Good/Bad are consensus votes. Its hard to escape their use just because of how deeply ingrained the programming is. We just think it makes "sense" and is "obvious" because its a meme that is already in our head. There is nothing inherently evil or good about any past/present/future animal on this planet.
>The little-known structures in tax-friendly destinations have contributed to the 15 pharmaceutical firms amassing profits of €580 billion in the last five years.
>This amount outweighs their research and development (R&D) costs, despite the industry's frequent claim that high drug prices allow them to innovate and design new drugs.
R&D is one thing, but most drugs fail at the clinical trial stage. This money (hundreds of millions per drug) is just gone, unlike R&D which might result in new tech or at least a patent. For Oncology its even worse, its close to a 95% failure rate. Simply taxing companies won't make their drugs successful. Large pharma companies rely on a few blockbuster products for their profits and they milk them dry. This is standard corporate greed/behavior, but it certainly seems offputting because we're dealing with peoples lives. Personally, I think its inevitable that there is going to be some form of nationalization for a protected class of life-saving medication.
> R&D is one thing, but most drugs fail at the clinical trial stage. This money (hundreds of millions per drug) is just gone, unlike R&D which might result in new tech or at least a patent.
Are you saying the clinical trial stage is somehow not part of R&D spending? It sounds like quite obviously research to me, but I'm not familiar with how it's actually reported.
The industry may learn from failures, but the shareholders of private companies want a return on their investment. I sure as heck don't want my 401k tied to oncology.
The WTO/TRIPS treaty already allows for governments to issue compulsory licenses in case of a national emergency. This also explains why some firms where so "nice" during the pandemic, as they feared it.
See https://www.wto.org/english/tratop_e/trips_e/public_health_f... for details.
Obviously, it does not suffice but at least it's something.
You could concentrate on patent expired drugs that cost the most on an annual usage basis. Yeah, that leaves out the newest meds, but you'd still make a dent, and every year, you get cooler stuff to make that's already been market tested.
No, because patients deserve the best or most effective drug, not the cheapest.
Unless you'd like to put a price on human life that is.
Also, meds don't hit the market before they are tested...
Testing meds is part of the resource allocation problems because cheap & effective off-label use won't be researched & tested because there is not enough money to be made...
The best or most effective isn't always in consideration. At least outside of the US, you get what has been approved for care and only if it's not too cost prohibitive.
If governments were making their own generic drugs and sold them at cost then it would help avoid situations like Martin Shkreli. The drug wasn't protected by a patent, it's just that nobody else produced that drug. In my opinion this seems like a reasonable thing for governments to do.
Also, the best treatments aren't always available in this first place. Eg (at least a few years ago, don't know today) Adderall was illegal as a treatment here. Some EU countries, until recently, even treated methylphenidate like that.
> Unless you'd like to put a price on human life that is.
It's been put at about $10 million, for the purposes of EPA regulations. Insurance companies have their own formula, from a couple hundred thousand into the millions, and healthcare users QUALYs - quality-adjusted life years, and those are priced at $50,000-$150,000.
I'd say non-optimized government corporation running on cca 0 profit churning out safe generics is still a massive good for all mankind, its type of inefficiency we can all somehow live with, compared to usual governmental fubars left and right.
Triple that in places like US (not for fubars but the costs).
And now it would be a political issue which drugs get manufactured for cheap and for whom. You can probably see how that would go in corrupted countries (i.e. all of them)
Profit is what is left over after R&D - they can choose to invest those profits into future R&D, or far more frequently, just pay out the profits to owners as share buybacks or dividends.
Much of the early development of novel new treatments is funded by venture capital and early IPOs. The big players buy them out when most of the risk is gone and it’s time to scale up.
Yeah, that makes the most sense. Take supplements if and when you need it. I think the real study would be if someone did trended vitamin deficiency data from all the subjects and then co-related the intake (for deficient persons) with health outcomes.
Why are you so upset that people derive a lot of value from Windows? Enough that they want to keep using it, and defend it because they don't agree with the "everything is broken" meme.
Because like industrial waste, Windows exports problems to other systems.
1. Windows has an absurdly short maximum path length of 260 characters.
2. On Windows, moving files to a temporary directory can fail, if the temporary directory has a longer prefix than the original path.
3. When uninstalling, the python utility "pip" first collects files into a temporary directory, then deletes that temporary directory.
4. To avoid running into MAX_PATH limits, pip doesn't use a normal temp directory. Instead, it makes a temporary directory adjacent to the directory it is removing. (https://github.com/pypa/pip/pull/6029)
5. If pip is interrupted while uninstalling, the adjacent temp directory is never deleted.
So, in order to work around a Windows-only problem, pip stopped using standard file locations, creating a new problem that only existed due to the workaround. And then I'm left trying to figure out why I'm running out of disk space.
The MAX_PATH limit is annoying legacy backwards compatible stuff, but can be avoided by prefixing paths with \\?\ before passing them into the Windows API.
