This bacterial strain was discovered in 1976, but has been in the environment for a long time. Resistance should have already emerged?
"As a toxic mechanism, cry proteins bind to specific receptors on the membranes of mid-gut (epithelial) cells of the targeted pests, resulting in their rupture. Other organisms (including humans, other animals and non-targeted insects) that lack the appropriate receptors in their gut cannot be affected by the cry protein, and therefore are not affected by Bt."
Edit: "Spores and crystalline insecticidal proteins produced by B. thuringiensis have been used to control insect pests since the 1920s and are often applied as liquid sprays and donut pellets."
It does exist in nature all over, but it generally isn't in such a high concentration. Its like releasing 1,000 goats in your backyard. Even if your backyard could support a few goats, it will certainly be destroyed by 1,000 of them.
Bacillus thuringiensis is the most used pesticide in existence and has been for nearly a century. We’ve sprayed it everywhere in huge quantities, including on pretty much anything labeled “organic.”
What I really want to know is how do I CULTURE Bacillus thuringiensis, sure its available readily on the web as a pesticide, but only at very high prices for anything beyond a small hobby garden; -what if I have an acre of crops to treat? ~20 usd a pound for a spray bottle liquid solution is not even remotely acceptable.
You can look up how to culture it in the academic research but not only is it relatively hard, culturing Bt isn't enough to turn it into an effective pesticide. Bt is a soil microbe and sticking it on a regular agar plate guarantees it will be outcompeted by contaminants so culturing it requires sterile lab conditions and expensive lab equipment. You're going to need a full blown fume hood with active ventilation just to handle the initial samples and that's the bare minimum. Chances are you'll need a sterilized glove box and commercial off the shelf bioreactor.
The end product is produced in large submersion fermentation tanks where the bacteria is first grown, then then forced to sporulate and allowed to ferment which produces the Cry/Cyt toxin crystals within the spores. Going through these phases requires precise control over nutrients, pH, agitation, and electron acceptors. Even activating the spores to create the inoculum and begin the reproductive cycle is nontrivial because of how sterile everything needs to be.
This is a confusing analogy because it's more like releasing 1,000 goats then wondering why they don't have an effect.
My non-SME answer to this thread is:
1) 100 years is not that long in evolutionary terms;
2) cropland is large but not all land;
3) evolution is complex;
4) resistance is actually already observed in some species.
In this day and age, it is almost impossible for certain businesses not to build on someone's else kingdom.
Facebook, instagram, uber, lyft, doordash, instacart, and hundreds of unicorn businesses are literally built on top of ecosystems that are controlled by 2 or 3 companies.
Thats not what other people's kingdoms means. Basically all non-trivial apps are already built on VPSs. The other people's kingdoms refers to how people interact with your app. Take Zynga - they were not literally running on FB servers, I'm sure they had their own VPS, but their games just had 0 reach outside of FB. When FB decided to change things Zynga just got fucked, VPS or not.
If you have a possible very high return for taking that risk (as the unicorn businesses do) then do it in full knowledge of the risk.
I am not convinced those businesses are good examples. Could they have redeployed elsewhere if they had to? Where they tied to one supplier? Did they have backups else where?
Most businesses and individuals do not have to take that risk and can avoid it.
The key insight is that mitigating risk isn’t a free action. No business has the bandwidth to avoid even most of the risks they could in principle avoid; you allocate some effort towards the ones that make sense to mitigate, and hope the others don’t come to pass.
But relying on other people's kingdom isn't free either, that comes with a cost.
The fallacy I hear often is that because something like AWS is sooo much more expensive than co-location or VPS, that it must be easier.
Yeah, it can be... sometimes. It almost never is. You trade off the complexity for new complexity. You replace your sysadmins with Dev Ops. It's not like it just magically gets better.
It's the same way with a lot of things. A more expensive car is not necessarily better. It can be, sometimes, if you're very careful and know what you're doing. More expensive clothes aren't better either. Popularity factors into this, too. Just because something is popular doesn't mean it's good. Plenty of really shitty things become popular.
You can absolutely build your application without relying on other companies too much. I'm not saying you need to go rogue, but you also don't need to use every single Google feature under the sun to, like, display some photos. And, if you do that, that's actually probably way worse and more expensive (effort + money) than if you didn't.
> But relying on other people's kingdom isn't free either
I know it sounds trite, but looking at what I'm responding to, it is not possible to not rely on others for anything but the most basic things where input(s), tool(s), action(s), and output(s) don't require anyone else. You need infrastructure, services, tools, materials provided by others for almost everything.
We are a highly networked species, and building a business is finding a spot in that network. You are always highly connected. Therefore I find that "kingdom" metaphor does not fit how our lives work. It's a network, not kingdoms, not even the richest country could afford to cut the connections.
Services like those from the big companies at least have the advantage that the companies cannot make them too bad, because that would backfire. Copying what most do is pretty safe compared to doing your own thing.
