 So I'm Ike, I mostly work on Kalk and a little bit in IED in end. I'm employed by Red Hat and Still give me some time to work on this even if not necessarily official, but well, I find some time Hello, I'm Kendi. I work for Kualbra and in the SCI I'm representing mostly the online stuff I'm Stefan, Red Hat, plumping Hello Quella McMurray Red Hat as well and I go with GNOME integration Yes, I guess I'm Michael Meeks and I represent the forces of darkness Now I don't know all sorts of different technical things Christian. Yeah, I'm Christian Loma or Glove and I'm doing the release builds a release engineering and that's why I have a seat on the ESC I can use this This way. Hello. I'm Miklo Schreiner. I'm very bad at updating minutes as you see Otherwise, usually I'm doing writer things and whatever else is necessary Yeah, I'm Olivier, I'm Olivier. I am with documentation Lionel, I work a bit on base in my free time I'm Gabriel Representative from one and one just for a month. I think member of this committee nice meeting Hi, I'm Michael W from the city of Munich and not really representing one specific topic Hey there, I'm Thorsten work for CIB Yeah, I'm pretty much on everything that nobody else wants to work on Oh Hi, so I'm Gisco. I work for TDF and I represent quality assurance in the ESC meeting I go from TDF and I'm here in the ESC for a user experience Cool, so we run a call every week. I guess and It's really very open for people to turn up or not turn up and plenty of people take that opportunity to not turn up So that's that's great. And we we write some minutes at the end and it's just really I guess the heartbeat or the rhythm of You know, some of the engineering that has to be done We just check in compare notes and see see where everyone's at and communicate. Hopefully So, yeah, everyone is welcome to join if they so wish Do you want to go through things? Nicholas, is there a Sure, so perhaps we can explain back. It's mystifying sections in minutes. You never understood So Regarding the Yeah, so we have this funding action item for Sophie to check with a little bit With the Libre Office conference heck best we could page a guest that's either done or no longer interesting Or we don't have Sophie or perhaps all three at the same time Yeah, and from last week there was this action item to backboard that Patch that was the result of the discussion of the regarding the is equal to be slash firebird migration That's his code. Did you do that? I can answer it has been done. Yeah, it's has been great. Thanks Then we would have the release engineering update So if you are walking away I know the status is that the six three two RC one is on the mirrors currently and the Tagging for the six three two RC two is going to happen the week after the conference so probably next Wednesday and The six to eight RC one is the week after that the last week of the timber basically and Yeah, and now we are also getting closer to the 64 alpha which is mid-October So, yeah, if you have experimental stuff then now is the time to get it integrated to have enough time to polish it for the alpha and Yeah, just not delay your major stuff What was the three point six star The is six three two RC two will be tagged next week May I remind you on the release plan 6.4 release plan is not online. It is online. It has been since two or three weeks Okay, sorry that case. Thanks Do you want to say anything about the remotes or the hundred year? Yeah, I know it I mentioned yesterday to me that the Android viewer master the arm-based one Crashes on every file operation. So something needs to be tracked down white crushes But yeah apart from that No updates And and I think with this new requirement to have us 64 bit version Yeah, and have a patch working to build a 64 bit version, but It'll be external the nss But need some tweaks and yeah, but it's in progress Okay, thanks. What else? I don't have any online updates Regarding the daily build Any news with the windows are the macOS thing Yeah, there has been some outage of the free desktop org and on shit service. So Some tinder boxes didn't build during the time frame As far as I know it's back online again. So it should be available for all platforms and Torsten also Restarted his own windows tinder box. So there's another source for for Windows daily builds again Yeah, when it finishes it does I Was there was an issue with the with the JDK Jre update It's not but it's now hitting or UI test problems Like every other build is failing now And I think the signing thing that's also a bit of an issue I don't know is that something you can fix or yeah, I fixed the signing thing basically I changed the signing operations to better allow for signing with a private key on a smart key on a crypto device So you don't have to enter the key so many times But I use the variable that is set to an explicit false and set to empty like other variables And this broke the packaging on the windows build, but this should also be fixed right now Let me check. So if it's succeeding then and no UI test is failing then it should actually upload something Are there other get mirrors that are hot of our Garrett mean is there a good lab or a get how we have an active replication to a get up to launch pad and To free desktop. Okay, so if free desktop starts to flake we can encourage people to migrate to something better potentially Yeah, yeah, or they can also just use a carrot. It's also Accessible without authentication And it's any other commands regarding real is engineering No, then we would have the documentation update Olivia you want to share us with something this week on the maintenance and Updates in terms of translation for the online version on master and six point three No more developments for the new help but not yet ready and On the guides, it's still starlet Google's season of dogs, but we have some work already started by the technical writer and With the outline of the documents so it's progressing on time Okay, thanks Any