nickmyersdt

Therefore we know that a proportion of paying grok subscribers will cause harm to real victims. This isn't an abstract debate about free expression.

Non-consensual intimate imagery harms real people.

CSAM normalizes and facilitates abuse of real children.

Grok, and everyone involved in it or similar endeavours, facilitate abuse.

The goal itself is flawed, not just the execution.

If you build a system explicitly designed to have no content boundaries, and it produces CSAM, that's not a failure of execution - that's the system working as designed. You don't get credit for noble intentions when the outcome was entirely foreseeable.

Deciding to place no limits on what an AI will generate is itself a value judgment. It's choosing to enable every possible use, including the worst ones. That's not principled neutrality; it's moral abdication dressed up as libertarianism.

A new one to me, will check it out, thanks.

Yeah, we're looking at webhook type solutions but that puts the clever stuff "outside the box" which we're paying for. Okta's group rules do exactly what we want but we can't do the multi-tenant stuff and it doesn't do the SAML Federation side of things.

The SAML Federation one is where all the modern SaaS fall short. Its still SAML but it involves:

All the metadata for 100s of IDPs being downloaded and made available to enable

Publishing the SP metadata to the federation(s) which may involve fees.

Specific rules around metadata (attributes) being released and adhered to.

And if your directory insists on having an email addresses for a user, that might be an issue.

There's a reason why Higher Education businesses have cropped up around doing SAML Federation.

I have had a trial of FusionAuth, and it was great, just didn't solve enough of our pain points to justify a migration.

Unfortunately not. It's capabilities are good but missing the same as the other popular options. So it can do everything our current platform can, but we need extra "stuff" to justify the migration.

Thank you, will keep an eye on it. Auth0 is definitely the closest I've seen to what my business needs (and I've spent waaay to much time one trials and demos of too many products so far).

If it would:

a.) Perform group assignment based on SAML attributes (like Okta's group rules), and, b.) "natively" support SAML Federations used in the Higher Education space (which Shibboleth appears to be the only thing that supports)

I'd sign up tomorrow and start migration of my 150k users.

One of the things my business needs, but I cannot find a SaaS solution for is:

Multi-tenant (each of my customers gets a fully separate directory, with access to all tenants for our admins)

SAML and OAuth (customers can set up SAML themselves via the SaaS interface, or we set the SP up for them)

Rule based group assignment based on SAML attribute evaluation (e.g. assign users to this group if the attribute X = Y)

APIs to manage users, groups, organisations (tenants)

We've built something using Okta, but all our customer users are in one Directory/Tenant.

Auth0 nearly gets there with Organisations but can't help with the sub-groups and rule based management.

For context, we have an education product and customers are districts or schools, and the sub-groups are typically schools and/or classes or groups of users (e.g. seniors or juniors).

We also need to support SAML Federations like InCommon, OpenAthens, UK Access Management Federation which makes the challenge harder (these federations want a single SP to which many IDPs authenticate) for Universities. None of the modern platforms support this.

If anyone has found an out of the box solution for this, I'd love to hear about it.

AFAIK all credible studies into the link between MMR vaccine and autism have concluded that there is no link, do you have new information?

I use my browser bookmarks bar for things which I want to access frequently, Notion for stuff that I want to keep "for one day, maybe" and send a link to my email for "stuff I want to read next".

Not a single app, but works for me (so far).

Agile works when everything in your business aligns with it as a way of working.

It fails (and therefore is loathed) when the people who need time, focus, room, safety to achieve the outcomes are forced into working with particular processes, procedures or tools in order to satisfy someone other than the team or the customer.

Awesome, thank you, I'll take a look at the things you've mentioned.

Knew I'd forget something. One other idea was instead of a physical bell, play the sound of a bell from a speaker (assuming a raspberry pi with a speaker or similar).

Depending on how much effort you want to put in, I'd suggest whatever the highest resolution you can buy, and rip to a video file.

You'll want to use an uncompressed and unencrypted format. You might want multiple formats, in multiple places.

Likely one of these will be network attached storage with RAID, however it's been 10 years since I did anything like this so things may have moved on.

You may also want to think about ensuring you have players that can play the formats.

IIRC Blu-ray discs are expected to be OK for 15-30 years, however I don't know enough to say whether this is specific to certain manufacturers or usage.

From my limited knowledge, most content is still released on disc, however with the move towards streaming, I can foresee this may diminish over time. IIRC the Blu-ray market (both releases and players) is shrinking year-on-year.

I too have encountered an immense amount of frustration due to content rights being sold to a single provider, who is not one of the three or four we already subscribe to.

Isn't there a GIF/Meme of Woody Harrelson dabbing tears with wads of cash? I've not read the article, however the subject makes me think of that.

Refresh. Frown. Close browser. Restart browser. Frown. Open command line. Ping. Frown. Switch from LAN (power line to router) to WiFI. Frown. Reboot router. Frown. Switch to mobile hotspot. Carry on.

You have too many managers and not enough developers. In this type of organisation, management goals are the only ones that matter.

Change organisations to one where you don't have teams starting with 1 dev, 1 prod manager, 1 scrum master and 1 manager. Move to one that has a clear way of developing and moving people up and on.

I take the attitude of "my only purpose is to help this person".

I don't care what they say, how they say it's how wrong (or right) they are or who (if anyone) made a mistake.

I only care about "what is the most helpful thing I can do for this person".

I have apologised, explained internal constraints, identified where our processes have failed, offered video calls, empathised if I felt it would help, however I am mainly looking to "solve the problem this person has" within the limits of my authority.