That's also because you want to concentrate on your core business idea. Sure, if you spend some effort doing your basic infrastructure different from others you may save a little, but overall for many companies it will be much better to copy what most others in your space do, so that you have the same basis, and not risk being different in an area that is far from your core competence.
For many areas in business you just hope for the best. You hope(d) Russia-Ukraine would not escalate, or later that it ends quickly. You hope that the latest US tariffs won't be too bad. The world is full of surprises, at least for businesses big company account access or cloud issues does not seem to be a big one, in comparison. I'm not saying this as a headline reader, but as someone dealing with a sanctioned country and other politics-related issues that have been impacting us for a while, despite doing business very far from anything critical (lifestyle consumer products).
We, for example, use one big company to host our DNS (due to history, they also used to host our emails when we began), but we have Microsoft host files (OneDrive) and Email (Office 365) for our entire business domain. I would like to not have to rely on US companies, nor more than ever (we are German), but that stuff just works. And not just as individual pieces, but also together. For example, when I open an Excel file stored in a shared OneDrive folder, and a colleague does the same, we get automatic shared editing and see each other's cursor position. Many small conveniences like that. AND, very important, emails just work - with rarely ever any issues because of rejected emails.
Whenever I see a discussion on reddit or here "why Microsoft (Office)", soooo many people only know the most superficial of arguments. They talk about "but LibreOffice", "but [insert other mail- or cloud file- service), but the apps are not even all that important. It is everything, the huge amount of infrastructure and methods they provide, automatically or manually usable, that ties everything together. The "glue" vastly outshines just Excel or just Word itself, either one of which could indeed be replaced.
When you do more than just simple email, when you have to administer a few dozen employees and their devices, you will find that the big US companies are very hard to beat.
What would be the alternative anyway? Having your own server is a nightmare in comparison! Even just making sure my emails won't be spam-rejected by at least some providers (where my customers and business partners sit) is too much. Making sure the dedicated server is always up to date with patches - that's a lot of work that I'd rather not do. In any case I still have to rely on the hosting provider, who may cut off my wonderful 100% self-owned and administered server at any time because I ended up on some anti-spam list because I did not react to some new threat that I did not even know about in time.
Overall, relying on the big mainstream providers is a prudent choice. You can't avoid trouble, and ceteris paribus choosing them IMHO makes sense.
> You can absolutely build your application without relying on other companies too much.
Obviously, since they do it! I could also bake my own bread. The point is I prefer to specialize and deal with things that don't differentiate me from others in my line of business as little as possible.
I started with computers at a time of 8 bit CPUs, when I knew every relevant memory address by heart. The many many layers these days are not something I enjoy, but I will still use the mainstream stuff nevertheless! Because I am not in the business of basic infrastructure IT and every minute I spend on it is a minute not spent where it actually matters in my business. Everybody, the few dozen employees, the partners, the customers(resellers, they all know the big companies and their products. The only thing we do ourselves is EDI messages on top - on cloud servers. But we sure don't want to come up with OneDrive and Ms Office alternatives, even if politically I'd like that.
Right, but theres levels of vendor lock in. Going up to more lock in doesn't necessarily mean an easier experience, that's what I'm trying to point out. So you have to do the risk analysis.
The fallacy I'm trying to point out is that if you outsource some functionality to locked in vendor implementation, then your life is easier. It can be, but Its often not.
> In this day and age, it is almost impossible for certain businesses not to build on someone's else kingdom.
No it is not. It is only the greed for bigger profits. If a company can work with Microsoft, it can also work with LibreOffice. But LibreOffice doesn't promise them the moon, while sucking every cent out of them.
There was a fleeting moment where the Internet was the "Wild West" but we are long past that. The GP's idiom is about as practical as "don't be a citizen of any state".
The only counter point I can think of is that you can always choose to build upon multiple tech platforms simultaneously, and depending on the technology you need, it might not even add all that much additional cost to do so.
I mean, technically you don't have to use those ecosystems and could roll your own stuff, including infrastructure instead of AWS but it's definitely going to be expensive.
Which is why we need regulation for those big players (gatekeepers as the EU has taken to calling them). If you're going to be so huge that you essentially operate your own market and economy, then you need to be regulated like one, and forced to play nice, interoperate, and not favor your own services.
Not so technically, your business needs customers and efficiencies. Big tech strategically positions themselves at those choke points.
But realistically , if Google and Apple both for whatever reason banned you from all their services, idk why, then you would not have access to a phone. So then you say, well, it was just one person in ten million, and they probably did something wrong-- and now you share the same perspective as Big Tech on this specific issue.
Yeah. I would still recommend that people create separate account for dev purposes just in case. But still I have not heard of Apple locking people out like Microsoft or others do
CP-KCS merger was approved in 2023. Created a single line connecting Canada, US and Mexico. So far we have seen streamlined operations, reduced transit times and improved reliability for cross-border freight. Some smaller shippers have reported concerns about rate increases and fewer service options, but it's been limited so far
https://m.youtube.com/watch?v=aeWyp2vXxqA
reply