5 ms·
The biggest drawback of sites from this era is they don't reflow on mobile screens. On a desktop they still work as well as they ever did. I'm still searching f
by Santosh83 7y ago
The biggest drawback of sites from this era is they don't reflow on mobile screens. On a desktop they still work as well as they ever did. I'm still searching for a good WYSIWYG HTML composer that can generate clean, responsive pages. Seems like this is a problem where there isn't sufficient incentive for the big tech companies to tackle, and the only s/w that seems to come close is BlueGriffon.
- jacquesm 7y agoThat's a client side problem, not a server side problem. The rendering and presentation of a webpage are entirely up to the client, absolutely nothing dictates that a page should look a certain way on a certain client.
- lultimouomo 7y agoThen again, the whole page is laid out in a table element, so I could see why mobile browsers won't reflow it perfectly.
- onion2k 7y agoThere are a ton of attributes in the HTML that dictates how the website should render - it has a bgcolor and a margin on the body tag, a width and height on the main table, center tags all over the place, etc. Suggesting that a browser should ignore the HTML spec and do something else would completely destroy the web.
- jimktrains2 7y agoUser agents already can ignore css and have user defined stylesheets and even JavaScript. None if this is new and it's a failure of our profession that that is abnormal and produces spectacularly poor results for users.
- onion2k 7y agoUsers should be able to override the default behaviour of their browser if they want to, but the default behaviour should still be defined by the HTML spec rather than the browser vendor. It's really frustrating when two browsers implement important parts the spec differently and push website developers to work around the behaviour one browser or the other with browser-specific code. It doesn't lead to sites being rendered differently as developers embrace the variety of user agents. It leads to people adding "This site is best viewed in Netscape Navigator" gif or "This application requires Chrome" on log in pages. Those are bad things.
- jimktrains2 7y agoMaybe the solution is to let go of this notion that every device needs to render exactly the way the designers want and embrace that difference. I've been on the internet since the mid 90s. I'm well aware of what was, andit was that way because people then as now wanted things to look and act exactly like they wanted and not embrace that not everyone wants or needs your carefully crafted graphical design.
- onion2k 7y agoI don't have any notion that sites should be exactly the same in every browser, but they should be approximately the same. Having two browsers render the same HTML in completely different ways would be very odd.
- jimktrains2 7y ago>Having two browsers render the same HTML in completely different ways would be very odd Or completely normal? Why shouldn't I bump up my minimum don't size to help read text and reduce eyestrain? Ditto for fixing low contrast text.
- onion2k 7y agoYou should, and the browser should let you. But if the text is set to 14px in the HTML or CSS and your browser is set to display it at the defined size, it should display at 14px. Browser vendors shouldn't decide to ignore the spec of using the defined size and do their own thing instead. That's what was being suggested.
- yjftsjthsd-h 7y agoOn the other hand, reader mode suggests that sometimes ignoring the way the site wants to be presented is a good thing.
- onion2k 7y agoThat's a user derived choice. I'm not saying user's shouldn't be able to change the way a site is laid out if they want to. I'm saying that by default it should use the HTML spec.
- thanatropism 7y ago> I'm still searching for a good WYSIWYG HTML composer that You need a dev team implementing Agile for react-native-ux with CI/CD capabilities and devops.
- verytrivial 7y agoWell, I'm not so sure -- have you tried viewing that site via a WAP proxy? That's the 2002 way to solve the problem.