The Web Design & Development Megathread

The Something Awful Forums > Discussion > Serious Hardware/Software Crap > The Cavern of COBOL > The Web Design & Development Megathread

«‹›362 »

drainpipe: May 17, 2004; AAHHHHHHH!!!!

Lumpy posted:

This should get you heading in the right direction:

https://jsfiddle.net/dh3wt4jo/1/

Right, that's sorta what I had, but notice if you change the 'width: 1200px' to 'height: 1200px' in #content, the content then overflows onto the page. The scroll will correctly work if you then put 'height: calc(100vh - 75px)' in #scrolly.

# ? Dec 25, 2019 15:29

Adbot: ADBOT LOVES YOU

# ? May 24, 2024 19:12

Lumpy: Apr 26, 2002; La! La! La! Laaaa!; College Slice

drainpipe posted:

Right, that's sorta what I had, but notice if you change the 'width: 1200px' to 'height: 1200px' in #content, the content then overflows onto the page. The scroll will correctly work if you then put 'height: calc(100vh - 75px)' in #scrolly.

remember how I mentioned setting min-height: 0? that's all you need to do: https://jsfiddle.net/6e4b23qm/ (but if calc works, then that works too!)

# ? Dec 26, 2019 02:03

drainpipe: May 17, 2004; AAHHHHHHH!!!!

Lumpy posted:

remember how I mentioned setting min-height: 0? that's all you need to do: https://jsfiddle.net/6e4b23qm/ (but if calc works, then that works too!)

Awesome, thanks! It does not make sense to me why setting min-height fixes this issue. It's much better than calc, though, since it makes it less fragile to resizing of the other components.

# ? Dec 26, 2019 17:18

Red Mike: Jul 11, 2011

So I just figured out there's been a new development on the CSS side, the "logical" properties for CSS that don't absolutely define top/bottom/left/right. Is it just me or have they done the absolute worst job implementing this?

Basically, they've set it up so there's a 'block' direction (english horizontal means top to bottom) and an 'inline' direction (english horizontal means left to right), and you define your margins/borders/paddings/floats/insets via those instead of via top/bottom/right/left. However, this was just slapped onto the current implementation with no switches or anything, so the old system and the new system exist simultaneously on every element.

https://jsfiddle.net/nj5tyvku/3/

So browsers that have implemented it now use default logical rules for stuff like <p>, and that means they've made the properties override (so margin-block-start overrides margin-top and vice versa, as if they were the same property) but only if the "environment" is set up correctly i.e. you're in english horizontal mode. It all makes sense, however it's so difficult to track generally.

It looks like dev tools don't show that override yet, which confused me for quite a bit but it makes sense when it wouldn't even be obvious that they'd override (because it's environmental). The "computed" properties list just shows both sets because they are independent (so don't technically override).

Is this the first set of rules that act like this? Or the first time this has had to happen?

However, the most confusing thing for me is that the CSS draft document is still an Editor's Draft, so surely they shouldn't be implementing it and actually replacing their default user stylesheet rules with it? Chrome at least already has though.

# ? Dec 28, 2019 13:46

Lily Catts: Oct 17, 2012; Show me the way to you
(Heavy Metal)

I have this weird Webpack error in trying to pack an Angular project wherein it works fine the first time, but starts to cannibalize its own output on succeeding tries (which obviously fails). Is this an actual thing? I tried excluding the output files (it has a main.ts and a polyfill.ts file) but it doesn't work.

# ? Jan 3, 2020 01:13

The Fool: Oct 16, 2003

When in doubt, delete and rebuild node_modules

# ? Jan 3, 2020 06:24

Empress Brosephine: Mar 31, 2012; by Jeffrey of YOSPOS

Do any of you have famiarity with Filemaker database hosting? My friend sent me a message asking about if I knew any recommendations and figured I'd query you guys. Reddit suggests setting up a windows vps on vultr and installing filemaker server on that but that seems a bit excessive for a low maintenance filemaker database.

