nice logo...

for Lispniks:

Travel Tipps for Analysts

Michael Stal proposed to send Market Analysts to some lonely island. I partly agree - but decision-makers in our industry (as in many others!) are NOT engineers, so they cannot evaluate technology based upon programming experience and technical risks. So we might end up sending decision-makers into a similar direction - what might end up in contraproductive vacuum, although programmers would surely love it (only until the end of the current project budget ... ).

I'll love Micha's conclusion:

"Why do you think, these guys have become market analysts and not software engineers?"

A little more on Lisp

I recently blogged about a simple Lisp IDE - now I'm back with more news on that topic
(why I care about Lisp? Read on...). Two more options for the motivated parenthesist:

• Lispworks (Personal Edition) is a real IDE, great documentation. Dozens of cool features (ORB, COM/Automation, SQL integration...) - personal edition is limited to 5 hours per session, which is ok to try it out.
• The real hands-on-feeling comes up with LispBox - you'll find emacs-based, hassle-free, out-of-the-box versions. They are companions to one of the best books I've read during the last month - Practical Common Lisp by Peter Seibel.

Why I care about the ancient language? Because its a highly productive language to develop prototypes, to explore requirements and to build mission-critical systems. And a few people got rich by using Lisp. Lisp is pretty comparable to Ruby (without the current hype, though).

And: Lisp has interstellar experience. Ruby doesn't.
Expect more on Lisp in the near future :-)

JavaSPEKTRUM Blog (German)

Die Zeitschrift JavaSPEKTRUM hat den Sprung in den Hyperraum der Blogosphäre erfolgreich eingeleitet - und Stefan Tilkov hat einen (seiner Blogger-Ehre würdigen) Startbeitrag dazu geschrieben. Well done!!

IBM on Software-Architecture

In IBM's RationalEdge online magazine I found a series of articles on software architecture and software architects. I liked it for its broad coverage - but do not expect too many details from it. Nice read on the train...

Critical Note on Hosted Services and Web 2.0

These days new Web 2.0 startups and services appear in large numbers - Ajax, Ruby and the like really started something new.

But let's sit back and reflect a little on potential drawbacks of such applications. Take Backpack or Google-Mail as examples: Do we really want them for our work?

Some arguments why we (oops - I personally!) might NOT want services like that:

• Security: I'm consulting highly sensitive clients from a variety of industries. Usually I have to sign a Non Disclosure Agreement (NDA), which disallows most communication about the project-at-hand. NDA's force me to apply tight security measures to the documents, models and other information processed and produces during the project. Just in case I trusted the provider of any hosted service (which I do not...), are the carriers reliable? Is my ISP reliable? Those guys did obviously NOT sign my NDA - so I'm not allowed to divulge any client- or project related information to them. To be more precise: I have to actively take measures so they cannot access the project-information. No hosted service provides the kind of security I need for sensitive data like that. On some of my machines I even keep parts of my harddisk crypted :-)


• Security, part 2: What, if the provider of my hosted service is located within the USA - and is beeing approached by the homeland-security guys? Nobody will care about my personal data, even the mere thought of keeping my data private will not cross the mind of my service providers' admin... Feel free to replace USA by any other country, rogue or not...


• Reliability: Can I rely on the data beeing stored in some obscure datacenter? Or will some unhappy employee of some Web 2.0 startup begin manipulating data one day? Do you include MD5 hashes in your documents when you upload it to your favorite hosted service? I bet you don't... (I did never do it)


• Availability: They all tell us about backups, redundant servers and the like. What if their business-model doesn't work out? Will there be backups even in the future, when I still need my data? Again, I trust my own backups (which I test regularly!).


• Availability, part 2: I'm pretty fond of my Internet-service-provider - they have approx 99% uptime of my DSL connection. But on a few occasions the net simply wasn't available at times when I needed to work.

My conclusion: There simply is no free lunch. And Web 2.0 will definitely NOT, unlike 42, be the answer to all our problems...

Weekend Fun: On Waterfall Projects...

(via Peter) I stumbled across a (fictious) IT conference called Waterfall 2006 - a really good April 1st. joke of some guys from the agile alliance. They wrote up summaries for pseudo-talks they won't ever give - dramatically funny, good read!

Get organized

I'm getting fond of backpackit, a hosted service to manage to-do lists, reminders, arbitrary notes and appoinments. I needed a while to get used to the idea to leave these info on the internet only, with no local copy on my notebook - but it works fine :-)

Backpackit has a cool mail-in feature - you send a mail to your backpackit-page and you're done. Just put a token like "todo:" in the subject line and backpackit automatically inserts it at the right place.

You can selectively share certain pages or items with others - making backpackit a simple collaboration tool. Try out the writeboard facility for group authoring - quite fancy.

Backpackit is imho an brilliant example of Web 2.0 software - no installation, just browser-based, neat and functional UI, both free and commercial accounts available. By the way, its written in Ruby by 37Signals.

Only drawback (or is it a feature?): It works only online. You can get used to this, believe me!

PS: Mac-Users: Go for packrat, a small app wich uses the backpackit-web-service-API to enable offline usage...