 Okay, so I'm going to be talking about a little bit about how open source is being used in South Asia, specifically India, though most of what I say can be also applied to, you know, the other countries in the Indian subcontinent, but most of my experiences with India. And I'm also going to be talking about Debian IN, which is a project that we're trying to get off the ground to increase the visibility of Debian specifically in India. So I think it has a lot of advantages for the people of India. So you know, why this is important is because if you saw, you know, Christian's map yesterday, you know, he's showing all the red areas where, you know, Debian has been localized and there's like a big white patch, the biggest one was over South Asia. And that's that's a concern because, you know, India has a population of like a billion. If you if you add all the other countries, some of them also have a very high population growth. There's, you know, almost a quarter of the human race is in that area. And, you know, in the past, it's had a lot of problems with poverty and so on. But things are looking better everywhere, really. You know, people are at least, you know, starving to death anymore. And increasingly, they're beginning to have a little extra income. They're beginning to think about more than just survival and, you know, actually making some progress and moving ahead in the world. And, you know, a big part of the story of modern civilization has been computers and computing. And this is something that's only going to become more and more important as the 21st century goes on. So South Asia also needs to be able to benefit from this. I can use PageUp and PageDown as one of the computers. Yeah, I don't want to seem to go past this slide. You can prepare more? I hope I do. And the next slide. Excuse me? I'm from the down there, it's the introduction title over me. Does it take a slide from there? Yeah, it should. Yeah, I'm sorry about that. All right. OK. It's going to suck if I have to keep doing that every time. All right, well, there's any number of people who are willing to help you use computers. I mean, Microsoft actually have been really investing a lot of effort into trying to, you know, get India and the countries around it to start using their software. But it's not a good fit all the time because proprietary software is inflexible. For the longest time, you know, Microsoft had as a checklist item support for Indian languages. It was horrible. It was not really working the way it was supposed to do. Plus, they didn't localize very much. I think it was basically just office and even then not properly. But, you know, if you're relying on proprietary software, you're stuck because you know, you have to wait until they decide to fix things. So there isn't that flexibility there. And as I'll be explaining later on, there's a lot of different languages in India. But typically when the commercial software vendors do bother localizing anything, they only pick like the top, you know, two or three. And that means you're out of luck if you speak a minority language. Furthermore, proprietary software companies are unaccountable. If they decide to discontinue their product lines or, you know, switch to something else, you don't have any recourse. You either have to take it or do without. And again, you know, you don't necessarily want to be, you know, at the mercy of the agenda of, you know, some other country miles and miles away if you have your own particular needs. Proprietary software is unstandardized. I mean, well, OK, so they do follow some basic standards. But many times, a proprietary software company will extend and embrace or, you know, just tinker with things or come up with their own pseudo standard, you know, just to get like some kind of market advantage. And that's not very good when you're trying to, you know, grow the whole market in the first place. Having commonly supported standards is vital to getting, you know, a free market going. And lastly, and I put this last for a reason, free, proprietary software costs a lot. And in a country where, you know, there are a lot of poor people, that places a huge burden, not as big a burden as you might think, because like in many other places in the Third World, there's rampant piracy going on. But if you want to be legal, you have to pay a lot of money. And actually, it's funny that now that the, you know, the big proprietary companies are putting more into, you know, like enforcing patents and copyrights and things, they're actually making free software look even more attractive because people know that they can't necessarily get away with just making, you know, bootleg copies anymore. But free software can solve all these problems. Let me try. See if I can go to the next slide. Yay! So the goals of Debian IN is to provide a free base for technical and social progress. Because they too often go hand in hand. Not everyone cares equally about either. Some people are more interested in the social side. Some people are more interested in the technical side. But still, you know, both areas can benefit from having a free foundation. That's, you know, reliable and usable for, you know, the things that they want to do, not what somebody thinks that they should do. We