# ? Jan 6, 2020 17:27

The Fool: Oct 16, 2003

It's been about a decade since I've touched anything with Filemaker but it didn't strike me as something that would lend itself well to cloud hosting. I feel like your reddit source is probably correct and the best path forward will be to have it hosted on a vm somewhere.

The Fool fucked around with this message at 20:40 on Jan 6, 2020

# ? Jan 6, 2020 18:46

Empress Brosephine: Mar 31, 2012; by Jeffrey of YOSPOS

Thanks I'll look into that then. If I have 0 experience with vm hosting am I in for a world of pain lol?

# ? Jan 6, 2020 20:08

The Fool: Oct 16, 2003

Empress Brosephine posted:

Thanks I'll look into that then. If I have 0 experience with vm hosting am I in for a world of pain lol?

Probably? You'd be looking at needing to manage, secure, and maintain a windows vm. Which isn't really hard, just has a lot of footguns if you don't have any experience with it.

It might be worthwhile to find a contracting service that does part of this for you. I remember QB hosting services that basically just setup a VM with remote desktop for you to use the application on.

e: https://www.soliantconsulting.com/blog/filemaker-software-service-saas/
something like these guys maybe? linking them is in no way an endorsement, they were just the first google result

e2: I also just realized that this is the web design thread. You may get more/better advice if you ask in one of the IT threads.

The Fool fucked around with this message at 22:04 on Jan 6, 2020

# ? Jan 6, 2020 22:02

Empress Brosephine: Mar 31, 2012; by Jeffrey of YOSPOS

Sweet will do, thank you.

# ? Jan 6, 2020 22:26

fuf: Sep 12, 2004; haha

Can anyone help me with a bit of reliable Apache code I can plop in .htaccess for every site I host that will:
1) redirect any http url to https
2) redirect any www url to non-www
including any combinations of the two.

This is what I've cobbled together from googling but it bugs me that the www redirect uses the actual site url (so I can't just reuse exactly the same snippet on each site) and I'm not sure whether there's a risk that the two rules might conflict with each other somehow.

Does the order matter?

code:

RewriteEngine On
RewriteCond %{HTTP_HOST} ^[url]www.mydomain.net[/url] [NC]
RewriteRule ^(.*)$ [url]https://mydomain.net/[/url]$1 [L,R=301]

RewriteEngine On
RewriteCond %{HTTPS} !on
RewriteCond %{REQUEST_URI} !^/[0-9]+\..+\.cpaneldcv$
RewriteCond %{REQUEST_URI} !^/\.well-known/pki-validation/[A-F0-9]{32}\.txt(?:\ Comodo\ DCV)?$
RewriteRule (.*) [url]https://%[/url]{HTTP_HOST}%{REQUEST_URI}

# ? Jan 9, 2020 14:08

bigmandan: Sep 11, 2001; lol internet; College Slice

fuf posted:

Can anyone help me with a bit of reliable Apache code I can plop in .htaccess for every site I host that will:
1) redirect any http url to https
2) redirect any www url to non-www
including any combinations of the two.

This is what I've cobbled together from googling but it bugs me that the www redirect uses the actual site url (so I can't just reuse exactly the same snippet on each site) and I'm not sure whether there's a risk that the two rules might conflict with each other somehow.

