Updates from January, 2010 Toggle Comment Threads | Keyboard Shortcuts

  • Bertrand 10:59 pm on January 24, 2010 Permalink | Reply
    Tags: , , , search,   

    New Search Engine 

    I have a question !

    I would like to know the actual possibilitys of making a “smart” search engine that could give you answers of your questions.

    For example, let say that i want to know the birth date of Napoleon, i’m going first on google and type Napoleon brithdate, i find few links and by habit, i click on Wikipedia. There, i look in first lines to find what i’m looking for.

    Is there a way of finding directly this information by asking a simple question like this? I mean noone care about looking stuff in search engine or encyclopedia but the answers.

    Even better, what if i speak directly in mic asking my questions and get replyed by voice. Avoiding typing and mousing gesture for earn time.  Imagine a children playing with such application for years! Or this could in reverse made the learning process almost unusefull for most people.

    Until where could we go about question complexity?

    John told me few years ago about ontologie and this really blow my mind but what now? Is there something like Wikipedia but ontologie oriented more open than a “simple” dictionary?

     
    • Daniel Radev 3:56 am on January 25, 2010 Permalink | Reply

      Currently WolframAlfa (http://www.wolframalpha.com ) is the best answer as far as I know.
      It would give you answers to questions like Napoleon birthdate and it has some other nice features (type AAPL vs GOOG for example)…

      • Neven Boyanov 8:21 pm on January 26, 2010 Permalink | Reply

        That is very interesting. I looked for various (and few very stupid) things but what surprised me was that: “Jesus Christ date of birth” – result ~ 4 BC, looks inaccurate. But it’s still better that nothing, i.e. as if the guy never existed. ;)

    • jyonkov 7:25 am on January 25, 2010 Permalink | Reply

      I agree with Daniel that WolframAlfa fits your description the best. Here are some other interesting once.
      http://www.clusty.com/ – categorizes/clusters searches.
      http://labs.hakia.com/hakia-lab.html – they are based in Turkey i think..
      http://www.powerset.com/ – from MS
      http://www.google.com/search?hl=en&esrch=FT1&tbo=1&tbs=ww:1&q=ontology&btnG=Search -Wonder Wheel
      I can’t skip WordNet http://wordnetweb.princeton.edu/perl/webwn

    • Bertrand 8:33 am on January 25, 2010 Permalink | Reply

      Ok guys, thanks for the tips. I’ll play around with.

    • Jean 4:55 pm on February 2, 2010 Permalink | Reply

      Aardvark (vark.com) seems something to check out too.
      The intelligence of the engine consist of routing your questions to experts in your social network.

      More human interaction, better?

      • jyonkov 2:18 pm on February 13, 2010 Permalink | Reply

        Thanks Jean, I signed up and i like it ! Even Answered a question trough GTalk :)

        I’ve been thinking about “universal” instant messenger with some good plugin framework that can make easy writing applications that utilize the FOAF contacts network since PMF days (PMF was a framework we wrote in Qt while working together…) but never got around to it. It would be cool if one can hack a network app with GUI in a week and publish to the world or some subset without effort… Ahh wait that sounds like Apple App Store?! … well not exactly.

  • Bertrand 7:39 pm on January 24, 2010 Permalink | Reply
    Tags: VFX, Visual Effect   

    Online VFX 

    As i’m working in the VFX industry since 12 years, i always thought that the tools that we use are totally unadapted and obsolete. This coming more real than ever with the years.

    2 years ago, i decided to rally my ideas from my own experiment around a concept. This concept is large and can’t really be describe in few lines. But there is the main ideas.

    But let’s talk first about the software industry. I hate this old school business model of making desktop software, selling it, and justify annual upgrades for making money, even if we don’t know what new functionality to put, we let few new bugs, increase the binary size and make it run slower to obligate clients to buy a new computer.

    In 2001, i was using a compositing software called Shake. This software was really perfect. Very fast, binary of 8Mb, impossible to crash it, very new and open philosophy, etc. It was used by the major VFX studios for make movies like Starwars or Lord of the Rings. What’s append, in 2002, Apple acquired Shake’s company and stop the smart development, stop the Windows and Irix version and make a MacOS bugy and slow one, and now this software is no longer sold and updated by Apple. But a huge part of the industry continue to work with, even with the 2001 version on Windows. For me, this example is very good, users don’t need economic upgrades but just smart tools. Another example is about the Adobe softwares, Photoshop is the leading pictures manipulation software on the market now, but what if i’m looking a little around it. Nothing new since 5 years, only attractive words around little improvement, still bugy, binary is about 1Gb now, and you need a fucking Nasa computer to work correctly. So most of the Photoshop’s operators that i know are still using older Photoshop versions. The tools was good 5 years ago, with everything to work properly, we know where are bugs and computers are fast enough now to made it run correctly.