Look for the facts, ignore the rest, and be helpful and kind.

Also, being sole developer and supporting customers is very hard. I'd be trying to separate me from the software and the customer issues, see them as independent from one another but nevertheless influencing each other (customer issues help you identify what to improve!)

We use Contentful. We did an evaluation of a load of headless CMS and found that Contentful was the only one that matched our needs and (for cloud hosted versions) didn't have bugs that were easy to run into. We also make use of its content management APIs to improve automation of certain tasks not supported by functionality already in the product.

> Other viruses don't do that.

...yet

1. Track actuals for work items 2. Compare new work items to actuals 3. Give range based on risk (known unknowns, unknown unknowns)

Fundamentally establish trust through transparency.

Remember most humans are trying to justify or find reasons to do something. Finding out it will take too long or be too expensive (in their eyes) is a reason not to do it. If those people have already made "positive noises" to other people, then your reality is going to run head on to theirs.

Help people, perhaps by asking up front "how much effort do you want to spend solving this".

It's not disrespectful at all. The engineering team does not exist in a vacuum independent from everyone else. What does happen is that engineers with a lack of maturity hate to have their judgment questioned.

Without knowing specifics of context, perhaps by understanding the constraints, needs of the business or limitation on budget, then any estimate you provide is going to be wrong in the eyes of the reciever.

IMO anyway.

What happened in 1997?

I've joined the discord.

I'm in a similar situation.

I would like a serverless and distributed solution, that supports: username & password, SSO discovery, SAML 2, federated SAML, OIDC, OAuth2, magic links, permissions, rules for assigning users to groups based on attributes, deletion automation, multi-tenancy and - of course - excellent security and monitoring.

We pay a lot of money to Okta at the moment, and have to use another solution "on top" to support federated SAML.

We'd rather all this be in our own AWS account and globally distributed so personal data never leaves the users country of origin.

Not asking for much surely? </s>

The muscle memory that kept opening HN was strong today. Thank you for bringing it back.

I think the core problem is that there isn't a reliable way to assess a person's SWE skills. Two people can look the same on paper but have wildly different capabilities for solving problems with software.

So many companies try so many different ways to establish competence during application and interview but they are all badly flawed in some way.

Add to that the fact that no two SWE teams are the same, and to be effective as an IC you need to be effective as a team member, and it gets really hard.

The true test really comes during the probation period (which I think of as the real interview) and if all things were possible, I'd hire many more people off the back of resume and interview with intent to retain those that are able to perform at the level they pitched themselves. There's always a bit of "leveling up" as people climb the ladder, so a little bit of over-stating one's own skills is to be expected, and that's where the core capability of "being able to learn quickly" gets proven.

Unfortunately budgets and my own capacity don't allow me to take this approach.

I'd also say that as someone who has led or managed SWE teams for a long time now, SWEs do love to argue for their favourite bit of tech. After all, if they've used it before, they'll be more productive with it (perhaps). However that drops us right back to "the list of technologies you've worked with" mattering more than "your ability to pick up skills".

I do wish there was a better quicker way, that could not be games or bluffed, to hire people. Many startups are trying to solve it, but none (AFAIK) have scratches the surface.

I think (IME) that getting a job is complicated because (a) it's very hard to determine a SWE skills by resume and interview (b) lots of people over represent their skills (c) hiring is probably the most critical thing an organisation does (d) SWE has become more complicated over time, rather than less, leading to proliferation and speed of introduction of new languages, frameworks, tools and libraries.

As a hiring manager, it's probably nothing you're doing wrong. There are so many applications for roles, that most get a cursory glance in the first instance. I try and do all the initial application reviews myself, and to look properly at each applicant however (a) I know I'm in a minority and (b) there's such variability in how candidates present themselves that despite my best efforts I go a bit blind after too many.

If it would help, I'd happily take a look at your resume and see if I can offer any advice on that would have worked on me (whether that advice is more universally useful I cannot tell you).

There are many reasons for rejecting applications, and the specifics about you probably aren't even in the top 5. Don't take it personally and play the numbers game (high volume of applications) or target hiring managers to get traction (otherwise you're likely to have your paperwork reviewes by a recruiter, agent or "hiring team" who are keyword matching at best).

Two choices:

Iterative and incremental, or, "Rebuild projects"

Tech isn't useful because it's new, and it isn't bad because it's old.

It's a business investment to "solve problems" or "create opportunities" for someone.

"If it ain't broke" yada

I would find what's broken (I can't get enough time to look for potentially better ways for our systems to work because I'm always building forms) and recommend prioritising some of your time per week to identify a solution (I'll spend 1 day per week investigating ways to use greater automation and or lower bespoke effort to build forms) meaning I won't have as much time to do x, y, z.

Then if you find a solution, work out what the first interation would be and guess how long it will take and seek agreement to invest the same or more time per week in implementing it.

Rinse and repeat

Steps 1-4 plus probation period is sufficient. The rest is bs.

Microsoft Money.

Whilst other products do similar things in terms of tracking accounts, I miss the forecast graph that would model how your money would perform over time.

Google Reader. I can't stand video or audio, and strongly prefer written materials so email subscriptions and RSS are how I keept up.

You are focusing too much on things that are external to you and do not directly affect you.

Life has always been hard for most people most of the time. There is nothing about the premise of existence that promises it will be easy or comfortable.

The past 50-80 years have been incredible in terms of advancement, poverty (lowest ever globally), health.

Turn off the news, get off social media. These things are not designed to give you a balanced perspective. They are designed to drive advertising revenue.

Go and live.

If you want to feel better about the world, go and do something that improves the lives of others.