Why?

H

hasbeen

Why manufacturers of electronic computing and communication devices, supported by various IT and software development groups, are pushing the so called "latest and greatest technologies" that require never ending updates, upgrades and fixes of the software which controls their gadgets? And then after few months of useful life those personal communication/computing gadgets become outdated so the makers can sell you more of the same, but called newer and better, which will need immediate software updates or upgrades.

Why software developers by themselves, sometimes called engineers for some strange reason, do the the same things as the electronic gadget makers - continuously push or offer the latest or greatest software solutions that don't provide much value or improvements over the the prior?

I have been involved in various aspects of Real Life (visible and tangible) industrial, mechanical, aircraft and jet engine developmental engineering, as a technician, for almost 40 years. But, I had never experienced the level and frequency of continues changes, updates, upgrades and fixes that I noticed in IT and software development. Moreover, I cannot see the value to society in general that the electronic computing and communication gadgets with their software applications bring to improve daily life of an average individual.
 
Beastie7 said:
Man.. why make software and hardware at all? We should just stick to typewriters and pigeons. ?
Click to expand...
Man,

You missed my point. Progress and improvements have little to do with never ending updates and upgrades to fix the prior updates and upgrades or to promote something as newer and better that is not.

Btw,
I'd do better with a piece of paper, pencil or pen and a flock of pigeons that can deliver my air mail, in case of emergency, when my electric power gets turned off for few weeks or longer. I bet, in that situation your tech gadgets with all the software apps would not help you much with getting help if you need it. :)
 
Because they want to sell stuff.

And so they adapt to how selling stuff works. And it does work different for B2B vs. B2C.
In B2B it is engineers talking to engineers, and finally a big machine is bought that must be suitable for a specific work.
In B2C one must at first make the customer believe that he would need the stuff (which he doesn't), by telling him it is the newest and greatest (which it probably isn't).
 
Is part of the bussines,they create the problem and sell the solution

I'am not a developer,in the past I made
Something in php+mysql(a ticket system)
And c + sockets for hobby
But now I see the new generation of developers and how they are abducted to the new "model"
Example..back end..front end..left end..up..down..developer?
Ok, is.not good have one guy for all
(Like the jurasic park it guy)
But..years before there is only one developer for all
And the bussines besides RH new generations..they create words..tell what is good for you..what is bad...
Na....I dont buy any of it
 
The art of Marketing is making people buy things they don't need with money they don't have to impress people they don't like.

Hardware is currently so fast we could stop making new things and simply carry on with the current state. Who needs more than we currently have. Not want - need. And what for.
 
PMc said:
Because they want to sell stuff.
Click to expand...
Indeed. No-one makes serious money from writing good software or manufacturing good hardware that lasts for decades.

It is our job as the consumer to not be taken for a ride (thus why we are on a forum dedicated to an open-source operating system).

In some ways, I feel elements of open-source has been hit by this constant compulsion to update. It leads to some major churn within Linux communities of constant change (often sideways rather than upwards). Perhaps simply because the Microsoft/Apple converts "expect it" and bring this kind of crap into open-source?
 
kpedersen said:
In some ways, I feel elements of open-source has been hit by this constant compulsion to update. It leads to some major churn within Linux communities of constant change (often sideways rather than upwards). Perhaps simply because the Microsoft/Apple converts "expect it" and bring this kind of crap into open-source?
Click to expand...
There exists a classic paper on the subject: https://www.jwz.org/doc/cadt.html
 
Crivens said:
Hardware is currently so fast we could stop making new things and simply carry on with the current state. Who needs more than we currently have. Not want - need.
Click to expand...
You would think so maybe because you're not Japanese, Chinese nor Korean.
People who need to input/edit CJK languages (except pure Hangul characters) are still severely struggling for inproper conversions, especially in Japanese. Japanese has too few (written) pronounces used, making things much more difficult. For example if we Japanese input "hashi", it can be converted into at least 3 different Kanji, each means bridge, edge and chopstics in English (this is limited with quite tipical cases only).
The conversion software must understand not only the grammer, but also sentences, backgrounds, attributes of the writer and even much more.

My recent guess is that AI with LLM which never be achieved is mandatory to satisfy this. For security and privacy, the whole bunch of conversions SHALL be done locally without external data connection.
And the conversion must be quite fast. Otherwise stressful.
 