Does the order matter?
code:
RewriteEngine On
RewriteCond %{HTTP_HOST} ^[url]www.mydomain.net[/url] [NC]
RewriteRule ^(.*)$ [url]https://mydomain.net/[/url]$1 [L,R=301]

RewriteEngine On
RewriteCond %{HTTPS} !on
RewriteCond %{REQUEST_URI} !^/[0-9]+\..+\.cpaneldcv$
RewriteCond %{REQUEST_URI} !^/\.well-known/pki-validation/[A-F0-9]{32}\.txt(?:\ Comodo\ DCV)?$
RewriteRule (.*) [url]https://%[/url]{HTTP_HOST}%{REQUEST_URI}

Looks fine to me. If I remember correctly the order for RewriteRule and RewriteCond does matter. Might want to split you your conditions in different <VirtualHost> sections if possible (if you have access or control to it)

Ie:

code:

<VirtualHost *:80>
    #handle http to https
    RewriteEngine On
    RewriteRule ^(.*)$ [url]https://%[/url]{HTTP_HOST}$1 [R=301,L]
</VirtualHost>
<VirtualHost *:443>
	#handle everything else
</VirtualHost>

Also check out the Mozilla SSL Config generator: https://ssl-config.mozilla.org/ if you ever need a quick config generated.

bigmandan fucked around with this message at 14:53 on Jan 9, 2020

# ? Jan 9, 2020 14:44

Maleh-Vor: Oct 26, 2003; Artificial difficulty.

Edge is going to be Chromium-based next week. How long will you offer support for non-Chromium Edge?

# ? Jan 15, 2020 16:02

Lumpy: Apr 26, 2002; La! La! La! Laaaa!; College Slice

Maleh-Vor posted:

Edge is going to be Chromium-based next week. How long will you offer support for non-Chromium Edge?

As long as customers say they need support for non-Chromuim Edge to continue paying us.

# ? Jan 15, 2020 16:10

The Merkinman: Apr 22, 2007; I sell only quality merkins. What is a merkin you ask? Why, it's a wig for your genitals!

I just hate how they killed and removed (replaced) Edge but kept Internet Explorer

They should have removed that too and made it only accessible via a per domain admin setting that opens an Edge tab using the older Trident engine.

# ? Jan 15, 2020 16:19

Thermopyle: Jul 1, 2003; ...the stupid are cocksure while the intelligent are full of doubt. �Bertrand Russell

Maleh-Vor posted:

Edge is going to be Chromium-based next week. How long will you offer support for non-Chromium Edge?

Lumpy posted:

As long as customers say they need support for non-Chromuim Edge to continue paying us.

This, or the related "As long as our metrics show X% of people are still using old Edge".

# ? Jan 15, 2020 17:38

kedo: Nov 27, 2007

Not one second longer than I absolutely must.

# ? Jan 15, 2020 17:54

Maleh-Vor: Oct 26, 2003; Artificial difficulty.

Thermopyle posted:

This, or the related "As long as our metrics show X% of people are still using old Edge".

I work in a corporate environment, and we're usually the ones to define this. I'm curious what % you would generally place this cutoff point for. We have a pretty large website, so our QA can get pretty tedious and expensive. The design team usually has to support QA and we have to spend a week or so just going through most of the website every single deployment. Being able to remove one more vertex would be amazing, and I'm often tempted to use 2% as our number and cut out Edge and Internet Explorer, which have 1.9% and 1.4% usage respectively in our case. Hell, Firefox is barely at 2.2%. Honestly I just wish we could just trust our QA department and not have to re-check it ourselves. Our process generally involves deployment every 3 months or so, with about a week dedicated to checking and tuning on our staging environment, followed by a similar amount of time once it hits production. I'm usually incredibly burnt out at the end of those cycles, I have no idea how QA testers retain their sanity. It's been about 6 months since I pitched an automatic visual regression testing platform and it hasn't even been confirmed that it will be implemented, since QA is another department.

