news from many major sources in one spot, what do you think?

that concludes my presence on this forum for the foreseeable future.

i'll monitor the bugreports and will fix any bugs that i can fix.

 

Didn't say it did, I listed what I was going to replicate, then I did it.

Does now in both blink and webkit (or at least has A custom scrollbar). No loving for FF or IE, but really that's NOT something you should be doing on a website as it's considered an accessibility violation. Seriously, it -- much like your colour contrasts -- is just a way to tell visitors to your site to go plow themselves.

MORE so if you're going to have broken, useless, unusable scrolling like what that scripting you have vomits up. Seriously, it's agonizing to even try and use.

Though if building for a full stack application -- aka what running on electron, nw.js, etc -- then you would consider if viable since accessibility rules don't generally apply there.

.scrollSection::-webkit-scrollbar,
.scrollSection::-webkit-scrollbar-button {
	width:1.5em;
	height:1.5em;
}

.scrollSection::-webkit-scrollbar-track-piece,
.scrollSection::-webkit-scrollbar-thumb,
.scrollSection::-webkit-scrollbar-button {
	border:solid #000;
	box-shadow:
		inset 0.375em 0 0.75em rgba(0,0,0,0.7),
		inset -0.375em 0 0.75em rgba(0,0,0,0.7);
}

.scrollSection::-webkit-scrollbar-track-piece {
	border-width:0 0.1em;
	background:#888;
}

.scrollSection::-webkit-scrollbar-thumb {
	border-radius:0.75em;
	background:#0A0;
}

.scrollSection::-webkit-scrollbar-button {
	background:#0F0;
}

.scrollSection::-webkit-scrollbar-button:vertical:decrement {
	border-radius:0.75em 0.75em 0 0;
	border-bottom-width:0;
}

.scrollSection::-webkit-scrollbar-button:vertical:increment {
	border-radius:0 0 0.75em 0.75em;
	border-top-width:0;
}

.scrollSection::-webkit-scrollbar-button:horizontal:decrement {
	border-radius:0.75em 0 0 0.75em;
	border-left-width:0;
}

.scrollSection::-webkit-scrollbar-button:horizontal:increment {
	border-radius:0 0.75em 0.75em 0;
	border-right-width:0;
}

Code (markup):

Not rocket science.

Didn't say I was replicating that. Though it's CSS, specificity override if you're organized correctly makes that too a breeze as all you have to do is change the ruleset.

If you have that, much like the rest of your site I've never seen it work. Not even sure HOW you would go about that... well, transition IS pretty powerful.

<div id="modalName" class="modal">
	<a href="#" class="modalCloseOuter" title="close this dialog"></a>
	<div id="modalName_content" class="modalInner">
		<a href="#" class="modalClose" title="close this dialog"></a>
		<h2>Modal title</h2>
		<p>
			Your Modal Content Goes Here
		</p>
	<!-- #modalName_content.modalInner --></div>
<!-- #modalName.modal --></div>

Code (markup):

Is the template by which all modals -- the slide in or fade-in -- can be created and inherit the associated style. I didn't do the goofy inner-scroll thing on them since that's typically broken methodology. They also would exist static on the DOM so you could in theory have them 'minimized' since when not active they're just hidden via "position:absolute; left:-999em;"

If you needed them dynamically on the fly for CSR, see that nice little make function?

	function makeModalDialog(id, title, innerClass) {
		var inner = make('div', {
			class : 'modalInner ' + (
				innerClass ? ' ' + innerClass : ''
			),
			content : [
				[ 'a', {
					class : 'modalClose',
					href : '#',
					title : 'close dialog'
				} ],
				[ 'h2', { content : title } ]
			],
			id : id + '_content'
		});
		return {
			inner : inner,
			outer : make('div', {
				class : 'modal',
				content : [
					[ 'a', {
						class : 'modalCloseOuter',
						href : '#',
						title : 'close this dialog'
					} ],
					inner
				],
				id : id,
				last : d.body
			})
		};
	}

Code (markup):

60 wouldn't be me then, 61.0.2 here.

Though I very much doubt it "loads in seconds" unless your testing off a lan with an i9 that's liquid cooled and oc'd, and even then. Well, unless you mean dozens of seconds.

What I'm seeing on your site is incapable of said performance on any internet connection or computer I'm aware of. The CODE is utterly unable to do what you describe. NOT BY A LONGSHOT! Not even the same PLANET!

That's what's so frustrating about your claims, techniques, and excuses. Everything I know from almost two decades of working with websites and four decades of programming calls bullshit on this ever even working ANYWHERE!!!

As @mmerlin, @sarahk, @qwikad.com, and pretty much anyone else you ask will likely attest.

Ok, now you're just flat out LYING!!!. Your scripting is incapable of even running on IE... as evidenced by the generic "SCRIPT438" error typical of "fatal" modern ECMAScript that Microsoft's JSCRIPT cannot even run because it doesn't even recognize it!

Even your home page won't load on IE or Edge -- it basically does what qwikad's screenshot did.

