He summarizes it well on Twitter:
The fundamental issue: IFTTT is a roll of duct tape that has decided to no longer be sticky, and instead license its recipe for glue
Haha I love that.
I'm having a hard time grasping the change here, can someone clarify?
It sounds to me like IFTTT used to do all the API work in-house: they'd write the code to, essentially, translate from one service's API to another service's API. But now they're switching to a model where each service needs to implement this new IFTTT API.
Is that right?
If yes, it's a remarkably radical shift in how they do business — now you'd need an IFTTIFTTT service that can now do the 'shim' work that IFTTT used to do!
Yes, I believe you have the right idea.
Now that they're well established, they demand that service providers integrate with them (instead of the other way around). It sounds a bit like a case of embrace, extend, exterminate to me.
That was unfortunate.
Stick it to the man Maciej!
I switched my Pinboard recipes over to Zapier yesterday. Hopefully it works just as well. I was even able to get GitHub stars -> Pinboard working again after GitHub killed that on IFTTT.
I wonder how many other products have the same sentiments or how many other developers have abandoned the projects and will just stop working?
Does this come across as whiny to anyone else?
IFTTT has every reason and justification to drop a platform that doesn't use it's new API.
I think IFTTT has a reason, yes. But I think I'm with him in the T&C. It nuts.
I didn't read over their entire T&C, but the points he made were 100% valid in that regard.
I just think that (given the new size of IFTTT), app developers stand to benefit substantially more than IFTTT would. This is of course not true in every circumstance, and I'm willing to bet that applications with a large user base are catered to differently.
It was a free integration, and initially they relied on applications to prop them up, so they spent dev time integrating with those applications.
Now that they're providing more of a service and need those smaller applications less, they're off loading the development time to them as they have more to gain.
I don't believe it's incredibly underhanded, or even morally wrong. If the owner of that application doesn't believe that it's worth the burden of reintegrating their product, they don't have to do it.
I reckon that IFTTT can do whatever they deem best for their future, but I totally agree that T&C are well beyond sensible levels.
Why should a service write a separate API from the one they already have just to cater to another service that doesn't do anything for them in return?
If I want to use data from a service, I use their API, not ask them to write a new one that works the way I want.
Exactly right. Pinboard is a one-person shop and already has a solid API. IFTTT worked with the Pinboard API for years and their model changed while Pinboard is still providing the same great service.
Pinboard is a bootstrapped business with paying customers. IFTTT, unfortunately, has venture capitalists as customers so their interests do not necessarily align with the users of their service – who want an interface to an API for Pinboard and other services.
If anyone thinks this is just about Pinboard complaining (which Maciej does love to do!), take a look at some other recent services gone from IFTTT:
App.net, BuzzFeed, Campfire, Etsy, ffffound, Last.fm, Readability, Yahoo Fantasy Sports, Yammer
Saying that IFTTT does nothing for Pinboard is disingenuous. The relationship greatly benefits smaller platforms like Pinboard, because IFTTT has such a large userbase.
I agree with the points me makes about the T&C, they're far-reaching and should be reeled in, but the fact is that IFTTT no longer needs/wants to get data from Pinboard if it requires them committing development time.
In my opinion, Pinboard stands to gain more from being on IFTTT than vise versa, and if IFTTT is requiring platforms to have API's that follow a certain model, that's their decision and it's up to the owner of the service to determine if it's worth their time to build it.
If it isn't, then it isn't and you certainly don't have to, but expecting IFTTT to make decisions against their best interest is unreasonable at best.
It sounds like you may be talking partly about interests of Pinboard users and not the interest of Pinboard, the business. Not that they aren’t closely connected, but if he invests $10,000 (for the sake of argument) of development time and server costs in IFTTT, is he likely to get more than that back in user subscriptions? Maybe – but I doubt it. And what if it's $20,000? He’s serious when he talks about being lazy, but the implication is that writing custom code, even for a really useful feature, costs him significant time and server resources.
Just like IFTTT’s interest is in getting people to make APIs work with them, Pinboard’s interest is in keeping existing customers and getting new subscribers. It is possible that IFTTT support doesn’t really figure into Pinboard’s bottom line at all. He isn’t asking them to do something against their interest – they’re asking him to do something against his interest. In this article, he credited them for doing hard work in the first place in creating a connecting API, but he’s not suggesting that they should keep it going forever. In fact, he is critical of their business model and would likely suggest that they don’t keep it going. Pinboard even offers IFTTT-like Twitter archiving integration that most users probably don’t even know about.
As a long-term Pinboard user and a long-time regular IFTTT user, I can say that I get value out of the Pinboard API every day, but I can take or leave IFTTT support. After all, Pinboard still has an API.
I completely agree, but I believe that Pinboard would have a much better chance of recouping their investment in terms of new users from IFTTT than vise versa.
If IFTTT invested their time to support Pinboard, would they recoup the costs in terms of new users who signed up solely because of the Pinboard integration? Very unlikely.
I don't think they're asking him to do something against his interest as much as they're saying they're not going to do something against theirs, and that hopefully Pinboard believes that it is in their best interest to integrate with them.
Otherwise it's a mutual decision that neither party benefits enough from the integration of one another to dedicate time to it, and we have the situation we're currently in.
I just feel like painting IFTTT as the "bad guy" is inaccurate and reactionary. For the users, it's unfortunate that it won't work anymore. But if neither IFTTT or Pinboard believes that it's worth their time, you cannot fault them for not doing it.
This will of course change if either platform sees a drop in user base as a result of this decision, which in a simplistic view, is how the market is supposed to work.
Very reasonable. I agree that IFTTT is not the bad guy. They will suffer for shutting down channels though – and Pinboard is one of many.
Maciej gets passionate about these matters, and he likes to pick fights with perceived competitors and antagonists, particularly if they are venture capital funded.
He's annoyed because IFTTT used to do all the work to tie services together themselves (ie Pinboard never actually integrated/wrote any code for IFTTT's V1 API). Now, IFTTT has written a new API and instead of doing the same thing in-house again, they want to get developers to integrate. However, they're being passive-aggressive about it, and now they have shitty terms of service.
They're sending out emails that are making it seem as if companies like Pinboard are being obstinate when in reality IFTTT is being manipulative. I think Maciej is perfectly justified in writing this.
I think it's also the way that IFTTT has let the customer know about this change and implied that it's Pinboard's fault, rather than the fact that IFTTT has changed.
It's also that they are doing a hard cut over. Other larger companies have legacy APIs and just recommend people upgrade (or only support the latest).