# ? Jan 15, 2020 18:26

Thermopyle: Jul 1, 2003; ...the stupid are cocksure while the intelligent are full of doubt. �Bertrand Russell

Maleh-Vor posted:

I work in a corporate environment, and we're usually the ones to define this. I'm curious what % you would generally place this cutoff point for. We have a pretty large website, so our QA can get pretty tedious and expensive. The design team usually has to support QA and we have to spend a week or so just going through most of the website every single deployment. Being able to remove one more vertex would be amazing, and I'm often tempted to use 2% as our number and cut out Edge and Internet Explorer, which have 1.9% and 1.4% usage respectively in our case. Hell, Firefox is barely at 2.2%. Honestly I just wish we could just trust our QA department and not have to re-check it ourselves. Our process generally involves deployment every 3 months or so, with about a week dedicated to checking and tuning on our staging environment, followed by a similar amount of time once it hits production. I'm usually incredibly burnt out at the end of those cycles, I have no idea how QA testers retain their sanity. It's been about 6 months since I pitched an automatic visual regression testing platform and it hasn't even been confirmed that it will be implemented, since QA is another department.

That really should be something determined by marketing or someone else other than the developers.

I've been in places where they'd say 10% and others where they'd say .1%.

For something personal I say gently caress you to everything except chrome/firefox and, very rarely, safari.

# ? Jan 15, 2020 19:52

Tei: Feb 19, 2011

I am scared of another monoculture, the ie6 monoculture was a disaster and I could see webkit going that direction. I already see some of that with they disabling autocomplete=off because some credi-cards related use case.

# ? Jan 16, 2020 09:10

Thermopyle: Jul 1, 2003; ...the stupid are cocksure while the intelligent are full of doubt. �Bertrand Russell

Tei posted:

I am scared of another monoculture, the ie6 monoculture was a disaster and I could see webkit going that direction. I already see some of that with they disabling autocomplete=off because some credi-cards related use case.

On the one hand I agree. More people should use Firefox.

On the other hand I'm not sure Edge/IE was improving the situation.

# ? Jan 16, 2020 15:37

Opulent Ceremony: Feb 22, 2012

Just got the notice from Google that Chrome Apps are being shut down (in 2022), so looking for alternative platforms for our kiosk media-streaming app is going to be ramping up. We are currently looking at the Amazon Fire Stick.

If I continue to use a hosted app where we rely on the functionality of the built-in browser to the stick, we are dependent on Amazon modifying that browser (I think a version of chromium) in the future. Does anyone know if Amazon issues automatic updates for the hosted app browser on the sticks? We've had plenty of problems with Google auto-updating our Chrome devices to new Chrome OS/browser versions that would break some of our app's functionality until we bought their enterprise licenses that allowed us to pin devices to a Chrome OS version, and I'm hoping we don't have a similar problem here.

# ? Jan 16, 2020 17:55

CarForumPoster: Jun 26, 2013; â¡POWERâ¡

Opulent Ceremony posted:

Just got the notice from Google that Chrome Apps are being shut down (in 2022), so looking for alternative platforms for our kiosk media-streaming app is going to be ramping up. We are currently looking at the Amazon Fire Stick.

If I continue to use a hosted app where we rely on the functionality of the built-in browser to the stick, we are dependent on Amazon modifying that browser (I think a version of chromium) in the future. Does anyone know if Amazon issues automatic updates for the hosted app browser on the sticks? We've had plenty of problems with Google auto-updating our Chrome devices to new Chrome OS/browser versions that would break some of our app's functionality until we bought their enterprise licenses that allowed us to pin devices to a Chrome OS version, and I'm hoping we don't have a similar problem here.

No idea but the Fire TV Stick I had (this version) was truly the worst media streaming device I've ever used or could imagine. A 2016 Chinese Smart TV would likely suck less, even if it was stealing my data.

# ? Jan 16, 2020 20:40

The Fool: Oct 16, 2003

Consider trying these guys out: https://www.appspace.com/

# ? Jan 16, 2020 20:42

Opulent Ceremony: Feb 22, 2012

CarForumPoster posted:

No idea but the Fire TV Stick I had (this version) was truly the worst media streaming device I've ever used or could imagine. A 2016 Chinese Smart TV would likely suck less, even if it was stealing my data.

We've been doing some testing with the 4k stick with good results, but I'll keep this in mind, thanks!

