Friday, March 28, 2003

Opinions or FUD

The most abused word in the industry these days. What is FUD? What is an opinion? Are we still entitled to our own opinions, or will we be accused of spreading FUD everytime some other person (whose fears, uncertainties and doubts may be already present and latent) disagrees? Interesting question, if you ask me - and I don't think I am the only one. Nevertheless, this BEA article may just be a little bit over the top - I would never accept all of the points that Dietzen's is making - but this one relates to our earlier blogversations:

In comparing these open-source projects to WebLogic (read: any CSS vendor's tools), you may find yourself thinking that even if they don't provide all of J2EE or WebLogic's features, at least they're free. However, the cost of the application platform is usually just a tiny proportion of the TCO for an enterprise software project. The majority of the TCO comes from managing and maintaining the application once it has been deployed, followed by the cost of developing the application, and then more distantly by the cost of the application platform software itself. Many of the critical features needed to reduce TCO are not available in open-source projects, so free often turns out to be more expensive than it seems.

Axis of Just As Evil

SatireWire apparently wrote this really funny story. Love it! Strange that somehow it got attributed to John Cleese - it was NOT written by him ...
Posted on Friday, March 28, 2003, , Permalink #

Thursday, March 27, 2003

Who cares if you're smart? I don't.

Marc has returned from outer space, re-entered blogspace. And immediately a post about posts to answer the remaining question: "How to discriminate between smart and 'average' ?". Is that really the question? It wasn't the heart of the matter for me. I really don't think that IQ- or any other Q really matters, nor do I really care. The end result is what matters to me: OSS being for the happy few, for a very tiny minority of developers that somehow are able to dig into the open source, grasp what is going on, and work on their own solutions from there. I am absolutely convinced that most OSS technology (in its current form) doesn't lend itself very well to deployment beyond this group. It is just "not fluffy enough".
Besides that, I think it is HIGH TIME we had another quarterly meeting ;-) ...
Posted on Thursday, March 27, 2003, , Permalink #

Americans at war

Never knew this: the American Armed Forces have outsourced some of their core capabilities (in earlier wars in Croatia etc, but also now in Iraq) to companies like MPRI. Nothing to be overly surprised about: mercenaries have ALWAYS existed in any type of warfare, professional armies have only existed for a century or two ... But as I learned from one of my closest friends yesterday, the really strange thing about it is that their legal positions according to international law are so disputable. Nevertheless, for those of you interested: they are still hiring ...
Posted on Thursday, March 27, 2003, , Permalink #

Wednesday, March 26, 2003

The heart of the matter

Is this going to be another day of blogging discussions? Don't think so - most of us do not have the time, and I think that most of the points have been made. I will try to convert Steven over lunch (yeah right!), but wanted to touch upon one final remark by Werner:

you can not compare a level A1 computer graduate with a guy who has 2 engineering degrees from 2 different universities because you don't compare nuts to apples do you

Tuesday, March 25, 2003

Discussing permalinks

By popular request, added permalinks to my blog - so we can discuss the Meaning of Liff till we drop ;-) ...
Posted on Tuesday, March 25, 2003, , Permalink #

Devil's advocate

Ok, let's push the argument a little further. I am going to play Devil's Advocate in this post, making an argument that I KNOW to be too extreme. Let's talk about the dramatic situation where either your CSS vendor ceases to exist, your CSS product gets cancelled, or your OSS project looses key people, looses focus, or whatever. The OSS/CSS product development grinds to a halt.

Let's now read this excerpt from Andy's post:

OpenSource gives you an "assured fix". You can take it over and fix it yourself. There is no risk there.

Flawed arguments

Without knowing it, Werner actually gave me some more arguments to doubt the viability of OSS. I should add them to the list ;-) ...

• Is their support staff manned with expensice experienced developers or with cheap juniors ?: Ok, so how is OSS different here? How can you guarantee that the people "supporting" OSS have the experience that you need. Let's take the JBoss example. I believe (don't know the details) that there is a sort of "distributed support program" available there: a network of partner companies that offer JBoss support at a price. It seems to me like it will be VERY difficult to do quality assurance on a distributed, loosely coupled set of companies that may no longer be there next year, than in a setup where a company (yeah, yeah, like Novell) has a limited amount of official support centers. The JBoss model may work for now (for sure as long as Werner feels like it), but it may not work forever and it certainly will not scale. I think.


