 release, which will let you explore pipelines up and down Kind of the way that they were triggered We did do a decent amount of work on this in the 11.6 release. So we do have the basic functionality working Which you can see here in the next screen Where you can go up and down just one What we'll be finishing up in this release is making sure that you can go You can continue Up and down the whole pipeline tree to see how your job was started and what was eventually triggered when you got to the end And that's it for verify. So I'll turn things over to you josh Thanks, jason. So uh, for package team, we'll have two major deliverables here for this coming release The first one is we'll be working to finish our Support for our npm package management. Uh, so similar to how we implemented maven support Oh, I think it led a couple releases ago. I will also be adding npm to that Support as well. And so you'll be able to upload npm packages to git lab And share them without your colleagues and within your organization As well as publicly if you'd like to so that's really exciting there as far as generation towards that vision Um, and the next item we have also is a group based maven api endpoint This will make it a lot easier for folks to consume Maven libraries across different projects within their organization Previously we had a project based Repository url where you essentially had to have Essentially a repository per project and the goal here is with this group based repository is to remove that need You could have one sort of uh repository for a top level group And then you can then leverage all the packages that can be contained in there And so setup should be easier and should be easier to consume packages going forward Once you've already added a repository for a single Maven that your instance that you're using within your organization So two great features there for our package and with that back to jason to talk about release right, um, so uh on the release side We're going to be introducing the releases page mpc Uh, which is a feature that we're really excited to bring and it's part of our product vision for 2019 What this is going to allow you to do is to have a page that looks something like this Within your project that will allow you to publish releases Sharing what's changed in them allowing the source code to be downloaded And then linking to the actual asset that's associated with that release Um, so this is a great feature and one that we would love your feedback on so please Feel free to jump into this issue for 1766 and let us know if you have any feedback on how this feature can work for you And with that i'll turn things over to you daniel. Thank you jason. Let me go ahead and share my screen here Okay, so the configure team is going to be working on a couple of exciting things so the first one of them is going to be api support for our kubernetes integration and uh, basically This is going to allow users to do all the activities that we do from the GUI today such as creating clusters Deleting clusters or making edits all through the through an api And basically they'll um will allow you to scale um activities that were done per project Of course with the introduction of group level clusters um This is going to give you kind of more latitude to uh operate in your clusters So that's kind of the first thing that we're working on moving on To upgrading kubernetes runner via our integration So currently when we install the runner application using our kubernetes integration There is no easy way to upgrade the runner version. Let's say if you want to take advantage of new features That have been added newer versions So basically what we're going to look to do here is show you what version is currently running On that um application and give you an easy way to upgrade it Moving on this is one that we're carrying over from 11 6 and this is configuring autodub ops With secrets that are not in your repo So let's say if you have a rails app and you need to store your rails secret We really don't want to store that sort of thing in your repo So what we're going to do is we're going to have a way for users to specify a variable That's going to be prepended With this syntax here of k8s underscore secret and that's basically going to tell The integration that this is to be treated a as a kubernetes secret Basically, you will not have to store it in your repo any longer Uh, the next one here is on the serverless side. So basically uh with 11 6 Users are now able to deploy functions to their k native um application And currently we're only showing you a list a flat list of functions And what we want to do here is show you more detail around those functions So basically we want to show you things like scale like how many pots have been scaled up or down or to zero For that particular function and how many times the function has been has been uh invoked So basically just show you a graph of volume over time And uh, that will be along with the things that we're showing you in the flat list today such as um Where the function resides in your repo along with the url where you can hit that function And then the last thing that we're going to focus on is going to be moving auto DevOps domain away from cict settings So currently um the auto DevOps domain, uh can be Configure as part of uh of your project settings But let's say when you're using multiple clusters, it's not immediately obvious how we need to configure each one of those Domains, so what we're going to do is we're going to move it away from the project settings And we're going to move it into the cluster settings. So this way when you create a cluster You're going to have a simple way to right there specify the domain for that particular cluster And that is what the configure team is going to focus on Uh, josh, how about monitor? Thanks, daniel That's uh, some super exciting things there from the configure team on the monitoring side We're working on two major items here. The first one is our first iteration towards uh error tracking or making it easier for users to work with uh consume and in general utilize errors And so we'll be working to improve the general workflow Between git lab and also a sentry as a great another open source tool for managing errors and out and exceptions