A worthwhile experiment, absolutely, but I can’t see myself ever using this for a serious production application. I hope we keep iterating on this before settling for a non-turing complete, directive-driven language that gives Angular 1.0 PTSD flashbacks
OK, I get what's happening here. Example goes hard out of scope than what HTMX serves. No wonder we get confussed Pikachus around this breathe of fresh air.
My #webDev dudes, dudx and dudettes we really need to figure out smartphone font sizes.
I'm only a little bit old and I've only popped my android phone system font up by one level, but it messes up so many links and buttons all over the web.
@sarajw if you are reporting through the feedback tool, then you can likely add them as attachments. Otherwise, anything you can give me, repro pages, screenshots, code samples, etc… would be great, and I’ll report the issue for you.
@evan well, for me, I've got a synology NAS that's got a webserver built in. I open up port 80, 443 and set the firewall to port forward there and that's the end of it.
Personally, I have an amazon account and I use cloudfront to proxy my sites, that way I get free ssl and I can scope my port opening to CloudFront's IP
@evan I personally use netlify since it's nice to combine with CI/CD, in my experience nicer than GitHub pages.
Best feature: preview deployments of PRs.
It is 2023. XHTML is dead and buried. HTML is a "living standard" with billions of users. So what kind of idiot would want to build a website using XML? Me. I am that idiot. Last year, I launched a "web page" which didn't use HTML. Called, appropriately enough, "YOU DON'T NEED HTML!" That (ab)used […]
Combined with CSS modules to scope styles to particular components only (while still being able to take advantage of the cascade if needed), that feels like there's no real difference in terms of composability? Anything I'm overlooking?
@VincentTunru@frontend yes, css modules handle the scoping quite well in something like jsx. One difference with shadow css is that even element rules from the light dom (e.g., anything in the page’s css reset) is not applied. A common conflict with css component libraries is that they rely on different resets.
In the flip side i find that i end up needing to apply a simple reset in every web component’s css. Would be nice to be able to inherit css rules from another shadow root.
I wonder how much more performant modern websites would be if people had to be explicit about what gets rendered when and updated how. presses a key in an edit box the entire website explodes and redraws
I've spent a few days playing around with #HTMX, and I'd like some insights from people who are using it in production. Doesn't need to be a paid project, but more than just "my pet experiment".
What are you using it for?
Like, to me it feels as if either the backend needs to be really tailored to HTMX, with HTML fragments and custom headers and stuff, or you need to start writing non-trivial amounts of JS in the frontend for anything but the most basic tasks.
@scy@odrotbohm sorry, didn’t mean to splain you :)
I was at a similar moment, fuzzing a lot around how to build requests and how to treat the requests from htmx, hoping to find a way that feels elegant.
If you continue to try out things and find a way, I’d be happy to see your approach whatever you find.
@kubikpixel@fedora KDE Plasma also works, but I was never a KDE guy. Personally like gnome more. And this is the great thing about Linux ... you have the choice
If there was an #HTML element that changes it's content when users interact with other elements on the page, what name would it have?
PLEASE NOTE: I am not suggesting that this element needs to exist; I am only asking what it would be called. I'm building a CustomElement, I just want it to have a name that makes sense.
Vote and suggest others in replies. Please boost for reach!
Round 2; because I've gotten some good suggestions
This <HTML> element updates itself with the result of an HTTP request that some other element on the page made. It can replace, append or prepend it's contents.
Do you have a favourite non-obvious use of #css#grid? I'm looking for nice examples for my Grid chapter that will expand beyond "here's how you make a responsive four-column grid"
@joelanman But those aren't foundationally breaking changes. Deprecating <center> means text looks funny. Deprecating <a> means the very foundation of the web—hyperlinked pages—break.
I need to convert #HTML to #Markdown and I'm looking for a tool to do that.
The output should
• preserve line breaks in paragraphs
• not contain additional, unnecessary linebreaks (e.g. 4 empty lines between paragraphs)
• be configurable (e.g. whether to use * or _ for emphasis, or * vs - for unordered lists)
• if possible, allow me to hook into details (e.g. to convert <pre class="shell"> to ```sh)
#Python or #CLI. Alternatively, what's a really configurable prettifier?
@sarajw for web, I don't really check much since browsers use page zoom instead of font resize.
But I understand that there's OS-level font-resize. Weirdly I had to do font size check a lot when building mobile apps (React Native) due to user base catered towards older folks who changes to very huge font sizes and breaks literally everything. Had to talk with the designers often 😂
@chriskirknielsen I've stripped all of my styling from the one that says test link. It's still the same size. Maybe it's a bug and all <a> tags just don't resize with the rest of the text on Android chromium (in my case, Edge)...
Have tried styling a {} with font-size: inherit and 1rem too, no dice.
Anyone else following the #CSS tag seen this? Links not following increased system font size on mobile (android, chromium)?
In the last few days I’m experimenting with substituting CRUD API code with Stored Procedures which directly produce the endpoints JSON as a single-row scalar value. API is then just a wrapper that authenticates, validates input and streams the DB’s JSON directly to the client.
No ORMs, no SQL generators etc.
All SQL is where it should belong: in the database
API does only single „CALL myfunc(…)“ db calls
A simple centralised error handler can accurately report errors from the database
No weird mixed row/json columns scanning into structs and re-marshalling everything to JSON
Codebase is collapsing to 20% (by LOCs)
Stored Procedures can use wonderfully declarative SQL code
Response times in the microseconds, even for multiple queries, all happens inside the DB
More side effects:
the data model can change and evolve without touching the API at all
Zero deploys mean zero downtime
the API application is so tiny, I could easily switch it to any programming language I want (yes, even Common Lisp) without worrying about available databases libraries, type mapping and rewriting tens of thousands of lines of intermixed language/SQL-code.
