Hacker News Reader

Comment by badloginagain | Hacker News Reader

I'm willing to give Spartan a chance, honestly. The ModernIE initiative taken by Microsoft the past few years feels like Microsoft gave IE to a team of competent engineers and told to run with it. They dropped support for older versions of IE and upgraded the technology behind it.

Firefox is too slow. As a power user I notice millisecond differences in page render times. I may be an outlier but I think most people who spend a lot of time on browsers will notice even small differences in load times. I've long switched from Chrome as my main browser, disliking the direction Google has been going with several of their products.

I might be an outlier, but I'll give Spartan a chance to impress. [As a side note, it's amazing how crippling to your reputation bad technology decisions are. IE 6-9 gutted respect for Microsoft]

View on HN

I too will give Spartan a chance, and in likelihood will use it predominantly on my tablets and laptop (since I use IE 11 in those contexts today).

On the desktop, I presently remain a Mozilla partisan, sticking with Firefox through thick and thin. But I have acknowledged that in terms of pure page user interface and interaction speed on a Windows PC, nothing can compete with IE. Its use of GPU acceleration for 2D page rendering makes things like CSS transitions shockingly smooth and fast. The CSS transitions for gradients are particularly delicious; I still have an optional subtle animation in the background of my blog using the gradient transitions. I used to have more animation in my blog (before readers convinced me it was annoying), and IE made Firefox and Chrome look as if they were running on clunky hardware from 5 years ago. IE is the only browser that—even when rendering typical web sites—reminds me that my workstation is a 3770K with a high-spec nVidia GPU. In Firefox and Chrome, I can easily forget how powerful my PC is until I fire up something that kicks off some 3D.

I am hopeful that Spartan will continue to lead in that area of so much importance to me: perceived performance. If it improves its actual JavaScript performance and CSS capability versus IE 11, which I expect it will, I'll likely be quite impressed.

That's actually a really good insight. I've always owned mid-high range computers, but browsing has always been much slower than my computers relative processing power. It's odd how I've come to accept that slow browsing is acceptable just because no one has "figured it out" yet.

Everything gets faster: computers have 4x the memory and processing power of 10 years ago, browsers are easily 10x faster than they were 10 years ago.

However it appears that the average web page has gotten more than 40x slower. (They're also not really pages, they're "apps".)

People my age or a bit younger (early 20s) say craigslist and hackernews look "hideous" but dammit this is how web pages are supposed to be.

I'm 40 and grew up online with an acoustic modem. There is no fucking way craigslist is the way the web should look. Many sites go to far the other direction but even CL is begging to update their look.

I really hope they don't change, I love the simplistic design and never having to worry about major layout changes is great. The downside, their massive legal team has a larger budget.

Even the small layout changes Reddit has done the past few years has started to drive me away.

The big problem that I see, and it applies to any browser, is that every page is loading tons of crap from ad or tracking sites including twitter or facebook. When one of these sites stalls the whole page stalls and it is a chronic problem. As an interim fix I block these domains using my host file pointing to localhost 127.0.0.1 but sometimes that stalls the loading because it doesn't get a reply so it waits for the timeout. I am thinking of setting up a web server on my LAN that replies 404 or 204 to all queries and pointing the worst ad/tracking/social domains there so I can load the content not the dross.

Eh, not totally. For instance, the backoffice UI of Umbraco CMS was rewritten using Angular or something, and now it's just...slow. Unreasonably sluggish. I've seen other work from agencies with no 3rd party dependencies that use all this crazy JS and animations that unless you're on a top line MacBook, it's just sluggish.

I've been using the web for a long time, and while loading was slow, I don't remember UIs feeling so sluggish and unresponsive.

> People my age or a bit younger (early 20s) say craigslist and hackernews look "hideous" but dammit this is how web pages are supposed to be.

Not for me. I'm 18, and I think HN is doing great with this webdesign. As well with other sites.

I prefer an application for e.g. Twitter, because the website is way too slow in Firefox. Especially when you follow over 1000 people.

I also really liked http://motherfuckingwebsite.com as an example explaining what is wrong with websites these days.

You haven't seen Spartan yet so I wouldn't make any promises on what you will do with it.

I think you're being too mild in your side note.

