Honestly? We need not a successor of FHS but of filesystems, who are intimately tied package managers and installers. Zfs timidly start the change, with IPS (Image Package System) and BE (Boot Environments, as zfs clones), and we need to go much beyond that instead of wasting resources keeping up an '80s model like some do from btrfs to stratis.
We need:
- query-able storage, because search&narrow is the current way of accessing information and collecting/transcluding data is the way to go;
- easy storage management, the "rampant layer violation" of zfs we really need;
- integration of such storage to the software stack, from the OS to single packages, it's a nonsense having to "spread" archives in a taxonomy to deploy them or downloading archives to be unpacked as well for updates when we have send-able filesystems (zfs send of snapthots) and binary diff (from a snapshot "tagged version" of a fs-package to another, sent over internet).
Unfortunately we need operation people together with devs and nowadays operation is nearly disappeared. Devs alone can't understand what we need, they can't go beyond their desktops in a mass large enough to avoid a positive evolution.
Who is "we?" I certainly don't need those things. If you need to add a bunch of complexity for your use case then feel free, but for most of us it's unnecessary.
> - easy storage management, the "rampant layer violation" of zfs we really need;
Except in zfs you have to think if you really want that device in that pool or that vdev. I use btrfs, slow and kinda unsafe, specifically because you just specify raid1c2/raid1c3/raid1c4 and it kind of survives c-1 dead disks (until you run out of disk space and everything goes to flames).
> - integration of such storage to the software stack, from the OS to single packages, it's a nonsense having to "spread" archives in a taxonomy to deploy them or downloading archives to be unpacked as well for updates when we have send-able filesystems (zfs send of snapthots) and binary diff (from a snapshot "tagged version" of a fs-package to another, sent over internet).
We (kinda, for some very generous definitions of) have that in composefs? But I still sense even with that, you still want some resemblance of sanity in your indivual layers.
It's for SEPA area BUT it's mandatory open only for PSPs/financial institutions cutting out end customers... Also most banks have a terribly limited support. Personally I use OpenBank APUs (PSD2/DSP2) for some banks (ironically via a Canadian operator, GoCardless) and well... generic accounts are supported, stocks are not, bank cards are not etc, most exports have very poor quality.
Long story short I can auto-import in Firefly III from EU banks only via a Canadian company and the quality of the process NOT due to GoCardless but due to local banks is terrible...
Me personally I'm very satisfied by org-mode, but the main point is not org-mode itself but Emacs, or an integrated, end-user programmable environment. Org-agenda handle todos, but in the same notes I handle attachments, runnable live code, links to mails/threads, ... because of that and that's the point: we have a single brain, we need systems who are integrated as well.
Not a gazillion of independent apps lacking also unix CLI IPCs (only cut&paste), but a single integrate, moldable one. Modern software have ERPs as best integrated solution, Emacs is the most integrated one still alive and kicking, Smalltalk workstations and LispM was the best from the past.
Maybe in some more decades, a step at a time, we will be back at such evolutive levels...
We've been talking about learning organisations for years, and the importance of a documental culture has been clear for a long time. However, most companies have never truly implemented these models. Remote work forces them to do so, or they'll fail. So, even though it's been "theorised for a long time" and widely dissected, remote work it's effectively a new thing, and it's sadly normal to encounter many problems, especially with management that lacks even the most basic substantial IT skills to truly work in a virtual company, and not just management, but a large part of all the staff too.
I'm one of those who are very productive working remotely, but not out of loyalty or some "ethics/morals". It's simply because, as an engineer, I seek efficiency, and remote work is efficient. I suffer from inefficiencies, rituals created to placate bipedal cattle, and senseless reactionary attitudes. We are few, I imagine, but we are also the cohort that innovates, at various levels and in various sectors, without whom I don't know how much the West could hold on, and this, indeed, is not free.
Today, with RTO, a social rift is forming between those who want to truly advance and those who merely muddle through, while the world moves forward. If this trend isn't reversed, the West, which has already lost so much, will lose what little remains and discover that its residual military strength amounts to little more than a Romans legio fantasma. At that point, the cost of "saving money" by not innovating will be so high that it will lead to bankruptcy.
I can understand your sentiment in the first two paragraphs, even though I think remote work was bound to fail for most of the companies simply because it is hard to update the culture. Broadcom mandated in-office work even during COVID as soon as the strictest lockdowns were lifted. And yet Broadcom stock is up 5x since 2022, so they are not missing out on anything by embracing the office. Same with Meta (mandated RTO, stock up 7x).
> the West, which has already lost so much, will lose what little remains
The West already advanced a lot until 2020 by working from office. If anything, fully remote conditions are a death knell for software jobs in the rich parts of the world. Most of the SW jobs do not require special talent and can be done for a fraction of the cost from Latin America, Eastern Europe, Africa or India.
The world is changing, yes, once upon a time, most innovation happened in person, because with the technology of the time, it was necessary to be there in person. But today's technology has made being there in person superfluous and inefficient. This means that some will continue remotely and they will be the ones who overtake those who have stuck to the old model.