questions to Olivia If not, then we would do you the UX update haiku are you you are here? Great? Yeah, thanks. We have a long list of contributions contributors Numbers of tickets that require input from the music experience or design perspective And what I do every week is to copy all the new tickets just to draw a bit attention on what people are interested in I Pull the data out of the RSS feed. That means it's not only new tickets that are listed here, but also tickets from the past that get a flag needs your eggs and Since I'm very really badly prepared. I don't have a list for this week. So Yaks is on hold what I did here is to put in two examples from That's the latest two tickets that have abses flag and usually I highlight things that are interesting from my perspective that I think it's Something worse to consider in a larger scale if it's Crazy idea or if it concerns just Visual changes I don't highlight it. So sometimes it's nothing and I have no idea what has been Was what came in this week? Okay, thanks Any questions to haiku So then we would have the crash reporting section. This is done Yeah, I don't have anything particularly up to date this week. I launched the crash testing script, but I can't As I say chin to see if it's still in progress or if it has failed So the numbers are the same there as last week to import failures to export failures There are better numbers I have seen better numbers I've seen all zero, but I think it's fairly a random that happens Converting that 35 which is the same as it was last week So no great change there and the Google fuzzing hasn't found anything at all this week What's new this week then is that I've announced the two latest CVs 2019 285 4 and 285 5 Which gives us eight? CVs that we have had since we became a CV numbering authority They asked us initially how many CVs we need and seems we only ever used two previously I felt it was foolish to ask for two so I decided asked for ten and it looks like we're going to have to go back and ask for another block of ten With these most recent round to see these there are fixes for our previous efforts, which in themselves were fixes for previous efforts so additionally in the latest releases we've changed The behavior of the core issue which is these scripts that can be run without prompting the user that they're being run I've been changed to be covered by the same macro security Handling that we do for ordinary macros so any documents that hopefully documents that contain calls to these scripts Will themselves be flagged as containing macros even though the macros that they refer to are actually the pre-installed ones So they'll fall underneath the same macro security, which will hopefully give us an extra layer of protection there So even if we have failed yet again to block these issues then either way we should fall back to the next layer Which is that this document is an untrusted document and could do anything Are you sure you want to do that? So hopefully that will bring this saga to an end and that's right, but Okay, thanks Then we would have the crush reporting Section Cisco, is there anything unusual that you would like to point out or? Hello What I said last week in 6.3.0. We had many classes related to the Dynamic number of columns worked on by Noel. It seems it speaks now in 6.3.1 At least first week the numbers are much better. So we have 800 Crushes while in 6.3.0. We had the first week 2000 crushes, so we are getting better there Okay, thanks We still don't have Sophie, right? Okay So we normally have a section for hackfasts and events and trying to discuss how to distribute people and so that There is always at least one or two mentors on each event So I think Torsten knows about the Google Mentor Summit hackfest. Ah, you did You want to say anything about that? Maybe again the call for Or the invitation to participate at the potential Munich hackfest There's a helpful link in the minutes. If you click on that you find a registration table And if you will be at all interested in potentially coming there's no There's no it's no commitment from your side But so that we can gorge the interest and either hold the event or not Putting your name there that would be really quite great If you could do that by the end of the week that will be even better As well with a bit of lead time also we're pondering perhaps to do some hacks Hackfest ng which requires some mentoring On-site and it has a bit of a lead time there so I would think probably Everything below 10 people doesn't make that much sense that will be my Mike my aim To get 10 by the end of the week and it looks like for the moment. That's probably gonna be before the mentor summit That will be 17 18 of October I This date So did the summit is on the weekend Saturday Sunday, but it's the 19th and 20 exactly So 17 18 will then be Thursday Friday before that Which sounds ideal? Michael, can you can you make it? And now I've got a corporate event that unfortunately so me make large candy and rush count But it's possible. So neither before nor after Before I can't offer have personally something but others might be able to do after but I think there'll be a gap Okay, so no harm. I mean this was a bit of a crazy idea with the And also there's quite a lot of other open source people in Munich at that time So from if we get some critical mass there and perhaps making doing some outreach and inviting other people over To fill the ranks, but it doesn't make sense if there's just two to the bro for suckers there on site to do that, obviously Okay, thanks Then we would have the G socks section. There will be a separate g-sock panel so just Torsten or Cisco or I grow you want to give us an update on just in summary where we are Everything is done finished passes down So maybe then I'm just a reminder that all the mentors, please if you don't have a student here Demoing the project. Please send me some slides me or Heiko Before tomorrow when we have the G-sock panel Before tomorrow means today