 So hello, my name is Emilio Ocala, and I'm here to talk to you again about the right to left status in the legal office. We'll go over a few of the basics. We'll start with the brief, what's the right to left, and a bit of the history, because that part of the presentation doesn't change from year to year. I'll talk about it quite quickly. So, one of the first writing systems in the world was Qunfo script, which you actually used the chisel on some clay or something like that. So it was used to be written from right to left, because usually you would put the hammer in your right hand and the chisel on your left hand and do something like this. And this is obviously easier for right to left and not the other way around. Later on in history, we moved to Phoenician, which is something that was around 1000 B.C. And they invented the English called Amchal script. It's actually a translation of the language in Hebrew, which the idea is that they actually use letters instead of little pictures. That's what, for example, is the difference between Egyptian hieroglyphs and modern writing systems. Later on, another development was Aramaic and they moved to Square Street, something we're mostly known today, that can be a rough estimate compared to C.G.K. or some Indian languages. When you see little, it's still letters or, of course, it's not a picture, but it's... I'm not sure I discovered correctly, nobody can correct me. But the idea is that the work is used from some signs, which are not necessary letters. So every sign has a meaning. The movement in Aramaic was to Square ones, which roughly are fixed space, which is today obviously is important for computers. Both Hebrew and Arabic are a continuation of this note, and Hebrew especially is relatively similar to Aramaic. In later period, Greek and Latin switched from right to left to left to right. In one of them, that's a modern example, which happened around a thousand years ago, was Hungarian, which you can see the communication on the Hungarian which are right to left, versus the new Hungarian which is left to right. For the last classrooms, we actually fixed the lab in the Hebrew office to support all Hungarian. I have no idea what it is, but why not? Just a reminder about languages, because these whole few talks are both languages. There's a difference between language that can be written, spoken language, can be written in a few scripts, and a script that can be used for some languages. A few examples are mainly in the area of, let's say, India and Central Asia. And also scripts, Latin, Arabic and Spanish are being used to do the language. A few examples of Hebrew office interfaces and interfaces you all know, simply the English interface. Then you can see the interface in English, but in right to left mode. You can see the menus are on the other side. All the properties and windows are moved to the left side. The new file menu or icon is over here. So everything is changed. A misconception is that everything is mirror. It's true for a lot of objects, but not for all of them. Some of the logic used to be amended, so it's not just replaced mix and while. We're going for basic idea, but it's not always good. A semi-interference test of Hebrew translation is more significant. Same for far, same for other languages. If you haven't seen previous lectures on the way to Hebrew right to left support, it's not a very important thing. This icon I hope you can see. And then after restarting the Hebrew office, you get a few more options. These options, for example, are the way to set parameters of originality. Both in the other menu, both properties in the menu. And that's quite important. You can recognize the language, so if I start typing Hebrew in a lot of cases, it will be the directionality will change to right to left. A line that will be from to the right, which is good, but sometimes you need to hit them with the buttons. Sometimes you want to force them to be in the other direction, so the buttons are very useful. The slide is based on what I showed last year and obviously the fixes. So, impressed in general is our weakest point regarding right to left support. We'll see it in a few minutes. It's very nice that since the last two prints do talk about spot fix. Okay, one, two, three, we're numbering. You have one, two, nine, then ten, then eleven, the same two digits, and then you have twenty-one, thirty-two, because it switched with thirty-two, thirty-one, fourteen to forty-one. It was completely weird. That got fixed in version six, zero. And then another problem with weak characters was fixed for six to one, and that helps me in some of the text more really. We still have two issues. The first one, I'll show you an example of it. And, well, I'll show you the other one. Okay, so this is the first one. This is the animation one, and this is about being important. Look. So this is about being important in 2012. This is the English option. No problem. This goes well. This is the animation in Hebrew. So some of the text appears like the logic is weird, but you do it again. But the only thing is the second animation effect. Okay? So it's a circle in the left-to-right line text, and then you see the effect in the middle of the text, and then something is ended and subscribed. No idea what was going on. So in last year, the effect was a little different. The text entered into the screen, and while the direction of the entrance was okay, the letters appeared in the wrong way. So the last letter of the sentence actually appeared first, then the one before last. And when the animation ended, the order of the whole sentence changed, and it could be readable. So each animation has its own problem. So you have to go to the base logic, and think that you get presentation from whoever, whatever. If the animation looks like this, you think the software is crap, and don't try to do anything else. And this one, brand-new this bug, is one of the top bugs for Impress, because it gives the user a very bad impression, and makes it not consider this the problem very more. And it's not only a technical problem, it's a very good marketing issue in the other side. With this, I can't go to any company, or any marketing person, and send them