The Fool posted:

Consider trying these guys out: https://www.appspace.com/

Might not have the flexibility we need but worth looking into more, thanks.

Edit: VVV I appreciate the additional suggestions!

Opulent Ceremony fucked around with this message at 23:21 on Jan 16, 2020

# ? Jan 16, 2020 21:02

MrMoo: Sep 14, 2000

Opulent Ceremony posted:

Just got the notice from Google that Chrome Apps are being shut down (in 2022), so looking for alternative platforms for our kiosk media-streaming app is going to be ramping up. We are currently looking at the Amazon Fire Stick.

Have a look at Kramer VIA, so many vendors are awful.

https://www.kramerav.com/us/product/via%20connect%20pro

# ? Jan 16, 2020 22:42

duz: Jul 11, 2005; Come on Ilhan, lets go bag us a shitpost

Opulent Ceremony posted:

Just got the notice from Google that Chrome Apps are being shut down (in 2022), so looking for alternative platforms for our kiosk media-streaming app is going to be ramping up. We are currently looking at the Amazon Fire Stick.

If I continue to use a hosted app where we rely on the functionality of the built-in browser to the stick, we are dependent on Amazon modifying that browser (I think a version of chromium) in the future. Does anyone know if Amazon issues automatic updates for the hosted app browser on the sticks? We've had plenty of problems with Google auto-updating our Chrome devices to new Chrome OS/browser versions that would break some of our app's functionality until we bought their enterprise licenses that allowed us to pin devices to a Chrome OS version, and I'm hoping we don't have a similar problem here.

You can install Firefox on the Fire Stick if you're not tied to Chromium. Either way, I'm pretty sure I saw an option somewhere to disable all automatic updates.

# ? Jan 17, 2020 16:56

teen phone cutie: Jun 18, 2012; last year i rewrote something awful from scratch because i hate myself

Spent my morning reading some of the OAuth 2 RFC.

Can anyone recommend me tutorials that walk you through implementing an authentication server from start to finish, preferably in Python?

# ? Jan 17, 2020 17:02

Tei: Feb 19, 2011

Grump posted:

Spent my morning reading some of the OAuth 2 RFC.

Can anyone recommend me tutorials that walk you through implementing an authentication server from start to finish, preferably in Python?

I feel bad for you.

These things are designed in environments where they can have 1 person the expert of technology X, so they see no problem making things "corporate" to ridiculous levels. They really don't understand the mantra KISS.

# ? Jan 17, 2020 17:52

The Fool: Oct 16, 2003

Grump posted:

Spent my morning reading some of the OAuth 2 RFC.

Can anyone recommend me tutorials that walk you through implementing an authentication server from start to finish, preferably in Python?

I don't have anything to recommend, except that if this is for anything other than the fun of learning, just use AuthLib instead.

Tei posted:

I feel bad for you.

These things are designed in environments where they can have 1 person the expert of technology X, so they see no problem making things "corporate" to ridiculous levels. They really don't understand the mantra KISS.

OAuth plays a very important role in security and identity management, and it is worth learning about the basics if either of those things are of the littlest importance to you.

# ? Jan 17, 2020 18:17

Nolgthorn: Jan 30, 2001; The pendulum of the mind alternates between sense and nonsense

OAuth is a bastard of an authorisation protocol which was adapted to serve authentication after the fact.

Designed by a committee of the worst possible kind. By the time they started trying to bolt authentication on to the spec senior members of the team quit over how absurd the "new standard" had become. It's no good in my humble opinion and the opinion of probably half of people who have been tasked with implementing any such system.

If you don't need yours to be compatible with third party systems then save yourself the brain damage and do it your own way.

