Monday, December 9, 2019

Language beyond C++

C/C++ is among the oldest languages, and still widely used. Many competitor languages came, to overcome drawbacks of Cpp, but none succeeded as much as Cpp. Those trying to fix Cpp, forgot to understand why C/Cpp stood the test of time from grandpa's generation.
It has a strong & great legacy. Most contemporary OS'es are written in C. So is the middleware for most hardware platforms. All this cannot be transitioned, by anyone, by whatever grand plan. So whatever new programmers say, C shall live forever.
But how did C++ overtake C? by being a superset, that only extended C with features. So also very constrained AnsiC, gave way to classes, C++, STL, C++98, Boost, C++11, and ongoing.
How about adding security, by removing error-prone constructs from historical C++. The approaches to create new languages were doomed from the start and could not overtake C++. So we need a construct that can sit within C/C++, to be easily absorbed not just by its programmers and also the existing code.
To this end, we can create a C+- language specification, which uses a feature.cfg file to define the features supported/rejected by the compiler. It'll have a line for each major standalone feature, each marked as unknown/supported/deprecated, where compiler must accept but generate warnings on deprecated features. C lang would just disable the classes feature. C++11.cfg file will define the entire C++11 standard, incl auto type feature. Disabling pointers feature will restrict directly accessing pointers outside the standard STL libraries.

Tuesday, July 2, 2019

Submerged Mumbai & Dry Maharashtra

Submerged Mumbai & Dry Maharashtra

India lies under Monsoon umbrella. Every year, Indian Ocean sends rain laden clouds over India. They travel till the mountains, and give heavy rains, while a dry shadow lays across the mountains.

The journey of these clouds used to be till Himalaya, but recent years has seen it partially abrupt at Western ghats, leaving a contrast of dry spells in most of Maharashtra while regions around Mumbai get submerged.

What is causing this change. High Urbanization & Industrialization is causing 'Cloud Seeding' that is intercepting Monsoon clouds beyond normal. Pollutants from various sources act as chemical nuclie that condense moisture around it, forming heavy clouds that can no longer fly over the low West Ghats and pour heavily on the Coast itself. In the past, when our coastal air was not so polluted, these clouds used to be fly high over the Ghats, bringing uniform rains across the country.

This climate change affect is not only causing submerged coastal cities, but also dry monsoons over rest of entire Country. To curb this changing climatic affects, we must restrict industrial pollution and vehicular emissions, for atleast 30 day period from onset of Monsoon, in Mumbai and surrounding region West of the Ghats. During the monsoon period, Polluting Industries must be shut, and carbon emitting vehicles must be controlled with zonal restrictions like odd-even.

Saturday, March 2, 2019

Political Alliances: To, or Not To

The author recommends to Form alliances, and addresses with who, how, and how much to give in alliance forming. 

In a vast and diverse democracy like India, alliances form an integral part of the election campaign. With a diverse electoral base, there are bound to be multitudes of opinions, and one set of opinions across policy fields cannot cater to all. The dominant and popular opinion in each policy field evolves as a result of the election process, amidst give and take of alliances.

In a national common electorate with proportional representation, each opinion is considered in this post-election alliance forming, after component parties have proven their strengths in the election.

In first-past-the-post elections, like India's local constituencies, or US's state representatives, elections in each constituency becomes bipolar (though technically multipolar, voters equate voting for 3rd and lower candidates as a NOTA or vote waste), and alliance formations need to be done before elections. Therein forming Alliances becomes a game theory play.

Game theory predisposes towards those with better information, giving rise to effective election analysts and backroom poll strategists in India. US system has evolved where both leading parties conduct primaries to measure backing for opinions and form alliances before the final election. Republican party which was then week, allying with and absorbing tea-party candidates was one such example.

In India, with so many small constituencies influenced by local effects, it's impossible to accurately measure voter pulse, and alliance formations depend on successful negotiations in a game theory. Against a strong candidate or party, it is important for other parties or candidates to form mutual benefit alliances. However, when both parties are optimistic about their prospects, compromises can't be achieved. Even if one is over optimistic, reasonable benefit cannot be provided to the other to form the alliance. It is this that I like to discuss and address.

The first aspect is who alliances can be formed with. Everyone. Nobody is untouchable in political democracy. Though alliances may be termed as opportunistic alliances of convenience to retain power, they are real alliances with those closest to your ideological and opinion spectrum on most important policies. Such alliances, it must be noted and understood, are objective compromises to meet end goals of least bad and most favored option, when public opinion is divided in free society. Two such parties are never enemies of each other, as they both represent people of the same constituencies. Nor is it ever a full policy alignment, otherwise these parties would not be distinct.

