|
Brb, I’m going to make an email client that uses webkit
|
# ? Nov 13, 2019 20:59 |
|
|
# ? Jun 13, 2024 06:44 |
|
Email is bad because I get too much of it.The Merkinman posted:Email is fine, there just need to be a push for the email clients to actually render with standards and modern compatibility instead of like, using the rendering engine from Word. Agreed. I know there are concerns about ~~~security~~~ but the fact that half of all CSS barely works in most email clients is absolutely idiotic. However on the bright side, if anyone wants to code emails for a living I know a half dozen companies that are desperate to find someone to handle it for them.
|
# ? Nov 13, 2019 21:02 |
|
The Fool posted:Brb, I’m going to make an email client that uses webkit Ten weeks later "Hey guys I figured out why everybody hates POP3, SMTP and IMAP."
|
# ? Nov 13, 2019 21:28 |
|
Nolgthorn posted:Ten weeks later "... so I came up with a *new* standard!"
|
# ? Nov 13, 2019 21:29 |
|
The only problem with email is the html hack
|
# ? Nov 13, 2019 21:48 |
|
Lumpy posted:"... so I came up with a *new* standard!" I built something and I love it but I'm not sure what everyone else thinks. The problem I was trying to solve is that most of the time email isn't used as it was intended. I get almost 0 messages from people, almost 100% of emails I receive from machines and they are almost 100% of the time something I would not respond to or even could if I wanted. It's always newsletters, ads, and email address confirmations. Sending all of this email out is fairly expensive, if you've ever worked with bulk mailing apis. So I built something that's basically just a sidebar filled with email inbox's, you can make as many as you want, and they are piped in there from whatever services you're subscribed to. Email confirmation is no longer necessary (assuming companies don't care about having a drat email address to send their marketing to which they always do), because they can just hit the api and it says "yes that id exists and is accepting messages from you". So basically, you send bulk websocket messages to my api instead of heading off to a bulk email api. You get to have the same fields, text, html, attachments and so on. So it's dirt cheap or free to use, users no longer validate their email they just follow a flow similar to single sign on (do you want to receive messages from x yes or no) and it generates an inbox. The company gets nothing like the information they get when you sign up via Facebook for example (assuming companies don't care about having access to every drat bit of information about you that they can get their hands on which they always do). All they get is an id, which the user can revoke at any time. It's got a couple of other bits and bobs like for example it will send the user an actual email on a schedule telling them about what's shown up in their inboxes if they want. Or forwards it to them as an email immediately. I also accepted that since absolutely no services would adopt such a thing, I added actual email address generation. So instead of the company doing things the easy way, you could sign up using a generated email address that would link into this entire system and work exactly the same way. They send you an email, it gets picked up by this external module that parses it and forwards it on to the regular api I designed via websockets. Bada bing bada boom you've got your messages from whatever service that is. I use it all the time, it's practically along side my password manager now. Generate email address, generate password, for everything I sign up on. I built it a while ago I haven't touched it in a year or two, should I pick that back up again and turn it into a product, would anyone care about it? I was told by an investor that companies would never adopt because they don't get any personal data out of it and may block my email domain.
