 Okay, so please anonymize my document by me. So why do I even want to anonymize documents? Well, I think that you're looking at a bug report that some user has made. So you're a developer or a QA person and you're sitting on the IRC channel and then you see that somebody submitted this report. So you go to see, you visit the bug page and they describe the bug. I have this document, I'm doing this thing, I'm doing that thing and this is what I get. They're even nice so they provide some build information, which is really great. But even though they see the problem, you might not see it. So you try to follow the instructions, you don't see anything wrong. Even if you're playing with the document a little, you know, putting effort into it, you still don't see it. So then you go back to the bug page and tell the user, please give me a reproducer, a document I could use to which exhibits the bug. Or link to me or link to that or attach it. And then they tell you, I would really love to do that, but I just cannot upload this document to a public venue. And this is the main motivation. And why can't they just upload their document? Maybe it's personal or it's a draft and publishing the draft will be a problem. Maybe it's the draft of the great American novel that they're writing or it's a business contract. Or maybe they're under obligation not to disclose. So maybe it's a personal promise or it was created when they're working for some company. It doesn't have to be big corp, it can be small corp. And actually, these are most of the documents that people are working on. Because when do we discover bugs? When we're working on a document, that's a draft. And most of the drafts are things that you don't want other people to see. So what are we going to do about that? The poor solution is have the user remove all the private information themselves, which is what they're supposed to do today. But this is actually difficult because it could make the bug disappear. And if they want to remove things selectively, that's hard and it's a lot of work. And if the document is large, then just forget about it, that's not going to happen. So bottom line with that is that the fraction of bugs that have documents attached to them, exhibiting the bug, is low. At least as far as I can tell. So what's the better solution? Do this automatically. So I have my personal letter here and I want some tool to replace this text. By the way, this does supposedly have a bug. Never mind that it's a false bug, but the lines seem to overlap. So I want to submit a bug and then instead I get some sort of replacement text instead. And hopefully the replacement text will have this similar features in some senses to the original, like in a number of words on the line or whatever. And then I can upload that instead. And this would be great. And it's actually not that difficult, at least not the basic form of anonymization because we don't actually need to run LibreOffice in order to do this. You can actually operate on the ODT file or whatever file format we support. We just traverse text nodes in the context, content.xml and also meta.xml for the metadata. And that means that you could have like a web form where you submit the document and some script run on the server and do that for you. And that means that you can offer this to people when they're reporting bugs on the bugzilla. Won't you also consider anonymizing your document and then uploading, like attaching it to your bug? And that would be like the real win. And of course there are lots of, and this can be made simpler but more complex and more featureful. So of course you want to scrub the metadata, that's not hard. But if you want support from multiple languages, that's another complication. Maybe multiple languages in the same paragraph which get different text. Do you want to replace it with some generated text or gibberish or like fixed text? And then you want to try to maybe maintain as many effects as you can due to the choice of specific characters. And enter character effects like connections in Arabic where the letters have connecting forms. And of course, and another challenge is replacing graphics. So you want to, maybe that has private information you don't want to disclose. So it could just be like a rectangle with nothing in it. Or maybe like blurring, that could also be considered. The even more significant challenge is if you want to do something with vector graphics. But let's just start, if somebody is willing to implement this for us, us meaning QA people, then let's just start with the simplest version that just takes care of the text and the meta fields like the document description. And then we can worry about these things later. So that would really help us, will help developers as well. I believe it should not be that difficult, so please do this for us. So I'm done, yeah, question.