4 ms·
(warning rambling rant) I feel like this was super true in the pre Facebook days but they were humbled A LOT by the massive flop of Google plus. I honestly do
by wh-uws 6y ago
(warning rambling rant)
I feel like this was super true in the pre Facebook days but they were humbled A LOT by the massive flop of Google plus.
I honestly don't feel like I get the my s * doesn't stink vibe from them anymore.
Especially because everyone knows their Achilles heel now.
But for their ad revenue what would they be?
Everything else material from came from an acquisition or is infra stuff.
It's cool but they also had to learn the hard lesson of Hadoop too.
Yeah they built big table... But the rest of the industry standardized on Hadoop (at the time) because they never released an implementation and so they started losing out on good hires because they didn't want to get locked into proprietary Google infra.
I think you see them recognizing this with intiatives like all of the stuff around kubernetes.
But hey I wouldn't build anything on their (non open) stuff and get more scared by the day over my dependence on El Goog because they've shown time an again how little they care for us common plebs.
But I feel like they get better when they lose but they've only started losing more recently
(Rant over)
- thethethethe 6y ago> Yeah they built big table... But the rest of the industry standardized on Hadoop (at the time) because they never released an implementation and so they started losing out on good hires because they didn't want to get locked into proprietary Google infra. Hadoop and Bigtable are not the same thing. Bigtable is a noSQL database. Hadoop is a big data processing framework. Hadoop is actually an open source implementation of MapReduce, which was developed at Google and which has since been replaced by Flume
- atombender 6y agoOn the other hand, Hadoop does have HBase, which is their version of Bigtable. It was, for a number of years, a very popular columnar K/V store.
- manigandham 6y agoHadoop is a collection of things: HDFS (filesystem), Hbase (keyvalue database), MapReduce (data processing), and YARN (resource scheduling and job management). Hbase was open-source implementation of Bigtable while HDFS was a copy of Google File System. The divergence appeared with other processing frameworks like Spark but now there's Apache Beam to unify it all.
- andromeduck 6y agoLol Stadia launch was anything but humble. Absolute farce compared to Geforce now and with no compelling titles or experiences to give it a leg up.