Linux Unless It Can't Be - Novell (R)
I was just reading through a whitepaper written by Novell(R), entitled "The Novell Migration to Linux : Real Business Value". (Sorry, I can't find where I downloaded it, it's been sitting on my desktop waiting to be read for a better half of 1-3 months)

I would say that we're more or less understood where Novell(R) was heading before they started to acquire SUSE Linux and Ximian. It was like a cry for help that drove them to re-invent themselves from being a largely proprietary company with proprietary tools running on it's Netware(c) platform to a company that runs Linux.

Hence their motto - "Linux Unless It Can't Be"

The whitepaper details quite a lot of numbers and facts in which they state their case for going down the Linux/Open Source Path. Though one of the numbers they state is the cost savings in moving from Microsoft Office(c) to OpenOffice.org (approximately USD$750,000 per year), i would have to say that cost shouldn't be the only reason for the switch. Unfortunately, the mere mention of using Linux or any Open Source Tools will undoubtedly bring forth this as one of the main arguments for going Open Source.

There are a whole set of reasons to not only cite that because without a clear business case/path, the initial push based on the $$ amount will only end in regret. Such was the case of "Peng Hong Hardware" (which was highlighted in the April edition of Asia Computer Weekly) which went the Linux/Open Source Path only to turn back to proprietary systems in a short 10 months.

I'm not saying that Linux isn't ready, it's actually the businesses and upper management which fails to see the bigger picture when deploying Linux/Open Source. It shouldn't be regarded as something which can turn businesses over in just 10 short months. In addition to that, Linux shouldn't be the path to head when there just isn't employees which are Linux/Open Source competent. (A reason they cited for the abrupt "U" turn was due to a lack of expertise when dealing with Linux. Quoting from ACW's article :

Peng Hong Hardware managing director T H Thong realised a switch back to Windows was necessary when he had to call in third party support to handle simple tasks such as adding new e-mail user accounts and new printers.

I can't say I don't understand this because, for the people on the street, Linux may just be like an alien from Mars. There just isn't enough exposure.)

Switching to Linux/Open Source also must not be done cold turkey. Careful planning and migration in stages also must be done.

In the Novell(R) whitepaper, they stated 2 main criterias (and I quote) :

1. Reduce licensing costs (mostly in the form of Microsoft licenses)

2. Dont break the business

While item #1 seems to be the (ultimate) goal, #2 actually should be the priority in any and all migration task.

The company stated above, I believe actually broke rule #2(due to the lure of #1) and hence had to revert back to using proprietary systems. I can't fully speak for that particular company because I know nothing of their migration plan nor their in-house expertise in performing the migration.

Which brings me towards the second issue which I see to be a cause of failed migration.

Lack of Expertise

What's one reason for this? Lack of Exposure? (Yes!) Lack of Education (Yes!) Lack of Interest (Not necessarily true.) Let's tackle this one by one.


1. Lack of Exposure

When I speak of lack of exposure, what do I really mean? Do I mean people on the street don't know what is Linux/Open Source? Do I mean, IT Managers and upper management of companies have not heard of Linux? Do I mean Linux isn't mainstream enough?

Actually all of the above encompasses what I meant. Even though the level of competency or literacy in computing is high among Malaysians, this does not necessarily mean that they know or understand the meaning of Open Source and what it stands for. And even if they do understand it, the other question would be would they want to try it out and why would they?

It's typical for upper management to look at the bottom line; and that should always be the case when it comes to any company. They do have to take shareholders (if any) interest at heart don't they? And I don't blame them for it.

What I do blame is the misconception that Linux/Open Source is the "Saviour". It really isn't, at least not until these companies knock some sense into whoever that is suggesting it is without doing the necessary homework and the necessary feasibility study into whether they have the expertise "in-house" to be able to undertake this task or if they're willing to enlist the help of outside vendors and do some retraining. Gargantuan it might not be, but if there isn't any expertise "in-house", you can bet it's leaning on the edge of stupidity.

