raghava

> Their big break came in mid-2010, when Neil Currie, then an analyst at the investment bank UBS, bought parking-lot counts for 100 representative Walmart stores and published the results in a quarterly earnings preview. The number of cars in the parking lots, he wrote, suggested that Walmart stock was undervalued.

> Currie’s prediction proved correct. As word spread that satellite images were a reliable predictor of corporate profits, a range of investment funds began buying retail-traffic data from RS Metrics.

Many cargo-culting orgs can use this as an example to learn "corelation != causation".

1. Buy shares of X; or bet a high-stake deal with X's board to up the stock price by N percent within a period T. 2. Run some offers/programs - which are intended at the theater - not real revenue. Instead of usual 200 cars a day, store might see 400 a day, sales numbers and volume remaining unchanged 3. Objective achieved; use propaganda marketing to sell the idea to others 4. This ups the stock price anyway - cuz data! 5. Collect bounty from X's board for making good on promise 6. Profit!

The theater might not be as simple as what's seen in this example, but this is exactly what startups in one way or other call "hustle" - and possibly the art of faking it till you make it.

Why not a Raspi Zero W with a USB ready monitor? The price point is almost similar and the build will be more modular. Or even just a buncha USB/HDMI based screencasting devices, synced to a cast-server which casts content based on recipient nodes?

Had seen a simple conf room notification system on those lines (which just mentioned who has blocked the room, when is the next blocked meeting by whom, calendar etc)

Cool project.

Need to see if this works with PostmarketOS and off to buy one old tablet.

> Researchers monitored the known repeater FRB 121102 with the Lovell Telescope, a 250-foot-wide (76 meters) radio dish at Jodrell Bank Observatory in England, over the course of five years. They found strong indications of a 157-day activity cycle; 121102 seems to flare up for 90 days and then go silent for 67, the team reported in a new study.

> It's unclear what's behind such cyclic activity, though scientists do have a few ideas. For example, periodic flare-ups could be caused by a wobble in the rotational axis of a highly magnetized neutron star known as a magnetar. Or they could be linked to the orbital motions of a neutron star in a binary system.

Could this be the Dzhanibekov Effect ? Funny I just stumbled upon it today!

https://www.youtube.com/watch?v=1VPfZ_XzisU

If Earth can maintain a stable axis, such wobbly oscillating axis could be that star's "normal". Just guessing! Or may be am too wrong. I am not a physicist!

Had seen https://medium.com/@Regulatory/gaia-x-5-reasons-why-the-euro... on this topic. Few of the points made sense.

Isn't https://docs.microsoft.com/en-us/windows/iot-core/tutorials/... IOT-Core for that purpose, or am I missing it?

In fact, the author could actually dig further and look at the potential losses an "AI-fied" solution could bring forth.

1. Unexplainable algorithms that cannot demonstrate fairness and biased algorithms - causing firms to be dragged to court for discrimination - where AI was used for decision making which impacted lives/careers (lending, credit, recruitment, medical procedure suggestion, financial modeling etc - just to name a few)

2. Biased algorithms resulting in small tainted outputs that could later snowball into a larger loss that get built over slow leaks over time. (Few AI based cloud app/infra monitoring systems ending up deciding the wrong scaleout factor/sizing - based on past history but not considering real situational context/need - resulting in a net loss over a larger time)

3. Some AIfied solution just outright denying users the level of control that's really warranted. ("full automatic , no manual" mode). This mostly happens where the buyer never uses it firsthand but buys based on brochure/ppt walkthroughs, and real users are disconnected from the decision making ivory towers. The risk ibeing these systems getting into the way, instead of aiding productivity, they end up being another JIRA - a hassle one could really do without.

Off late (since 2018) Dell Latitude has been pretty troublesome.

https://www.dell.com/community/Latitude/Major-Latitude-7390-...

https://www.reddit.com/r/sysadmin/comments/8wbhhv/high_failu...

https://www.reddit.com/r/Dell/comments/ar1mas/the_latitude_7...

Ah, finally, one could build a PC that runs Slack, MS Teams, VSCode, ~200 Chrome Tabs and other necessary stuff.

    Der8auer recently conducted an experiment where he pushed an FX-8350 to 8.1GHz (on liquid nitrogen), and dropped it to 7.5GHz, then ran Cinebench R15 – but even at this level, the old processor still couldn’t match the single-core performance of modern Ryzen chips, being slightly slower than a Ryzen 5 2600X (not even a current-gen or high-end Ryzen).