what you should consider Impress for. He grew out of it all. The second bug is an example of text. A sentence starts with weak characters, sorry, with strong left to right characters. In this case, the beginning of the sentence is the word English, a hyphen and then text in right class. This is Hebrew, this is Arabic. The problem is, at least for Hebrew, it's okay, but for Arabic, you see the text is not really good at all. What I do is I just remove them, so you can see it. If I remove the English word in the beginning, suddenly the Arabic text looks okay. So this is an important issue, it's not critical, because usually the sentence will begin with text in Arabic, not in English, but in the few cases it would. It's not completely confirmable, you can compare the letters, and especially in this word, which is completely unreal, it will be safe. Okay, so this was the bottom one. The bottom one, newly reported, so I think it was reported in March, but the guy who reported it, you could reduce it back to version 3.5. So it's not a new one, but just reported recently. And that's a powerful power, whether you're right to left T, or you're right to left the speakers, in the whatever community, that we didn't report in time. On the other hand, even back six years ago, our students fixed, but we're very happy for what was to be fixed. You can see what there is in general, there are some issues, but none of them is critical. The students fixed in the previous version, one in 6.0, one in 6.1, which makes us again very happy. With writer, we actually offer our tool for users to use it. With confidence that it would be okay, it's not perfect, but the support is good enough. For day-to-day use, there's nothing critical about it. This is from the previous version, that's far, so this was like a recent perspective about the amount of what was to be fixed. I have a bigger picture, you can see. For the version 6.0, which was one after previous versions, we had a lot of, relatively a lot of what was fixed. Usually we see this increase because we talk to the developers, because of talks like this, that highlight some of the major problems. And again, it's important to say thank you very much. And then usually after a pattern, we have a high blood scale fixed, and then the next version has a lower amount of what's been fixed, which is okay, but we're trying to get more and more what's fixed. The benefit is that with time passes, the criticalness of the box gets lower and lower. So ready for 6.1 to be not too much fixed, and we'll see about the next version. Everything you see here is on the weak page, on the document foundation, on the wiki system. So we document all the bugs and the risk of what was fixed where, and also what got backward and what occurred. So if the bugs made fixed on master, we'd like to make sure it's being backward. Then usually that's my job, first to track things, second to make sure changes go to other branches. Just to give you a general idea about categories, we have a few. Some of them are language support, they happen in one language, sorry, not the left language, happens only with languages, they don't have languages. Then we have directionality problems, which are regardless of which language, in English, but in right-right directionality, it has a problem. Then we have interface problems. Sometimes things happen only when used right-right interface, but most of the language or document that we use, that might be stuff that the positioning is wrong, the logic behind the graphics is wrong, maybe an item which is not shown or hidden by some algorithm, and sometimes things which are more complicated like logic, which are different, depends on the right-right directionality. And things which are a mix of the above, like the stuff that we used to have about like this, that if you save a document in right-to-left interface, some of the defaults in the files get changed to right-to-left, and then when you open it in another interface, some of the same things which are really weird. And you don't want the document contents to be irrelevant, to get changed depending on the interface. It might be a little effort different, that's it. There are some bugs which are specific for just one languages, as I show you sometimes Arabic has many problems with people, because it's a little bit more complicated with the content of the words and the fonts. That's why we separate things in different methods actually in the next. Where do you focus for the next versions? Because our top bugs get reduced, which we're very happy for. Some of the next issues we think we've seen with our users is mostly interoperability with Microsoft Office in Microsoft. We just entered into a lot of RTF and some other filters, fixes for us, which was great. We still see some issues, but there are problems with those lower and lower, but as we saw, you can see in this one, there used to be a problem when we put regular text in comments, it's got reverse on export. When you see it in our Microsoft Office, it can be read or it can be read from the other direction. And that's a big problem, it's not much fixed for 6.1. We still see some of these issues, but again, the direction we see is very good and we're happy for it. Another thing we need to focus is in press and roll by a fact of it, because they have a different editing engine, on the rendering engine than CalCAN writer, and then sometimes we have to fix bugs twice, one for each engine. And that's another issue which probably you need to do with other aspects, but for us, we need to double check each time the logic and everything else, and then obviously sometimes you fix, it's like a black and blue one side, it can be called the other side, and we saw a few of these as well, but in press and roll, as we saw earlier, it's not well supported enough for write-all languages. In another issue, that's probably where I need communities back up and help, is how to create a better infrastructure to test write-all out of the box. So when the developers run his, when the developer commits and runs his test, there will be more stuff that run automatically and give the feedback and then we'll save from being reported, or noticing problems when the feature is