So what we're doing is we'll be having working to add a a list of the errors And you can now see within git lab and you can then open up Uh those within sentry for more information and then we'll also be working to try and build sort of the the framework for Viewing an error as far as the key information goes within git lab Which you can then use in the future release of git lab to embed within an issue or an incident or merge request and see some The really important details right there Within git lab and of course you can always dive into sentry from there For the full workflow experience, but this year again, hopefully make it easier folks to Kind of work both the tools and consume information Wherever they may be in their workflow. So we're really excited about that Next up is we've been working to switch our Monitoring dashboard over from our previous charting solution over to our new one And we're working on just finishing up a few of the last features It's a hit parity for our our justin dashboard and so we'll be doing that this release and then we'll be going ahead and I shouldn't that for all to consume with a more robust Charting library that'll improve the experience today and also make it easier for us to Add more features more quickly and more efficiently in the future And so we're also very excited about that as well from a velocity standpoint on the marinering team With that over to you Fabio for secure Thank you, Josh and In the security for 11.7. We are continuing to improve the security dashboard at the group level So the first item that I want to talk about is the filter Option for the group security dashboard This is very important because the more we add data into the security dashboard The more is important to be able to get information you really want to look at So we already shipped The security dashboard mvc, but we are going to add a new Option here at the top you can see in the mock up Dropdowns to filter by severity by project or report type Data that is actually shown in the list So the list will be filtered and you will focus on what is really important to you You can see an anticipation also of the The charts that will be shipped as part of 11.6. So this part is already there in the next release On the same topic We also want to show the penises scanning results in the group security dashboard at the moment We just show sas results But it's not the only thing that we want to show there So the penises scanning is the next step because it's very important and the information is very useful To security people and data will be shown Consistently in the same dashboard in the same list Where the sas data is currently put in and the drill down capacity and Filter options will allow you to select if you want to see both sas and the penises scanning That's a default option or if you want to filter for one or two kind in the future We're going to add more and more types of results into the dashboard We are also working on something very very interesting that is Part of the auto remediate feature and want to create a code diff to fix the penises scanning vulnerabilities It means that in the pipeline when the penises scanning is running The process will be able to find exactly which changes are needed to your code In order to fix a possible dependency scan the vulnerability for example Not dated the npm package that must be upgraded to the latest version This information is collected in the security report It is sent to the gilab system and the gilab system will show to the users The code diff that is needed For the fix so this information is shown in the model window for the vulnerability vulnerability itself And from here people can see what they need to change for this release Changes should be done manually by the user So users can look at the diff and can reproduce the same changes in a merge request in a commit to their code But this is just a step towards auto remediate in the future in the next iterations We want to automate this process and make these changes to automatically land in your code base We also want to work on another issue that is the inclusion of predefined templates in gilayaml file This will allow you to include any kind of template that is already defined and available in the gilab For example the autodevops template or any template that is listed in the dropdown For gilayaml files into your own Configuration this could be done in a similar way You can include any external file with a url or local file to your repo This is very important for the secure team in particular because we want to Provide a standard definition for the sas job for the desk job and all the other jobs that we are Supporting for our features so people can include them in their pipelines Instead of copying pasting examples from the gilab website That's very important because we want to keep these definitions up to date And so when you upgrade gilab to a new version The new templates will be automatically available and includes will target those templates as well So it's a very good way to keep your projects up to date With the security features definitions This issue specifically is part of devops verify, but the secure team is going to work on it and that's why it's mentioned here And that's all for a secure team in 11.7. So james back to you for giddily Thanks, fabio So continuing the work that we've been doing for the past few releases in the giddily team We're working on the de-duplication of git objects for fork networks So forks a really handy way of collaborating on a single Project, but everyone works instead of on a shared repository and pushes and pulls to the exact same repository forking allows you to Have copies of that repository and then merge between them And so when you've got all these copies of the same repository That means that you've got a lot of duplication of the underlying data because all the historical data for A repository will be the same So for a large project like git lab or other large open source projects. This means a git lab instance Quickly accumulates a lot of data that really doesn't need to be there And so what we've been working on is implementing the first iteration of this which uses git alternates to Share a common pool of data And if you saw the release notes for git 2.20 We've been