    This aptly illustrates that there is a lot more to a CPU than pure clock speed, and underlines the kind of advances we get from progressing through different CPU architectures – with major IPC (Instructions Per Clock) increases – as time goes on.

Wow!

Guess the keywords { orchestration, AI/ML, distributed, streaming } are left out for future enhancement of this beautiful prose.

And when one does an RCA on why that's happening across our industry - the cause could be just one of

- severe craving for resume pimping - severe craving for ego-stroking - severe incompetence and inability to recognize and control complexity.

With https://news.ycombinator.com/item?id=22796017 being last month, I guess the people are now ready call this out in a more bolder way, and that's good!

    Some of these "cloud" places, though, don't have that kind of oversight. They build their systems to just auto-scale up, and up, and up. Soon, there are dozens, then hundreds, then even thousands of these crazy things. If they write bad code, and their footprint goes up every day, they don't even realize it. Odds are, the teams themselves have no idea how many "machines" (VMs in reality) they are on, or how quickly it's changing.

    There's no equivalent of my friend to stand there and tell them to first optimize their terrible code before they can get more hardware. The provisioning and deploy systems will give it away, and so they do.

    What usually happens is that someone will periodically run a report and find out something mind-blowing, like "oh wow, service X now runs on 1200 machines". How did this happen? That's easy. Nobody stopped them from needing 1200 machines in their current setup. People wring their hands and promise to do better, but still those stupid instances sit there, burning cash all day and all night.

Most accurate picture of shops struggling with the problem of doing cloud well.

Not only "cloud", but 3rd party APIs/managed services as well. Slab based pricing, pricing on various rates, pricing on data. what not!

I have come to believe that most of the cloud/API vendors make bill tracking so convoluted and hard on a definitive purpose, just so that people can't get easy visibility and early-warning of impending disasters.

Any firm/outfit that says "we want to use cloud for its elasticity" must be prepared for that "elasticity in the billing" too. If they are not, they will find themselves in deep troubles.

That's what am saying as well. Such thing works for SFDC. Not for any ordinary B2B firm that's struggling to gain their 10th/100th customer.

> The infrastructure used to run our Salesforce instance probably costs somewhere on the order of $100 per month. Yet, they charge us tens of thousands of dollars per month… how?

Not really. Unless the subscription is in tens of hundreds, that benefit-of-volume is not possible. Even more so when the solution involves bleeding edge stuff (AI/ML/real-time analytics etc). Easy to say, very very very hard to even achieve decent margins on such infrastructural cost. It's always a tradeoff between { feature-richness, time-to-market } and cost-efficient SaaS. Only the players who got a solid customer base would ever get to enjoy that privilege.

> Though it’s far down in the acronym, the identified pain is really where you should start with the customer. You should be able to answer in great detail why their life is currently painful, and how they handle the problems you solve today.

There's a slippery slope. Any "solution" won't help, in the era of "disruptive solutions". During the initial stages of industrial revolution, cities were scared of all the horseshit they thought they'd have to deal with. A 'con'sultant would've typically come up with an efficient way to dispose off all that manure into the sea. Only a mad scientist working on an automobile somewhere ended up solving that poop of a problem really. And in hindsight, every mildly lucky person would consider themselves "disruptive" without actually having to be one though. Such is our world! :D

[edit] just noticed that all comments in this thread that are critical of the narrative are voted down. one is fine, all of them?

> everyone hopes to docker-ify everything

It's actually k8s-ifying everything (docker-ifying was 2015-2018).

> nobody really embraces the zen of devops

I know exactly what you mean, as am struggling to get people around me understand the value of tooling, developer delight but to no avail.

Management, other non-tech teams (sales, product, Ops etc) that have learn/worked mostly in regimented setups have a long long way to go before they even start to understand what DevOps really was supposed to be - all about actionable faster feedback loops.

But, faster, actionable feedback loops also bring out the the orgs's / dept's systemic rot/muck in policies/deficiencies/politics and make it visible in-the-face. That's very threatening to many who put up a make-believe fakeshow that "our stuff's all cool, that other team/dept has all the problems". Hence, all that pushback against the new or throwing the tool in (docker/k8s/cloud) and claim that it would act as deep magic to fix the fuckery happening all over.

I learnt it over my exp at big, small, medium sized firms - over last 15+ yrs.

Only way to get to achieve DevOps (yeah, not "adopt"; one can't just 'adopt' that) is to get CTO/COO/CFO to really really understand the value. Else, lost causes!