Let's say I need I need some hour to paste it all together. So if you would give me that Much time that would be appreciated and then of course if your student is not here It will be really quite helpful if you could then be on stage as a mentor and present the work otherwise I'm not sure I would do justice or Heiko to the project So so I realized that there might be conflicts So of course, obviously then we'll find another way Okay, thanks Okay, then we would have the mentoring or easy hack update. This is actually up-to-date stats, which also shows that this week we did not have a first-time contributor But then we go double clap on the new contributor next week Regarding commit access anybody want to propose somebody to have commit access or None this week. So normally the way we do this is that in case there is somebody who does already quite a nice work And there is little value in forcing somebody to be reviewed for each and every commit for master then He or she is proposed in this meeting and then we discuss if there is anybody objecting If that's not the case then we send some welcome mail to the developer and I Teach that actually this is this means that now it is necessary to take part of the review process and how about its reviews Developer certification, I guess that's still sleeping for four more weeks Christian any young kids are see I update and why I Didn't prepare any stats this week, but like more like if there is anything unusual that that is worse nothing No, I don't think so At least I'm not aware of any problems currently Okay for the third time. We don't have Sophie. So Yeah, I can say that your web light is in the in the public testing Now and I will have a talk on on Friday morning at 10 o'clock Basically giving a summary and being there to basically summarize the feedback we have gotten of so far but the short Summary is that we will likely switch to web late So there are no real blockers that were raised so far. So it seems to be the tool of choice going forward I Great thanks. There was this topic regarding the fire burn migration from it's equal to be from last week But that's apart from doing that but poor that you already did Cisco There is nothing to to be discussed further than right, okay So then we would have the QA update Regarding most pressing box this is meant to be up to the stats so we You did that. Okay, so we did the work twice But the good news is that it matches what I pasted so It's good to have things reviewed So we don't have a new most pressing buzz bug this week, which is great Regarding other most pressing blocks the first one is Has no name much to that. That's some macOS problem Now what why is this tag this macOS is it normally what we try to do is Review these bugs and find out who might be a good person to look at that Given that it's probably serious enough that it needs some action And The the problem we have here is that now in six dot three we use Xcode 10 to build for Mac and It's giving Problems with a high resolution screens If you scroll up, you'll see that there are many duplicates already so many users are Having this problem Then we did a build with Xcode 9 and the problem is not happening, but we cannot use that We cannot build with Xcode 9 because Christian well Because of the notarization feature the additional Signing stuff that Apple introduced and enforces for the new versions of Mac OS those require Mac OS 10 So while it might be possible to build the software with Xcode 9 and then do the signing with Xcode 10 I don't want to Mitch and mix and match those tool chains So you rather just want to build with Xcode 10, but unfortunately, I don't have a High DPI Monitor or Mac, so I cannot really reproduce the problem myself Christian the hardware budget for TDF for the year is under spent by several thousand My suggestion would be that the board would Approve something here, and it will be worth you having high DPI screen I don't know if other board members. Yeah, I think we're probably quarreling here to have a board meeting and by you But even having a high DPI screen wouldn't mean that I automatically knew where the problem lies So that's very true, but you want to change But at least I would have a You at least have a high DPI screen which would be nice So just to add in 6.2. We had a similar problem Initially we build it with Xcode 10 and then There were many crosses and hangs in on Mac so then we decided to use Xcode 9 and then Tomas Fixed those crosses and hangs, but we still have this remaining issue with high definition screens Is there any hope that the new toy chain will be a hard requirement for the app store version and then Whoever does the app store version of Mac would be interested in looking into that That's us so yes It just depends how much time and quick he has and his managers can give him a way to fix Stuff like that, which at the moment is Not a lot, but okay So in just to sum it up if somebody has the needed hardware here and could look into that That would be great otherwise and or later We should get cloth to have a high res high DPI screen just to smoke test the builds because Increasingly it behaves differently I would thought Give him an action item Well, I don't think I need to test it personally. There's people from QA team already So it's very fine this issue So release engineering would traditionally smoke test something by running it to see if it runs Usually a good idea Okay And there are two more bucks here, but at least there are some ideas who could look into those and those are not new Then we have the new high severity box of the week. So For these If it's something that's in your area, then it is worth having a look For next week, I would try to do this usual annotation where We can have some idea who would it would be a domain of what developer Yeah, normally we don't have such a long list of high severity box I normally double check that they are indeed high severity box, but I didn't do it this week So someone it's increasing the