|
# ? Nov 13, 2019 22:35 |
|
why cant i use a bloated 28MB javascript bundle with multiple tracking/fingerprinting software and autoplay video on my emails
|
# ? Nov 13, 2019 23:48 |
|
email should be replaced with youtube reaction videos
|
# ? Nov 13, 2019 23:49 |
|
kedo posted:the fact that half of all CSS barely works in most email clients is absolutely idiotic. agree’d. the amount should be zero
|
# ? Nov 14, 2019 00:58 |
|
Text-only email would be a great start, but what would also be great is a system with guaranteed delivery and actual workable spam prevention features at the protocol level rather than a cobbled together mess of poo poo and piss and bizarre architecture that rarely works properly, is a massive pain in the rear end to set up, and still doesn’t prevent spam. Part of the problem is that a lot of email, at least from the development side of things, is both automated and wanted, but it’s very hard to filter a legit password reset email from your brand new app from a phishing attempt without explicit whitelisting, and while that may not cause problems for you as a developer, it will mean your client whinges at you and you get to explain why there’s nothing you can do beyond saying a prayer to the dark gods of email. Bonus points if the customer has some third-party email and/or DNS provider so you can’t really set up SPF or DKIM yourself and are left with the unenviable task of walking them through that poo poo over the phone or something. And that’s before you address the problem of people who think email ought to be the answer to the question “how can I do advertising with no budget?” Email marketing can be an excellent tool, but it relies (like most marketing) on providing people the opportunity to buy poo poo they already are motivated to buy. My local liquor store reminding me I can buy wines I like on sale is getting a solid return on investment; the shoe store I once ill-advisedly provided my email address too, much less so.
|
# ? Nov 14, 2019 05:11 |
|
Has there been any research into stopping spam at the protocol level though? At first blush it seems like that would require some form of pre-shared information or whitelisting or something... which could be annoying to the users. Part of me thinks an email replacement will never happen because the end users don't care enough about how it works to justify switching to a new service. Any replacement would have to be backwards compatible (yuck).
|
# ? Nov 14, 2019 17:04 |
|
My Rhythmic Crotch posted:Has there been any research into stopping spam at the protocol level though? At first blush it seems like that would require some form of pre-shared information or whitelisting or something... which could be annoying to the users. Well, there's SPF and DMARC, both are efforts to verify the identity of the server sending mail.
|
# ? Nov 14, 2019 17:51 |
|
From what I know, DMARC seems to be pretty good at ensuring the sender's address hasn't been forged, which is good. But for anti-spam in the broadest sense... ie, determining if an unsolicited message was junk or not, it seems almost impossible to know if an unsolicited message was junk without some kind of pre-shared info or whitelist or something. Also the current system seems to rely on good old fashioned trust to an alarming extent. Like google will flag your messages as spam if you try to setup your own email server, forcing you to use a service like mailgun. That's especially annoying because who wants to pay just to be able to send email? Seems kinda broken.
|
# ? Nov 14, 2019 18:22 |
|
My Rhythmic Crotch posted:Part of me thinks an email replacement will never happen because the end users don't care enough about how it works to justify switching to a new service. Any replacement would have to be backwards compatible (yuck). One could argue that in a way, it's slowly being replaced by SMS and other short chat formats (eg. Slack). A lot of the transactional messages I receive from companies as an end user, or communications from coworkers/clients happens on those platforms these days. Granted I'm not sure that's entirely a good thing. I have a part time job in the summer that's not web-related and 100% of my communications with my employer are via text which makes it drat near impossible to find anything if I need to reference a text a few weeks after I received it. It's a mess. kedo fucked around with this message at 18:41 on Nov 14, 2019 |
# ? Nov 14, 2019 18:38 |
|
Modern big email providers do as good a job with ML blocking spam as any sort of other solution would do. Gmail, for example, blocks well over 99.9% of spam.
|
# ? Nov 14, 2019 18:42 |
|
Google does that because it is trivial to spin up an smtp server and fire off 100k emails and then shut it down in less than 5 minutes. As for unsolicited e-mails, I'm not sure what you would do that wouldn't require user input. My Rhythmic Crotch posted:Seems kinda broken. Not just kinda.