I have just used calver (https://calver.org/) to keep it simple - and I use the date[ddmmyyyy]hour,min,second stamps of build time as a marker in the version. I have had to deal with the build basically being a recipe of taking multiple ingredients (libs, apps - each with their own lifecycles) and putting'em together in one image to get stuff running.

The image version in my case is a tracker of the last-known-good-set (of various ingredients) - as validated by the quick, pre-build tests.

I haven't had any major problems till now; may be I am just lucky to not have encountered the hassles w.r.t calver till now.

I stumbled upon one similar approach too, recently. https://worklifenotes.com/2020/02/27/automatic-version-incre...

Simple, functional!

When trying to save settings, I guess the backend is validating Personal Profile Url for uniqueness and it fails [ Settings -> Personal Profile Url ]; but uniqueness must be checked in the context of a different user only!

> It’s not a great world to be raising kids right now.

This seems to be like one absolutely weird and baseless moral panic.

What was good time then? During medieval times/WW1/WW2/60s/80/90s? If any, world's been the same all the time. The have-nots being looted by those who have 'em. Just that the tech makes it a lot visible, is all.

All of us have dealt with those kinds.

1. Overzealous ActionJockeys, who would just do shit - bothered only about local maxima ("I did mine! His thing is not yet done!" or "My code is superb; others' is total poop") Few variants are BrilliantJerk, who is supersmart but nobody wants to work with him. Also on same lines are some appreciation junkies and hero-role-playing characters , who always want to "save the day" - the LastMinuteHero types; but they'll never ever train others to hold their own as that could pose a peril to their heroism.

2. Dr AnalysisParalysis and his gang of bikeshedders. Could spend a week to choose the color of a button and while doing so earn a PhD in color theory.

3. Clueless Jimmy who just does what the leads ask him to do. He's a sure shot follower (in the lead-follow-or-get-outta-way kind) who just does NOT want to know the big picture.

4. Albatross managers. They come, they scream, they poop and fly away. Caw-caw caw Ca-caw!

5. IndifferentManagement - Anything goes as long as money comes in timely.

"All doers, no talkers" is just absolute peak Silicon Valley - may be the does should talk often and properly, to do things well. And do it well.

Collaboration over competition - it's 2020 FFS!

One approach is building a team that actively prohibits bad characters (like those listed above) and empower them by treating them like adults, and things would just work? But very few would do it - cause "the SV prophets said otherwise! "

HR Tech is where many VCs have invested in, with the hopes of selling big riding on the AI wave.

However, many HR Tech solutions (which claim to use AI/ML/DL under the hood) are now being forced to answer how they counter hard-coded biases in the AI models, how is explainability implemented, how they adhere to norms set for algorithmic accountability (ex. Ethica etc)

Most of them neither have answers nor the mindset to even bother about these.

> some of the admins don't have internet at home means they're not able to support anything

Yes, this is the reality. Most of these shops do not believe in issuing a laptop along with a secure connectivity option, simply because they deem all of that as a sap on procurements.

And they still use shitty 2005 era desktop since work gets done on VDIs (citing info sec and DLP) and desktops are mere dumb terminals. But even to operate those dumb terminals, they are ready to spend cost on electricity and logistics but will actively avoid issuing good equipment just to avoid cost on licenses and manageability.

Bunch of old coots that do not ever learn.

software engineers use "containers" to densify deployments. But in cases like this, real containers can't densify or offer mobility. Indian Railways engineers have submitted a proposal to mobilize trains for quarantine and mobile hospitals. Retrofitting is not too difficult and a prototype with designs have been shared. (even to SAARC nations, I guess).

https://theprint.in/india/governance/rail-coach-as-icu-how-m...

Most related excerpts below:

‘A good idea’ According to media reports, a Kochi-based firm called Asset Homes had submitted a proposal to the Prime Minister’s Office (PMO) where it offered to fashion hospitals out of trains.

“We have 12,617 trains with 23-30 coaches in our country. We can easily convert them into ‘mobile hospitals’ with facilities like consultation rooms, medical store, ICU and pantry,” the firm’s managing director reportedly wrote in the letter.

“Each train can accommodate at least 1,000 beds. Using the 7,500+ railway stations, the patients can be admitted to the trains.”

A senior railway official said it was good idea to convert coaches into mobile isolation wards, but added that it would be easier to first convert the railways’ Accident Relief Medical Equipment Vans (ARME) — or rail ambulances — into isolation wards.