# ? Jan 18, 2020 00:38

Nolgthorn: Jan 30, 2001; The pendulum of the mind alternates between sense and nonsense

https://hueniverse.com/oauth-2-0-and-the-road-to-hell-8eec45921529

This is a good read

# ? Jan 18, 2020 00:44

Thermopyle: Jul 1, 2003; ...the stupid are cocksure while the intelligent are full of doubt. �Bertrand Russell

Nolgthorn posted:

If you don't need yours to be compatible with third party systems then save yourself the brain damage and do it your own way.

Yeah, this is the correct take. It's bad because it's hard to implement correctly.

# ? Jan 18, 2020 02:05

teen phone cutie: Jun 18, 2012; last year i rewrote something awful from scratch because i hate myself

welp I had no idea what I was getting myself into.

this is all just for fun because I wanted to learn, but Authlib does seem pretty cool. tbh, all forms of backend development makes my head spin, and reading that spec did nothing for my imposter syndrome. Sometimes I like to venture out from my front-end bubble and every single time I do, I end up getting too overwhelmed to ever accomplish anything.

teen phone cutie fucked around with this message at 02:27 on Jan 18, 2020

# ? Jan 18, 2020 02:25

PT6A: Jan 5, 2006; Public school teachers are callous dictators who won't lift a finger to stop children from peeing in my plane

Grump posted:

welp I had no idea what I was getting myself into.

this is all just for fun because I wanted to learn, but Authlib does seem pretty cool. tbh, all forms of backend development makes my head spin, and reading that spec did nothing for my imposter syndrome. Sometimes I like to venture out from my front-end bubble and every single time I do, I end up getting too overwhelmed to ever accomplish anything.

Start smaller with backend stuff. If you want to try using a third-party API, start with something that's designed to be easy to use instead of a very complex system. Try something like integrated with Stripe for payments, or Twilio for SMS messages or something similar in terms of complexity.

# ? Jan 18, 2020 03:48

marumaru: May 20, 2013

applied for a job, they sent a HackerRank test. 5 questions, 2 hours. couldn't get 2 of them to pass (had several failing test cases).
i feel incredibly inadequate, and the usual defense of "you'd never need to use any of this in real-life front-end development" isn't helping. :saddowns:

# ? Jan 20, 2020 19:29

Thermopyle: Jul 1, 2003; ...the stupid are cocksure while the intelligent are full of doubt. �Bertrand Russell

Inacio posted:

applied for a job, they sent a HackerRank test. 5 questions, 2 hours. couldn't get 2 of them to pass (had several failing test cases).
i feel incredibly inadequate, and the usual defense of "you'd never need to use any of this in real-life front-end development" isn't helping.

Apply for lots of jobs, figure out why you failed the tests and learn from it. Also, remember that lots of the tests are bad and that what you're learning is how to get past stupid gatekeeping, but that's the game.

(yeah, it sucks, but it's a numbers game)

# ? Jan 20, 2020 19:38

fsif: Jul 18, 2003

Inacio posted:

applied for a job, they sent a HackerRank test. 5 questions, 2 hours. couldn't get 2 of them to pass (had several failing test cases).
i feel incredibly inadequate, and the usual defense of "you'd never need to use any of this in real-life front-end development" isn't helping.

The company is probably poorly run if that's their screening process. gently caress em.

# ? Jan 20, 2020 21:56

Adbot: ADBOT LOVES YOU

# ? May 24, 2024 19:12

Nolgthorn: Jan 30, 2001; The pendulum of the mind alternates between sense and nonsense

Inacio posted:

applied for a job, they sent a HackerRank test. 5 questions, 2 hours. couldn't get 2 of them to pass (had several failing test cases).
i feel incredibly inadequate, and the usual defense of "you'd never need to use any of this in real-life front-end development" isn't helping.

Being able to do 3/5 things isn't bad, that's 3 things more than most of the people I've worked with.

# ? Jan 21, 2020 13:48

The Something Awful Forums > Discussion > Serious Hardware/Software Crap > The Cavern of COBOL > The Web Design & Development Megathread

«‹›362 »