Crazy Days

FOSS, where are we...

FOSS adoption has been in my priority list for a long time, most probably since the day I joined Carson. From a academia, to a big corporate, was some what drastic change to an inexperienced personality like mine. Administering Linux (and only Linux) servers and designing networks to hearts content was the luxury I missed after joining Carson, Sometimes I feel that was one of my stupidest mistakes that I would regret my whole life. Pity is that, that is not the only one.

For the longest time I tried pushing for FOSS in to Carson in many ways, which I did not consider very successful. One reason would've been the local culture and Carson being a typical, some what bureaucratic organization, which I would consider INERT. In Sri Lanka most of the companies can be considered inert, which explains why after three years of service I'm a good fire fighter, which I believe explain most my working background.

I never had faith in formal education, thanx to my parents I ended up as a graduate and I always thought that (still I think some times) starting an MBA was the next biggest mistake to joining Carson in this millennium. But recently I realize how ones experience and education affects the thinking behaviour, (most probably my meme pool is getting reshaped) because now I tend to see things in a different way compared to three yeas back. That might be what people call maturity, which different people tend to see in different ways.

[End of MPHOT]

FOSS adoption is my latest obsession (time to time these obsessions change). Until recently I was under the impression that adopting FOSS is just a matter of single step. Which in other terms “a theme park ride”. But after selecting the FOSS adoption as the final year research topic for my MBA, I was forced to look at it from a broader perspective (may be with a PEST). Sleeping with the concept for many nights I slowly started realizing the bitter truth of adopting FOSS, in a typical corporate environment, which lead me to write this crap.

Best place to start would be from the magnificent company called M$. which at one point managed to own 85+ % market share for PC operating systems. Almost all the corporates embraced the “windows” to look in to their companies future. I think that was the perfect timing, a perfect economic respond, people had a need....somebody replied to that need. Result was some guy from down town was named the richest man in the world and created a giant unbeatable by any corporate in the world. This journey wouldn't have been so easy as I say but in a summary that is what happened.

Some one other than M$ said same thing in a similar presentation, in marketing (manufacturing) sense, the best place to be (from a economic perspective) is in a monopoly situation. If you have 95% market share of a product stream I see why not someone should think like that. It was the case until Netscape filed the antitrust case against M$, which ruled in favour of Netscape (NS), which ultimately lead NS to open up the source code for their NS Navigator (and some other things too). This is the most significant incident out of series of unpleasant harassments to software industry.

During all these silly events happening in Redmond, few weird looking geeks, who forgot to shave in the morning were continuing another crusade (crusade of 20th century, crusade for holy (free) software), and FOSS came to attention. So the story goes like, few Sri Lankan geeks followed these geeks in elsewhere and supported the crusade, and I happened to be one of them....but still I shave in the morning (at least once a week).

[FOSS Adoption]

I love saying this....I would like to accept any level of survival. Likewise there are several levels of FOSS Adoption we can think of. In business automation analogy, totally automated to brick and mortar situation applies to FOSS adoption as well. If you're confused, in simple terms, full FOSS adoption or the partial FOSS adoption would be the common options. This is in realistic terms, a grey scale in between full adoption and no adoption. That is how simple it is.

Partial FOSS adoption is a better entry point for anyone who is willing to move into FOSS. Partial adoption can be defined in many ways. The most successful adoption is/was, migrating back end services to FOSS and keep front-ends proprietary (that is Windows/M$). For most companies this was proven successful and many organizations already finding this as a good way of getting down the TCO for most of the services.

Emails, web servers, database servers and etc. are good example of similar applications. Key success factor in these applications are that, they're transparent to end users. Some times not even the higher management is knowing what the technology behind the services are, generally they don't need to know. In my belief technical decisions should made by technical people, not the management people, it is up to management to decide the solutions are beneficial (enough) to the business. This is a utopian dream I've been having.

Other side of the coin of FOSS adoption is the partial adoption involving client ends. A classic example would be replacing Open Office with, MS Office. Which sounds like a perfect plan, with few hiccups. First thing, though Open Office supports MS Office documents (DOCS), DOCS are not a open format. I don't know how this support got included, I assume this has some thing similar to samba situation in the early years (reverse engineer – which some people think unethical). Because of this situation it arises the problem that DOCS don't render properly in Open Office (OOo), at least most of the complex documents. Specially some one like my self trying a funny ActiveX crap in a DOC, would lead some one whose trying to view in OOo to real trouble. With some difficulty OOo will have the capability of running them under windows, but elsewhere, I rather not comment.

Therefore innocent (_asumption_) corporates like Carson, or any in that matter, would have to go through a bitter process, which we call a document conversion project. This was not something new. The same situation was experience some time ago when people switched to MS Word from Word Perfect. I would consider this as a classic example of COST of CHANGE. As I mentioned earlier in one of my blogs, “cost of change would be much higher than the loss of using”. Larger the amount of data painful the process is.

In order to conduct a successful document conversion project, we need to have a conversion path. The worst case scenario would be non of the software support each others formats properly.

The way I see there can be two solutions to this scenario. One is that either M$ to come to open grounds with their DOC format, which I don't think will happen even in another several months of Sundays, or M$ to support/adopt ODF format. To amuse me, recent post by M$ published that they're going to provide a Plugin to support Open Office Document (ODF) files. At the same time killing the enthusiasm, this is going to happen in 2007, in their next release. Assuming that this is not going to be a half cooked pie, like in many cases, one can keep hope of a possibility of a document conversion. (the cost of this scenario will be discussed later), in case of a migration.

There is a buzz about the whole scenario, some parties willing to provide plug-ins for existing versions of MS Word, which will decide the conversion paths in different ways, yet we need to wait and see at the outcomes.

This the distance you need to travel to get a simple application to run on a PC. These scenarios would be more complex, when number of users and amounts of data generated are large. The way I believe these migrations would have a cost involved, but is it higher than the price we pay for Proprietary Software? Or at least M$ will make sure that is the case. This is some thing I'm waiting to find out.

This is just a single factor of the FOSS adoption, more scenarios will be discussed later.......

All about change....

Just a thought, why we stick to a single application to perform the same task. Why not organizations shift application the way I do? That is one of the most interestng influences I had from FOSS, earlier I used Mozilla and later conviniently switched to FireFox, and from Gnome -> KDE to IceWM. I'm sure in a singular context it is possible to act like that but in a business environment, things aren't happening that way.

Change is the most feared factor in common organizations today (atleast in Sri Lanka). Managing change is given the highest priority in most companies (including where I work for the moment) and, in auditors perspective unable to manage change is a bigger crime than leaving a password blank. For the longest time I've been trying to figure out this equation and finally it started to make sense.

In a business perspective "Cost of change would be much higher than the loss of using". That is why selecting an ERP system is a life time decision, for most of the companies. In a simple example, if you implement a buggy ERP system and used it for several years, scrapping it might not be a feasible option. It doesn't have to be expensive software. In simple terms that could be one of the reasons why FOSS has a cost involve.

Now I'm started thinking in the direction of FOSS, could this be one reasons why organizations are hesitant to move in to FOSS based solution? Known facts, could support be a problem? nope...we have enough vendors operating in Sri Lanka in my belief. Could awareness be a problem? impossible....our guys make the biggest noice possible...even biger than M$ noice in these parts of the world...anyway M$ is only loud FUD, most of the times. Then what could it be? Any comment accepted...