The first half of the IE6-9 era was part of a deliberate scorched-earth strategy to harm the open Internet, and the second half was them furiously backpedalling from that when it failed.

It wasn't a technical decision, it was cynical business strategy, and having their IE brand get trashed is the least of what they deserve. I'm kind of annoyed with how quickly people are willing to forgive them, considering what bad actors they've been up until recently.

I don't care who does it or why - I just want higher quality software platforms. I'm glad that Microsoft didn't immediately give into the "open web" because it's a horrible mess of designed-by-committee compromises.

I wish someone would just make a new browser that lets people browse a mix of native apps and web pages so we can stop bolting more crap on top of HTML/JS/CSS.

Sure, you can decide to just spectate, but like it or not we all have to share this tech ecosystem. If there's no consequence for companies or people that behave in ways that harm the whole industry, we're all going to suffer.

What gives you the idea that am I spectating? I'm a developer who votes with my choices and I've been in this industry for over 20 years. I just don't have a problem with Microsoft's tactics because I'm not an idealogue like you.

You on the other hand are being completely impractical by expecting Microsoft to just roll over and work towards a goal that they don't agree with technically and philosophically. I wonder how you feel about Apple and their choices?

So go use your other browsers and don't use anything from Microsoft. It's funny that you even have a choice since Microsoft is apparently in control of the whole industry!

Leaving aside the greater Windows thing, let's be honest - all the major browsers in that era basically said FU to the standards and raced ahead deploying their own stuff. There was greater interplay between the non-IE browsers, but that was because of marketshare and grim realities. Every browser was guilty of releasing features and functions that were "Best viewed in Browser X"

Sure, Netscape Corp. did, and good riddance.

But the Mozilla Foundation was founded in between IE6 and IE7 and, if I recall correctly, their stand on a standards-based web was there from the get-go. They fought a war, against Microsoft in particular, and were starting to win by the time Chrome came in and cemented the victory.

It certainly wasn't an act of altruism, but the entire modern web can trace it's AJAXy roots back to Microsoft's decision to eschew web standards. Microsoft added XMLHttpRequest to Internet Explorer so that they could develop Outlook Web Access.

IE may have had utterly stank HTML compatibility, but adding XMLHttpRequest and inventing AJAX was probably the single most profound changeset in the history of the web.

Also isn't the css `box-sizing: border-box;` the box model IE wanted to go with in the first place?

[I think this proves what i said right](http://www.paulirish.com/2012/box-sizing-border-box-ftw/)

"Inventing Ajax" does not excuse the behavior though. Plus, I'm sure someone else would have come up with a similar idea not long after.

Such as not supporting XHTML until IE9.

I agree with much of what you are saying, personally the slew of bugs are the unforgivable bit as they are the parts that continue to persist. Legacy features tend to die

> Firefox is too slow.

Really? I didn't notice. Mozilla are working on the new generation of browser engine anyway (Servo), which is much more performant. However Firefox itself is nowhere slow.

I think "firefox is slow" partialy comes from people using css3 animations eg. this http://bartaz.github.io/impress.js/#/bored works smooth on my MBP as well as on my Ubuntu desktop whereas seems sluggish on FF (on both platforms)

Looks pretty smooth to me in the latest Firefox (Debian testing), but if somewhere it's a bottleneck, to improve that Firefox should fully implement OMTC.

Also try using apps which use flexbox intensively — FF triggers exponential slowdown when handling deeply nested flexbox layouts[0]...

[0]: https://github.com/philipwalton/flexbugs/issues/8

I'm not sure why it's reported somewhere on Github and not in Firefox Bugzilla. If reporter cared about the bug he had to report it in the right place.

> As a power user I notice millisecond differences in page render times.

No you don't. Or if you do, it's a strange definition of notice. A millisecond is one tenth of how fast the centisecond counter on stopwatches change -- while one can obviously see that change, could you tell the difference between two different counters going ten times as fast, but at a difference of a couple of milliseconds? I don't think so.

One light-millisecond is ~300 km, or roughly the distance to the horizon at a height of 1000 moh, for another frame of reference.

I'll give it a chance as soon as I can add simple extensions (like ublock, vim bindings, etc). And when they make the UI not look unfinished. Maybe it's silly, but those ugly square tab things look like a prototype before the design team had their turn. Petty as that is, I really find it annoying. Also not a fan of sharing the address bar with tabs, but that's minor in comparison. (I used IE for a long time, until well after Chrome was shipped, so I'm not inherently biased.)

