Convex的动态

Microsoft Veteran Ditches Team Tabs, Blaming Storage Trauma of Yesteryear: Veteran Microsoft engineer Larry Osterman is the latest to throw his hat into the "tabs versus spaces" ring. From a report: The debate has vexed engineers for decades -- is it best to indent code with tabs or spaces? Osterman, a four-decade veteran of Microsoft, was Team Tabs when storage was tight, but has since become Team Spaces with the advent of terabytes of relatively inexpensive storage. "Here's the thing," he said. "When you've got 512 kilobytes, and you're writing a program in Pascal with lots of indentation, if you're taking eight bytes for every one of those indentations, for eight spaces, you could save seven bytes in your program by using a tab character." It all added up, even when floppy disks were part of the equation. However, according to Osterman, things have changed. Storage is less of an issue, so why not use spaces? A cynic might wonder if that sort of attitude has led to the bloatware of today, where software requires ever-increasing amounts of storage in return for precious little extra functionality and a never-ending stream of patches. Any decent compiler should strip out any extraneous characters, assuming the code is indeed being compiled beforehand and not interpreted at run-time. For his part, Osterman is now a member of team spaces. "I like spaces simply because it always works and it's always consistent," he said. Read more of this story at Slashdot.

A decade back, I took ownership for a large distributed system inside Microsoft. The codebase created a shared library that ran on 10s of thousands of machines. It was constantly improved on for several years and had accumulated a lot of details in it. Not to mention, the codebase also came with a gift of several thousand bugs / issues / suggestions. With a team of just about 50 people, we were definitely drowning in this technical debt. Reading the article https://lnkd.in/ggScM3VQ, I recall several of these steps we had to take for taming the technical debt. Collectively with the team we took several steps a. Review and triage the issues and bugs b. Continue to innovate while reserving capacity to fix bugs c. Reduce or throw own old systems; embrace new open source systems when/where possible d. Rewrite completely busted components (especially the ones that had single points of failure) e. Improve test cases to enhance productivity One thing was certain: despite the theory and practice of #ZEROBUGS, there are issues that need to be addressed. We just have to take pragmatic steps to continuously improve the systems.

"Engaging with customers and understanding their requirements and pain points is vital when evolving or reconstructing a major platform component. This approach was instrumental in shaping the blueprint for MSM. By applying first principles, we deconstructed the problem into its smallest components and envisioned a system that delivers the flexibility and extensibility required for the platform. This solid foundation enabled us to rebuild from the ground up with clarity and purpose, ensuring the platform meets current demands while remaining adaptable to future challenges. This blog has only scratched the surface of the asynchronous platform we’ve built over the past few years, and we’re constantly looking for new ways to improve our infrastructure. We’re excited to, in the future, dive deeper into other critical design decisions that help us build a more efficient and useful Dropbox!" https://lnkd.in/e76wPfQ6

Ramping up our #snowflake project here and, recalling its original key value proposition, I see a parallel with how we all have to navigate the modern business world: separating storage & compute! I've written about my favorite notes app before, and without that & other tools to store & organize our knowledge it would be a heck of a lot harder to accomplish our actual #data work.

Loop is stepping up its game in the database space by adding a feature that Notion had out of the box. This move is expected to give Loop a significant advantage over its competitors. Stay tuned for more updates on this exciting development! #database #technews

Patience is a Virtue! Understanding Your Storage Charts in Microsoft Fabric Capacity Metrics. #MicrosoftAzure #AzureFabric #DevOps #CloudComputing #DataStorage #Metrics #AzureMetrics #CloudManagement #ITInfrastructure #TechCommunity #CloudStrategy #AzureDevOps #StorageSolutions #DataAnalytics #CloudEngineering #Onelake #FabricCapacityMetrics

Hey Coders! Did you know what caused the recent Microsoft outage? Let's dive into the technical details! We are excited to share the latest research from our talented member, Dipanshu! Check out Dipanshu's research on Major Microsoft Outage. Explore the root cause of the outage, its impact on different systems, and potential solution. ?#MicrosoftOutage #techbreakdown #cloudcomputing #business #economy #dataprotection

I've been intrigued and curious about companies who *could* set up a Rhino Compute server shared across multiple teams (and accessed using Hops) but choose not to - if you work in such a context, could you outline why this is so in your case? If it's at least partly due to difficulty in measuring the demands on the server and tracing them back to each team/project, then perhaps something like this could help: https://lnkd.in/gJirvaaH

Just stop the Bulls**t... Yay! more outages, when faced with telling your customers what is going on, this is the typical word salad response crap we are being shoveled and gladly ask for more like its the socially acceptable response. From Microsoft today - 'We've identified multiple workstreams and are working to mitigate impacted workstreams by performing failover operations.' Enough, not just from Microsoft, but others, with their half ass rushed error handling in applications like the vague errors Outlook can give you. It's time for developers and big companies to actually talk straight to us with outages and errors in the apps, that way perhaps we can get a better handle on fixing issues and not looking like a complete idiot when standing in front of a client. Why is my business down and not running?!.... umm.... PC Load Letter....

While not untrue in general, there are two factors to consider: 1) Can you actually efficiently use a big server without virtualization? It's far harder to do than the advice implies. Usually system bottlenecks on something but not everything at once, especially if you are dealing with a single physical server. In the cloud you can often use extra bandwidth or disk IO which is above your fair share if box you are on is not loaded to the max. 2) What is your starting point? Is writing custom scripts for local server worth the time or you have good recipes how to deploy and manage your app with little to no effort already?