also want to advocate the use of Debian Moulinix in the Indian subcontinent. What I'm finding is I explore this further myself. By the way, although I'm of Indian origin and I'm heavily involved in, you know, Indian culture, I actually live in the United States. So I'm kind of at an, you know, arm's distance from what's actually going on. But I try and keep up with these things. But, you know, we're finding that Debian is still a very unknown quantity. Because the thing is that with, you know, the communications not being that well, most of the information that people have about Linus comes from marketing. And so if you're a company that does marketing, like Red Hat does, or like Son does, or IBM does, then you can get your particular message out there. Debian, you know, historically hasn't really cared much about marketing even. You know, it's not even a question of whether we're good at it or not. We just, you know, our attitude has always been, you know, we'll let the distribution speak for itself. You know, which is fine when, you know, people know, you know, more about the Linux world than the pre-software world. But when they're completely new, it just, you know, it's not even an option that they think about. So, you know, we just want to, you know, maybe, you know, speak at Lugs, or, you know, get like articles in magazines, or things like that. You know, just get the name of Debian out there. We're not gonna, you know, do any kind of very glitzy, you know, you know, endorsed by supermodels, kind of TV advertisement. But just, you know, just get the word out there, that there is this thing for Debian. And if you try it, you may find that, you know, you really like it. And also, we want to get the pre-software hackers that are in India, you know, to get more involved in the wider pre-software movement. Because that's another thing I'm gonna be talking about later, is, you know, the problems of like bandwidth, and you know, just geographic isolation that people have. There's, you know, I don't know if you remember seeing it, but like a couple of weeks ago, there was a message on, I think it was Debian Project, by somebody from the eastern state of Orissa. It's like on the east coast just below Calcutta. But, you know, he was saying that, you know, I'm a poor student, but I'm interested in Linux, but can you send me some CDs or something? So I actually wrote back to him privately and said that, well, do you know that in the capital of your state, there's a Linux users group? And you can go there and they can help you, you know, burn CDs. It would be more convenient, I think, than having foreign people send them to you. And also they can help you, you know, install and get started. And he replied that, well, one, no, I didn't know that there was such a thing. And two, even if there was, that's still a good like 250 miles away from me. So that's not somewhere I can necessarily go, you know, like every weekend or something. But so, you know, this is a problem. If you're in a big cosmopolitan city, like Mumbai or Chennai or Delhi, then you don't have so much problems because, you know, there's more outside contact, you know, people are beginning to get like DSL, cable modems, that kind of thing. But even then, they have more problems than, you know, you would if you were living in Europe or North America. So, you know, we want to do things that we can to help people, you know, know what else is going on out there as well. So the situation in India is that it's very diverse, has a very ancient history, so there's been all kinds of movements of people in and out. And there are 18 languages which are recognized by the Indian constitution. However, there are many more languages than that which don't have official status but still have, you know, maybe a million or more speakers. So, you know, we're talking about a lot of diversity. I'm reminded, I think it was Alexander Pope who defined a language as a dialect with an Arminian navy. You know, and while that's not literally true in this case, it is true in the sense that, you know, the definition of a language versus a dialect or where one language starts and another one begins, ends and so on, that is actually, you know, involves a bit of politics as well as just, you know, simple linguistics. I'll give you an example. Hindi is the main language. It's the most widely spoken language in India and it's like, out of the 18s, it's the official one. But in Pakistan, they speak Urdu. Now, the partisans of each language would tell you that these are totally different languages. In fact, a Hindi speaker can understand 99% of Urdu and vice versa. The only difference is Hindi is written in one script. Urdu is written in an Arabic script. Hindi has been more influenced by Hindu ideas. Urdu has been more influenced by Muslim ideas. So, you know, there's definite differences there, but the core is actually something that's, you know, fairly common. But, you know, you could try to get people to admit that and sometimes you run into resistance because it's obviously not patriotic or whatever. But actually, Urdu is actually one of the languages that are recognized in India as well because there are still people who speak it, but I'll go on more about that later on. These languages belong