Slashdot Mirror


Google's Summer of Code Over

yootje writes "The Summer of Code ('Google's program designed to introduce students to the world of open source software development.') is now over. The result: 410 participants helping 38 projects suchs as Apache, KDE and FreeBSD. 'Among the project awards are both complex and simple innovations spanning the width and breadth of everything that the open source world has to offer. There are projects dealing with security, networking, VoIP, Java, mono, IP-PBX, online picture galleries, instant messaging and content management. There is even a game that Google's summer internship helped to pay for.'" Update: 09/11 17:15 GMT by Z : Added the story link at submittor's request.

6 of 171 comments (clear)

  1. Re:Where is the story? by yootje · · Score: 5, Informative

    LOL, I forgot to put my source in it: http://www.internetnews.com/dev-news/article.php/3 547611 Slashdot editors, please put it in the story?

  2. Re:awesome! by Anonymous Coward · · Score: 5, Informative

    This is nothing more than a shot across Microsoft's bow. They are warning them that they are ready to make a big play in open source and also have managed to basically, buy the hearts and minds of nearly every single FOSS developer for merely $2million - an absolute bargain.

    Google still isn't that open source friendly. Sure, they use Linux for servers but that's simply because there is little competition in the market for 100,000 PC 'clusters' with zero license fees.

    Where is Google Desktop Search for Linux? Where is an official Google Talk client for Linux (yes, I know about Jabber, but it doesn't integrate with your gmail contact list)? Where is the Gmail notifier for Linux?

    Until that changes, I don't really think that you can consider Google anything more than an open source user, not a contributor.

    Finally, Google knows that open source is a huge catalyst. For the $5,000/student they paid it would of probably cost Microsoft $50,000+ to do the same work. That basically means that for every million dollars Google puts in, they cost Microsoft $10million. It's a bit like the CIA funding the mujahadeen to fight the soviets, it probably cost them $1 to inflict $10,000 worth of damage onto the soviet economy (eg a $10,000 stinger missile launcher taking down a $10,000,000 helicopter).

  3. Journalling for BSD filesystems? by Alwin+Henseler · · Score: 4, Informative
    Checking out FreeBSD's ideas page, my eye was caught by this: "UFS Journaling: Add transaction journaling and playback to the UFS filesystem. The goal is to increase the reliability of the filesystem and greatly reduce the need for a full 'fsck' after a crash or power loss."

    If I'm not mistaken, all major BSD's (Free, Open, and Net) support a feature called 'soft updates'. Basically, re-ordering filesystem updates in such a way, that the filesystem remains in a consistent state, even in the event of a badly-timed crash or powerout. All this to avoid the need for a full fsck on reboot.

    Quote from the FreeBSD features page: "Soft Updates allows improved filesystem performance without sacrificing safety and reliability. It analyzes meta-data filesystem operations to avoid having to perform all of those operations synchronously. Instead, it maintains internal state about pending meta-data operations and uses this information to cache meta-data, rewrite meta-data operations to combine subsequent operations on the same files, and reorder meta-data operations so that they may be processed more efficiently. Features such as background filesystem checking and file system snapshots are built on the consistency and performance foundations of soft updates."

    From the NetBSD site: "Soft Updates permit metadata writes to be ordered to achieve close to asynchronous disk performance without risk of metadata corruption. This significantly improves the performance of FFS file systems."

    You might still do a full fsck later (as regular maintenance), perhaps even as background task, but it wouldn't be needed for a reliable restart.

    Journaling is another way to do this, by adding an extra 'log' of the latest updates to a filesystem. Then in the event of a crash, you don't need to check the entire filesystem, but can bring it back into a consistent state by 'replaying' those latest updates from the journal.

    Now here's what I don't understand: why add journalling to a filesystem, when you're already updating it in a 'crash-proof' manner (soft updates)? What's the point? Seems rather like a step back to me, with soft updates looking like a smarter way to archieve crash-proof filesystem handling.

    I assume that this soft updates feature is limited to certain OS/filesystem combo's. And maybe journaling provides some thing(s) that soft updates doesn't? Can some knowledgable BSD user shed some light on all this?

    -- This sig just wasted another 0.x seconds of your precious time. Supporting banning sigs!
  4. The article forgot NetBSD by hubertf · · Score: 4, Informative
    NetBSD got 7 slots assigned, see the NetBSD-SoC webpage for all the details. One of the projects (tmpfs) is already integrated into NetBSD, even.

    - Hubert

  5. Blenders Summer of Code projects by LetterRip · · Score: 4, Informative

    Here are SOC projects done for Blender,

    http://wiki.blender.org/bin/view.pl/Blenderdev/Sum merOfCode2005

    We had some really awesome projects happen (fluid simulation, high quality boolean tools, improved nurbs, 'Verse network integration, animation constraints improvements, and a drawing tool, alas two projects - ODE integration, and a live tutorial didn't happen).

    LetterRip

  6. Nmap Project Results by fv · · Score: 5, Informative
    such a great idea but no content on their site re: the actual work. They should have paid someone $4500 to maintain their summer of code page!

    Yeah, for a $2 million dollar project it was ridiculously understaffed on the Google side. But Googlers like Chris DiBona and Greg Stein worked extraordinarily hard to keep things flowing relatively smoothly. So it still turned out to be a huge success for Nmap and most/all of the other participating projects. Thanks, Chris and Greg!

    So what did we (Nmap project) accomplish in those two months? The sponsored students and their credentials/projects are listed here. Much of their work can be found in Nmap 3.90, which was released on Thursday. SoC changes include:

    • Doug Hoyte nearly tripled the size of the version detection database, added OS/device type/hostname detection using the version detection DB. He made numerous other improvements as well.
    • Zhao Lei added more than 350 OS detection fingerprints to Nmap, bringing the total to 1684. He also helped design a 2nd generation OS detection (stack fingerprinting) system.
    • Adriano Monteiro designed and implemented an advanced Nmap GUI and results viewer named http://sourceforge.net/projects/umit">UMIT (screenshots).
    • Ole Morten Grodaas designed and implemented another advanced Nmap GUI and results viewer (its nice to have choices in open source!) named NmapGUI. Details and download here)
    • Chris Gibson has written a sweet little network tool named Ncat, which takes the venerable Netcat in an interesting and extremely useful direction with features such as connection brokering, socks proxying, and much more.

    It has been a crazy two months, but I'm very pleased to see so much accomplished! If you're using an older version of Nmap, you really should consider upgrading to 3.90 to see the difference.

    Cheers,
    Fyodor