    What’s new now? Since few years are coming online apps. Google doc is a very good example of what to do. Simple app, light, easy and handy for free user and more advanced with support for clients. I love this business model.

    My idea is very simple, made a complete picture manipulating online apps based on a media repository and a single rendering engine.

    1 – Media Repository:

    No new ideas here! I was thinking about such thing few years ago. Now such tools are very usual. Except that movie industry users need the medias to be in local network, remote web access for uncompressed pictures or video are not really possible yet. Also, a real user interface, no complexes setup, think artist way, not programmer’s way. Mesh is a very good example. With a versioning system, it could be perfect.

    2- Media Manipulating Tools:

    What are pictures tools? There is a bunch of software’s kind that use different GUI system for do the same things. Editing, color grading and compositing are very similar softwares. Only GUI allow you to do different things and work on different parts of the media more or less comfortably.  The main idea is to setup all those differents GUI all around the same rendering engine.

    Advantages are users can define them own way of working, all clients are always up to date so no compatibility problems exchanging projects, an online plugins and scripts repository will allow users to send rendering remotly all over the world without installing new plugins, new version, etc.

    A simple version could be free to let new users work and train with the app, and a professional versions could be rent with extra GUI an functionality depend of clients. More precise tools and dedicated hardware control.

    The renting tools idea is really great for industry. Rather than buying very expensive tools, they could only rent software even for a day in some cases. It’s also better for manage production pipeline and operator’s licenses. It can be more precise in prices range over a project.

    The Open fX project is also very interesting for making plugins inter compatible. It already exist every possible functionality. The use of GPU in 2D software still rare and not often used in smart way. GPU for make sound processing just even don’t exist yet.

    Off course, the greatest probleme making such application is to find a way to make the browser’s GUI and the local media communicate together. It’s not really possible to work remotly on a web server for profesional that use a lot of uncompressed contents.

    My google presentaion

    3 – Media Project Tools:

    No finished online tools exist for managing film, tv, print or any media projects. This could be a very good starting point making money. Free simple and graphical tools to let people get addict, and more complete and professional tools  in rent.

    Conclusion:

    I think that the software industry is really deeply changing. All artists that i know are really tired of existing old school softwares and they are always looking for something lighter and smarter. Of course i can talk about this for weeks. But all what i just described don’t exist yet. It should.

     
    • Daniel Radev 4:29 am on January 25, 2010 Permalink | Reply

      HTML 5 specification would allow processing local data from a browser to some extent (both SQLStorage and Local Storage APIs).
      The problem however is that VFX industry works with huge amount of data (in Avatar case 1.5 Petabytes row material, I read somewhere) and speed is critical.
      However it worth to discuss what and how can be done

      As for the harnessing GPU,I think OpenCL is very promising (http://en.wikipedia.org/wiki/OpenCL)

      • Bertrand 9:32 am on January 25, 2010 Permalink | Reply

        Yes, you’re right, local storage is critical for such project. It’s very common to exceed 1 Gb of data by picture on a 3D movie.
        OpenFX is based on OpenGL. Now there is a lot of advanced VRML/OpenGL plugins for browser but no general standard. Flash is the most interesting GPU based platform today for 2D rendering.
        Processing(http://processing.org/) also is really cool! It’s java based and it’s relatively fast.
        But the point of this article is more about the fact that there is no graphic app web standard yet.

    • Neven Boyanov 7:57 pm on January 26, 2010 Permalink | Reply

      A lot of the tools you’re using today are old-fashion because they’re inherited by much older products and that’s because very few large companies exist in that area of business that produce such software.

      The ideas that you propose here are great but they are beyond the capabilities of a small to mid sized company, there’s too much work that needs to be done. The problem is the time. It could take too long and by the time of its completion it will be old fashion again.

      The other approach is the community, a.k.a. open source. The problem in that case though is that you need a (very) large company to back this project. Because, the open source projects do not exist by themselves, they exist because companies of various sizes fund them. And here comes the problem – the companies that would be interested in such project are not those that would support any open-source project, they are all proprietary. May be they have their own reasons not to like the openness, that’s not our concern.