to two basic linguistic families, the Indo-European or Indo-Aryan as it's called sometimes, which is very strongly related to the European languages as well, for the most part, except for Finnish. And Dravidian, which is the four South Indian languages, Tamil, Telugu, Malayalam and Kannadi. And of course, there's been a lot of cross-pollination, but there's definite linguistic differences between the two groups. The Indo-European languages in India are descended from Sanskrit for the most part. While the South Indian languages, they've been influenced by Sanskrit, but they're not directly linguistic descendants of it. And I'm mostly going to be discussing the Indo-European languages mainly because I'm not actually that familiar with the Dravidian ones, but much of what I'm gonna say is in common to both. And all these languages, for the most part, use scripts which are on the same basic model, which I'll also explain in a bit. The exceptions are Kashmiri, Sindhi and Urdu, which use the Arabic script. So while, you know, our idea with Devin Ayan is to welcome anyone who speaks those languages to join the project, I think just as a practical matter, they might feel more at home with the Persian and Arabic localization people, because they have more in common technological problems and so on, like right to left, right and then we do. Yes? I just wanted to say that when several people in Israel tried to get BD support, we're trying to get involvement from Arabic people and several from, especially, Urdu people, but we don't have people to be of point of contact and so on. But mostly the support is there, we just need someone to translate nothing more. I mean, you know, I don't wanna get into, you know, a possible flame war about this, but you know, a big difference has been that India has been a democracy and, you know, fairly stable while, you know, Pakistan has had more, you know, military dictatorships and so on, fundamentalism and all that. So it's kind of behind technologically. There are actually several people in India who are interested in doing the Urdu localization and I'm trying to get them maybe to get involved if they decide to get involved with Demian to like contribute to some of these things. But, and I'm sure that there are some people in Pakistan too, but you know, their whole situation is like, I would say about 10, 20 years behind, we're in the A's even, but I'm not gonna get too much into that because that way lights madness. You can send them to me if there are some B-directional problems who are already working on that. Yeah, okay, all right. Now, as I mentioned, there are different scripts used to write Indian languages. I think there's a few tribal languages in the far northeast where they use the Roman script because, you know, they were only reached by like missionaries or something early in the 20th century or something. So, they're not really part of the Indian cultural sphere, even though they might be in the geographical region of India, but most of the Indian scripts are related, whether they're from North Indian or South Indian languages. And of the languages in this list over here, Aswanese, Bengali and Manipuri use the Bengali script. They use the same script. Hindi, Konkani, Marathi, Nepali, and Sanskrit use Devanagari, which I'm gonna be explaining about. So, they all have the same script. And the other languages all have the same script. And if you look at them, you can see the similarities, but they diverged over time. And other countries that were influenced by Indian culture in the past often also have a similar kind of script structure. This include Sinhala, which is spoken in Sri Lanka, Thai and Tibetan. And of these, we have Thai and Tibetan fonts, but I don't think we have a Sinhala one. But for all the Indian scripts, we have fonts in Devian for them now. And I want to make the point that the languages and scripts are not necessarily corresponding one to one. There are several cases where you can, you know, write a language in different scripts. One example is Punjabi. Punjabi is the region in the Northwest of India. But after partition, half of it is in Pakistan. But Punjabi speakers in Pakistan use the Arabic script while Punjabi speakers in India use the Gurmukhi script is called. That's like the native script of Punjabi. Or they can sometimes use Devanagari as well. Some of the scripts have like extensions to cover sounds which are non-native, like North Indian languages don't have a native Z sound. Their J is the nearest. But that's why if you've heard an Indian speaker say zero instead of zero, that's because they don't have that Z sound. But there's been an extended Devanagari character. You can put a dot under the J and that signifies Z if you want to explicitly show that this is a different sound. And that's especially important for foreign loanwords. So, you know, that's the situation with scripts. In the wrong way. How many scripts are there? There's about ten major ones. I think as far as Unicode is concerned, Unicode is the attempt to consolidate all the world's scripts. There are currently, I think, ten index scripts which have code blocks. But that's not all the various scripts that are out there, even. But those are like the main ones. So this is Devanagari. And