“These are basically moving hospitals meant to provide medical treatment and assistance in case of rail accidents,” the official said. “It might be easier to convert them into isolation wards since they already have some medical facilities.”

Among other steps taken to check the pandemic, the railways have been instructed to introduce isolation wards at railway hospitals.

This is also the time that th underprivileged would get exploited more.

Work permit/visa holders will see an increase in workload (with a "work-more-or-eff-off" note, along with possibly a paycut too) Even pay negotiations on open offers, making it seem generous tha offer doesn't get cancelled.

Personally know a couple of friends who are in the US (with H1Bs awaitng transfer/extensions) facing this particular situation now.

Hiring is fucked. Tech hiring and ex\mployment more so.

Seems every employer treats employees as though the wage slaves are just so, or just above hamsters on wheels doing tricks.

Depressing, to say the least! :(

> This is looking 2001-dot-com-crash bad, but much, much faster.

"fail fast; fail often" seems ominously dark for the circumstance.

> Wouldn't be ethical, but wouldn't be surprised if some companies took advantage of the situation.

I know for the fact that few US, Indian firms with offices in India are just using this as an excuse to trim their workforce, enforce paycuts (upto 50%), hold back increments for the year (and even next) and restructure salaries drastically - by moving almost 60% to vaiable pay(which typically never gets paid).

Such a list is live at https://careervscovid19.com/insights

Traditional BigCo ITeS/outsourcing shops are using this as a great opportunity to trim the flab in the middle (claiming as "a dire measure to sustain shareholder value") while it was always about . These are the same bunch in middle-management whom they would pamper regularly by giving quarterly/monthly awards (quarterly Gem, Spot awards, Player of the month, etc bullshit accolades)

And the markets also work that way. On a regular day, if there's firing, stocks of a BigCo go wonky sensing something is amiss. During a crisis, if they do not fire people by the thousands, markets deem it as lack of financial prudence and stocks go down.

Everything (and anything) in the name of profits and shareholder value. There's a saying dated thousands of years, goes like "vyaparam droha chintanam” meaning "Business is all about profits most often via undue advantages".

Related article - http://www.informaticsjournals.com/index.php/dbijb/article/d...

Brilliant!

Now, I know that there are many companies that force people to be on with a live video feed, and that many don't really like it.

How about recording a 3-min clip and playing that in an infinite loop - creating a fake feed (remember Keanu Reeves' Speed?) - so that people can avoid not being seen, but still get things done better? A mask on the face is a simple addition to avoid detection. As the saying goes, modern problems require modern solutions!

Oh, spot instances to save lives!

There is also https://covid19-hpc-consortium.org/

  COVID-19 High Performance Computing Consortium

  Members: a lot of universities, federal agencies, NASA, US DoE, IBM, Amazon Web Services, AMD, Google Cloud, Hewlett Packard Enterprise, Microsoft, NVIDIA,

compute muscle -> Over 402 petaflops, 105,334 nodes, 3,539,044 CPU cores, 41,286 GPUs, and counting.

To join - go to https://www.xsede.org/covid19-hpc-consortium

Just curious - are Apple/FB doing anything about it?

EDIT: Apple is helping in a different way (https://www.theverge.com/2020/4/5/21209270/apple-face-shield...)

As an engineer, I am finding i very difficult to decide on things like capability/modularity boundaries, layering constraints etc w.r.t architectural decisions for a B2B, cloud native, SaaS software solution. No matter what resource I refer to, there's a lot of conflicting info.

Is there an equivalent of "cyclomatic complexity"(McCabe count of code) for cloud/distributed systems/μServices architecture? Cant find any such objective measure, apart from a few dated research papers.

IMO, it could be useful for many of us. Sometime back, I had tweeted tagging @copyconstruct and @kelseyhightower asking this question, but haven't got a response - may be it went unnoticed.

Would love to get to know if there's any such method/framework to measure and manage architectural complexity, for the world in cloud.

Is there a method/framework to measure/manage architectural complexity?

Please share any pointers you might have!

Neat project!

I was hoping of building a similar one using a simple LCD [https://robu.in/product/3-5-touch-screen-lcd-raspberry-pi/] display, with a way to even showcase urgency/priority - and list at least next three events lined up for the day.

Possibly even add a color code / simple snooze button as the LCD is touch responsive. Found https://github.com/monitoror/monitoror and hoped to use it for this purpose.