Approximity blog home
835 to 844 of 910 articles

My LinuxTag 2004 photos   25 Sep 04
[print link all ]
Some photos from LinuxTag 2004 in Karlsruhe. I especially liked the Xbox booting Linux screenshots. pics

Software for Slackers   25 Sep 04
[print link all ]
I need this program to stop my internet addiction.

Are you a slacker? So am I. Do you browse the Web, read the news, and write email all day in stead of working? So do I. Does it make you feel miserable and apathetic? Do you tell yourself to stop browsing the fucking Web and get some bloody work done? Do you have absolutely no discipline? I know your pain.

But recent technological advancements have made it possible… There is a cure for your disease!

Years of slacking at the renowned Massachusetts Institute of Technology have resulted in a brilliant 461-line Perl script (which includes 130 lines of comments for free!) that makes it all possible! Your productivity will dramatically increase!

Today, I present Lockout: The Self-imposed, Computer-aided Work Enforcer. This program will help you get some work done by not allowing you to browse the Web. It won’t allow you to do anything but work. It’s a miracle! Your colleagues will respect you, your Ph.D. adviser will compliment you, and your boss, if you have one, will probably not notice the difference! It’s amazing! Scroll down! Read more!

Get the program

The Real Point of Oracle10g Manageability   25 Sep 04
[print link all ]
Curt Monash has written a very good article about Oracle 10g. The article argues that the real focus is on manageability, which makes perfect strategic sense. TCO (Total Cost of Ownership) is king. And with hardware getting cheaper, software getting cheaper, and custom programming being outsourced to cheap countries, administrative costs are an ever bigger part of TCO. Whats more, manageability is historically a major competitive challenge for Oracle; 10g is designed to neutralize that issue.

Ender's Game and Software Development   25 Sep 04
[print link all ]
Very interesting entry by /\ndy Hunt. Ender is in reference to a novel by Orson Scott Card called ‘Ender’s Game’. Its part of a series of three books, all of which are well worth reading.

Test-Driven Writing   25 Sep 04
[print link all ]
(Source: Stefan Schmiedl)
 >An activity that I /do/ still have trouble with, however, is writing.
 > When faced with having to compse anything more substantial than an
 > email response, I feel the fear start to creep in and I get myself all
 > tied in knots.  Even after I start to put some words down, I often
 > find myself getting stuck because the thing isn't flowing and the task
 > of finishing seems overwhelming.

Yup, writers block definitively, as John Roth diagnosed already. But if you’re able to describe it in such flowering detail as above, there’s no need to have it.

 > So on my way home last night (after another frustrating couple of
 > hours trying to get some thoughts on paper), I was thinking about how
 > I could make my prose writing come as easily as my code writing.  I
 > started wondering if I couldn't somehow employ a TDD-like cycle in my
 > writing process.

I am often writing articles with my business partner, who’s<br>especially good at collecting lots of nice stuff on the web. The first thing I have do with the "drafts" I get from him, is to find the<br>structure fitting best to the available data. This is currently donein a Mindmap using freemind (freely available at, IIRC). For some<br>time I also tried vimoutliner ( for this, but found that for this process, the two-dimensional display of a mindmap is better suited to my brain.

When the outline is finished, I start to grow the flesh on the bones. That’s relatively easy, as I confine my work strictly to the current paragraph.>

The next step is easy, if I have the time: I let the stuff settle for a few days, then go over it once more and clean up the unbelievable mess I created then. If I don’t have the time, I need to play about two hours nethack, which erases my brain just as well…

So the steps are:

 - data collection
  - gradually by experience
  - by force (coauthor delivery)
 - data organization
   - mindmap
   - outline
 - draft
   - follow the map
   - work local
 - refactor or polish
   - grammar, spelling, rhetoric
   - present line of thought more clearly

I think that there’s a difference between code and prose showing here. You expect your code to give certain results for a given input, and you are free to not care about the implementation at all. With prose, implementation is almost everything. So the cost of providing a "working release" is higher with prose than with code. At least for me.

 >find myself getting stuck because the thing isn't flowing and the task
 >of finishing seems overwhelming.

Writing is like every other kind of art. It is never finished. Feeling better now?

Writing is like dealing with animals. Don’t be afraid of it, and it won’t hurt you.

Your fellow author in pain, S.

Test First, by Intention   25 Sep 04
[print link all ]
A code and culture translation from the original Smalltalk to Ruby Original by Ronald Jeffries, translation by Aleksi Niemela and Dave Thomas.

In this document we show you the Ruby version of the Smalltalk code published in the pink book.

Selling XP   25 Sep 04
[print link all ]
Alistair Cockburn has a very interesting paper on "The Costs and Benefits of Pair Programming". Of course Pair Programming is not the only "extreme" aspect of extreme programming but Alistair’s article contains some very interesting metrics (seems a lot less "extreme" after reading Alistair’s article).

XP success story: Sabre takes extreme measures   25 Sep 04
[print link all ]
(Source: Computerworld) Using extreme programming practices, Sabre Airline Solutions has reduced bugs and development times for its software products.

Sabre Airline Solutions had many years of experience with its AirFlite Profit Manager, a big modeling and forecasting package that many airlines use to wring more income out of flight schedules. Even so, Release 8 of the software was four months late in 2000 after final system testing turned up 300 bugs. The first customer found 26 more bugs in the first three days of its acceptance testing, and subsequent joint testing by Sabre and the customer uncovered an additional 200 defects.,10801,91646,00.html

IT WON'T WORK HERE doesn't work here   25 Sep 04
[print link all ]
(Source: Kent Beck posted this to the XP mailinglist) This came up in a discussion of how to handle long-lead-time materials. The OP basically said,
 "I can't do all that stuff you say I should do, but how do I handle the situation ..."
  The response:
  -- IT WON'T WORK HERE doesn't work here.

Wine-Migration   25 Sep 04
[print link all ]
At Linuxtag in Karlsruhe I met during lunch one of the authors of ganymede. I am the last person in the world to promote the use of the Win32-API, but it can come in handy, when you have to use some legacy software on Linux and do not want to pay license fees for the usual suspects like vmware. Having to maintain only one code-base is sexy, too.

I asked David Guembel, one of the fathers of the software to email me a short describtion:

 On an abstract level, the idea behind our ganymede system is simple: To make
 an application run under Wine (a free Win32-API layer and Windows
 executable loader), it is neccessary to know what parts of the Windows API
 are actually required by that particular piece of software. Software has a
 modular structure - in this context, a module is a Windows executable
 (.DLL, .EXE, .OCX etc.) - and every module provides (exports) certain
 functionality and requires (imports) functionality from other modules.
 Thus, ganymede internally creates a dependency graph of an application's
 binaries. This method is static and does not require anything but a fresh
 installation of the software to be analyzed.

 Before x-raying a Windows application, ganymede parses and stores an
 analysis of the soure code of one or more Wine versions. It automagically
 determines the implementation state of the APIs provided by Wine. During a
 software analysis, the functionality required by the Windows application is
 compared to what Wine provides, and missing or incomplete APIs are
 reported. By storing Wine versions and the dependency structure of the
 analyzed applications in a database, automatic or manual re-analysis with
 different Wine versions is possible. Via the API ganymede itself provides,
 the collected data is accessible in several ways. One application of that
 API is our tool named sysiphus, which uses ganymede and a GUI-based
 approach to semi-automatically determine the best possible Wine
 configuration while providing for the possibility to re-use already
 licensed Windows DLLs to fill the gaps Wine still leaves.


powered by RubLog
835 to 844 of 910 articles Syndicate: full/short