I suddenly remembered this after being gone from my previous company for nearly a year.

So, I worked there as a tech supporter and Linux engineer.
What would often happen was clients calling with an issue regarding software of some sorts and about half the time, instead of LOOKING AT THE GODDAMN ERROR MESSAGE they'd just click it away fast and complain shit wasn't working.

I specifically remember this one case:

*big client mails complained that one of their clients' email isn't working. Screenshots weren't possible apparently so after emailing back and forth for way too long, we decide to do a screen sharing session (which we never do).*
(for the record, already emailing for hours, client very frustrated, me as well because the behavior of the software sounds impossible)

Me: alright, close everything, then open it again so I can see what happens.
Client: *opens mail client, error appears, client clicks error away faster than an arch user being able to mention they use arch*
Me: uhm.... I assume you already know what that message said and that it has nothing to do with the issue?
Client: it has nothing to do with the issue.
Me: okay... But have you at least looked the message?
Client: no but it has nothing to do with the issue.
Me: but, how'd you know if you won't look at it?
Client: it has nothing to do with the issue, okay?
Me: okay.... so, what's happening here?
Client: the user isn't receiving email anymore at this point!
Me: alright, have you checked the settings and everything?
Client: of course, all good
Me: okay but can we at least restart the software again to at least check the error message?
Client: FINE. *restarts client (pun intended, of course)*
Error message: username or password incorrect, can't connect to the server.
Client:..........
Client:............
Client:...............
Client:..................
Client:.....................
Client:..................
Client:...............
Client:............
Client:.........
Client: 😐
Client: 😶
Client: 😅
Client: 😬
Client:..... Right, I changed the password...
Client: *sets correct password*
*poof, error message gone*
Client:..... Thanks 💀
Me: you're welcome 😄

💀

I'm sorry, why is this a problem again?

It's finally happened. I've used my mail servers for about a year to give out different email addresses on my domain to things I sign up for online, and only used my "actual" email address that received all this email for the whole domain but the single one that I used outbound for private communications.

This worked well for a long time as I could see when spam comes in, where it came from by looking at the email address I designated it. Each company's email would be sent not only from an email address that they choose, but also to an email address that I choose. It allowed me to easily determine where there were problems. For example, on Freenode IRC my vhost happened to make my username@host there a valid email address. It eventually got blacklisted due to too much incoming spam as crawlers started detecting it. Another one was "nickname"@my.domain as I posted it a few times here. Got crawled as well. But it allowed me to easily blacklist each.

I'd never thought my actual outbound email address, my real one, to get crawled though. That would require the mail server of a company I explicitly communicated with to get hacked. But today that happened. I wonder whose it is, but I can't tell.

Time to make my outgoing email bound to a designated email address as well. I want to know which companies this happens to, even if they don't disclose it.

I just tried to sign up to Instagram. I made a big mistake.

First up with Facebook related stuff is data. Data, data and more data. Initially when you sign up (with a new account, not login with Facebook) you're asked your real name, email address and phone number. And finally the username you'd like to have on the service. I gave them a phone number that I actually own, that is in my iPhone, my daily driver right now (and yes I have 3 Androids which all run custom ROMs, hold your keyboards). The email address is a usual for me, instagram at my domain. I am a postmaster after all, and my mail server is a catch-all one. For a setup like that, this is perfectly reasonable. And here it's no different, devrant at my domain. On Facebook even, I use fb at my domain. I'm sure you're starting to see a pattern here. And on Facebook the username, real name and email domain are actually the same.

So I signed up, with - as far as I'm aware - perfectly valid data. I submitted the data and was told that someone at Instagram will review the data within 24 hours. That's already pretty dystopian to me. It is now how you block bots. It is not how Facebook does it either, at least since last time I checked. But whatever. You'd imagine that regardless of the result, they'd let you know. Cool, you're in, or sorry, you're rejected and here's why. Nope.

Fast-forward to today when I recalled that I wanted to sign up to Instagram to see my girlfriend's pictures. So I opened Chromium again that I already use only for the rancid Facebook shit.. and it was rejected. Apparently the mere act of signing up is a Terms of Service violation. I have read them. I do not know which section I have violated with the heinous act of signing up. But I do have a hunch.

Many times now have I been told by ignorant organizations that I would be "stealing" their intellectual property, or business assets or whatever, just because I sent them an email from their name on my domain. It is fucking retarded. That is MY domain, not yours. Learn how email works before you go educate a postmaster. Always funny to tell them how that works. But I think that in this case, that is what happened.