It is useless even once it loads on a i7 4770k in either Win10, Mint, or Ubuntu (though the last was only tested in a VM). It sure as shine-ola doesn't work in Safari (again, OSX in VM). or on any mobile device.

Hell just trying to load that on mobile would drain half the battery life.

Have you encountered ANY forum where people have it WORK?!? That are telling you you're doing fine? I doubt it...

well, unless they're trying to sell you something.

 

Well again, I specialize in helping site owners who are knee deep in those types of failures, often long after the original creators have left them high and dry.

Given that on certain types of sites these failings can land you in court, with fines, or unpaid due to breach of contract, there's a REASON actual professional systems are built with semantic markup, separation of concerns, paying attention to the WCAG and usability studies like those at NNGroup; and most importantly keep it in their pants when it comes to scripting with a full scripting off graceful degradation plan.

Aka all the things react, angular, vue, bootcrap, turdpress, and "JavaScript for everything" fanboys scoff at or just plain make lame excuses for.

 

ok folks, round 2..

http://seductiveapps.com
https://gitlab.com/seductiveapps/seductiveapps

- i've improved efficiency, error handling and code-structures by quite a bit, which is clearly visible on http://seductiveapps.com/news(section'English_News')

- i've MIT-licensed my entire platform (i used to ask for 1% of your revenue, but it is free for commercial usage since today)

- i've added seductiveapps.com/search, which points to http://seductiveapps.com/googleSearch(query'') and which is a google search query app that directly goes to google, but which shows no ads and is set to "no personal ads" mode so maybe you're tracked a little less while using it than using the google homepage. (not sure about that last bit).

- if you login with login-name "Guest", password "Guest" (without the quotes ofcourse), then 2 icons at the very top-left of the site in the #siteQuickbar become available.
the green icon allows you to upload entire folder structures worth of pictures, reliably, and use tinyMCE to publish rich-texts.
the icon to the right of it, the hammer and wrench icon, allows you to put a custom user-interface look on :
-- (as Administrator only) : a site-wide default, or a default for a certain page
-- as logged in user : your personal site-wide default (works today), the default for all your pages (doesnt work quite yet), the default for a page of yours (same thing), and your personal default for some page. (this is also still being worked on, but will all get polished within the next week).
After that is implemented, i'll make sure these settings get applied to however you browse into and around a seductiveapps site.

i do value @deathshadow 's remarks, thanks to him it was easy to plug in a minifier-concatenator that does wonders for people behind VPNs.
thanks to someone else on this list, i was able to fix the bug that didn't allow any news to be seen from timezones other than my own.

thanks to earlier messages in this thread about my software not working at all, i'm now getting emails of ajax and javascript errors by viewers.

what i can't do easily however, @deathshadow, is go all CSS instead of that custom approach of mine that you loathe publicly so often.
(A) it takes a ton of lines just to get CSS to work in all *modern* browsers, and having custom javascripts do such tasks ensures that i was for instance able to load up my site on windows 7 with a recent firefox browser.
(B) you can't always get things done in pure CSS, especially if you rely on CSS settings to guide javascript execution.
(C) there are always multiple ways to implement something. if someone has a certain coding style, please don't go dissing it without explaining the need to change, and preferably also offer me a quick effective change that doesn't break other use-cases.

 

after sleeping on it, going MIT Licensed just doesnt feel right.
i changed the license back to it's old version.

 

FYI : the #siteManager (photo album manager and rich text editor linked to a tree that supports now both users and groups), and the Site & Page Settings dialog,
both work much better now. The intuitive applying of settings to certain sections/pages of the site, works now too.
"Look of the site"-settings are now applied when browsing between pages on the site, regardless of what initiated the page change.
Feel free to experiment under the Guest/Guest account. You can just click on the login link and hit the button in the dialog that will appear, right away.
If you have any comments, even negative ones, i'd like to hear them here.
And both dialogs are just about smartphone and pad-compatible too, a few small bugs remain here and there, which will get fixed within the next week or so.

For those of you interested in my business-plan, i thought about it some more, polished the old license, and feel much happier to keep the license as is for the foreseeable future.
I also published the rest of my business- and technical-plans on the front page of my site at http://seductiveapps.com today.
I think some of you might find it interesting. Do read the entire front-page again if you have before, i've added much relevant information today.

 

The scripted scrolling is still painfully slow and agonizing to use, alternative navigation remains nonexistent, you still have large sections of text far below usability/accessibility minimums (your footer text of white on orange effectively being invisibile), and it looks like ghostery is telling your CSR for subnavigation to go suck an egg. The date is blowing out of it's container up top as you've got scripttard controlled fixed heights with dynamic/elastic content...

Though that menu appearing/disappearing willy-nilly certainly isn't helping. You mouse over it, it takes so long you click... then it disappears and does nothing...

ALL because again, you're using JavaScript for things that are NONE of JS' blasted business doing!

Oh, and a max-width might help a lot since your long lines of text are hard to follow -- upping the line-height wouldn't hurt either.

