A plan to set aside places at some of India’s best-known professional colleges for low-caste Indians has bitterly divided the country - BBC News
Archive for May, 2006
Anyone who remembers the 70s and 80s two-wheeler landscape across India, would recollect all those Lambrettas and Vijay Pushpaks that tot-totted the roads alongside dug-dugging Chetaks made by folks at Bajaj. Ofcourse there were the Vespas as well, but to me, they were all yuck and clunky.
Anyways, the point is that I never thought these two-wheelers to be cool(who did!?), until I saw an old blue/white lamby parked by one of the grachts in Amsterdam. For a moment, the restored piece of junk ‘actually’ looked cool. I brushed it off as nostalgia (maybe its me), until I saw a full coloured billboard with some hunk pitching a premium clothing line, sitting on one of these (so it wasn’t just me). The advert seemed a bit surreal, but hey, the fact is that the tot-tot is a cool thing to own.
And, I am not kidding, there seem to be many ‘whites’ picking up ‘imported from India’ lambys for anywhere around 1000-2000 bucks in this part of the world (NL/DE/UK). This makes me wonder, if I should pick one before they really become cool and expensive in India? I am thinking.thinking.thinking…..then again, there’s no harm in shipping out Indian metal scrap and junk to the west eh, they send us their ships, and we send them our scooters ![]()

How do you determine the cost of developing and maintaining an OSS platform versus buying a licensed product and customising it?
If I as a client buy a product from a vendor, say Oracle, and want to customise it, I give it to a Oracle partner company (service provider) who would build upon the application. Now, during the development and subsequent maintenance phases, Oracle would constantly be patching the application, and further more, when the service provider faces a problem, Oracle will provide support for free (well, I as a client shell out 20,000 USD for a sinle processor license, I pretty much have paid Oracle to support my service provider).
Now, if I contract an open source service provider, how am I to ensure that my service provider is billing me accurately for the application development and maintenance phases? Is the service provider liable for problems that are inherent to the free software that I downloaded from the Internet. One way to do this might be to ensure that my service provider logs bug fixes in two seperate ticket trackers 1. Problems solved, that belong to the free software 2. Problems solved, that arose from work done by the service provider. Now there is a third category here as well, problems that arise in the developed part that due to issues in the core (default). Well, there’s many more.
Now, here’s more. If I picked up a software of say version 1.0 and it has 40 bugs, and the only way forward for me is to ask my provider to fix them, how can I ask the provider to provide a quality guarantee if I or my service provider have no clue what’s stable and what is not in the software we are working with? One way is to ask for an RFP where you tell your provider, here’s an open source product, deploy it and maintain it, give me a blanket estimate and take care of all problems. This would be an Annual Maintenance Contract. So in effect, what I could do is ask my service provider is for a blanket AMC offer and a development per project offer. Now, do I know what buffer my provider will put in his estimates? My guess is as good as yours!
As it happens, sometimes when there are 40 bugs in version 1.0 and the development time is 6 months, and I know for a fact that version 1.1 is being released in 8 months and it pretty much resolves 30 of the 40 bugs by default, what would my decisions be?
1. If fixing 40 bugs cost more than the cost to upgrade + fix the 10 bugs, I might want to upgrade. It’s like killing two birds at the same shot
2. If fixing 40 bugs costs far less than the upgrade, then I might let go the upgrade decision
3. If fixing 40 bugs costs less than the upgrade, but I have many new future developments that need the new upgraded software’s new features, I might want to analyse the situation further based on what my future needs might be and what development hours I am saving.
Well, this list can add on, but you get the drift.
Now, here’s the catch. Does upgrading mean no bugs? Ofcourse not, when my provider upgrades, I cannot expect new bugs to not crop up. Because of the point I discussed earlier on. The software in open source, so I still face the same issues. The upgrade would have solved 30 bugs I had, but it never meant that the software is bug-free. It now has bugs inherent to its new release.
So, if open source is so buggy and requires ongoing maintenance and fixes, how is it justified?
This depends. In my opinion, if the cost of maintaining the system with an annual cost that includes fixig bugs etc, is less than the license fees of a commercial product, the the open source product is better off. This is a very naive statement, but it isn’t totally off the mark. Think about it.
Entries (RSS)