devnull3

I think an in-between approach could be:

1. Identify top-N tenants

2. Separate the DB for these tenants

The top-N could be based on mix of IOPS, importance (revenue wise), etc.

The data model should be designed in such a way that from rows pertaining to each tenant can be extracted.

To be fair, if we know something will happen with high certainty then its not much of a prediction.

The fact that no one really knows how LLMS will pan-out means every projection in future is a prediction.

It will still be much lesser than perma Cambrian explosion in js frameworks.

Infact, a lot of the patterns in the likes of HTMX will be standardised.

On thing about datastar vs htmx: Datastar embraces htmx's OOB swap and makes it a first class concept.

HTMX's OOB swap appears to be an after thought.

I don't know V but mostly what I have read about V is that it is vaporware [1]. Is there truth to this impression?

[1] https://www.google.com/search?client=firefox-b-d&channel=ent...

mentioned in the article at [1]

[1] https://thephd.dev/c2y-the-defer-technical-specification-its...

New and non-plagiarized text is essential for all the AI models to train on. Without these the AI models will reach the saturation point or worse the output will not be as effective as it can be.

If AI trains on generated text it is equivalent of an incestuous relationship and I don't have elaborate on analogy further.

This assertion needs a lawsuit and deep pockets. Most startups will not be able to do this.

Infact the patent holder can file a lawsuit against a startup. Now even if the startup can prove that the patent is invalid, it has to spend precious money on the lawsuit.

Having filed multiple patents, it does not leave a good taste in your mouth (esp software patents). You know you are bullshitting but unfortunately it needs to be done because others will do it and deny you to use the invention.

> Does this mean if I disclose an invention, that nobody can patent it after a year?

Others can. But during their application who is going to tell that there is a prior art? Patent offices only do basic checks on existing patents and not on the wider internet.

Once the others get their patent, you can appeal [1] it.

[1] I am not sure its the right word

On the other hand its published on open-std.org which adds to its credibility

This appears to be a out of season April Fool's joke.

Am I the only one who thinks the this sacrifices the readability?

Highly doubt it. It is increasingly becoming difficult to get (or renew) H1B visa.

The less said about Gandhi the better. He had an amazing PR in the age when there was no concept of PR.

The world has always been divide on: Owners and workers. This transcends communism and capitalism. Owners always wants more work out of workers. Owners unknowingly or knowingly behave as if they own the people. This was literally true when slavery was widespread.

Even in western countries we find the unwritten understanding between Steve Jobs and Eric Schmidt not to poach each other's engineers [1]. Where is a the "free hand of markets" here? Does it not limit the career opportunities for employees?

Activision Blizzard, one of the top AAA game developer had huge culture problems. There is a lawsuit as well [2]

How is it in the land of free and ideals of Martin Luther King and Abraham Lincoln get to a point where humans are treated like corporate property?

[1] https://www.theverge.com/2012/1/27/2753701/no-poach-scandal-...

[2] https://en.wikipedia.org/wiki/California_Department_of_Fair_...

At a coarser level it works in India, but not at granular level. If you are involved in a lawsuit then it can drag for years. So essentially it becomes "justice delayed is justice denied".

There are 2 types of companies in India at present:

1. Services

2. Products

The product companies's culture and environment are on-par with silicon valley and this includes a good pay as well.

Services is another matter. The fundamental issue is the business model. These companies earn revenue per person and hence the incentive is to over-staff or over-sell the people needed. There is no incentive to solve or be pro-active about customer's problems.

Not a law per-se but I am not aware of any firm which allows dual employment. Moonlighting is not allowed and is mentioned in the employment contract.

Hence the "leaving certificate" serves multiple purpose:

1. Person is not working elsewhere

2. Person has gracefully exited the employement and there not pending dues or conflicts.

In India your new employer demands "leaving certificate". This is an official letter from previous letter that you are no longer employed with them.

An existing employer can weaponize this and simply not issue this letter.

This does not happen in India but I am not sure if this the case in Japan though.

> my product should find problems in 3rd party components.

Does that mean that say a security vulnerability in openssl is a responsibility of all software which uses them? I think its unreasonable to expect software projects/products to find things like heartbleed.

What about bugs in kernel/OS? How many user-mode software can find bugs in kernel/OS?

Who decides the definition of "what was suppose to work" in the context of a given software product?

There are times when a feature is used in a way which was not intended by the developers. Now do the developers have to publish their test scenarios?

What if the bug is in 3rd party library? Add to it the complexity of open-source code.

Few points about India

1. Anyone claiming to know India well is either foolish or lying? (This includes Indians). India is diverse on all counts imaginable (color, language, wealth, urbanization, etc). Every permutation will throw up large number of people.

2. India consistently disappoints the pessimistic and the optimistic. If you see the bad side (poverty, chaos, etc) it is easy to bet that this country is going nowhere but this is just few scenes of long movie. There are other scenes and its easy to be absolutely bullish.

3. Think in absolute numbers as well and not just percentages. 3% in India translates to lot of people. It is more than the adult population of many countries. Apply this to any metric you will get large numbers on either side of the bell curve. So in India you will find 3% of the smartest and also the dumbest (in IQ terms) both in huge numbers. You will see these (only a subset of) 3% migrating to western countries and be hugely successful.

As per me the only way to deal with India is to accept its all hues and colors.

I think the main attraction of object storage like S3 is that it already solves the hard problems (distributed-ness, scalability, durability, etc) with a simple(r) API.

Also, apis like S3 are defacto standard with many vendors supporting a descent subset of it.

I stand corrected! Thanks!

Agree but my point is when you see a new and better rival then instead of pivoting SF became even worse and became malware-ised.

Also SF was based on SVN. They failed to understand and capitalize on a better tech on the market i.e. git.

The rise of github also coincided with enshitification of sourceforge.net. SF although was not git based at that time but it had the mindshare of lot of open source projects and it went complete downhill.

So, a downfall of a potential alternative was also a factor IMO.

Edit: after I commented I realized that SF was already mentioned in other comment

It is/was an April Fool's joke

In the beginning, I was trying to use Rust like C (which I think lot of people do) and was struggling with the borrow checker. I had a light bulb in my head one day to read the function signature and see if its '&self' or '&mut self'.

If any one API of a struct/enum has '&mut self' then its instance cannot be shared across threads without using a mutex (which brings its own friend along i.e Arc). This led me to structure the code in my mind much in advance to prevent borrow checker issues.

Another use of this is while embedding an object in a struct. When I realized that '&mut self' infects everything up the call chain it was a big learning moment. That is the embedding struct also needs a mutable borrow as well.

Under pressure, I suggested the following:

1. Only one solution (or path in this case) should exist

2. No loops (meaning the valid solution does not cross it-self)

3. The length of false path from correct path should not be more than 4

Maze(s) triggers PTSD in me. I was asked to write a code to generate one using a 100x100 grid in an Google interview in 40 min such that it was "fair". It was for L6/L7 (IC) position. I wrote a working code and the interviewer even acknowledged that I had atleast have a working code compared to others. The actual time I had was 30 min as 10 min included upfront discussion about what "fair" means and some time after-the-code discussion on improvements.

I was rejected with lean-no-hire (or weak-no-hire .. I don't remember the exact term but basically not a strong reject).

How is this priced? Is it Get + Put or only Put?