HNHacker News
TopNewBestAskShowJobs

dayjah

943 karma · joined February 20, 2007

meh.
submissionscomments
dayjah··on Ask HN: Guild/Studio structure for Software Engineers?
My understanding of "guilds" is the notion of people coalescing around projects based on needs and interest vs a top down hierarchy. Like Valve vs Amazon, etc.

At a prior employer (~1000 person company) we used the word "guild" for a cross organizational team that are tackling a short term, low-medium value problem - with the notion you go back to your existing team when you're done. Or you handle 2x load while the guild work is necessary. These are similar to "tiger teams", but those are generally top down endeavors vs guilds are bottom up.

Zappos tried this out a few years ago, so did some other well known tech co. though the name is escaping me right now. It seems to be called Holacracy and there's this webpage about it: https://www.holacracy.org/

dayjah··on Apple: Apps should not require users to opt into tracking to access content
All visas work this way afaik, at least in the few places I’ve traveled where I’ve needed a visa; notably Africa.

I met a Italian man at immigration control in Ethiopia, at the Addis Ababa airport, he had been traveling Africa and this was his third attempt to enter Ethiopia. The first two attempts were via land borders with Kenya and he had been turned around both times, despite having a visa, because border control felt it would be too dangerous for a white man to travel through southern Ethiopia at the time. So he had to travel to Nairobi and get a flight; he was admitted this time, but he had to pay for another visa at the airport. Unclear why? He wasn’t best pleased about the whole thing.

dayjah··on Restream: Stream live to many social platforms at once
Oh, right! I was obvi missing something simple. You’re right, my b: they only relay to a few services so it’s keep their costs low.

I guess I’m just so conditioned to think about video going direct to consumer that I missed that.

Sorry for the noise :)

dayjah··on Restream: Stream live to many social platforms at once
Context: ex-early-twitch engineer.

The economics on this are confusing to me; if they’re hosting a “repeater”, they’re paying for bandwidth in/out, and that’s expensive.

If the streaming software is sending to multiple locations; the stability will be low (assuming folks become outbound b/w limited from their home conn).

Am I missing something obvious here?

At twitch we needed a heck of a lot of interconnecting with various providers to make our service viable.

Edit: https://restream.io/blog/restream-series-a-funding/ - goes someway toward answering my q.

dayjah··on Cloudflare was down
First place I went to also. Then HN.
dayjah··on Twitter internal panel linked to account hijackings
For reasons I cannot fully remember my voicemail broke many years ago. It goes something like: I’d switched to google voice for vm, where T-Mobile handled my line generally. Then google did something to google voice, some sort of discontinuation + merging with gmail and my vm broke. This occurred in tandem with me moving to a house with terrible cell reception and before wide spread WiFi calling support. The result was that many folks figured I’d had my number disconnected because calls which couldn’t connect to my phone would get the “disconnected number tone” when being directed to my voicemail box.

Every month or two I’d fill in the google support text area explaining the problem. No response for ~4 years. Just this Feb, for whatever reason, I decided to call T-Mobile and report it. Problem was fixed by a higher up tech that described the problem as “very strange” and the “first time” he’d seen something like this. It took approx. an hour.

Upon rumination I full accept that I took the “easy way out” by filling in the text box vs trying to talk to someone. End result is that google lost a gvoice customer and no one calls me anymore. meh

dayjah··on Twitter internal panel linked to account hijackings
I’ve long considered why apps don’t have some VOIP client in them; if one can Face ID into their account, and use the VOIP client to connect to a rep - then the metadata associated with the call can inform the rep you are who you say you are. Seems E* is almost there!
dayjah··on Twitter internal panel linked to account hijackings
> “close enough”

I guess one, admittedly brutal, solution is for customers to act like the immune system. Call up, fudge your way through to something that should be protected, and then escalate to a manager and report that you got access to your own account with vague details; they can listen to the call log to verify.