severity and yeah It needs a double check Yeah, if you could double check if those are really high severity that would be very much welcome Okay, so that's about the high severity box and Other than that, we only have the QA starts so Is there any Top-level topic that we would like to discuss here or is there any question from the audience that You hope that there is an answer from the ESC or any comments on the work we do or I Thought Ashley's question earlier was good in but much too generic. I wanted to know what specific architectural problem So That's a very good question. I should elaborate So the things that I have in mind Take a concrete example would be something like configuration infrastructure for example, I think it's it's quite old right and it underwent some changes and some It took some some punching around I guess, you know during the years and I don't think there is an expert that I know of who can Explain it maybe you can okay And I'm not sure this is one this is just one example and I can I can think of others I think conflict is quite good. I mean compared but what do you think? More than ten years ago, maybe But Maybe it's just one part of the overall experience of what you mean with configuration. So So my experience the bug generic maybe For example for me the pain point is if I need to debug something there. There's a there's a whole Mechanism that kicks in whenever you change something that triggers a Config update that the config update then would enable a feature or disable a feature Right and based on that you would have UI items show up like sidebar something an area that I worked on quite a bit On and it's almost it's so opaque that it's almost impossible to trace it to figure out essentially what changed here At runtime maybe depending on the config files that ultimately triggered the enablement of a feature or a UI element Or something or another and this whole thing happens Because of you have the notification Subscriptions and listeners that would update the config that then would trigger Other listeners right including UI etc. Etc. So it did the whole thing is almost impossible to Okay, so that's the layers atop the configuration. That's what I meant That's just one piece of the of the puzzle probably. Yeah, that might be true I'm not I'm not right and similarly familiar with the layers above So you say luckily almost nothing changes in reaction to configuration keys changing, right? There's also some manual horrors to make these things actually have effect, right? So it's easy to imagine a world where you know You set the gconf key and then everyone reacts to the event listeners and makes them happen, but it's Typically not the case in the game Yeah, she's kind of sucky because when you want the demo you want to do your LDAP Okay, so that doesn't have notification either, but you want to do some kind of Change and then see all the machines go power power power down the road as they as they change because Demes are everything right and that's not where we're at another good example would be the some of the the way we Managed the UI and how how we define them all these I mean some of it it seems to me that a review I'm not I'm not saying they're necessarily so bad that they need to rewrite that that wasn't my point my point is How do we manage the problem of revisiting major features and components just to at least have a you know a core developers? Opinion and input as to whether there needs to be some You know major investment in improving that infrastructure Because going forward it's becoming very costly to maintain or add features or you know We're spending a lot of time tracing Bugs because it's just so opaque. No documentation. No code comments or minimal or misleading because they've changed You know what I mean my my point that I'm trying to raise is how do we deal with the problem of having a very very large piece of software That the expertise in the different areas is is constantly changing for lack of a better word People leave people come on board and and then they do touch parts and pieces not necessarily understanding it fully, right? Which is fine. We encourage that right But ultimately how do we deal with this problem so that going forward we make sure that the project stays in a healthy Form so that when somebody new comes on board is able or she's able to pick up And and go on without spending months on end trying to wrap your head around something To make a change. Hmm. That's not hackish. That's not horribly So so my personal response will be that I guess we're doing quite well on on that end and I don't think there's any Any money we can throw on anything that would improve things substantially in that regard So I mean we all know there's areas that are lacking Maintenance and then if somebody comes up with a problem there we collectively try to to to find out a way To do better there or to to to invest energy into into certain areas And sometimes it works to to make a tender out of it other times somebody Heroically finds the right idea to to to get things working again And then there's other areas that are more more maintained So I think on a large-scale overall we're doing quite well and we're quite healthy I would say so there's nothing That we would fear If something broke in some area that we would be totally helpless I mean we demonstrate that all the time that we together are able to to address the the issues that come in even for areas that are under under maintained for years and Documentation of course, but nobody writes documentation and and even if you throw money at that issue I'm not convinced that what would come out of that would be any any any better What we have so if there's an issue It's always I I think it's always this reactive thing if there is an issue cropping up somewhere You start to ask around on on IRC on mailing lists in the ESC What other people's opinion is there and if that's an area where we are lacking then we try to to build our Our ideas around that