Live data from Hacker News

FarmLogs

blog.samaltman.com

81–90 of 92 posts

Re: FarmLogs

#81
post #35

The problem with FarmLogs is the same problem with anyone wanting to provide cloud-based services to agriculture: who owns the data? And specifically, FarmLogs Terms of Service clause 7 "Content" does not leave one feeling warm and fuzzy on this topic. This has become more of an issue with farmers over the past year, as they realize their data is a commodity just like what they raise--whether crops or livestock. How…

This is an excellent point. During university, I was part of a team that published a business case centred around the story of Vidalia Sweet Onions. This guy had developed a technology that took a sample of onions from a farmer's field, crushed the onions, and analyzed them for the characteristics that made for sweetness. The batches were tagged with grid numbers to enable farmers to confirm which plots of land had the type of soil necessary to make truly sweet onions. This then enabled the farmers to guarantee the sweetness level of various onions. The old guard was absolutely against it, as the collective of farmers that sold under the brand of Vidalia onions partly depended on the mysticism of not knowing which plots of land were able to generate sweet onions. Without this testing technology, the farmers were all on a level playing field, irregardless of their actual soil composition. The testing technology was able to clarify which emperors had no clothes. Whether a farmer can or should be able to own the data that is critical to his success is a question people never really had to consider before.

I liken it to mining. The companies who dig for various mineral deposits make significant investments into exploration of the ground. Their exploration generates a ton of data. They use that data to determine where to dig further and estimate the financial potential of the deposits. This then determines company stock values. It's ludicrous that such data may not be owned by the mining company that generated it (or acquired the rights for it). Obviously, what's happening with farms and FarmLogs is not a perfect analogy. But the parallels are interesting.

Re: FarmLogs

#82

As an early user of FarmLogs (and as a programmer), I found it really useful. All I wanted to do was find out average rainfall on my land and make a way to denote the prime spots for crop planting on the land (purely as a hobby). It was super useful and intuitive for that, something that I couldn't get to make happen easily trying all sorts of things with Google Maps, various mapping APIs, and Google Earth.

As another farmer and programmer, who isn't based in the US, I have found all of the most desirable features (like rainfall data) are not available. I have tried it out, but there is nothing that keeps me coming back. 20% of American row crop farmers is still fairly impressive, but it will be interesting to see how they try to approach the rest of world.

Re: FarmLogs

#83
post #78

I worked on a sugarcane farm management company a while ago and also built a backbone.js frontend site that tracked sugar cane harvesters and estimated the "outcome" to send trucks to offload it on demand with some colleagues (closed source unfortunately) in resume I like this area very much. When I saw this post I went running to the Jobs page, and there my heart got smashed when I saw Clojure on backend positions..…

It's Clojure not Closure. Why don't you like it?

> I am a plain Java dev with learning skills (worked with Groovy/Python/JS a while ago)

Perhaps he's promoting some other JVM language -- he did mention Java and Groovy. If you type "Groovy" into HN search, you'll notice more than the usual amount of Groovy stories being submitted over the past week, and the recent comments mentioning "Groovy" are scattered around different submissions rather than clustered as usual. Groovy has more than doubled its percentage on Tiobe between January and February (if you get what I mean), and its project manager has just finished running a campaign among his Twitter followers to double its Github stars from under 600 a month ago. Because the Groovy and Grails project managers are now competing for their funding, expect to see more comments of that nature.

Re: FarmLogs

#84
post #83

Earlier quoted context omitted.

It's Clojure not Closure. Why don't you like it?

> I am a plain Java dev with learning skills (worked with Groovy/Python/JS a while ago) Perhaps he's promoting some other JVM language -- he did mention Java and Groovy. If you type "Groovy" into HN search, you'll notice more than the usual amount of Groovy stories being submitted over the past week, and the recent comments mentioning "Groovy" are scattered around different submissions rather than clustered as usual.…

I think that's cynical (if I read it correctly), but maybe true? When money gets involved all kinds of weird things happen.

Rich Hickey is on the record as saying that he doesn't want to promote Clojure, and if people like it they like it. He'd rather have people who came to it voluntarily than who were marketed to.

I would like to know real answers to why people don't like it. I've used a (normal) number of languages over the years but I think Clojure is by far the best for all kinds of reasons.

It's interesting to talk to people who have gone as far as trying it but found they preferred something else. I can understand you might be put off from a distance by the parens, or the immutability or whatever, but opinions change (or don't) once you actually try something.

Re: FarmLogs

#85
post #83

Earlier quoted context omitted.

