• software as Commodity
• network-enabled Collaboration
• customizability and software-as-Service

link

We’d like to share what the Internet sounds like when it sleeps, and in its current highly agitated state, we think it’s worth sharing. www.theregister.co.uk/content/6/34227.html

rubyforge.org/projects/net-ssh

Version 0.0.2 brings the implementation to full compliance with the SSH2 protocol, since you can now use ssh-dss key types.

The most significant new feature is a limited implementation of the SFTP protocol. Only a subset of the features of SFTP are implemented, namely directory enumeration, and getting and storing files. More features are coming.

The SSH protocol itself is asynchronous, so the "core" implementation of the SFTP protocol (Net::SSH::SFTP::Session) is also asynchronous. However, a synchronous version (useful when you don’t need multiple channels open simultaneously) is also available (Net::SSH::SFTP::Simple).

Until Ruby 1.8.2 is released, you need to also install the patched version of the OpenSSL module for Ruby (also available from the Net::SSH site). Ruby 1.8.2 will include the patched version of OpenSSL, though, so once you have installed you’ll need nothing else to run Net::SSH.

One thing I found interesting about his talk that I managed to relate to XP was when he talked about how humans have "normal" level to categorize things. For example he showed a picture of a tree. Whenever people see a picture of a tree and you ask them what it is, they say "tree", not "maple" or "plant". There seems to be a "middle" category that the mind tends toward.

Traditional software development takes either a top-down or bottom-up approach to categorizing things. That is we don’t start at the natural middle abstraction and work our way up or down the hierarchy.

I was wondering if when we take a TDD approach to design, we can manage to start at the natural middle level and then refactor to generalize or specialize as we need to.

A Cisco purchase gone wrong has cost San Jose, California’s CIO Wandzia Grycz her job. Grycz exited her CIO post earlier this week just ahead of an audit release detailing the city’s findings on a recent computer and phone network installation proposal. Grycz has publicly denied that she allowed Cisco to craft the nature of the IT contact.

…

A new $51m computer system has had so many bugs that city officials can’t get the technology up and running at all. And the culprit looks like SAP.

"We find problems on a daily basis, and part of that is getting the (computer) system to work for us," Diane Supler, budget director in Tacoma told the Associated Press. "Every time we think we’ve identified all of the issues, something else happens in SAP (the system software)."

link

So, lets talk about SQLites handsome features:

• SQLite is swift. In my own testing, I have found it to be speedy. Some speed comparisons with MySQL and PostgreSQL are here.
• SQLite is not a large database server, such as MySQL. You dont connect to the database. Using SQLite, you access a database file. Everything happens in-process.
• SQLite is an ACID database. Supports transactions, triggers.
• SQLite is public domain. Absolutely no licensing issues.
• SQLite is typeless. Any type or length of data may be stored in a column, regardless of the declared type. This allows extreme flexibility and avoidance of type errors.
• SQLite allows custom functions and aggregates. This is my favorite feature of SQLite, which we will explore shortly.

link

Why does the world need Nutch, when search engines are free? Search engines are free to use like television is free to watch, but, like television programming, search results are subject to manipulation by the interests that control them. The only way one can be certain that search results are unbiased is if the technology which computes them is public. Nutch seeks to make high-quality search technology freely available.

How can a non-profit afford to run a search engine?

Nutch is primarily a software project, not a service. Large scale deployments of Nutch will probably be run by commerical interests separate from Nutch, funded by advertising or somesuch. If the Nutch software is good enough, perhaps existing major search engines will use it in place of their current closed source code.

The Nutch project itself may choose to host small-scale demo system, so that folks can see that it really works. This will require only moderate funding. The Nutch project may never host a full-scale deployment for folks to use as their everyday search engine. We’ll leave that to commercial ventures that can afford it.

Will Nutch ever be as good as other search engines?

We hope it will be better. With developers and researchers from around the world helping out, we hope to be able to surpass the quality of what any single company can do.

Nutch

Make sure you also read the story about Crypto AG and the Iraq-Iran Conflict.

 The really weird twist to this story is that the U.S. has already
 been accused of doing that to Iran. In 1992, Iran arrested Hans Buehler,
 a Crypto AG employee, on suspicion that Crypto AG had installed back doors
 in the encryption machines it sold to Iran -- at the request of the NSA.
 He proclaimed his innocence through repeated interrogations, and was finally
 released nine months later in 1993 when Crypto AG paid a million dollars for
 his freedom -- then promptly fired him and billed him for the release money.
 At this point Buehler started asking inconvenient questions about the
 relationship between Crypto AG and the NSA.

link

  "Are you still using a database?"

    Madeleine is a Ruby implementation of Object Prevalence:
    Transparent persistence of business objects using command
    logging and snapshots.

     http://madeleine.sourceforge.net/

          Hi,

          Just figured it was a good time to release all the good stuff I and
          Stephen Sykes have been preparing in the Madeleine CVS. YAML marshalling
          and snapshot compression should be the highlights for our existing
          users.

          Madeleine 0.7 (July 23, 2004):

         * Broken clock unit test on win32 fixed.
         * AutomaticSnapshotMadeleine detects snapshot format on recovery
         * Snapshot compression with Madeleine::ZMarshal
         * YAML snapshots supported for automatic commands
         * SOAP snapshots supported for automatic commands
         * Read-only methods for automatic commands

        If you're planning to use either YAML or SOAP marshalling, beware that
        there are objects and classes that Ruby's own Marshal can handle but
        these can't. You will have to try for yourself if your application
        works, both to make a snapshot and to read it back, with the marshaller
        you want to use.

                                  cheers

                                  /Anders