Worst case is that the rep gets fired (which sucks..) but if enough reps get fired then future reps will be hired and trained more diligently.

dayjah··on New fuzzing tool finds USB bugs in Linux, Windows, macOS, and FreeBSD
Too soon! I’m currently debugging an issue where a USB 3.2 Gen 2 scientific camera is unable to negotiate a connection with Linux. Under closer inspection it seems the USB 3.2 Gen 2 port on this relatively modern Intel NUC is being identified by the kernel as being USB 2.0. No such issues under Windows; but we’re not deploying that to the field.

So I go a bit deeper: the chip is actually a thunderbolt chip (aka USB 4.0), which is able to emulate all of the USB specs. For reasons unknown, the two controller chips on this NUC present as four controller chips; each chip presents as one USB 2 root hub and one 3 root hub. One of the chips is only available via an internal header on the board. All of the externally available USB ports (3x USB 3 ports, 1x USB 2 port, 1x Thunderbolt port) all result in connectivity to the USB 2 root hub.

Every fiber in my being wants to work out wtf is happening here; but I’m resisting because fortunately we have a vendor that can replace this NUC with one from 2016 instead of this one from 2018.

I did, of course, peek at the spec and experienced the bone-shuddering realization of how complex this whole system is. No wonder Windows BSODd when Gates first demoed USB on stage ...

dayjah··on Tell HN: Triplebyte reverses, emails apology
Unfair. I’ve worked with Ammon IRL at a different co, he’s one of the most genuine people in the valley. Did PR consult? Probably. Is it just spin? Definitely not.
dayjah··on AWS services to avoid
Yeah, missing the proxy pass option will definitely burn a prospective Lambda user.

I also felt a similar tinge when the post talks about KCL. Yes, you use it, and yes there are rules about output streams - but they all make sense when deployed in fairly complex environments. I can’t remember when I last used stdout for logging outside of greenfield development. Once it’s on prod everything goes to stderr and is highly structured.

Similar sentiments about Cognito. At some point of complexity you must have a server in between using the Admin* range of commands. If you want their “get going quickly” then yes, there are trade offs like WebViews. That said; I’ve never used Auth0 - so insert a Luddite warning here.

Finally, one hundred percent agree on CF.

dayjah··on U.S. airlines want a $50B bailout. They spent $45B buying back their stock
As I understand it the issue here is cash flow; the assets the companies have on their books are still valuable, but there is a cash flow issue right now. Every normal company would approach various lenders to take out a loan - I'm not quite sure why airlines are being treated any differently, or why the government or public needs to receive stock as a make weight.

If their businesses are solid, sound businesses, major lenders will not have an issue with offering good terms on whatever it is that they need to bridge the gap.

Other than a firm understanding of what that gap is, am I missing something obvious which makes bank loans not the go to solution here?

dayjah··on Building a static serverless website using S3 and CloudFront
I’ve had to do this multiple times over the past two years. It’s always a shock to me how hard it is.

The Lambda@Edge functions especially. They have 4 phases they can be injected into: viewer-request, origin-request, origin-response, and viewer-response. If you put your code in the wrong place you get hard to debug issues. Additionally you cannot “annotate” a request, so a decision at say “viewer-request” cannot pass information to later stages.

Also, deployments take 15 minutes at least which just further frustrates the debugging process.

dayjah··on You Can Now Tell Facebook to Delete Its Internal Record of Your Face
At Amazon new products have a press release written during the spec phase (to help frame the pitch internally); I presume, at this point, fb writes an apology during spec phase...
dayjah··on Ask HN: What is the state-of-the-art environment for developing Erlang code?
I’m currently on a project which uses Elixir and the Phoenix framework. Both are pretty good.

It uses “hex” for package management, which has a tool called `mix` to call build related commands.

Last I was in erlang was 2008; and I’m quite impressed with where it’s at now in terms of the above.

I’ve found the vscode elixir/erlang defaults to also be quite helpful.

dayjah··on GitHub open sources action parser
Of late there has been a lot of discussion about vendor lock in (notably wrt Lambda on AWS). I’m interested in how folks feel about this.
dayjah··on On Redis master-slave terminology
Context: I'm in the "source and replica"-camp. I'm English and white passing.