complete, and there's something we'll probably find out how to do it. Our side, the community side, at least what we did last year was to create a telegram channel for both right and left end Hebrew, Arabic, I'm sorry we didn't have enough users, someday we'll turn to the right and left group, but didn't have the long. First, we really had to be trash and bugging bugs, reporting bugs, checking them, getting some Windows user into the group and checking them on Windows as well and the interoperability bugs with Microsoft. It was really helpful. We saw the traffic in these areas and for example, in one presentation I had to work three days checking it up checking the status, updating things and this was just a basic work for the presentation. This year I sent a virtual meetup on the group, we decided to date the time. We had about 8 people at hand we covered about 50 bucks in something like 3 hours and a few reputations in the next few days and I got the status it's easy to report it's easy to report I also got feedback from others what do they think are the problems or where should the focus be or if there is something that really prevents them from working and they would like to meet and present to you and it's the differences of the community when you have teamwork even if it's a small team it's still completely different that I go separated by balance of work and things much faster than we saw it in the last year also we had a user who did a lot of work on QA connecting us to other parts of the community especially in the QA who helped with metabots of other parts of the UI or engine that helped to get a lot of focus on developers by triaging all the bugs this week what we started following what I told you about the meeting with trial the bugs yet again is trying to assess what would be the 5 top bugs we needed for the community to fix as you saw we have in the last year we started to get fixed so we need to fill the place with what's our top priority and with classic clearance from if I tell them I have 3 bugs I don't need to fix it, it's easier please I'll fix these 3 bugs it's easier than to tell them well I have a weak pain about another bug please take care of it okay, focusing people really helps for the result of this process we'll go back to the weak patch and the plan is to repeat this process every 6 months after the release so we can reassess things at the moment it happens in good times every year in bad times every 2 years we want to increase the frequency I come back after last session we added one for people and one for our work and it really helps to triage bugs and then send them to let's say a sub community or sub group and said okay this is just a human related okay the other guys double check that or I find about in Hebrew well I have an example from Hebrew I go to the other community as well if yes it's a right to left or CPL bug if not it's just a Hebrew bug and we saw it to be a very efficient process when you need just one person from each sub community and then you know if it's a general issue or a very specific issue some of the other things we want to get done and this is a bigger effort and if you have leads or ideas where to start and we have to get the ideas is first raising the workers and this is one of the goals of this talk the other one is we are taking test scenarios for automation as I thought earlier one of the things we need in the session I mentioned earlier is to test documents to bugs which we are linking to these documents and keep doing it so even people without the ability to read people or Arabic could have a test document picture of how it looks or the problem even with our red circle or something like that and then when they fix or want to check if the regression or sometimes another fix made in effect they can do it easily even without reading the language or at least noticing there is a change and then flagging that for us and then we double check in our channel and that is the point that is also relevant for some of the stuff I saw in other sessions mainly by is that how we bring more people from leading speakers with languages which are right to left and have millions of speakers but they are not part of our community there is 350 million Arabic speakers 160 million ULTU speakers 120 ULTU speakers and 1 million Foxy speakers and I think there is no presentation for them in our community and that is more than 500 million people which we could reach or have possible as users and of course from the viewers and it is very hard sometimes it is technological sometimes it is more of cultural gap or to reach people how do you work with them if you have any class maybe some of the Asian guys that might need these people in conferences please refer them to the slides to the telegram group to the weekly everything that will make them be a little bit involved will be very helpful and of course email to me I will be happy to be the bridge I guess that in some places there might be politics and the fact that I am from Israel there might be a problem then I am happy to provide other people providing solutions I think there is a lot is much more important for me to be the connection and if some likes to bring a better result I am proud of that the goal is to get better in the office whoever the person is we still try to find ways to train local developers specifically I am talking about Israel but it could be in other countries there might be opportunities from local companies if you like vacation in summer it probably will be waiting to pay for your travels and it will change your workshop about how to have a new office might be an interesting deal yes obviously there is commercial options from our sponsors but we also have independent developers that would be a very different experience especially with communities who might prefer to work with an individual and not a commercial company and there is always weird nights and whatever in that side of the world just say hi we try to cooperate with the city exchange speakers both with meetings then I have at least a few representatives from the groups because some issues might be crossing between languages and problem known languages groups so that would be complex type of languages