> I am a plain Java dev with learning skills (worked with Groovy/Python/JS a while ago) Perhaps he's promoting some other JVM language -- he did mention Java and Groovy. If you type "Groovy" into HN search, you'll notice more than the usual amount of Groovy stories being submitted over the past week, and the recent comments mentioning "Groovy" are scattered around different submissions rather than clustered as usual.…

I think that's cynical (if I read it correctly), but maybe true? When money gets involved all kinds of weird things happen. Rich Hickey is on the record as saying that he doesn't want to promote Clojure, and if people like it they like it. He'd rather have people who came to it voluntarily than who were marketed to. I would like to know real answers to why people don't like it. I've used a (normal) number of language…

I used Clojure a lot as a hobbyist language for 2 years, then changed to another one. I found I missed the immutability the most, even though I hardly wrote multi-threaded code. Once I was thinking "concurrently", going back to having to think about whether a list or map was in the right state took some getting used to.

As for what I couldn't get used to in Clojure, even after 2 years the circumfix parentheses still seemed unnatural, even though I know they're necessary for macros. I guess using f(x,y) notation in school and one's first programming language is a bond too hard to break.

Re: FarmLogs

#86

This type of story is so motivating to build vertical software, but for people like me who are trying to identify a vertical software opportunity, it's discouraging that these developers needed to spend a couple years in the industry to truly understand the pain point behind FarmLogs. Does anyone have thoughts / strategies around identifying these opportunities, ideally without the need to spend years in the target i…

Unfortunately, you need to gain a lot of tacit knowledge about an industry to really solve the right problems for it. That takes time and dedication. There are some ways to speed it up by talking to lots of experienced people, reading analyst papers, researching, or even shadowing a person in the industry for a week. However, nothing beats first hand knowledge.

I think you get at a great point that the onky way to solve the right problems is to be committed to learning about an industry through all available information channels. To be comfortable spending that much time on one specific industry though I think means you need to come in with a thesis to begin with, which to avoid circular idea justification means you probably need to come into it with a broader, more abstract thesis independent of industry knowledge.

Re: FarmLogs

#87
post #83

Earlier quoted context omitted.

It's Clojure not Closure. Why don't you like it?

> I am a plain Java dev with learning skills (worked with Groovy/Python/JS a while ago) Perhaps he's promoting some other JVM language -- he did mention Java and Groovy. If you type "Groovy" into HN search, you'll notice more than the usual amount of Groovy stories being submitted over the past week, and the recent comments mentioning "Groovy" are scattered around different submissions rather than clustered as usual.…

Oh shit what a conspiracy theory...

I played with Scala too... Maybe I am receiving a paycheck from all other langs besides Clojure to kill the damn thing.

Come on...

I just don't like the LISP like syntax and the general idea of it and I think that other languages like can provide the same nice pieces it does... Simple... Simple... No aliens, JFK Zombie killers, Umbrela corps or anything.

Re: FarmLogs

#88
post #78

I worked on a sugarcane farm management company a while ago and also built a backbone.js frontend site that tracked sugar cane harvesters and estimated the "outcome" to send trucks to offload it on demand with some colleagues (closed source unfortunately) in resume I like this area very much. When I saw this post I went running to the Jobs page, and there my heart got smashed when I saw Clojure on backend positions..…

It's Clojure not Closure. Why don't you like it?

Thanks I've corrected it.

Lisp like syntax all those (((())))))(()() drive me crazy and I think Scala has the features of Clojure that I like without the LISPness.

But I have nothing against people using I just don't like it.

Re: FarmLogs

#90
post #88

Earlier quoted context omitted.

It's Clojure not Closure. Why don't you like it?

Thanks I've corrected it. Lisp like syntax all those (((())))))(()() drive me crazy and I think Scala has the features of Clojure that I like without the LISPness. But I have nothing against people using I just don't like it.

I thought you'd say something like that. Have you counted how many brackets though? It's often very close.

    (defn blub-extra [a b]
        (blub (inc a) (inc b))) 
8 parens + 2 brackets

    def blubExtra(a: Int, b: Int): Int {
       blub(inc(a), inc(b))
    }
8 parens + 2 braces

    Integer blubExtra(Integer a, Integer b) {
        return blub(inc(a), inc(b));
    }
8 parens + 2 braces

It's roughly the same numbers of brackets (or equivalent) in Clojure, Scala and Java. They're just in slightly different places.

(excuse syntax I've not written the last two for years)

Post reply on HN