Devanagari, very early on in the history of India, linguistics became important because, you know, language has always been thought of as like the supreme human invention. Like, you know, how we say fire is like the original invention in western culture, that's like the image. So speech was a big preoccupation of ancient Indians. So they did a lot of analysis of speech and so on. So the alphabets are typically laid out in a linguistic pattern. Yeah, they're not alphabets, they're scripts. Use the words interchangeably. But each glyph represents a sound. And the sounds are organized. First we have the vowels. Then you have, you know, compound vowels or diphthongs. Then the consonants are arranged in terms of, you know, valors, paletals, retroflex and so on. And you have semi vowels and you have fricatives, as they're called. So, you know, it's a very musical sounding thing if you say the whole thing. But all the index scripts are based on the same model. And in fact, you could, well, there are a few minor variations, but you could even, you know, just algorithmically, you know, convert from like one script to another, you know, just by, in unicode, they're arranged in blocks of 128 code points. So, you know, by adding and subtracting, you could do a mechanical translation like that. It might not necessarily mean anything, but at least in terms of pronunciation, it would be the same. And each of the vowels here, this is the, you know, full form of the vowel. And for the consonants, you have an inherent er vowel because er is the most common vowel. So, let me actually go over to... This is my name in Devanagari. And it's just the, you know, the characters, the characters, the characters. I mean, it's not entirely phonetic in some cases because, for example, in my language, Gujarati, we tend to drop the er at the end of a syllable. So, it should actually be in Sanskrit Jalada, but I pronounce it Jalada. But those are basic consonants. But in my surname, Vyas, we have a situation where we have a compound consonant. And so, like, this is a v and this is yeah, but it's not vyas, it's Vyas. So, I combine them by... There's a symbol for this called holland, which is like that, which makes it, which, you know, removes the vowel from a consonant or consonant cluster. So, this is now just v. Now, I can combine it with yeah, like this. So, now, this is yeah. Now, together, along our sound, I do this. This is the joined version of that ah vowel that came on the list before. And then, so, Jalada Vyas, this is how I would write my name in Sanskrit or Hindi or whatever. So, the problem with this, as far as computers are concerned, is that, as you can see, you need... There's already, you know, way more than 26 glyphs in the original set. When you count all the combinations, you know, you can easily, to have a complete font, need like between, I would say, 800 to 1,000 different symbols. Because there's actually even three consonant clusters as well, and possibly even four, but those are very rare and only really in like Sanskrit words, mostly it's just two. But still, you need to have that. So, any system that's going to display that's going to display, you know, Indic languages has to have that shaping ability. And, you know, sometimes also the order changes of the symbols as well. So, this has historically been, you know, a huge problem in getting any kind of Indic support into computers. But, you know, now it is possible under X, at least, but it's still impossible to really display Indian languages on any kind of console mode program. It can actually show up, it doesn't look pretty good, yes. Well, it's just something I remember. I understand this script is way more complicated, but I remember still from the early days before graphical interfaces were common. Seeing also in console mode written Arabic that each letter also modified the previous one. Right. I was at the border point just getting the steering at the infestination. But even, you know, even Arabic at least has the benefit of the characters are mostly like the same size, you know. But here like a proper representation of a character can be variable in width. And that's been, you know, the thing that's really hard to do. There are actually, like even in the days of DOS there were like add-ons you could get to display Indian languages, but they just didn't look very good at all. You couldn't imagine, you know, most people using that kind of thing. So, typically what they did was just kind of transliteration scheme into, you know, the Latin alphabet. So there are fonts out there, but unfortunately another problem that we've had is that the people who do the fonts have just, you know, put characters all over the place. So, you know, there's no real encoding in a lot of these fonts. And that's a big problem for especially my parents like to visit like newspaper websites to get the latest news about what's going on. But it seems like every newspaper's website has its own font that you have to install because if you have the wrong font it just looks like gibberish because they didn't match up, you know, the characters in the same places. There are some standards. Previously the main standard in use was called ISKI, the Indian Standard Code for Information Interchange, which