The TCO for the software licensing was a no-brainer, but the hidden cost of development for new applications, migration, training, maintenance, and other standard operating requirements eroded the cost savings a little bit. - Joshua Tong, Director of Technology at Zuji - ACW March Issue

Note : He mentioned "A Bit" and based on the article at ACW, the final cost savings even after taking all those into account was still significant.

Novell(R) is one of these companies which understands it based on their whitepaper. No doubt they are much better off than some of the other companies which attempts to migrate because they acquired SUSE Linux and Ximian. They have all of those expertise "in-house" if not a phone call/email away.


2. Lack of Education

Another possible reason for this state of affairs which brought Peng Hong Hardware back to using proprietary systems is (as cited by them) a lack of expertise and the need to call for outside vendors. Again, this is true, most companies that look towards lowering of TCO via software licenses will be surprised that Linux only pays off when used for a couple of years or when there is adequate support "in-house" for Linux/Open Source Solutions. The latter is most certainly true for the the case of Peng Hong Hardware.

But the question remains, why are there so few persons who are experienced in Linux/Open Source? If you look/ask around, you'll find that most persons who are experienced in Linux/Open Source are so based on their own initiatives. They found that Linux/Open Source interests them and they embarked on a journey to learn it. That's how I did it and that's how most of the Malaysian Open Source Community learned it. Sure, there were some amount of hand-holding and knowledge networking, but obviously not enough. (I have to also state that while hand-holding is a good thing in some regard, experience is always the best teacher, esp. when one has to shift through a lot of FREE documentation to find out an answer to a specific question.)

Now, if this is the case and > 90% of all Open Source persons are learning on their own, doesn't this mean that's something wrong with the education system? Tiertary education should (remember this, I said "should") be the place where this is all taught or exposed to. Students at higher education is where the focus should be. They are the next in line for "taking" over the country aren't they?

Honestly, I find it appalling that the very place which is supposed (again, supposed) to be the repository for free(?) knowledge turns to to actually alienate Open Source/Linux in the first place. There is an over reliance or over-dependence on proprietary solutions because of a few factors.

i) Uni/Inst. are Cash Strapped and relies on the generosity off companies which provide them with their proprietary tools.

ii) Lecturers themselves aren't well versed (if at all) in Open Source/Linux hence, how can the be expected to be able to educate those yearning to learn?

While (i) can be achieved by using Free Tools/Open Source/Linux, it doesn't help if (ii) is the obstacle.

Yes, Linux/Open Source can no doubt help lower TCO, but there's a catch, it can only be when there is already a framework for educating the persons necessary to support these companies/industries.

On another note, I was at a dinner yesterday and sat next to a lecturer from one of the local Universities and he was commenting on the rate of decline of graduates(at some of the Universities/Inst of higher learning). He also notes that by far, some of the most sought after graduates are those coming from Universities which has taken to teaching the subjects in English rather than in Bahasa Malaysia. (which is the National Language of Malaysia) If you do recall, proficiency in English(or lack of) is one of the main reason for the high number of jobless graduates.


3. Lack of Interest

At the current juncture in most IT Person's lives, they are mainly interested in a few things, money being one of them. (I can't say I'm don't have an affinity towards it though)

When it comes to getting experienced persons in Open Source/Linux, one of the few reasons for them to not be interested in Open Source/Linux is the fact that it's not widespread and there's not a lot of demand for it.

So, even though a particular person does indeed have an interest, it's often clouded by other priorities and more often than not, this interest wanes and eventually disappears altogether. The lack of a local Open Source/Linux community in Malaysia no doubt accelerate this.(lack of interest)

The more mainstream proprietary systems, on the other hand, have a larger number of "so-called" experts, (I term them so-called because how else can one differentiate between an expert and a normal user when almost everyone is an expert?) and the interest is fueled by the natural desire to earn more money.

So, a lack of interest isn't the cause, it's more towards, what creates the interest. As of right now, and again, without the necessary framework in place at the education level, all this is futile, Linux/Open Source isn't gonna go anywhere.

So what if the Malaysian government pushes for Open Source? It ain't gonna happen in a big way till they get the local Universities/Institutes of Higher Learning into account.

That's my 2 cents. But bravo to Novell(R). (but I've yet to find out how much exactly is the local Novell(R) office has progressed in this respect)