and and see how to fix the the issues that do come up and not Ours has proactively tried to to lay the grounds for for future contributors That would make it easier for them because that's always a thing that doesn't happen in reality anyway So you you do what you think needs to be done, but you don't cater for this hypothetical newcomer that you want to a handhold in some area and Just to add to that like we have done this several times in the past which was a mistake Yes, so take the config manager case it was a second system Rewrite that we started with and it was designed heavily designed and it was even documented and All the modern technologies 80 locks for every single key you took are not thread safe You know and so just simplifying it and focusing on the real need can really help The other thing is that I think Chris Sherlock was trying to write some overview documentation And I think that would be extremely helpful So I think if you look at the Linux kernel There are people who write books that explain give you a map And we have some pictures of the code that are helpful, but I think having having the basic concepts written in a book Might help if people still read books So just the the classical answer to that question I think is and it's It's a cop-out, but it's still An answer if you're dealing with like huge monster of legacy code that nobody understands or that you find hard to to find your way around with is start writing tests and Whenever you discover an invariant or something that you reverse engineered You reverse engineer to be maybe a design decision in the beginning and you can write a test for it and the good thing about tests is they are documentation and They are documentation that is that drops harder than other documentation because it breaks if if you break the software that makes Breaks the invariant that you have so And that that can be incrementally helping to get a grip of what this old thing is Once you want to kill it off and replace it with something saying again So I was One particular answer so there was there was this I'm not saying you said that but there's this idea and then Michael You were also referring to that That if you don't understand something you need to rewrite that the problem that comes from there Is that unless you have not completely understood the problem very likely the the solution We're coming up with is either the same same like like Baddest of the earlier one or worse because you you have to learn what the other person implemented that already knew so I'm not at all enthusiastic and Entering that sort of Turn especially because we are kind of under resourced anyway So we're much usually much better things to do than preparing for the hypothetical If you look at some of the larger refactoring's we've done inside modules They've been I think immensely successful when quailons UI rework Takes us to just such a an immensely better place for Well, let's not say it's a small part, but I mean, you know, what would you lay out and rearing? There's still plenty of problems in VCL that need fixing You know from rendering and drawing and things but taking one part and made it Nominally clean and much more beautiful, right and and all the calc core being rewritten or you know life cycle problems being fixed It does happen. It's not I do agree if you give me a minute to Add a point. That's very important. I think I completely agree with every everybody's Feedback, so let me clarify. I'm not suggesting that we rewrite parts that are hard to understand That's absolutely not what I suggested what I am asking is when do we know if a part of the system has become a So messed up because of multiple hacks that maintaining it making a change is Extremely costly meaning you need to make a change and the only sensible way to do it either is to change a lot or you break features or behaviors and change those test cases to adapt to the new behavior and Potentially make a lot of users very very angry at you So how I mean this is this is a little bit of a hypothetical problem the examples that I've given Are examples that I felt that there were too many hacks that changing it is almost impossible You can't go around and change it because if the system is fragile enough, right? There is no clear way of doing it and the next thing you have to do is invest a lot of time How do we manage that? That's that's it's an open-ended question. I apologize, but otherwise I agree with you I wouldn't go around rewriting things Any other comments to that are probably we discussed that to dust now So I think actually we're relatively blessed with people who are experts in areas and also an understanding of who is expert in what area Say that broadly we have reasonably minimal conflict and it's reasonably obvious what needs fixing and usually how The only thing that's not obvious is how to fund any of it So, you know Any other questions? Yeah, Mike. Hello My question is about the situation the question from Possible contributors that I face very often when they want to do a one-time Contribution for instance fix some built-on some system some Operating system for instance and they Struggle well, it's not too complicated, but usually they complain that they are busy people already having account at github or something. Why do they need another account and They usually not only need one another account, but multiple one for filing a bug at bugzilla one to have some access to Garrett and so on My question is not that we should for instance move to github or something, but could we have at least At first stage a single Sion sign on to bugzilla and Garrett and further possibly to have some way to create a change from bugzilla to Make it easier for such contributors to create bug and then from then create a Change possibly using use web UI. It's often desired from my experience when talking to such Developers So that make it easier for them. So for the first question I hear there might be news on Friday on that So for your first question