is an 8-bit encoding. Lower 128 characters are the same as ISKI, and in the upper 128 characters they have all the different characters that are needed for an index script. And because all the scripts are on the same basic model, you just need one byte in advance to identify which script it is, then you can use like, for instance, this bias, like, you know, I showed it in Devon Agri, in Gujati or Punjabi, it would look different, but they would still be the same half-v, year, and a. So, you know, you can have a neutral encoding for those and just have one byte in front saying, you know, it's this language as opposed to that language. But that's not, still not really that, you know, good, and it's not even really been that widely adopted. I think it's mainly the government that used it. The future is Unicode, which, you know, when the ISO people were getting that together, they actually sat down and actually thought about, you know, the issues involved with index scripts. So, you know, as far as the Linux world is concerned, I think really Microsoft's on all the operating system people that are just not bothering with ISKI, and they're just using Unicode in UTFA. So, the fonts that we have are typically in the open type format, which is a superset of true type, but allows for the conjunct characters and things. It's a bit more descriptive, but it's basically the same as that. So, you know, there does have to be some special support in the application because of the need to shape the characters, but it's still, you know, most programs that use true type fonts can at least show something. But there is shaping support provided in the toolkits, such as QT or GTK. Pango is the actual library that does that. And it's pretty usable, and I would say it's at least as good and in some ways even further ahead than what Microsoft has. I think Microsoft is only going to have like really 100% accurate support for Indian languages in Longhorn whenever, you know, that's supposed to come out. And even then, it will probably only be for like the top six or seven Indian languages. And so, and of course, you know, who knows when like Adobe or, you know, YouTube or media or anyone will get around to it. So, you know, I think we have a good window of opportunity to really get, you know, Linux into India, not just as, you know, a, you know, me too, but as really an innovation, as really ahead of the competition. And I think if that happens and Linux becomes entrenched, no matter what kind of schemes or, you know, price promotions or whatever the proprietary vendors do, they won't be able to get past. But that depends on a lot of work getting done and, you know, it remains to be seen if we can actually do it. But I hope that we can. The input of characters is done via input methods. There's many different ones. Wow. You know, I was talking about them. They're done by input methods such as SCIM, Triple IMF, QT has its own. And the translations are just done using get text and profiles, just like anywhere else. So I'm going to skip ahead a few. I was going to talk about some of the political issues, but if you attended the talk on Latin America, you kind of know some of that, because it's the same basic things. A lot of countries tend to have this really dead weight bureaucracy that just doesn't do anything useful, but interferes with things. There are some good signs, too. The president of India, the president is a very ceremonial position, actually. The real power is in the prime minister's hands, but has a lot of respect. And the president, Abdul Kalam, has been spoken out favorably about free software and Linux in the past. So even if he can't actually achieve anything by himself, at least he gives prestige to the idea of using Linux. And that's very useful in countering the kind of people who say that it's just a toy, or it's not supported or whatever. There are some cultural issues involved in getting free software adopted. Some of them good, and some of them not so good. One is that language is a marker of social status. And so people want to see their language, you know, become involved with computers, not because necessarily they have any idea of using it themselves, but then they can say that, see, our language is so modern because it's used on computers and so on. So that's actually made people kind of receptive to having all these localization projects, you know, just for their prestige value. Also, the good thing is that Indians are very pro-science in technology. You know, you don't have that problem of... Because, like, for instance, in Hindu holy books, like, the earth is shown as being like a flat disk supported on four elephants, supported on a turtle. But, like, there's no one, not even the most, you know, fundamentalist person, and I'm a very devout Hindu myself, who would try and, you know, like, in America there's an issue right now that people are trying to get biblical creationism into the textbooks, that the earth was created in six days, and so on. There's nothing like that going on in India. And people really value, you know, technology and so on. So they're receptive to new ideas in that way. One problem I'm seeing is that Indians I work with, they like to be told what to do. They like to have, like, a master plan