      Here I should say that it is common misunderstanding that the open source and the openness in general is created by enthusiasts that work for free. No. It’s a philosophy. You either have business that makes money anyways and you’re (self-) confident enough to open-source your creations, or you just fund yourself with your personal savings. There’s another, thirds, case where you’re helped by other people (your rich uncle on your mother’s side or your wife who’s working while you’re in the lab) who may believe that you’re genius and you cannot take care of yourself but you should be given a chance.

      So, you have to (1) find a company that already has such functionality or is willing to build such by your recommendation, or (2) find a company/organization/person that will support (i.e. fund) such open-source project.

      I know that this may sound discouraging or skeptical but it is not. It works the same way everything works in the capitalism. It’s a free market. And we have high hopes for the humanity of the richest capitalists.

    • Eisneim 12:58 pm on June 14, 2011 Permalink | Reply

      awesome stuff!!

  • Neven Boyanov 7:06 pm on January 12, 2010 Permalink | Reply
    Tags: community, everything, life, universe   

    What we should do with our (small) community site? 

    I think before we start “just-another-place-to-share-everything-we-do” or “the-ultimate-place-for-doing-cool-stuff” we should try to get the maximum out of what we have right now, which is WP.com for one reason or another. I’m afraid that we will start wasting time, resources and much effort creating one place after another instead of doing things.

    Does anyone have clear idea what exactly we would like to do?
    Let’s answer this first.

    OR

    Let’s just let everyone talk here and may be something will come out of it. The Universe is a perfect place, it’s built of random (although the existence of the randomness is a controversial philosophical question) stuff and events. Let’s see what the chance will bring to us.

     
    • Daniel Radev 10:12 pm on January 12, 2010 Permalink | Reply

      Well, in general I think the whole idea of the site is to share some ideas, that we might have and eventually develop them. The thing is, I do not expect that we will generate ideas every other day, so it seems unreasonable to expect heavy site traffic.
      On the other hand letting everyone talk about their ideas will eventually generate traffic/ideas/potential, but it also means some unwanted traffic, so we probably need some way to filter content if we let everyone talk…

    • jyonkov 11:59 pm on January 12, 2010 Permalink | Reply

      I agree that we have plenty here…It would be nice to have other languages and adds but ROI might be insignificant… Personally I’m exited about a DevCamp with at least a Green Energy theme and a space/page where people could write about what they are hoping to develop. if more than 1 team emerges we can have a competition and the winer can be decided based on the energy produced. About other developments we can always showcase stuff here…

    • Neven Boyanov 9:26 pm on January 13, 2010 Permalink | Reply

      It is not appropriate to filter content, everyone should be able to at least write a comment. The posts though are started by the members only. The only thing we should be doing in regards to DJR note above is to encourage one thing and discourage another. But I think we have no right to judge anyones opinion.

      The biology is all in Latin, a good part of ancient philosophy in Greek and just a few other languages, so why not the hi-tec in 21-st century in English, It’s not the native language for (almost) anyone of us, but it’s a good common language. The problem is for people like Hristo Minkov (from Kozarevec) who’s English vocabulary is very limited and he cannot participate here for the same reason, but at the same time he is far ahead of anyone here in terms of practical implementation of alternative/green sources of energy.

      • jyonkov 6:33 am on January 14, 2010 Permalink | Reply

        I agree with everything about filtering, nicely put! About Hristo and some other potential members, they can always use google translate with google toolbar… and if willing to contribute they can do that in Bulgarian if they like…

      • Daniel Radev 6:57 am on January 14, 2010 Permalink | Reply

        By filtering content I meant avoiding spam/scum/phishing attacks, which are inevitable if we open the site for all. Otherwise I totally agree that we must encourage any opinion (even if we might not like it by any reason) and we have no right to judge….
        As for the language let’s just say that English is the “official” as the common denominator here, but anyone is free to write in language he/she can express his ideas best…

    • Neven Boyanov 8:06 pm on January 26, 2010 Permalink | Reply

      I think that the best thing about our website is that it’s a good place where we can share our thoughts and expect our friends to share their opinion.

c
compose new post
j
next post/next comment
k
previous post/previous comment
r
reply
e
edit
o
show/hide comments
t
go to top
l
go to login
h
show/hide help
shift + esc
cancel
Follow

Get every new post delivered to your Inbox.