|
# ? Nov 14, 2019 18:47 |
|
Thermopyle posted:I never used Google Wave, but my understanding was that people who used it really loved it...the problem was less that everyone rejected it and more that it's really hard to bypass network effects. I was a huge fan of Google Wave, and the reason it failed is because it was shapeless. It allowed people to communicate withouth a structure. Trying to use pre-existing ways to communicate in a organized fashion was chaotic. Like using it like a chat or forum was chaos because everyone could (and would) edit their original posts all the time in real time. Using it like a forum was chaotic. The thing never really feel like a product, but a experiment. Is possible that part of the reason Google Wave failed is because his grown was so quick, that It failed to find a use for with it was the "killer-app". Withouth a killer-app user case it was technology withouth a purpose, technology for the purpose of technology, so it was dropped after everyone had fun with it. Then google killed it, and this is another error in a serie of errors. Killing the oxygen support for the technology stopped it from maybe spawning other ideas or having other effects on the web. Prior to use Wave, I had experience using . "shared community walls" where people could write and read in real time in a shared chalkboard. These things where incredible useful to teach people computer programming. Wave could have ended has THE way to write code in a shared enviroment and teach people. But it was tryiing to be a tool for no-programmers, so had bold text, embeded images and what not. Somebody was reinventing the OLE technology and that "everything-technology". Google used to be a engineers company, but it changed, and now is a mix of scientist and bean counters with some weird work culture. The original Google company would have never killed Wave or maybe created it the way it was shaped. Do this sound like I am still angry?, yes, I am. It was a cool technology.
|
# ? Nov 14, 2019 18:50 |
|
kedo posted:One could argue that in a way, it's slowly being replaced by SMS and other short chat formats (eg. Slack). A lot of the transactional messages I receive from companies as an end user, or communications from coworkers/clients happens on those platforms these days. Yeah, that's a good point. As these alternatives continue to get better and more popular, people just stop caring about the legacy stuff like email. It's not worth the hassle of fixing.
|
# ? Nov 14, 2019 18:56 |
|
My Rhythmic Crotch posted:Also the current system seems to rely on good old fashioned trust to an alarming extent. Like google will flag your messages as spam if you try to setup your own email server, forcing you to use a service like mailgun. That's especially annoying because who wants to pay just to be able to send email? Seems kinda broken. If you don't care enough about your email to be willing to pay to send it, odds are way higher than normal that your email is spam.
|
# ? Nov 15, 2019 00:21 |
|
Weird hill to die on but okay
|
# ? Nov 15, 2019 00:53 |
|
Jabor posted:If you don't care enough about your email to be willing to pay to send it, odds are way higher than normal that your email is spam. Ultimately that’s true, but it’s still frustrating that the only thing a paid solution offers in a lot of cases is a higher trust rating. It seems like there should be a solution. Look at what Let’s Encrypt did for SSL; used to be that you’d have to pay money for a trusted SSL certificate, and ultimately it’s not a lot of money, but it’s much nicer not having to worry about that until your project is “established.”
|
# ? Nov 15, 2019 03:58 |
|
I'm following a tutorial for a React app and have just added a new node module and getting an error I don't fully understand. I've previosly implemented and started using 'styled-components'. Then I added 'react-beautiful-dnd' and now the app won't run. code:
|
# ? Nov 15, 2019 15:36 |
|
I’m learning JS and nodejs with trial by fire using one of the udemy lessons you guys reccomend we. I created a little app and was wondering if there’s a site like python anywhere that I can push the little app up to live for free?
|
# ? Nov 15, 2019 17:25 |
|
Heroku, Azure App Service has a free tier DigitalOcean would let you spin up a vm $5
|
# ? Nov 15, 2019 17:39 |
|
Have you tried the free tier at zeit.co (aka now)?
|
# ? Nov 15, 2019 18:55 |
|
uncle blog posted:I'm following a tutorial for a React app and have just added a new node module and getting an error I don't fully understand. I've previosly implemented and started using 'styled-components'. Then I added 'react-beautiful-dnd' and now the app won't run. My suspicion is that you weren't in the right directory when you ran yarn add. But rather than spend time troubleshooting this, my suggestion would be to blow away your entire node_modules directory and yarn install to install them all afresh.
|
# ? Nov 15, 2019 19:16 |
|
A pretty good rule of thumb is any time you see any error in a node modules directory, delete node modules in it's entirety and reinstall.