and somebody gives them a task, and then they'll go and do it, and, you know, they're not lazy about doing it. But the free software community is more about, you know, finding a niche and scratching it. So, you know, we need to get them more, you know, acclimatized to a different way of doing it. There's a different way of doing things. You know, they've done this way for me to tell you what to do. Find something you're interested in doing and doing. The other problem is that the education system is so specialized. Like, if you're an arts person, chances are you know nothing about science. If you're a science person, you can barely, you know, sign your name. You know, and with translations especially, you know, translations can't just be mechanical. You have to have, like, a kind of poetic flair as well. But, like, you know, I'm trying to get, like, for instance, my mother is a poet in Gujarati in her language. But she's, like, utterly computer phobic. I'm trying to get her to get involved with, you know, translations and so on. But the idea is that, you know, that's why I have a son to do computers for me. You know, I mean, what do I need to bother with any of this stuff? Or I'm, you know, up in the clouds somewhere. But so, you know, that's very frustrating because, you know, if there was a bit more mix of skills, I think that would help a lot. Infrastructure, like I mentioned, is still a problem. In big cities, you get cable and DSL and so on, but it's not always reliable. If you're in a village or something, then forget it. Actually, an interesting thing I found out, I was talking to somebody and he said that practically everyone he knows has a DVD burner. I don't even have one myself in my computer, so I asked why that wasn't. He said, because sometimes it's so much easier to just get, like, a distro, burn it on a DVD and then, you know, like, take it to somebody's house rather than try and, you know, FTP stuff and, you know, send things around like that. And, you know, this is a problem because, like, you know, God forbid trying to, you know, follow Devin DeVal on the 28-8 modem. You know, it's just like, it's way too much to keep up with. So this means that the people who are interested in free software, who are hacking, you know, are not really often in, you know, tuned with what's going on in the rest of the world. And, you know, more and more people have disposable income now. You know, it's not the case that India is like poverty-stricken. Even poor people often have, like, some money, but, you know, they're more concerned with, you know, making a living than with, you know, doing charity work, which is what free software really is. And the other thing is that, you know, they don't want to just, you know, have, you know, some kind of cut-priced bargain solution. They want to spend their money wisely. But it looks like I'm out of time, but... There's a talk after this. That's why it starts in five minutes and we need preparation for it. Can I maybe talk until people start coming in? They haven't started coming in yet. Just a few other things I wanted to mention. The other thing is that many of the Indian languages, you know, you actually have to invent the technical vocabulary, because they've just never been used. You know, some languages have prestige, but others are just like, you know, for the peasants. So they've never really developed any kind of, you know, high terminology. And so, you know, and also we have the problem of making translations, you know, culturally coherent. Because not all, you know, like all the little Western in-jokes make sense, you know, to people in India. Some free software groups. So what has Debbie and I done so far? Well, we have a logo. And, you know, that's been our major achievement so far. But we have also packaged fonts on the scripts. But there's more that needs to be done to have, you know, a properly usable system. So in the future, we're going to do things like, especially when we get graphical install, we're going to do localization of that. Debtconf messages, some documentation. There's other things that need to be packaged. Like I said, we want to help Indian hackers join the global communities. And maybe bank like a live CD or a CDD, something like that. And so how can Debian help? Well, you can feel free to join us if you're interested. I was thinking maybe one day we can have, you know, Debtconf in Bangalore or at least, you know, somewhere near like Singapore, someplace where it's easier for Indians to visit. Because I have a bunch of people who are interested in becoming Debian developers, but the sticking point is getting the key sign because there's no one around. There's two developers currently living in India, but they're all very far away from each other. And so if you're in India on a holiday or something or on some business trip and, you know, you know of some like log meeting or something, you can go by and sign keys. That's a way to help. Just keep on being Debian because, you know, even the stuff that you do which is not related to India whatsoever does help because by increasing the pool of free software out there, you're creating something that everyone can use. So thank you very much. Feel free to talk to me about any of these issues at any time.