The general direction of the dev industry is heading in the opposite direction. More ORMs, more layers, more database abstraction. More weird proprietary cloud databases with each their own limited capabilities and query language.
So you tell me: Is it crazy? Is it wrong? Why do I have doubts despite everything working out beautifully?
Intermediate report on my "Stored Procedure" project (long post).
I've now written 35 stored procedures and functions ( in MySQL ) replacing a massive amount of application code.
MySQL's stored procedure syntax is very capable, I had to write only a few functions to add some JSON conversion comfort. I can use multi-result sets and had to modify the MySQL driver to Go to support OUT parameters, not a big issue.
What I can say is that a really good SQL editor that "understands" your code and provides some intellisense while writing helps a lot. So far I stick with #DataGrip, which also supports syncing the DB schema with Git, so every change can be properly tracked. It has a price tag though.
I've tried MySQL Workbench, DBWeaver, TablePlus, Querious. None of those help with writing stored procedures like DataGrip does.
#DevArt's MySQL Studio could be another candidate but is Windows-only and painfully slow running with #Crossover on macOS.
I've replaced really super-ugly dynamically generated SQL code with beautifully pure SQL queries, discovering a lot of bugs during the process. Now, looking at the code, I can see what it does, not just hoping that it will work out. Reminder: I have to work with a complex data model spanning many tables, so dynamically adding JOINs and ORDER BY clauses is not easy and amount to potentially hundreds of test cases. Not with Stored Procedures.
If you are really interested in a big-boy query involving spatial queries (distance from point, bounding box etc.), LATERAL CROSS JOINs and parameterised ordering, check this out:
What surprised me most is that MySQL is fast and I mean really fast. In most cases involving complex JOINs over many tables much faster than PostgreSQL (which is everybody's darling, I understand). Having this performance at hand I could replace all hourly-updating MATERIALIZED VIEWs with real-time queries.
Also, MySQL has excellent Spatial functions and a decent FULLTEXT search, I don't miss anything from PostgreSQL right now. Also its native Event scheduler (a cron for SQL), makes table maintenance a breeze. No more separate SQL scripts running in a crontab anymore.
PostgreSQL pg/plsql syntax is bulky, ugly, and hard to read (with::all::those::automatically added::type::casts). MySQL doesn't need this.
What MySQL is not good at: error messages are super-crappy and don't help most of the time ("you have a syntax error, check that your code is correct ..."). So a high level of SQL confidence helps, but documentation is very accessible and will structured.
There is still a long way ahead, I think I've migrated about 50% of the application right now - there are still some complex use cases to solve but I feel confident about this approach.
If you still think about MySQL capabilities in terms of versions 5 or earlier, you will be surprised how far it came with the most recent version 8.
I know Oracle is universally hated in the Fediverse, but: don't forget that most of PostgreSQL recent "let's add hundreds of features every 6 months" party is also mostly driven (funded, developed) by commercial parties looking to lure away Oracle and SQL Server clients.
I can't wait to finish this project and will keep you all posted.
#mysql#sql
(#)PostgreSQL (let's not make half of the Fediverse angry... 🙂 )
I see this as indictment of browsers as a "user agent". They seem so uninterested in exposing semantics in any interesting ways. Many A11y problems come from browsers failing to expose semantic info for readers. But beyond that: why do we still need a TOC?! The browser has that info, and could expose it on all sites!
If the browser showed us the semantics that would help us use the semantics.
(On the other hand, it's funny to imagine "solving semantics"…)
While I think filesystems could have better UX, I reckon understanding that files are a thing is empowering. The only reason it is arcane knowledge is because it has been made so. If people can understand the concept of books they can understand files.
HTML/CSS is incredibly versatile & can work as well for the commandline as it is for GUIs. JS & an over focus on visuals fights against this.
When I searched on: cell phone follow property lines
I got 4.7+ billion hits
The first hit is a PDF for my consulting service.
New personal record for number of hits where one of my things is at the top. #WebDev#SEO
Does anyone have recommendations for a lightweight self hosted alternative to wordpress? I'm done with WP.... I'm at the stage where I'm about to bust out notepad and write some HTML 😂 #CMS#WebDev Please boost for more EYES 👀
@SeaFury I've had some success with Laravel + TwillCMS.com . Prior to that I ran an OctoberCMS install (it's great), but I think their pricing model changed recently.
@voxpelli indeed and our experiments have shown behavior is still divergent and, well, just not there yet. I do not like it when tech ppl champion an unimplemented solution to current irl problems so I try to avoid that sandpit.
Anyone else feels #html heading ranks (h1 to h6) are meant for documents like MS Word documents/essays and that the HTML spec should have a generic <caption>/<legend> element for UI components like product lists, widgets, article lists, etc.?
In my interpretation <ol><li><a><article><h1>(product name... is the correct structure for a list of product thumbnails #semantics wise, but I doubt this is right for #accessibility in practice. #webdev
A long time ago, as HTML5 and the living standard was coming together, it was proposed that heading levels would reset within an article or a section. This would make sections more "portable" - you could simply extract a section and put it in its own page without having to modify its heading tags.
But for some reason browsers and search engines pushed back on having this - and so we kept the h1-h6 hierarchy as it is.