• product lifetime for a closed source appliaction is not necessary longer than for an opensource project ...Apche for example ? Apache is a good counterexample. Some parts of Apache have indeed matured and survived - most likely because of the LARGE contributions by CSS vendors. But what about the numerous other projects that simply implode as soon as the "natural born leader" goes off and does something else? I am not a specialist myself, but I have definitely heard war stories about this type of stuff.


• version upgrades and migration paths : hahaha, come one let's be serious, the migration paths for IBM WebSphere are not always paved with success .. been there to prove it. Every single vendor in the Java space that has been there for more than 2-3 years has had MAJOR upgrade issues. Many of the vendors (Sun and IBM for sure) have litterally thrown away their codebases and started again from scratch to build their J2EE containers. And I seem to remember that the same goed for JBoss (the 2.x->3.x migration was horrific, was it not?). So the question is: where does OSS offer improvements here? I can only see real-life negative impacts about the fact that the organic group of hackers does not have a vested interest in providing migration tools and utilities to go from one version to the next. Maybe CSS vendors have a bad reputation here, but CERTAINLY the risk of the same thing happening in OSS is greater? Or am I missing something here?


• Can't find the staff to support OSS developments ?? give me a call ! I have NO doubt whatsoever that Werner's team is VERY capable and offers tremendous value for his clients. But how many of these "shiftatlookalikes" are there on the market, not to mention on the local Belgian market? Or even worse: what is the POTENTIAL number of companies that could handle the type of stuff that Werner is actually doing? In my experience, OSS "products" are typically more difficult to use than CSS products. The initial rampup in intellectual capital is far greater for OSS than for CSS - CSS products typically have these protective "shells" around them that make life easier for the average coder. The above-average coder HATES THESE SHELLS, but is it not possible that average coders cannot do without them???


• Is your staff not qualified enough and will they never be ? what makes you think they will be qualified enough to support the applications on a closed source product ? Because a CSS vendor knows that - besides developing the actual product, he will need to make these investments in his market. Steven already noted this, but draws the wrong conclusions. CSS vendors often offer free training initiatives to partners of which they think that they can bring them business. CSS Vendors understand that products need resources/partners/integrators to be succesful, and I am very convinced that it is easier to track those people in the CSS world than in the OSS world. This is of course reflected in the final price of the product - but it is very normal that it happens like this.


• compliance : well how can you be sure when you use a close source version ? by knowing the API's yourself and then coding a test application that you can use to test, check out the demo's ... hey, you might even have a look at the code because it's opensource :-)): this is the only argument that I actually like. The current J2EE compliance testing needs a lot of work. And I do not see why companies could not disclose their sources more for this very reason. Especially in the Java world, it is pretty pointless to "hide" sources from your customers. But that is still very different from the OSS model: OSS says that sources should be free (as in fries AND as in speech). In my opinion, this is not very realistic.

Subjective Decisions

I love blogging!!! This is a great discussion going on between myself, Werner, Marc and Steven... A couple of points to make here:

• I really did not want to define the war in Iraq as being a religious one. Nor did I want to imply that the OSS vs. CSS discussion one is a religious one. No. Religious wars typically have very little or no arguments to support or refute them. You either believe in it, or you don't. In both cases (Iraq and xSS), there are plenty of rational arguments for completely binary opposed views.


• My real point wasn't about supporting or not supporting either of the viewpoints, but about the difficulty of choosing, and the uncertainty of knowing what is right or wrong. That's were I have the most problems with. Those of you that know me a little know that I am a very liberal person - live and let live please - but I am also quite fond of rationalising different points of view. I like to know what I choose, and THAT's where the trouble is.


• As for the OSS issues, what I tried to suggest in the last post was the fact that I think that there are very valid counter-arguments for using OSS in Enterprise environments. Werner tried to argument against my "fud" but it is my personal point of view that - because there are so many pros AND cons, it is safer to choose CSS in mission-critical environments. Note that I specifically mentioned that I would consider OSS in other cases.


