It's really a question of what's the chance that any one package in the hundreds (thousands?) your solution is using directly or indirectly has

- A critical unfixed bug
- A dependency or code which will break on end of life of a language feature, library, system API call, ...
- The package will be taken over and malware injected into the code
- The package will become unmaintained or bump version number only zombie maintained
- The package license will be changed for commercialization or require server side runtime royalties
- A cloud vendor maintained package dies with end of life of one of the cloud services it uses

Deeper down in in the package dependency tree, the risk to the top level solution is higher since any one of those cases multiplies with the number of dependencies on that package.

Bleaching bone like skeletons of coral...... /s /s /humor

https://www.smithsonianmag.com...
Smithsonian Magazine Aug 20, 2013

From Gunpowder to Teeth Whitener: The Science Behind Historic Uses of Urine - Preindustrial workers built huge industries based on the liquid’s cleaning power and corrosiveness—and the staler the pee, the better

Microsoft will most likely replace Teams with something else in a few years.

The and other tech companies have shortened the production life span of many major technologies to less than 3 years, since it only needs to defer users finding blocking bugs for 1.5 years until an 'upgrade to the new version' answer is the only answer from the tech company.

Seems that the last X years of TIBOE rankings are mostly the same story, Python, C/C++, .net, Java, SQL and JavaScript in relatively the same rankings.

Why the editorial comments on Perl surging to the 10th spot when it's change in percentage of less than 1% is within any regular survey's margin of error?

We can just ignore these rankings and other clickbait articles as space fillers.

The only apparent trend is that the quality items of long-term language stability (not perpetual feature creep), library comprehensiveness and stability, tool set completeness and documentation keep a popular language near the top.

They are feature complete enough that they are just collecting micro features resulting in more feature bloat, less productivity.

A reduced complexity version of Java, C#, etc. would be much more easily for automated tools to process.

It is good to see an actor/public figure not going a worldwide apology tour just to legacy build late in the day.

Maybe we could convince "journalists" to stop writing "(famous singer X) hated these bands and would never tour with them..." articles.

>Python is actually a pretty good language in the hands of a competent (!) developer.

Agree. Decent developers write decent code in most any language. Then again, it is possible to write "Fortran in any language" as the saying goes.

Somewhere, we may discover that beyond a certain number of lines of code, going from an any/variant dynamic object based language (javascript, python, typescript) to one with declared types checked at compile time is worth doing. The major languages, beyond Rust/Go have been collecting new language features for decades resulting in languages trying to be any programming paradigm you want, as long as the language has all the good/bad parts of JavaScript.

World Bank - https://data.worldbank.org/ind...
- 8090 publicly listed in the USA in 1996 is now 4070 today

UK - https://www.theglobaleconomy.c...
- 2415 in 2008 and 1646 in 2021

Given that the USA is the only (?) major market allowing share buybacks, add in a few mergers, private equity buying out public companies, etc. there will be a shrinking set of companies with enough trading volume to sustain active stock markets.

That would be a shrinking pool for the retirement funds from people, pension funds, and sovereign wealth funds.

It would take a while and provide upward rising tide for (?) 20 years to the biggest stable dividend paying companies.

of the executives at the next shareholder conference call?

Same for LLM / AI companies such as "How much more new AI training data have you harvested and used in the last 3 months?"

If the web is in rapid decline, then will the knock-on effect be that the AI / LLM companies won't be able to have a sustainable amount of new AI training data?

Google stated in court that the web is in rapid decline.

Any shareholder in the next quarterly meeting which hears good things about web advertising can follow up with a "Are the number of web sites viable for advertising increasing, decreasing or about the same?"

If there is a 0.001 chance an individual NPM package has issues, using 10 packages gives a greater than 1% chance and using 20 packages give a greater than 2% chance.

A react hello world package uses 2839 packages https://medium.com/frontendweb...

Gartner, McKinsey and Booz Allen Hamilton, all business consulting strategy firms advised US corporations that offshoring was an effective long term business strategy.

It resulted in tens of thousands of layoffs, with every thousand laid off having a few end in divorce, broken homes, addiction and suicide.

From 2011, National Institute of Health - https://pmc.ncbi.nlm.nih.gov/a...

" using job histories of workers in Pennsylvania, Sullivan and von Wachter (2009) find that mortality risk doubles in the year of displacement; is 75% higher in the year after displacement; is 50% higher in years two and three after displacement; and eventually settles down to 15% higher in each subsequent year:"

AI model builders will just delay longer and longer knowing that those depending on click advertising revenue, book sales, music sales, etc. will either go out of business or settle for a much lower amount.

The question of AI model arms race for future self-defense and battlefield tactic will keep the government busy during this time while the licensing and royalty payments are figured out.

I'd agree that actual musicians are needed and that AI slop will be good enough so that the 75% of people who are mainly disinterested in what music they listen to will not bother to notice AI versus human musician produced music (not the X people listed as "composer" on most pop songs today).

Music will probably follow the Tomas Kinkade painter of light business model where a 'pop star' will fly in, sing a few words on top of a barely passable AI pop song and call it her own, splitting the money 90% to the pop star for her name and 10% for everyone else.

The odd thing is that AI will dramatically reduce the number of paid jobs for the music industry, affecting the nepo children's careers.

What is the line where news organizations determine where selling / talking your own book is not news and an advertisement?