or the first suggestion I hear there might be news on Friday on on that topic as a single sign on is something that Clough and Guilham has been working on for a very long time and as Torsten says it's not yet finished. I Think there is hope that in the in the near future That will come to an end and actually you can really just register on the user document foundation or again And use that account everyone Yeah, and even without having it in place already you could already use your GitHub account to look into Garrett you could also use your Google account to look into Garrett So you wouldn't have to create a completely new account For this and if you want to not use kit check out you can create a change in Garrett web interface and don't have to Yeah, use command line if you don't want to so for simple fixes or one line fixes you can create the Change request from within Garrett web UI But yeah, as Torsten already mentioned the single sign-on system is spread out of a multiple Services on the TDF infrastructure already and Garrett will be the next one to be switched over and then hopefully the problem Thanks with gun Okay, so the other thing is While it is possible to create Garrett changes just using the web UI it's In my view it's a relatively rare that that's the bottleneck usually. It's not the building time or the Or the code writing time does the bottleneck is you spend most of your time reading existing code And then it does not really matter if you at the very end you create a change locally or It's actually helpful to create the exchange locally so that you can quickly test it get some good Iteration once you are happy with that you can upload that for review And regarding the bugzilla integration. I also don't see how you would start the fix With your bugzilla ticket. So our integration works the other way around when you create a Change request and reference the bug ID then there will be an automatic notification in the bugzilla ticket that the Commit has been committed But if you have a concrete suggestion how we could integrate it the other way around. Yeah, please talk to a calamity or drop an email I suspect as well that I guess that so I Think right now the the problem we rather have is on the review mentoring and and that's side so Opening some fire hose for people posting random Changes to quote from I don't know if that is really the largest issue we have right now I agree that the sign-on thing is irritating and I wonder if I mean perhaps getting getting signed up then this is really just If people are already on bugzilla, then there's nothing else they need to do double my hope when this Everything works Because right now even if you have a github account, it's still a bit of clicking in the just ring and form filling Yeah, just One another thing of course like if you run into things that are like easier to done Then like it takes time to actually contribute that Please do file easy hacks for that because like I don't think that we have some good flow of easy hacks Now so if you find things like this that can scale to to other people That would be extremely appreciated Yeah, and just a closing remark. I think the most troubling Stuff for newcomers these days if they join the project is getting used to carrot at all because they have never seen it before No, the other stuff is probably easier But so I think it's a valid point and we should revisit this regularly to just see If if we move too far from the mainstream so that people are already scared off by our tooling But I don't think there's there's an easy solution to that now So I think back in the day we discouraged people from sending emails with patches in them for some reason But but a long time ago an email with a patch in it was the preferred means and if anyone is scared or worried or put off Then just email the patch to Mike who will help review it and get it into Garrett for you That's what I suggest in terms of Habitual workflows or muscle memory that people will be used to get hub So with both GNOME and KDE having migrated to GitLab There were probably a nice source of data on that whether that changes contribution on their side my I don't know so Personally, I guess that's probably if you want to master liberal first and Garrett is your smallest problem But I might be wrong. I think we're coming up to the end of our slot on the beginning of Stefan's which shirt You know, I know no janitor of sanity, you know, so Last last things got a few more minutes of two five three Anything else otherwise you can do QA QA gets relegated to the end tragically Cisco is there any any well, I just want to comment on that Be your mention Garrett, I think Now especially John and newcomers are not used to IRC And that might be another thing that they are they might be afraid of I think in that way at least in communication way Nowadays other tools are used. I'm not I don't have personally. I don't have anything against IRC But that's something that maybe we should consider as well So you can embed it in the web page so as you're filing a bug that's the bugzilla, you know QA channel You know, you know IRC widget on bugzilla, so you can ask Cisco which button do I press, you know, yeah, could be Hmm, I don't say anything about the QA bug stats. Let's go. I don't know if make if they're up to date That's the only section which is not up today. So at least depending on that, please don't draw a different conclusion compared to last Well, I updated it Okay, so do you is there anything unusual that that is worth nothing? Well to be honest, I Normally don't look at this stats, I think they are just more for information Probably the primary At least one purpose of this is that there are these top lists And if you are one of those top bug fixers, then finally somebody writes your name there and you can be happy Fine, so I guess that would be the end of this ESC slot unless anybody wants to talk about anything else have it Yes, see No, nobody Now this is done. Thanks for your attention