• Da FUD. I don't really see why - just because I was making a couple of points about general OSS opinions that many people can relate to - this should be considered to be FUD. It's just an opinion, right. With just a much reason, I could state that Werner's comments (Is their support staff manned with expensice experienced developers or with cheap juniors ?) is just as fuddy as my statements. So here's my personal invitation to Werner: come over to BXL and visit our Novell support center here in Brussels! You can then judge for yourself how many juniors/seniors work there. Point is that my statements are NOT FUD. They are valid arguments, just as valid as the pro-OSS arguments are valid. My whole point is about the fact that there ARE two sides to the story, and that it is REALLY hard to make an informed judgement call here. My final opinion (as expressed in the last post) is nothing more than gut-feeling-guesswork. The only thing I know, is that I do not know - what will be better in the long run...


• I am a little itched by the fact that Werner (who makes a living off OSS too, right?) says that I am supporting the CSS view because I work for Novell. That's a little off the mark if you ask me - it has very little to do with my personal, private, sometimes intimate thoughts that I share on this blog. My opinions are my own, not Novell's. I do not see why I could not support the OSS rationale and work for Novell at the same time - it seems so natural to me to separate business and pleasure ;-) ...

Monday, March 24, 2003

Welcome to the real world - religious wars are never a good thing ...

Werner's World of business ethics certainly is worth a blog entry. I really do understand his point of view, but I am very convinced that it is NOT the only one.

Now I'd like to abuse Werners's blog to make a point about these religious wars, and talk about the difficulty that I myself - and I imagine many other people too - have in judging about "the right thing to do". In some way, I actually think that there are some resemblances to the War-on-Iraq discussion and the war of OSS vs. Proprietary Software. To me, both are issues where there is no black or white.
Let me start with the real, and definitely most frightening war - in Iraq.There are quite a few VERY good arguments FOR a war, and there are as many VERY good arguments against it.

Why go to war?

Why refrain from war?

Saddam is a brutal dictator, abusing his own people Saddam has links with international terrorist organisations, threatening our lives and lifestyles Saddam has no respect for the United Nations and their resolutions The US needs to secure Iraq's oil supplies in order to sustain the US (and western) way of life The US needs to secure the middle-eastern region through the introduction of a new - hopefully democratic - regime. The US needs to boost its economy through massive defence and reconstruction projects

Unilateral action is not permitted by international law - the US need to stick to the UN path! The same results can be achieved with peaceful means - more time and resources for the weapon inspectors! You cannot take action against Iraq on the grounds of its disrespect for the UN resolutions, and do nothing about the situation in Israel and Palestine. Taking action in one of both matters, without dealing with the other, could have grave consequences for the entire region. We have to put human interests before economic interest! War never solves the real issues underneath. There are never winners in a war - just losers! The attack on Iraq will only trigger new terrorist attacks The war could be perceived as a war on Islam - thereby further dividing the two major religions of the world.

My reasoning to date is simple: as a precaution, let's take the no-war-option for now. But I am prepared to revise my position later: I certainly do not share the "no war" dogma - there IS such a thing as justified war ...

So let's now turn to the OSS vs. Proprietary SW-war. As with the war, there are lots of arguments FOR OSS, but there are just as many counterarguments.

Why use OSS?	Why not use OSS?
It is FREE as in Fries! It is FREE as in Speech I will have more money to train my staff I will never have to worry about vendor-lock-in I will never have to worry about a vendor's long-term viability It is great-quality software, very innovative in many cases!	Ok, it's free, but where do I get support? Ok, and how long will the product be supported? What about version upgrades and migration paths I can't find the staff to support OSS developments I found the staff, but I think they are not qualified enough, and I doubt that they will ever be How do I know that this OSS software is as compliant with the open standards that are on the market?

It's not a black and white thing, and there ARE good reasons to refuse to work with OSS. I know many of my clients do this. As far as I am concerned, the decision between the acceptation or rejection of OSS is a very subjective one. A question between believers and non believers. It's not because some (sales)people reject the idea of OSS in production environments that it cannot survive... As for my personal views: for mission-critical applications I would most likely prefer a proprietary vendor's infrastructure products. For other, more simple or short-term-oriented apps, I would not see why we could possibly object to OSS in production environments.
Posted on Monday, March 24, 2003, , Permalink #

Belgian elections: VLD programme

One of the first "real" election programme's that I have read to date for the May 18th elections here in Belgium, and it's an interesting one to start off with: our prime minister, Guy Verhofstadt, has written an Open Letter to every citizen with tangible results of the past and objectives for the future. Adminitedly a little sloganesque, but interesting nevertheless ...
Posted on Monday, March 24, 2003, , Permalink #