working on some Other elements to do with this performance improvements. We won't specifically be using delta islands in this next release 11.7, but we're working towards that as part of This this broader project of de-duplication of forks. So specifically in 11.7 What we hope to ship is the first iteration of this That you will be able to turn on and use So there are some requirements around using hashed storage. So both The parent project and any forks need to have hashed storage enabled and if that condition is met and this feature is enabled Your fork will be very small So a 200 megabyte project each fork might only be a megabyte rather than having a thousand forks of 200 megabytes a piece So this is a very nice improvement for instance administrators who operate larger open source projects particularly um also Customers like git lab comm as well. So we're very excited about this improvement and you can read more about I guess delta islands in the the recent release notes from git um over to you josh Thanks james. That is super exciting Can't wait for that to arrive Let me go ahead and share My screen here and talk about what we're doing for package So we'll be working on two Uh dependency updates here two key ones. We started those in our last release and we'll be continuing to do them in this release I essentially uh jumped to an interim release Because the deltas between the previous version and the new versions were quite large And so we'll we're continuing that workflow here with runits as well as also with a registry Um, for example, uh, the last release was about a year ago and the commit release was quite large Registry for the new version. I can't wait to get it out there a bunch of our improvements But we're trying to do it in a Manageable way and so we'll be essentially went halfway last time and we're looking to continue the process here with this release as well um, while still working to Publish a new version of of git lab on the gke marketplace This is a really nice improvement in that we can utilize helm tiller Completely on the on the deploy side And so when you go through and you kind of do your click click click deploy in the gke marketplace under the hood Uh, the tiller will be utilized Which means that you can continue to use helm and helm tiller if you'd like in the future Kind of maintenance of that deployment to make it much easier to consume manage update And control in the future and so really excited to get that out there for folks Who are deploying git lab through the gke marketplace? Um, the final work we're doing here as well is also to continue our efforts on our operator The operator is essentially a control plane that can live around in kubernetes That can then essentially look after the deployed instance of git lab and then perform some actions and automation The first use case we're using for this is downtime free upgrades, um, which requires some automation beyond that which helm can support And so, uh, we've been working on this project for some time and now at the point of essentially it's merged It's not on but a fault and before we kind of Turn on for everyone About a fault who want to make sure we had tests and other basic practices to make sure that we don't You know break anything back in the future and make sure we had a good coverage there Um, so that's what we're working on here in this release as we prepare to turn it on for everyone And make it much easier to update and deploy a git lab without downtime Which is really exciting for our kubernetes influence of git lab So with that that's what we're doing on the distribution side. Uh, dmitry wants to talk a little bit about what you're working on here for the ux team Thank you so much. Yeah, I should let me share my screen See So Hello, i'm dmitry one of you guys is saying that desires on git lab I'm gonna introduce you to some improvements on the git lab design system Which is dedicated to supporting and improving philosophy on developing Any of the product areas of git lab itself and as the first improvements For a certain will be restructuring the design system information architecture To include branding and marketing. So This way branding and marketing will be included on the design system And you'll be able to find that information there to give a little bit of an example This is our design system currently and with a work in progress improvement. You'll be able to see. Hey, there's brand product information research information and marketing is in there as well Going on We have some improvements towards the chart documentation and pattern library those will particularly be beneficial towards the security and monitoring team as well as Configurability of configure team and some other product areas Will be improving the The common charts library This way any decisions will be set in stone or at least be defined And we can we can know how to support these these charging library We know how to define those charging libraries inside of the application The same goes for line charts important as well And how to define certain areas of the chart to zoom in and get your Get your details out of them To go on with the last part For level seven of the GitLab design system is our branding efforts So we're branding our design system to get more eyes on it to be more inclusive to see to get everybody to use it The more people that use the design system the higher the velocity will come As more decisions will be defined in there and part of that is branding it with the logo Here are some work in progress iterations on that as well as it will Follow through in setting a fed icon for the design system And eventually promoting the principles that designs have set with illustrations And with that is the end of this GitLab kickoff and I'm Quite excited for the next iteration. There's some amazing improvements there and I can't wait to break them To you We'll stop sharing for now Great. Thanks Demetri. So that's it for today's kickoff. Just a couple of notes December 22nd in less than two weeks. We'll be releasing GitLab 11.6 That'll be the last minor release Of GitLab and what we just talked about will be GitLab 11.7 So that'll be very first release in the new year So I'll see everybody back on January 8th at exactly this time and have a happy new year. I know