This is literally throwing the baby out with the bath water. If we couldn't depend on other products to make our own product, then the computer industry is pretty much dead and we'll all have to go back to farming the land (and, if you're in SoCal like me, even then you'll have to depend on the mercy of our northern neighbors for water).
I think what you mean to say is "don't build a product that depends on LinkedIn", and to that, I absolutely agree. LinkedIn, with its chain email spamming and forced mass invites, is literally the post-child for internet dark patterns and unethical behavior. Furthermore, the sheer volume of people on there automatically force recruiting agencies to use bots (or at least copy-and-paste) to spam folks on there, and in general leads to bad returns for the recruiter and tons of spam for you the programmer.
No he did not mean just don’t depend on LinkedIn. He meant don’t place the fate of your software and company in the hands of any third party. The list of companies who have burned developers includes Twitter, Apple and Google just to name the three spectactular examples. Facebook has been somewhat better to their developers though there have been a few instances there too.
Unless you have an ironclad contract with whomever you are partnering with(whose API you are using) and know exactly what you get from them and they get from you and for how long, you should not be depending on them.
If you want to spend a few weekends writing to some API and don’t care if the rug gets pulled out from under it, and all your users then fine. If you want to support an API in a product that will survive without that integration that is also fine. Just don’t gamble your entire product and your future user base on API's that can disappear overnight when some dickhead MBA decides he isn’t making his numbers and the free loader third parties aren’t worth it any more.
Facebook cut their API heavy when they phased out v2 last month. No access to friend data and limited access to user profile. LinkedIn is following FB's lead.
But to your point, LinkedIn's practices can be a little pushy, but the site still provides decent value to me (it got me my current job). I can put up with ignoring mass invite requests, recruiter messages, and configuring my notification settings to maintain this professional network. It may get me my next job in 5 years, so for me it's totally worth it.
If a component from one manufacturer becomes unavailable (or that manuf. isn't playing "fair"), another one can usually be sourced.
That said, I think that: "don't make your service depend on other services without some sort of contract or agreement to provide" is a better rule of thumb than the one above.
Building your company around the core business of LinkedIn/Facebook/etc is dumb, because you are nothing more than a complement to their core mission. Best case, you become a commodity and achieve modest success, but the oxygen gets cut off at any time.. If you're very successful, you become a feature of the mother product.
When you're Apple buying memory from Samsung, that's different. Samsung's output is memory chips... that's the business. If nobody buys memory chips, they lose money.
On the other hand Apple, Google, Faceobook, Twitter, Amazon and LinkedIn are increasingly sucking all of the oxygen out of the rest of the software ecosytem. Software developers are increasingly being painted in to a corner where all the users are sitting on those platforms all the time and unless you are there too its really hard to get eyeballs to look at your software.
Developing for the open web and trying to get people to websites in the current mobile/social era is increasingly brutal unless you are willing to be completely dependent on a handful of big companies.
The difference is that the core competence of Intel is providing chips to other businesses. It's not a stretch to say everything that goes into an Apple product is made by companies that specialize in selling components to other companies. If some supplier doesn't work out, there are a dozen others to pick up the slack.
Basing your product on top of LinkedIn? The core of their business isn't anywhere near providing that API.
Comments
Don't make your product depend on other products. It's pretty simple, really.
Follow that rule and you won't need to worry.
This is literally throwing the baby out with the bath water. If we couldn't depend on other products to make our own product, then the computer industry is pretty much dead and we'll all have to go back to farming the land (and, if you're in SoCal like me, even then you'll have to depend on the mercy of our northern neighbors for water).
I think what you mean to say is "don't build a product that depends on LinkedIn", and to that, I absolutely agree. LinkedIn, with its chain email spamming and forced mass invites, is literally the post-child for internet dark patterns and unethical behavior. Furthermore, the sheer volume of people on there automatically force recruiting agencies to use bots (or at least copy-and-paste) to spam folks on there, and in general leads to bad returns for the recruiter and tons of spam for you the programmer.
No he did not mean just don’t depend on LinkedIn. He meant don’t place the fate of your software and company in the hands of any third party. The list of companies who have burned developers includes Twitter, Apple and Google just to name the three spectactular examples. Facebook has been somewhat better to their developers though there have been a few instances there too.
Unless you have an ironclad contract with whomever you are partnering with(whose API you are using) and know exactly what you get from them and they get from you and for how long, you should not be depending on them.
If you want to spend a few weekends writing to some API and don’t care if the rug gets pulled out from under it, and all your users then fine. If you want to support an API in a product that will survive without that integration that is also fine. Just don’t gamble your entire product and your future user base on API's that can disappear overnight when some dickhead MBA decides he isn’t making his numbers and the free loader third parties aren’t worth it any more.
Facebook cut their API heavy when they phased out v2 last month. No access to friend data and limited access to user profile. LinkedIn is following FB's lead.
Figuratively, not literally.
But to your point, LinkedIn's practices can be a little pushy, but the site still provides decent value to me (it got me my current job). I can put up with ignoring mass invite requests, recruiter messages, and configuring my notification settings to maintain this professional network. It may get me my next job in 5 years, so for me it's totally worth it.
Figuratively, not literally.
Maddeningly, at least one dictionary disagrees: http://www.merriam-webster.com/dictionary/literally (2)
Most products depend on other products. e.g. Apple computers use components from other manufacturers.
Name one product that doesn't rely on another product.
If a component from one manufacturer becomes unavailable (or that manuf. isn't playing "fair"), another one can usually be sourced.
That said, I think that: "don't make your service depend on other services without some sort of contract or agreement to provide" is a better rule of thumb than the one above.
Additionally, the same model may be sourcing parts from multiple vendors (for instance, laptop displays)
Don't rely on free APIs to stay free. : P
It's not the same concept, at all.
Building your company around the core business of LinkedIn/Facebook/etc is dumb, because you are nothing more than a complement to their core mission. Best case, you become a commodity and achieve modest success, but the oxygen gets cut off at any time.. If you're very successful, you become a feature of the mother product.
When you're Apple buying memory from Samsung, that's different. Samsung's output is memory chips... that's the business. If nobody buys memory chips, they lose money.
On the other hand Apple, Google, Faceobook, Twitter, Amazon and LinkedIn are increasingly sucking all of the oxygen out of the rest of the software ecosytem. Software developers are increasingly being painted in to a corner where all the users are sitting on those platforms all the time and unless you are there too its really hard to get eyeballs to look at your software.
Developing for the open web and trying to get people to websites in the current mobile/social era is increasingly brutal unless you are willing to be completely dependent on a handful of big companies.
The difference is that the core competence of Intel is providing chips to other businesses. It's not a stretch to say everything that goes into an Apple product is made by companies that specialize in selling components to other companies. If some supplier doesn't work out, there are a dozen others to pick up the slack.
Basing your product on top of LinkedIn? The core of their business isn't anywhere near providing that API.