Consider unity of all parties against INC's authoritarian rule and emergency in 1970s. And then 1990s considerations of centrist INC allying with rightist BJP to keep left and regional groups at bay and provide stable central government. And then far leftist Communist parties supporting centrist INC in 2004 to keep rightist BJP out. Such alliance formations based on social economic parameters are much more overt and publicly discussed in other countries.

The second aspect of resolving the alliance arithmetic is information, and collective gathering so agreement can be had based on it. This reduces the game play between partners, so they do not sabotage each other to assert each other in alliances thus formed. Consider breakup of BJP-SS, INC-NCP alliances before 2014 Maharashtra elections, and their uneasy alliance for next 5 years. Having shared surveys and voter pulse gathering would at least remove disagreement about facts and vote base estimate about each other.

The third and most important is how and what alliance arithmetic to form between similar yet different parties, that best serves to further their ideologies, instead of risking sabotaging each other towards the opposite ideology. Trying to address the what arithmetic first is prone to disagreements and show of strengths, resulting in sabotaging each other and breakdown of negotiations. I am of the strong opinion that how part should be addressed first, and the what arithmetic easily follows.

How to form alliance. India is a parliamentary government. It is not one election, but 543 local elections, formed of different groups of people, with different priorities. Priorities of drought prone Tamil Nadu, Jharkhand villages, Naxal affected Chattisgarh, and urban Mumbai, is and will be different, and alliance question in each should be treated on its own merits. So alliances should be formed regional and sub-regional, or 1-on-1 give in one constituency and take in another.

Treating each constituency independently, helps filter out the constituencies where either party is already confident of winning and does not want to compromise. Similarly, in multiparty alliances, constituencies where two parties are collectively comfortable need not include a third partner. This boils down to the constituencies where you are independently not sure and alliances can help prospects. Left over constituencies are subject to negotiations and game play. Here, alliance must be seen as a means to further your position and reject the opposing position. Where not competitive, not fighting and sitting out, is better than fighting and cutting votes of a partially aligned opinion in favor of a fully opposite position.

In this game of open seats, each looses if they don't ally. Collectively gain if ally. Individual gain is subject to bargaining. As both share equal benefit, rationally, both benefit from alliance, and it does not matter who has how much individual influence. But unlike BJP-INC-JDS trilogy, where numbers are known and dependency is forced, before an election, the exact numbers are not yet known, and optimistic/pessimistic possibilities exist. Because each is trying to achieve highest votes in first-past-the-post, the best normative recommendation is to share alliance seats in each sub-region according to their proportional vote share.

When voter opinion is much more uncertain, another approach is to divide some seats and contest some. Example, two parties A and B, can divide one seat each in an alliance and contest on the third. Optimistically, they had gain 2, but pessimistically they'd still gain 1 and not sabotage on the 3rd.

Hope i had see an election where our parties push for themselves rather than denouncing the other. After all, no one knows the outcome, and there are no untouchables in politics.
Behind Jumla govt
BJP..govt..HQ

Behind the jumlas of this govt are policies allowing various opportunities of corruption, that formed the huge building.

Selective beneficiaries means corruption in selection.
Launching of Medical care policy requiring lakhs of crores with just 10,000 crores.
Funding cuts to various schemes while still retaining the schemes.
Demonetisation was a classic case of what govts in India can do, taking away all liberties, for filling their coffers.
Poor Reservation is another such instance. With drastically increased threshold for poor, nearly all of Indians now fall under this 800000/yr threshold, and get only 10% quota. If all were really given this benefit, it would be equivalent to general quota and no different. If magistrate certificates for poor fall short like new notes during demonetisation, then this quota is a guise for reserving 10% for BJP workers and bribes from sponsors.

These are all instances where there will be huge bribes given to be part of the select beneficiary list, or benefits accrued to those ruling party workers, or worked from their affiliated organisations. Some of the Youth, instead of progressive career plans in building economy or nation, are taking advices on how to enter the core of ruling party, like the crave for govt jobs few decades back.

There is development too, unregulating some chores, and reducing discretionary powers of bureaucrats. Combination of few administrative cleansing along with launching of corruption enabling policies, is like, the ruling govt does not want competition in corruption.