I really appreciate antirez's point on this, "Mark" definitely seems to be a person with their heart in the right place but lacking tact and being intolerant to decisions not going his way; especially given the technical work this change would necessitate the change does not seem appropriate.

That said, I hope that in time the technical world will settle on some alternative to ensure we can talk freely without distracting or offending folks.

A few years ago I was in a meeting talking about database options, hitherto I had used "master/slave" to describe this database topology. In that meeting was an African American co-worker ("Frank") and as the words left my mouth I realized that completely irrespective of my intent I brought into the room the spectre of times past. I've no idea how Frank felt about it - but all of a sudden I had lost all ability to continue my point. I paused and said something that amounted to: "you know,... that phrase is fucked up - let's not use it - what words should we use?" and Frank suggested we look at what terms Amazon uses; they used "source and read replica" and that's what we went with. I've never regretted recoding my use of words on that one.

If there's anything that this post-Trump era has really brought into focus for me: it's that my privileges of being white, able bodied, and male results in me being least well positioned to say what is offensive or not. If you think it's offensive, open it up to the rest of the room and establish how to move forward.

dayjah··on Show HN: Glycemic load calculator
Also some form of typeahead prompt would be useful for a similar reason — for example: if one enters “bread”, being able to select the type of bread from a known list is helpful.
dayjah··on Full Guide to Developing REST API’s with AWS API Gateway and AWS Lambda
Oh, interesting - can you share some information as to why you think the monolithic approach is best?
dayjah··on Full Guide to Developing REST API’s with AWS API Gateway and AWS Lambda
IME the AWS GUI is not useful for all that much in a meaningful context. Thus this post is pretty misleading to a new person getting involved in cloud; it doesn’t qualify as “full” in my mind.

As a number of other comments have pointed out, there are packages that help with managing your infrastructure in code; which is critical outside of anything other than a play environment. Terraform, serverless, cloudformation, SAM, zappa, etc - are all essentially requirements for cloud usage.

At Amazon they’ve tooling which makes it relatively easy to set up an AWS account to develop against; the tool primarily configures the billing to the correct team within amazon. Additionally your manager gets “rights” to the account, IAM accounts can be linked to AD, etc, etc. This is all part of a push to get Amazon employees building infra on AWS proper. When you create the environment you have to denote whether it’s dev, or whether it’s production; in the former case it’s widely accepted that you use the GUI to set up your trial. However in the latter case you’re strongly encouraged to use something like cloudformation. It was always a shock to me that for prod accounts they don’t disable the GUI for anything other than viewing.

dayjah··on Apple needs to tackle the issue of smartphone addiction
Agree, outdoor magazine has a detailed piece about this [0], somewhat ironically the author, a friend of mine, was killed recently by a vehicle (an accident; the truck hydroplaned into him while he was fixing a flat) [1].

[0] https://www.outsideonline.com/1930211/bikes-vs-cars-deadly-w...

[1] https://www.outsideonline.com/2282551/author-and-cyclist-and...

dayjah··on Apple needs to tackle the issue of smartphone addiction
This can be turned off in settings; I’d be interested in knowing how many of us have done that (I have mine on).
dayjah··on Apple needs to tackle the issue of smartphone addiction
In general, I agree strongly with this sentiment:

> What does a healthy, moderate digital life look like? I think that manufacturers and app developers need to take on this responsibility, before government regulators decide to step in

As a bicyclist and driver I believe one of the best ways to enact this would be add a feature that allows investigators / insurers to answer the question of “whether the device was I use at the time of an accident”.

I’ve taken to yelling “get of your effing phone” at drivers obviously looking down at devices while I’m driving, walking or cycling around; I think the phone doing that would be more efficient.

dayjah··on I almost lost my hearing from the lid on the tank of a toilet
Now we’ll have stickers saying: “These toilet lids are known to cause deafness in the State of California”...

