Or "Why Architecture Matters." Google Pay’s disastrous year continues, promised bank account feature is dead.
And yes, even at a Tech Company, the bean counters are in charge. "Why can't you just..." is a bean counter question. There is no right way to answer that question. Or rather, they will ignore every answer. (Even though they know squat about tech.)
Let's build a system aimed at a specific country that has a very distinct way of doing things, that differs from the rest of the globe, and then we will launch that system globally. "What could go wrong?"
In March of this year, Google Pay launched a disastrous revamp Google shut down the perfectly good, mature service that had been around since the days of Google Wallet, and the company replaced it with a brand new codebase that was originally developed in India. The India-centric Google Pay (which was promoted to be the new, worldwide Google Pay) was modeled after WhatsApp and used your phone number as your identity. Requiring a SIM card for identity meant that the Google Pay website functionality had to be shut down, multiple accounts were no longer supported, and users had to re-build their contact lists to send money.
This is the 2nd time that Google has done this, because they didn't learn their lesson the first time. Do you think they will learn it this time, or will the bean counters refuse to admit that they had a hand in this? The new Google Pay repeats all the same mistakes of Google Allo. (That link is from March. But back to the present day...)
Google Pay is currently in an awful limbo state, and it's not clear what the service's future is. Plex [the promised digital banking feature] was supposed to be a major feature of the new Google Pay. But with it dead and the architects of this entire revamp no longer at Google, the division feels like a rudderless mess.
People who don't understand anything about tech, like executives, always ask, "Why can't you just do X?" When you tell them why you can't - or not for the amount of money they are willing to invest - they get angry and say "Do X anyway, because I said so." This is the kind of thing that results.
And while I never worked at Google, I have had that conversation (it isn't a conversation if you aren't listening to me) from any number of bosses. It is probably one of the main reasons that I left so many jobs. I have been yelled at for letting bosses do exactly what I told them not to do. They also don't like it when you have documentation of what you told them.
Personally, I think using SMS as an identifier in the face of all the problems with Signalling System 7 seems like a recipe for disaster. It isn't so much an issue of security, as the lack of security.
SS7 security, right. It was designed to run between known, trusted facilities over a specific trunk. It didn't occur to them back in the day that they needed some way to validate a sender. And it was never properly updated for VOIP over the open Internet. Spoofing stopped being funny a very long time ago. It has gotten people killed.
ReplyDeleteWell, I suppose it's more secure than in-band signalling.
Bean counters are one of the main reasons Compaq died and yes even after 30, 40 50 years of their doing really stupid things they never learn because you should always pay the abosolute lowest amount you can no matter what
ReplyDelete