T-Aoki once upon a time I have written a kanji speller input widget, without knowing the language behind it I have to admit. The symbol space is impressive, but nothing that could not be navigated by hardware 20 years old. Extracting the meaning, the context, is more work but the limit there is memory space. If it can be extracted at all - which I highly doubt. You will need to set a limit what is acceptable, and 100% can not be reached.

I might change my mind when I see some LLM correctly translate the text about the philosopher who liked to eat lions.
 
T-Aoki said:
You would think so maybe because you're not Japanese, Chinese nor Korean.
People who need to input/edit CJK languages (except pure Hangul characters) are still severely struggling for inproper conversions, especially in Japanese. Japanese has too few (written) pronounces used, making things much more difficult. For example if we Japanese input "hashi", it can be converted into at least 3 different Kanji, each means bridge, edge and chopstics in English (this is limited with quite tipical cases only).
Click to expand...
So how would a human being (foreigner) achieve to learn this?

Maybe here is a misconception happening: the idea that computers are not just there to process structured data, but are able to understand humans, to act as being on-par with them, to act as a friend (or girl-friend?) or such.
That idea seems quite common-place in the artistic works of the last decades, and some people seem to actually believe in this, but it is science-fiction nonsense. A machine cannot translate philosophy or poetry, where even well educated people struggle to do. Watch "Arrival" (2016).
 
Crivens said:
Extracting the meaning, the context, is more work but the limit there is memory space.
Click to expand...
would be correct. But not only memory "space", but also "band width". Fortunately, character (symbol) space itself is no longer a problem recently, in contrast with '80s hardwares. Even before, in '70s and before, handling Japanese charset itself was just a dream, at least for personal users.

Crivens said:
If it can be extracted at all - which I highly doubt. You will need to set a limit what is acceptable, and 100% can not be reached.
Click to expand...
Not requiring 100%. Just as free-from-stresses as English speakers.
Even English use-cases, with all characters on keyboards, there are typos.
Not 100%. Just want the same level of stresses. Not more. Is it a problem?
I don't think so. Just need more and more breakthroughs.
 
Crivens said:
Hardware is currently so fast we could stop making new things and simply carry on with the current state. Who needs more than we currently have. Not want - need. And what for.
Click to expand...
Back in the 80s, I wrote a dialup bulletin board system.
It was done in Turbo Pascal with ASM subroutines for the modem I/O.

I had a 386/40 at the time, and it took about 45 seconds to compile and link to the final .EXE
Just for grins, I dragged it out of mothballs to see if it would compile on my Win 7 i7-870 2.93 ghz workstation.

I hit "enter" and nothing happened.
Figured it just crapped out or something.
Until I saw it compiled and linked the EXE file in 0.1 second.

Just a wee bit faster than the old 386/40.
And my i7-870 is a slug compared to current processors today.
 
PMc said:
So how would a human being (foreigner) achieve to learn this?
Click to expand...
Japanese (language) seems to be considered as "most difficult to learn language". On the other hand, for Japanese (people), other languages are hard to hear. For example, in Japanese pronounciations, "L" and "R" are NOT distinguished. So if you hear "A lice please!" in a restaurant, don't afraid. The Japanese person just wanted to say, "A rice please!".

An science fictions just shows a goal, but not implementations.
Maybe in 10th century, many persons would dream if something calculate things for me. At the momet, it should be just a dream. But now we have calculators and computers. Hope dreams come true.
 
Consider this: the construction foreman gave me a big hammer and a spike, telling me "if I nod my head, you hit it". Court date is in two weeks.

To extract the meaning from something is hard. I mean, NP hard. I see no way to make a computer do this without also knowing about humor, history, psychology and an awful lot more.
 
Crivens said:
Consider this: the construction foreman gave me a big hammer and a spike, telling me "if I nod my head, you hit it". Court date is in two weeks.

To extract the meaning from something is hard. I mean, NP hard. I see no way to make a computer do this without also knowing about humor, history, psychology and an awful lot more.
Click to expand...
Yes. So converting Kana to Kanji without stressing the user is currently just a dream. But historically, many dreams came true, yes, just small part of a huge number of dreams, though.

Note that I've described Kana to Kanji because usually ASCII characters (Romaji) typed by the user is converted to Kana (regardless how they are displayed) internally, then, converted to Kanji.
Why Romaji to Kana? It's because JIS (stands for Japanese Industrial Standards) keyboard has Kana in its keytops in conjunction with ASCII characters and some users prefer typing in Kana, and for some Kana, multiple combinations of ASCII is allowed as Romaji. Using Kana to convert is simpler.
 
You must log in or register to reply here.
Back
Top