</s>

dayjah··on Using Technical Debt in Your Favor
I’ve always found the metaphors around Technical Debt really interesting. They simplify it to a point where everyone “gets it”; however I have often worried whether there’s a cult of “anti-technical-debt”, much like there are the cults of “absolutely no credit card debt / usage ever”. Sure, it works for some, but I’ve also seen product progress completely halt over the argument of “that’s adding to our technical debt”, surely it would be phrased better as “this is adding to our customers experience, at the cost of our chase for purity and technical perfection”?

As has been said a few times in this thread: there’s nuance to engineering, and a careful trade off. I think the metaphors are weaponized to a point where that nuance is lost in harsh memories of student-era credit card debt.

dayjah··on Introducing Uber Health, Removing Transportation as a Barrier to Care
What might not be super clear to most folks is that this is one of the last areas that medallion owners make bank. Typically, in many US cities local taxi operators have deals with the govt to ferry the elderly around. So this in effect is uber putting the final nail into the taxi company coffin. I've spoken to taxi drivers which have resisted switching to the gig economy because they made enough from this government tie up to not have to move; I guess that all changes from here on out.
dayjah··on The Business of SaaS
While running Heysan (YC W-07) we were having a hard time hiring; startups were a hard sell back then, and given we were building a product aimed at feature phones it was harder still to convince folks to join. We faced a meaningful decision: whether to buy a subscription to a product or build it ourselves. The notion of SaaS clearly existed back then, but it had not yet developed into what it is now; it was more akin to buying software licenses per CPU than the sort of products and pricing that we have now. Also, pertinently, I was a young and inexperienced CTO who had no real framework for how to approach this problem, so I took it to the board, the answer was resounding: pay for it.

In 2011, while deeply ingrained in converting justin.tv -> twitch.tv, the same sort of decision came up: should we build our analytics or should we buy it (mixpanel in this case). This time it was obvious: buy it.

In the time from that first decision and the later decision at Twitch I was able to fully appreciate why you buy, and Patrick nails it in the first few sentences:

> Unfortunately, many entrepreneurs discover this body of practice the hard way, by making mistakes that have been made before, rather than by spending their mistake budget on newer, better mistakes.

also:

> Customers love SaaS because it “just works.” There is typically nothing to install to access it. Hardware failures and operational errors, which are extraordinarily common among machines which are not maintained by professionals, do not result in meaningful data loss.

I've often retaught these points to folks: "Our job is to deliver a product to our users, they don't care about the ancillary product we use to deliver them our primary product. Buying Saas is like hiring pro's to do a job at the fraction of the cost of hiring pro's to do that job".

Or as Matt Brezina (Xobni & Sincerely) likes to say: Hire via API ( http://www.mattbrezina.com/blog/2011/08/hiring-via-api/ )

dayjah··on Robinhood Crypto – Invest in Bitcoin and other cryptocurrencies
As a very happy user of Robinhood financial, this really does look exciting. If you’ve not used robinhood, and have a little interest in investing, I’ve found their simple interface to be very nice to use relative to the more heavy weight options I’ve had experience with (etrade, schwab, fidelity).

I didn’t see how their crypto offering would allow me to transfer my wallet from Coinbase to them, do I create a new wallet, and then send money from one to another? If so it’s unfortunate that I’d need to incur transaction fees to onboard.

dayjah··on Twirp: A new RPC framework for Go
Unlikely? It's hosted on medium - I'm pretty sure they can handle the load. However here is the cached version: https://webcache.googleusercontent.com/search?q=cache:6vYOM9...
dayjah··on Takeovers Roar to Life as Companies Hear Footsteps from Tech Giants
I subscribe to some news sites (WaPo, The Guardian, NYT, The Economist) — they each have a much better hook than WSJ which just dumps you into this weird limbo. I wonder how successful this approach is versus the “5 free articles a month” approach others use or the “polite request” approach of The Gaurdian. Does anyone know?
← PreviousPage 5 of 8Next →