It starts out with that title trying to piggyback on Marc Andressen's famous article. Then by 1/3 of the way through, it's full on "Apache Kafka® and its uses".
I feel cheated. I was ready to defend the non-software companies of the world, but the article has very little to do with the title. It seems like just an ad.
Agreed. There should be law that forces all advertising to be within a dedicated HTML tag called <marketing> so that anyone can just opt in to a plugin that deletes all marketing tags.
Well somehow it doesn't seem quite as toxic as newsvertisements similar to original post in question, and despite my original statement, I still enjoy posting here, so clearly this is a spectrum and what you are describing and what I am talking about are not quite identical.
There's a difference between offering up a startup-oriented forum that still encourages people to post anything that is intellectually stimulating and interesting, while still allowing it to be an outlet for YC-backed startups to post and share things for discussion vs. posting a news article that piggybacks on Andreessen's well-publicized and becomes marketing spin about a specific product after a few introductory paragraphs. Just like how there's a difference between allowing a few "sponsored posts" on a search page, and having all of the front page be sponsored posts. Does that make more sense?
Do you disagree that there's a difference between those two things?
Sadly these days (probably has been true for a long time) the first thing to do when reading an article is checking the authors ties to a topic, and what agenda they might be trying to push.
Since I already knew the company, when I saw the headline and their domain, my immediate reaction was: "Of course they would say that. They are a software vendor!"
Only this time it almost failed me. The top comments were lamenting on how misleading the headline is. They were correct of course - but nonetheless it is in in-depth article on Apache Kafa discussing it's origin story, use cases, how it was used at Linkedin, and most interesting how it used to replace ACID applications with eventual consistency.
All in all, a very good article, despite the lead in waffle at the start.
Thank you thank you thank you. As soon as I saw your quote "Apache Kafka" the first thing that went through my mind was whether this was going to be a Confluent article and lo behold the title (and link) said it all, saving precious cycles!
> That is, the core processes a business executes—from how it produces a product, to how it interacts with customers, to how it delivers services—are increasingly specified, monitored, and executed in software.
When I think of businesses I think of the local auto repair shop, the restaurant around the corner, the paving company working on the roads, the local HVAC company around the corner from me, and the local gyms.
The core products at the boot camp workout place across the street from me don't need to be specified, monitored, or executed in software.
Somehow this reminded me that empathy is decreasing in society (you can search on it). Putting more computers between businesses and people isn't necessarily going to be better for people or the way we interact with each other.
The sad thing, or at minimum odd thing, is that we're few years away from the positive differentiation factor being "you'll be talking to a human".
In fact it should be made illegal to have people who's livelihood depend on a platform to be left at the hands of shit auto replies. If large scale comes at the cost of injustice, maybe large scale shouldn't be a thing.
Youtube, Amazon Seller Central, and many others are terrible examples of what future holds for a lot of people.
That's already the case in many industries - for many SaaS and telecom businesses the differentiator for expensive B2B plans is "24/7 support with a real human being", big IT companies like IBM and Oracle rely on "There's a human to blame when things go wrong" as their sole value prop, high-end retail like Nordstroms differentiates themselves from low-end retail like Walmart by the availability of personal sales assistants, and waiter-service restaurants differentiate themselves from McDonald's by the existence of a human who takes your order. The difference between Uber and a car service is whether the driver's waiting for you at baggage claim, greets you politely, and makes small talk on your way home.
Personally I like it - I usually pick the self-checkouts, kiosks, and apps over talking with a real human being, who frequently acts like a robot anyway and tends to be slower and make mistakes more often.
Also, I wonder if folks who say "if large scale comes at the cost of injustice, maybe large scale shouldn't be a thing" have really thought through the implications of this. I remember talking to a classmate in high school about Native American tribes, industrialization, American history, and how we really fucked over this idea of a tribe or village or community and the world was a lot less richer because of it. (This was pre-Internet - if only we could've guessed what the next 25 years would bring.) I agreed with all of his points, but then pointed out that the carrying capacity of the earth before modern agriculture, modern medicine, fossil fuels, etc. was roughly 1/20th what it is today, and did he really want to kill 95% of the people alive today? He responded, "Man, I feel shitty now." That's the nature of the scale vs. justice dilemma: you can't critically analyze scale & progress without also acknowledging the huge injustices it's created, but at the same time, reversing scale literally means death & suffering.
Not to be glib, but there's no indication that continuing to scale won't result in the same death & suffering.
And yeah, I wouldn't be alive today (and probably would've never been born in the first place) if not for hyperindustrialization, but... I really wonder if that's a bad thing.