The Bronze Age was characterised by bronze, which allowed for many advancements. But then we discovered iron, and those who chose it first massacred those who had remained with bronze. Bronze was no longer what had allowed for evolution; it had been surpassed.
This is my feeling and my point. These changes are obviously not that rapid; culture changes across generations, but they will happen, and those who are left bolted behind will not have the long and slow experimentation and refinement that allowed to master the new, and so they will simply be crushed. Software developed around the world matters little; quality software is so rare that only by working globally can we still create it. No single company, not even a single whole population, can truly do it on scale.
> But today's technology has made being there in person superfluous and inefficient.
Hard disagree. Lots of otherwise talented people still have problems with written communication or other norms of remote work. They thrive in an office though.
Mere proof of that is a lot of startups which are still flocking to SF and mandating in-person presence 5 (or 7) days a week so that they can move fast.
It's simply the proof anyone need personal ownership in a free internet instead of being dependent on giants on their own computer. It's curious how many in IT fails to see this...
The fact we have PSPs endanger Democracy, because we could technically exchange eCash P2P, there is exactly ZERO reasons to keep up the banking systems, and banksters knows that well, https://web.archive.org/web/20240213185758/https://www.cimb.... unfortunately people seems to be unaware of that...
We are many, search engines are the mean to discover things because even with usenet it's impossible for a human to discover via URLs and links enough information on the web, that's the real revolution: links are useful but not enough. Search engines are the best tool we have had so far to find knowledge around the web, now LLMs try to surpass traditional search engines milking knowledge from web contents, like we have many articles about wildfires in a region, but let's say not one about wildfire trends in that region, an LLM could try to spot a trend milking all articles in a significant timeframe. The Conrad Gessner's Biblioteca Universalis dream.
So well, LLMs do not kill the web, eat it. We are still almost the sole valid source of data for LLMs.
What really killed the web are social networks as proprietary walled gardens instead of an open Usenet with a web companion for stuff to be preserved for posterity or too long/complex for a mere post. What killed the web is the fact that ISPs do not offer an open homeserver instead of a closed box called "router" even if it's a limited homeserver. With an open version, with IPv6, anyone could buy a domain name and publish from his/shes own iron a blog with a ready-to-write software, with automatic RSS feeds, newsletters etc. If we give such tool to the masses the original web will be back but it would mean free speech and giants/politicians etc have free speech preferring ways to master public topics through their platforms to hide from most stuff they dislike and push ideas they like...
Search engine indexes being turned into copyright enforcement levers also significantly killed the net as it created scarcity in info dissemination for the sake of maintaining info asymmetry.
Go ahead and try to find JLG equipment/service manuals on the open net anymore. I'll wait.
they are anyway needed and we also have YaCy as an example, and other distributed search solution. The point is that most do not participate so only commercial one get enough resources to be useful.
It's a scale problem and a targets one: we are damn many, if we all want a fillet a day we haven't enough cows to satisfy such demand, and that's valid for essentially anything. So far we prove to be very skilled in doing identical stuff on scale, which enable industry hyper-growth but we can't feed production lines with enough raw materials, we can't produce most things in circular manner and even some production naturally renewable can't be completely renewable due to the scale of the demand. In an ideal world we cultivate and farm in proportions where the guano and manure from the species we raise provide enough fertiliser for what we grow. But there are many of us, and to feed everyone, this balance is impossible, so we must crush rocks to nourish plants sufficiently, which is obviously not renewable... We know how to makes wood-frame homes and trees grow up again, but again the demand much surpass the capacity of trees to grow up again and so on.
The result it's finding new way to do more with less, and finding them quickly. Some do works well, some do works a bit, many gives only the illusion to work enough and people buy them anyway because an illusion it's still something more than nothing.
The target issue is the model, capitalism, issue, in the past we have used money as a means to barter things counter something we all accept. Nowadays we use money to makes more money, so goods are just a mean not a target, and the result is that we do not care about quality, being just a mean if we can sell them it's enough to milk money. To solve this we need to makes money public, creating by governments without fractionary reserves and public debt mechanism, taxed to keep the supply limited enough following the availability of any specific resources, so essentially like Swiss we need to tax just VAT with continuously variables rates following nature and tech, while taxing local properties just to assure local consumption does not exceed a sustainable threshold of resources usage.
We need:
- query-able storage, because search&narrow is the current way of accessing information and collecting/transcluding data is the way to go;
- easy storage management, the "rampant layer violation" of zfs we really need;
- integration of such storage to the software stack, from the OS to single packages, it's a nonsense having to "spread" archives in a taxonomy to deploy them or downloading archives to be unpacked as well for updates when we have send-able filesystems (zfs send of snapthots) and binary diff (from a snapshot "tagged version" of a fs-package to another, sent over internet).
Unfortunately we need operation people together with devs and nowadays operation is nearly disappeared. Devs alone can't understand what we need, they can't go beyond their desktops in a mass large enough to avoid a positive evolution.
reply