Make a million, lose a million, who cares?. This is a good spin on the phenomenon we just went through the last couple of years. Datacom, this is the area in which I specialize in too. It is sad to hear that the entire datacom/telecom industry is regarded as a bad taste in the mouth like yesterday's undigested lunch. I have had my share of working in a 10 people company. I had spent a short time at Kestrel Solutions in Moutain View back in 1998. Each engineer had a Costco table for his computer, another one for his stack of books, and a rolling chair. It was a garage-like setting in a building which was broken down and no facilities. Conveniently Costco was located next door and twice a week the marketting person would take orders of supplies and make a trip to Costco. At first glance it seemed like a hellish place to work in, but the comradery and the spirit of the place tempered that view quite a bit.
I didn't stay long enough to see how the company fared. But for a period of 4 months, I was the software guy making a difference. Although it didn't seem like much, it felt good being part of something, creating something. Financially, the experience didn't go anywhere.... but I'm sure that's not why we all got into engineering in the first place. To use our minds to create, to be respected for what we do, and to earn a decent living... at least this was my intentions for entering the business coming out of college.
Here is a good read on the topic of software evolution. A unified theory of software evolution
I do agree that to get a good view, you only have to look at the evolution of the other aspects of life. Whereas the comment of Software not following Moore's law: Moore's Law only predicts the physical aspects - the number of transistors thrown into a chip die versus the actual complexity of the die itself. Looking at software, you can make the physical size of the software as big as you want by including and linking with as many libraries as you like.... however, that does not necessarily make it any more complex than a simple HELLO program.
It looks like the grace period for Internet is over. Soon you will have the privilege of paying for internet services that have been free (up to this point). ABC News reports: Is the free ride over? Starting April 1, 2002, Yahoo will be charging for email access from the POP interface. That means those who have been using Outlook, Eudora, or any other POP mail clients will either have to migrate to using the Web-Email interface or cough up the amount required by Yahoo. Furthermore, Yahoo has decided to SPAM your mailboxes by setting your option defaults to Opt-In. Because you have been such a devoted user of Yahoo.... they will now use that fact to send you lots of advertisements about things you don't want or need.
With Hotmail, Microsoft has already moved in that direction long ago. If you have a hotmail account, you will find yourself getting 10 to 15 ** DO YOU WANT SEX or ***HOW TO BE A BILLIONAIRE OVERNIGHT messages a day. Within 2 days, you get a message: *** YOU ARE OVER YOUR ALLOTTED DISK-SPACE, sign up with us for more SPAM space or be kicked out. Frustrated at yourself for putting up with this everytime you log-in for email and at them for doing this to you, you vow to drop the account by week's end. Then at the end of the week, you get one real message from a friend. And you decide to keep the account because your friend might send you another one.
The cycle of vicious ritual goes on.
I am taking the plunge into the Microsoft .NET framework. I hope everyone likes fashion because .NET is Microsoft's latest delivery of technology fashion.
It is indeed a revamp of how programmers do things. The plethora of new languages: C#, VB.NET, C++.NET, JScript is basically an affirmation of Object Oriented programming. Ok, VB'ers, kiss your "goto" days goodbye and good riddance. C#, well, finally C++ programmers can do internet programming without grimacing with disgust by being forced to resort to VB or a scripting language. Then there is MFC being reinvented once again.... this time to look alot like OWL, which Microsoft happen to kill about 8 years ago. The revitalized OWL is now called Forms (as in WebForms or WindowsForms).
Before you jump up and start praising "Microsoft, ... microsoft", understand that all these benefits come at a cost. Just like abstractions cost, .NET is a major platform abstraction that will cost the user. What these costs will be won't be apparent until .NET is fully cooked but an initial point of view is that performance of .NET applications will suffer slightly due to the JIT (Just-in-Time compiler) overhead.
Here is an article for those of you in the software producing business or profession. It's based on a MAC shareware producer but I'm sure you can abstract and generalize the experience.