Would the world be better off if after making a product more efficient, our first reaction wasn't "Great! Now we can scale up production!"?
> I wonder if folks who say "if large scale comes at the cost of injustice, maybe large scale shouldn't be a thing" have really thought through the implications of this.
I'm saying companies should be accountable for the loss they cause - if you're a Amazon Seller and you get your listing removed, you get an auto-reply and you try to talk to a human just to find a stream of auto-replies pasted together, because is in the sake of scale and progress the it's ok?
Same goes for copyright claims on Youtube. You end up talking to auto-replies.
Why should people livelihood be the cost to pay for scale? Why can't Google or Amazon pay people to supervise these replies and make amendments when the help is not only requested but required?
I don't buy that "cost of progress" argument. I'm pro progress, I'm just against half baked solutions that seem to be progress but end up being cost saving measures.
Why can't a user who got wrongfully harmed be compensated? Is that against scaling? Or shouldn't that be a cost of scale?
Currently in my opinion it's irresponsibility. Irresponsibility disguised as Policy & Terms Agreements.
At least in the EU Amazon is being pressured to cut this bullshit and act properly.
All the businesses you mentioned still have core products that are now reliant on software - customer service, scheduling appointments, marketing, analytics/business decision making and lots more.
The two auto shops I've used in recent years were cases where I called someone on the telephone. I don't know how they handle scheduling on their end. It wasn't part of the core interaction loop.
Now that I think of it, the place I now prefer to go isn't just because they do good work but because of my human interactions with the people there. They were more personable. The experience felt better because of the people.
None of the core part of my interaction required software.
To continue the auto shop example, sure they use software for their invoices and to detect problems in vehicles. They don't have to do that for invoices. It's a choice. It's not in the critical path.
Most small businesses shouldn't have to spend a ton of time marketing. That's outside their core work. Marketing takes on a variety of things including word of mouth. That will change over time and sure, people use software based ads. But, it's not required and not the core of the business.
> When I think of businesses I think of the local auto repair shop, the restaurant around the corner, the paving company working on the roads, the local HVAC company around the corner from me, and the local gyms.
That makes sense that you’re thinking of those local businesses. They’re the ones that are left after we nuked taxi companies, record stores, radio stations, bank tellers, and retail. We haven’t totally killed all those yet but we’re working on it, along with grocery stores and more.
I agree that putting more computers in between people everywhere won’t be an entirely good thing, but there’s no denying that this is happening and has been happening everywhere we can pull it off.
People hardly even talk to each other these days without computers in the middle.
Literally all of vsyu's posts are submission links to either confluent or buzzsprout talking about Apache Kafka. No comments, just banal advertisement.
While I mostly agree with Jay Kreps, I believe that the upfront costs of event sourcing systems are way higher than traditional ones.
Kafka itself seems to have minimum hardware requirements way above something like a single-node Mongo or Postgres installation, let alone the infrastructure to create projections.
True for a lot of industries. Unfortunately, there are two corporate responses to this: 1) Software is a strategic competitive advantage and market differentiator and 2) Software is a cheap commodity that can be outsourced and cobbled together from lowest-bidder parts.
Not only do both attitudes exist at different companies but, in my experience, a single company can flip between "software is our secret sauce!" to "let's focus on our core strengths and outsource the rest!" in a heartbeat depending on what the venture capitalists say that week.
Using some SAAS is often smart and useful but if you're outsourcing your core business processes and new development for the sake of short-term efficiency, you're handing the keys of innovation for your company over to somebody else.
36 comments
[ 2.7 ms ] story [ 77.8 ms ] threadBut they leave out the other big part of running a company, which is getting the low-paid worker drones to show up on time and make the company money.
I feel cheated. I was ready to defend the non-software companies of the world, but the article has very little to do with the title. It seems like just an ad.
Would that clean up the Web or what!
There's a difference between offering up a startup-oriented forum that still encourages people to post anything that is intellectually stimulating and interesting, while still allowing it to be an outlet for YC-backed startups to post and share things for discussion vs. posting a news article that piggybacks on Andreessen's well-publicized and becomes marketing spin about a specific product after a few introductory paragraphs. Just like how there's a difference between allowing a few "sponsored posts" on a search page, and having all of the front page be sponsored posts. Does that make more sense?
Do you disagree that there's a difference between those two things?
Yes. Although, I didn't make the assertion that they are the same.
Since I already knew the company, when I saw the headline and their domain, my immediate reaction was: "Of course they would say that. They are a software vendor!"
Only this time it almost failed me. The top comments were lamenting on how misleading the headline is. They were correct of course - but nonetheless it is in in-depth article on Apache Kafa discussing it's origin story, use cases, how it was used at Linkedin, and most interesting how it used to replace ACID applications with eventual consistency.