|
# ? Nov 15, 2019 19:35 |
|
Empress Brosephine posted:I’m learning JS and nodejs with trial by fire using one of the udemy lessons you guys reccomend we. I created a little app and was wondering if there’s a site like python anywhere that I can push the little app up to live for free? glitch.com
|
# ? Nov 16, 2019 04:07 |
|
Thanks everyone. Now to figure out why the app doesn’t work once it’s pushed live
|
# ? Nov 16, 2019 05:11 |
|
Has anyone found less awful ways to cleanup state in a SharedWorker? So far I have to have a webpage hook into the unload event and call a custom API to cleanup, seems so unreliable. Here with Comlink:JavaScript code:
|
# ? Nov 18, 2019 04:10 |
|
Have any of y'all come across clients who want, specifically, "an app" for god knows what reason? I've had a few clients now who are convinced they should have an app on the Apple App Store and Google Play, despite the entire functionality of the app being easily duplicated by 20 lines of HTML and CSS (not even Javascript, we are seriously pushing the definition of "app" here). What is their goal, and how do you convince them this is a horrid waste of money and Apple will tell them to get hosed because they don't want nonsense befouling the App Store? EDIT: Also I'm fully aware that 90% of my posts in this thread are about handling insane clients rather than anything development related, sorry if that's a bother. I don't know how or why I attract them.
|
# ? Nov 18, 2019 06:37 |
|
Build it with phonegap and charge them x300 times more that you would normally Apple won’t give a poo poo if your app isn’t scammy/malicious/circumventing their weird rules
|
# ? Nov 18, 2019 06:47 |
|
The Fool posted:Build it with phonegap and charge them x300 times more that you would normally No, they're literally talking about "a screen with several links on it." I've tried doing this before and it's absolutely been rejected as a waste of space. I can try, but it's just so much work for no loving point, and whether or not I make money, I just feel depressed about doing work that has no point to it. It does not spark joy. Perhaps I need to raise my hourly rate. Also: why do they want this? What problem do they believe it will solve for them or their users?
|
# ? Nov 18, 2019 07:01 |
|
Anony Mouse posted:glitch.com Much easier IMO
|
# ? Nov 18, 2019 07:54 |
|
PT6A posted:No, they're literally talking about "a screen with several links on it." I've tried doing this before and it's absolutely been rejected as a waste of space. Apps are sexy- you could try convincing them that progressive websites are the new, sexier, thing.
|
# ? Nov 18, 2019 13:15 |
|
I would just do the app and put on paper that you're not responsible for an eventual rejection from the App Store. I mean, unless you hate free money. Maybe even take it as an opportunity to learn a new technology like react-native!
|
# ? Nov 18, 2019 13:52 |
|
PT6A posted:Perhaps I need to raise my hourly rate. correct. go play outside Skyler posted:I would just do the app and put on paper that you're not responsible for an eventual rejection from the App Store. extremely correct
|
# ? Nov 18, 2019 14:34 |
|
With projects I hate the idea of doing, I need to stop thinking in terms of "it will take this many hours to complete, therefore it will cost this much" and start thinking in terms of "this is the amount of money I need before I will consider putting myself through this poo poo."
|
# ? Nov 18, 2019 16:19 |
|
Every trade has their 'gently caress off rate'. Yeah get lightning in a bottle when a dumb but rich client goes for your gently caress off rate and the work itself is pretty easy.
|
# ? Nov 18, 2019 16:54 |
|
|
# ? Jun 13, 2024 06:44 |
|
Me again. I'm getting an error about an uneven amount of hook renders in my app. I'm trying to update my context state in an Apollo based object, when I add the useEffect call to set the state, the errors start happening. "Uncaught Invariant Violation: Rendered more hooks than during the previous render." This seems to be the offending component: code:
|
# ? Nov 18, 2019 17:58 |