This is something that languages/runtimes with more effort put into portability already handle for you:
Because even if you don't touch Windows (or whatever mediocre malware Microsoft presently peddles) those folks come to you and say stuff like "skype won't start" and lo! it does not start, though after much clicking around and rebooting and trying the obvious things you discover that if you right-click and try "open with skype" on the skype icon then skype will start. That problem at some point disappeared as mysteriously as it appeared. Eh, who knows, it's Windows, and there's more science to reading tea leaves or goose entrails.
Then after za'o decades of stories like the above (it is merely the most recent of many) one might wonder how does Microsoft with so many programmers and so much money produce such kusogeware? That continues to waste my time?
You can have your own view. Nobody is taking it away or forcing you to believe otherwise. My point is why are people so upset when someone has a different view or doesn't agree with your personal view on Microsoft?
What? You shouldn't defend bad behaviour regardless of if you derive a lot of value from the same source. A good organization wants to be called out on shitty practices so they improve.
You can make an argument to convince people of your personal point of view, but there is no reason to be all upset if someone has a different viewpoint. Thankfully we are all at liberty to have our own view on this topic.
Just to put things in perspective - MacOS forces you to run on their update treadmill for the OS, and you can't even update your xcode dev environment if you refuse. Ditto with Linux, you can't refuse OS updates if you use a package manager because eventually you can't update anything without accepting all the dependencies. Windows for all its warts did do something well with binary compatibility. I am happy as a pig in poop with my work supplied LTSC install.
Yes it has an editor and a debugger, and some other stuff. So what? Its still an application. An application forcing you to update your OS would be laughed out of the room in any other situation. Personally I wouldn't defend such a broken design, but to each his own.
Sorry, but the Linux part doesn't make sense. If you want to update system components from the system repository, it makes sense that the dependencies are rolled into it. But if you don't want to upgrade everything, you don't have to. Install a specific app version from flatpak, download binaries, or compile it yourself for your current system. You have so many ways to do this - same as you're doing on windows.
>Install a specific app version from flatpak, download binaries, or compile it yourself for your current system.
Then you can't have those dependencies managed by a package manager, which means you'll constantly be compiling (and dealing with conflicts yourself). This is not a viable path forward as it seems to be presented here.
>You have so many ways to do this - same as you're doing on windows.
On Windows I copy a binary and run it w/o giving a crap about the OS updates. Vastly different workflows for it to be called "same".
> On Windows I copy a binary and run it w/o giving a crap about the OS updates.
You don't just copy the binary on windows. You copy the binary and its dependencies. Unless it's based on .Net-framework, then yeah, you still care about it being updated through windows.
> Then you can't have those dependencies managed by a package manager, which means you'll constantly be compiling (and dealing with conflicts yourself).
That's not right. You don't need to use dependencies from the system. If you're compiling yourself, you can either make them static, or use your custom location. In either case, there are no conflicts to deal with.
If you're using whole app packages, then you don't care about system deps at all. Flatpak apps ship their whole environment. AppImage apps do too, just without full isolation. (https://appimage.github.io/apps/ entries are all "download-and-run")
>You don't just copy the binary on windows. You copy the binary and its dependencies. Unless it's based on .Net-framework, then yeah, you still care about it being updated through windows.
Sure, but I can install s/w built for Windows XP into Windows 11 and it will just run 20 years later w/o modification - And I can do this with a large amount of commercial software, so its not a special case example which you can demonstrate on any OS. Simply not possible on any other desktop OS. Even if you have the source (which you won't for commercial s/w), its not just a simple recompile after 20 years.
>You don't need to use dependencies from the system. If you're compiling yourself, you can either make them static, or use your custom location. In either case, there are no conflicts to deal with
Static linking glibc.. yeah. :)
>If you're using whole app packages, then you don't care about system deps at all. Flatpak apps ship their whole environment. AppImage apps do too, just without full isolation. (https://appimage.github.io/apps/ entries are all "download-and-run")
Yet another "standard" that ultimately will fracture the deployment of software. These are just buggy half-baked standards anyway. Vendor releases 32bit AppImage and goes out of business, it stops working on 64bit OS.
I don't think you guys really get the insane backwards compatibility effort that Microsoft does to help keep line of business applications running.
> can of course install Arch or Gentoo or NixOS Minimal and then audit the packages that they're installing to see that there's no obvious security violations, but it's unrealistic to think that most non-software-engineer people are going to do that.
It's a fantasy to think that random devs can audit kernel/security code. No single person can. Too many lines of code to audit (that you didn't write yourself). Even if you hired a team, by the time the team does the audit, the goalposts have moved with new source code.
Sorry, I guess I didn't really mean to imply I was going to dissect everything line by line, but I can at least look to see if every package in there is directly open-source and if there are any packages that are being pulled in that are frequent security concerns.
ETA: I know I can technically do that with Ubuntu or Fedora or OpenSUSE as well, it's not like it's a secret which packages they include, but what I like about NixOS Minimal or Arch is that I have to explicitly add every package I want. There are transitive dependencies obviously, so there of course can still be stuff on my machine I'm not happy with, but I still think it's better.
> if there are any packages that are being pulled in that are frequent security concerns.
As an individual, do you think you can do that? I know a lot of packages with security concerns where CVEs are never issued. You just need to go to their PRs and luck into finding descriptions of a security fix. I doubt this would scale for a given individual.