All in all, a very good article, despite the lead in waffle at the start.
Do you mean this?
https://a16z.com/2011/08/20/why-software-is-eating-the-world...
When I think of businesses I think of the local auto repair shop, the restaurant around the corner, the paving company working on the roads, the local HVAC company around the corner from me, and the local gyms.
The core products at the boot camp workout place across the street from me don't need to be specified, monitored, or executed in software.
Somehow this reminded me that empathy is decreasing in society (you can search on it). Putting more computers between businesses and people isn't necessarily going to be better for people or the way we interact with each other.
In fact it should be made illegal to have people who's livelihood depend on a platform to be left at the hands of shit auto replies. If large scale comes at the cost of injustice, maybe large scale shouldn't be a thing.
Youtube, Amazon Seller Central, and many others are terrible examples of what future holds for a lot of people.
Personally I like it - I usually pick the self-checkouts, kiosks, and apps over talking with a real human being, who frequently acts like a robot anyway and tends to be slower and make mistakes more often.
Also, I wonder if folks who say "if large scale comes at the cost of injustice, maybe large scale shouldn't be a thing" have really thought through the implications of this. I remember talking to a classmate in high school about Native American tribes, industrialization, American history, and how we really fucked over this idea of a tribe or village or community and the world was a lot less richer because of it. (This was pre-Internet - if only we could've guessed what the next 25 years would bring.) I agreed with all of his points, but then pointed out that the carrying capacity of the earth before modern agriculture, modern medicine, fossil fuels, etc. was roughly 1/20th what it is today, and did he really want to kill 95% of the people alive today? He responded, "Man, I feel shitty now." That's the nature of the scale vs. justice dilemma: you can't critically analyze scale & progress without also acknowledging the huge injustices it's created, but at the same time, reversing scale literally means death & suffering.
And yeah, I wouldn't be alive today (and probably would've never been born in the first place) if not for hyperindustrialization, but... I really wonder if that's a bad thing.
Would the world be better off if after making a product more efficient, our first reaction wasn't "Great! Now we can scale up production!"?
I'm saying companies should be accountable for the loss they cause - if you're a Amazon Seller and you get your listing removed, you get an auto-reply and you try to talk to a human just to find a stream of auto-replies pasted together, because is in the sake of scale and progress the it's ok?
Same goes for copyright claims on Youtube. You end up talking to auto-replies.
Why should people livelihood be the cost to pay for scale? Why can't Google or Amazon pay people to supervise these replies and make amendments when the help is not only requested but required?
I don't buy that "cost of progress" argument. I'm pro progress, I'm just against half baked solutions that seem to be progress but end up being cost saving measures.
Why can't a user who got wrongfully harmed be compensated? Is that against scaling? Or shouldn't that be a cost of scale?
Currently in my opinion it's irresponsibility. Irresponsibility disguised as Policy & Terms Agreements.
At least in the EU Amazon is being pressured to cut this bullshit and act properly.
Why not guarantee proper customer service instead of forcing a half baked solution?
Now that I think of it, the place I now prefer to go isn't just because they do good work but because of my human interactions with the people there. They were more personable. The experience felt better because of the people.
None of the core part of my interaction required software.
To continue the auto shop example, sure they use software for their invoices and to detect problems in vehicles. They don't have to do that for invoices. It's a choice. It's not in the critical path.
Most small businesses shouldn't have to spend a ton of time marketing. That's outside their core work. Marketing takes on a variety of things including word of mouth. That will change over time and sure, people use software based ads. But, it's not required and not the core of the business.
That makes sense that you’re thinking of those local businesses. They’re the ones that are left after we nuked taxi companies, record stores, radio stations, bank tellers, and retail. We haven’t totally killed all those yet but we’re working on it, along with grocery stores and more.
I agree that putting more computers in between people everywhere won’t be an entirely good thing, but there’s no denying that this is happening and has been happening everywhere we can pull it off.
People hardly even talk to each other these days without computers in the middle.
Kafka itself seems to have minimum hardware requirements way above something like a single-node Mongo or Postgres installation, let alone the infrastructure to create projections.
2030 Every Company Is Becoming Smart Contract
Not only do both attitudes exist at different companies but, in my experience, a single company can flip between "software is our secret sauce!" to "let's focus on our core strengths and outsource the rest!" in a heartbeat depending on what the venture capitalists say that week.
Using some SAAS is often smart and useful but if you're outsourcing your core business processes and new development for the sake of short-term efficiency, you're handing the keys of innovation for your company over to somebody else.