So I appealed it, using a random link to something on Instagram's help section from a third-party blog. You know it's good when the third-party random blog is better. But I found the form and filled it in. Same shit all over again for info, prefilling be damned I guess. Minor convenience though, whatever.

I get sent an email in German, because apparently browsing through a VPS in Germany acting as a VPN means you're German. Whatever... After translating it, I found that it asks me to upload a picture of myself, holding a paper in my hands, on which I would have a confirmation code, my username, and my email address.. all hand-written. It must not be too dark, it must be clear, it must be in JPEG.. look, I just wanted to fucking sign up.

I sent them an email back asking them to fix all of this. While I was writing it and this rant, I thought to myself that they can shove that piece of paper up their ass. In fact I would gladly do it for them.

Long story short, do not use Instagram. And one final thing I have gripes with every time. You are not being told all the data you'll have to present from the get-go. You're not being told the process. Initially I thought it'd just be email, phone, username, and real name. Once signed up (instantly, not within 24 hours!) I would start setting up my account and adding a profile picture. The right way to ask for a picture of me! And just do it at my own pace, as I please.

And for God's sake, tackle abuse when it actually happens. You'll find out who's a bot and who isn't by their usage patterns soon enough. Do not do any of this at sign-up. Or hell, use a CAPTCHA or whatever, I don't fucking care. There's so many millions of ways to skin this cat.

Facebook and especially Instagram. Both of them are fucking retarded.

I am beginning to hate the relationship between email and my clients. I never thought it would come to the point where email is the worst communication platform I've ever used because some of my clients simply don't know how to use it properly.

I have one client who never uses the subject header in his emails. This makes conversational threads very difficult to follow, and I can't just scan the inbox I have for him. I have to actually do searches on my emails just to find recent conversations.

For some reason nobody knows how to start a new email thread. I have multiple clients that will just take the last email that I sent them, regardless of what it's about, and start a new conversation completely unrelated to the other email by hitting"reply". I end up with email threads that are 60 to 100 emails long and contain many different subjects, which again makes it hard to find anything. Never mind that they've usually put two or three important attachments, or username password combinations, or other valuable information in there amongst all the noise.

Worst of all, I have a few clients and co-workers who insist on starting a new email thread whenever anything about a particular issue comes up. This means that just today I have five separate email threads about the same goddamn issue from the same damn person. Am I supposed to respond to each thread with the same damned information? One of these people is supposed to be both a media consultant and an SEO expert and really should know better. Also, if you do actually send me an email with a subject like "the robot.txt error", please don't give me one sentence about that and five paragraphs about what color you'd like the background to be. That's ridiculous. How the hell am I supposed to find that later? Especially since we already discussed this in the other email that sitting in my inbox.

I swear I am setting up a bug tracking system simply so that my clients can log in and leave me bug reports, and feature requests, and will stop filling up my poor email boxes with what amounts to piles and piles threads that I have to sort through.

For a person who suffers with a form of ADD this is extremely frustrating. Why is it so difficult for my colleagues and clients to write good emails with good subject lines, and reply to the right damn emails?

Am I just being too anal, or does this bother others as well?

Guys I need to deploy a very simple authentication API service.

You register with a username (actually an ID with a determined format), a password and uuid. You login with your username and password and if credentials are correct you get back the uuid as a response (JSON or whatever the fuck).

If you forget your password, you can use your uuid (which is confidential, very long string) in some POST request to set a new password. If you forget your username, you use the uuid again in a GET request to get back your username.

I've been looking at a bunch of solutions online and I don't think they suit my purpose exactly and all require emails (Like Firebase, AUth0, etc.) So, let me get this straight: NO FUCKING EMAILS INVOLVED PLEASE.

The above are the EXACT requirements I need for my work (for a good cause too). I fucking hate 0-requirement exploratory research tasks and I'm plagued with those. Those requirements are the only way it should work. So again, NO EMAILS INVOLVED PLEASE.

Also, please note that I have never developed an API in my life. I feel like StackOverflow will be assholes about this so I am asking this here.

I know it is very easy to do and there are probably dozens of ways to do this. I just do not know how, documentations are vague and overwhelming (or I'm just a little stupid lately). Another thing is that I am not sure of how can I do this in the most secure way. Bonus if this can be dockerized.

I know I sound a little rude,so I am sorry. It is just my frustration and depressing times I am going through that's preventing from thinking straight.