Given all the JS for NOTHING and middle finger to the reason HTML and CSS even exist, hardly a shock it's blowing 129k of markup just for the base page to deliver 15.5k of plaintext and ZERO content media -- around 24k of code's job.

dude, you've got some seriously interesting ideas, but for the love of Christmas LEARN SOMETHING about HTML and CSS before you throw endless mindless pointless scripttardery at everything! You're making things -- even WITH the functionality you're aiming for -- be five to twenty times the code needed to do the job, and in the process having it so slow it behaves worse than IE5.01 on a 486/33 under windows 3.1!

 

ok, well, first of all :

i found a major bug in my news app's PHP code today, which i corrected here : https://gitlab.com/seductiveapps/seductiveapps/commit/0984c12ee40d06917c5f2cd374b5ac0b6f4a2c08 (lines 317-323 are the fix) (the problem was that most of the news on the around 280 RSS lists that i crawl every 5 minutes, was simply getting lost completely. never written to disk or displayed in the browser.).
I also cleaned out the 'World Headlines' section, and find the picture of the global news that it presents now a lot more useful than before today. see http://seductiveapps.com/news(section'English_News__World_Headlines')

feel free to use the JSON and PHP memory management techniques i use in newsApp in your own crawler apps. I've found them quite efficient on a core-i5 with a simple SSD and a simple 2MB/sec outgoing ADSL bandwidth. it goes to show not everything *has* to be done in something like couchdb (with the associated learning curve to keep it efficient).
feel free to just steal, with my permission, the walkArray() and chaseToPath() routines used by the PHP class for newsApp, in https://gitlab.com/seductiveapps/seductiveapps/blob/master/seductiveapps/sitewide/functions.php
they're fast and very handy to have in your toolkit.

i also changed the license for the news app today : you can now run your own copy for free, only if you run it or code from it commercially do you need to pay me a 1% of your revenue fee.
see
https://gitlab.com/seductiveapps/se...pps/news/appContent/newsApp/2.0.0/LICENSE.txt --> http://seductiveapps.com/LICENSE.txt
https://gitlab.com/seductiveapps/se...s/seductiveapps/news/appContent/newsApp/2.0.0

@deathshadow : for the record, loading up either the youtube API to display a youtube video in an iframe, or the facebook API to display their widget that shows content from someone's timeline, and the same for twitter's API, they're *all* around 500KB gzipped somehow.
so is my entire framework, gzipped.

you said : learn something about HTML and CSS. i did, long ago, when making a menu that supported video fragments as state indicators via PNG sprites.
i found that if you put settings in pure CSS, you can't access those properties via jQuery. At least i never found a way to do that. Have you?

also : i do rapid prototyping, and that means hardcoding some or a lot of CSS right into style="" properties in HTML files, is quicker coding than if i were to change back and forth between HTML and CSS files. it also saves a whole bunch of names inventing if you know what i mean.

i've also never been a fan of using specific class names to hold state/UI information via CSS. i prefer to use classnames as guides for javascript execution, and let the javascript operate on the style property of the element i'm working with.
it just gives me a lot more flexibility, and ends up being less frustrating (i find switching between files all the time frustrating as hell), and results in quicker development times, that, on rewrite or writing of the next major version, do undergo a check to see which CSS settings are so steady that they can be put in CSS files. Which once again, i do use, but are without browser-specific or resolution-specific CSS code.
i use javascript to do my resizing, not CSS. and i've got it fully working too, including : respect for the zoom factor of desktop browsers, a feature i predict future smartphones will support too.

so if you find the text too small, or the background too interfering, you can tweak those settings.
and if you open your own account, your settings are protected from changes that others might try to make (unlike the Guest/Guest account, which is a bit of an admin/release-party mode of my site currently).

and may i ask, what about the scripted scrolling do you find annoying? i deliberately made it so you can speed up and slow down the scrolling using quick mousewheel motions. is the scrolling set to continue for too long? should it screech to a halt? what do you think is annoying about it?

i'll also have another look one day soon (within a week) to the way the menu works. but the hiding of menu-items, that's to make it compatible with smartphone screens. i suppose i could turn it off for desktops. would you like me to?

 

FYI : fixed a major bug in the news fetching and storing.

Each upgrade of the sources to the live server stops and restarts the cron level php daemon that fetches and stores news (which is then fetched efficiently by js -> ajax -> php -> server-disk),
but the news was being incorrectly, even double, stored.
This caused the fetch routines to store all found news items (around 6700 for my 282 RSS lists) every time the server's sources were upgraded, and thus could end up showing news items twice or more times instead of just 1 time when upgrading the server more than once per 3 days (which happens a lot for on my server).

This bug has now been fixed. Server disk usage is almost slashed to about 25% of what it was, and double news items are now truly avoided (unless they get posted twice in a period spanning more than 3 days, but that's to prevent CPU abuse when checking for new news).

I've also improved the information in the output of the cron script, in case you're running it directly from the commandline instead of from cron (or, use cron to put it's output in a log file).

for the diffs, see https://gitlab.com/seductiveapps/seductiveapps/commit/d61d838824b3583d5ca8d7463ed3f46dc7d02514
and
https://gitlab.com/seductiveapps/seductiveapps/commit/46c0d03b7782f9b70ee15d45f736a52ba297a417