What's the word with Firefox performance? My company's app is critically slow in JUST Firefox. It's an Ember app, but we use D3 and Immutable.js. We aren't doing anything weird AFAIK

What has profiling turned up? Because in my experience, Firefox is faster on the apps I work on for Javascript performance, so I'm really curious what sort of performance profile your app has.

I've noticed that SVG performance in Firefox is about 10x slower than Chrome -- is that what d3 is using?

Yes, D3 uses SVG

My company's app is slowest in Firefox too. Most noticeably its the CSS3 transitions that are slow and chunky compared to all the other main browsers. It doesn't really affect the usability too much, but it still annoys me that I can't get the transitions to be nice and smooth like they are in Chrome or IE.

I'm willing to give Spartana a chance, as long as they release it as open source and i will be able to run on linux without wine.

What browser do you use now? I agree with your points, but I've stuck with Chrome because of the range of extensions.

I don't use much extensions, so that's never been too big of a deal for me. I'm using Opera although currently trying Vivaldi[0]

Vivaldi has a cool story, former Opera guys that want to make a browser for the power user, not your mom- which is basically the only browser that actually targets me as it's intended audience. The downsides are it's only in technical preview mode and they pretty openly admit to tracking your usage [at least their openly admitting it, everyone else does the same expect maybe Old Man Firefox]

[0]https://vivaldi.com

Try Otter Browser too, it also follows the philosophy of the old Opera but it's free software and cares about your privacy. It's also in an incomplete alpha stage though (it has quite a bunch of features but still doesn't have some convenient stuff, remarkably autocompletion in the address field).

http://www.otter-browser.org

I agree. Have you fired up the Windows 10 preview and enabled the new rendering engine in IE? It's a pretty good step forward so far, and I'd imagine the next build will be even better. AFAIK, Spartan itself should be in the next build.

>Firefox is too slow.

Despite all the evidence to the contrary?

>As a power user I notice millisecond differences in page render times.

No you don't. As a web developer, I personally can attest that's not true.

Oh you can easily figure out render performance differences in milliseconds with your eyes. Try that. For example, a difference of 300 milliseconds is noticeable by me. I don't know what are you 'attesting' there. I'm not a web-developer, but just 'a developer'.

Also, what evidence you have that says Firefox can't be slow(even with all the add-ons it supports). In fact, those hundreds of bugs will be quite a sufficient evidence that it indeed is for many people. (It's coming from someone who still hasn't converted to Chrome)

When you say “I notice millisecond differences”, that means “one millisecond”. If you meant something more on the order of hundreds of milliseconds, you should probably clarify that.

I can survive on milliseconds of sleep every night. Preferably 27 to 28 million milliseconds, if you have to be so picky.

Exactly, that's the point minitech was making. Your statement is accurate. Saying you survive on a single millisecond of sleep each night would be clearly impossible - as is noticing single-millisecond differences in page rendering times.

You might dismiss it as pedantry, but there is a clear linguistic difference between "I notice millisecond differences" and "I notice differences of 300ms".

Whoosh.

That's one of those places where common sense should take over literal English. When someone says 'millisecond differences', there are high chances that they mean 'milliseconds differences', and not exactly ONE millisecond difference.

Oh well, first, I was not the one who said that. Then, there are HIGH chances that the original comment meant 'milliseconds' differences and not one millisecond difference, which all of you didn't even try to consider.

> Despite all the evidence to the contrary?

It's fast in benchmarks… but it's either single-threaded or has horrible locking. One slow script somewhere will not just block the tab, but the entire browser. Good luck hunting it down with 40+ open tabs.

(I really miss Opera 12, it's still the only browser to decouple the UI to its own thread so a long-running script doesn't lock up the page. Or the whole browser…)

As an Opera 12 fan, have you tried https://vivaldi.net ? You're the target audience.

Yet another Chromium reskin, with semi-mandatory "social" features? No, I don't think I'm the target audience.

Chrome and Chromium are not Blink. Blink is the rendering engine alone. Opera uses Blink.

You may not consciously notice, but you sure as shit do. I've got some lovely scatter plots sampled across billions of user sessions that show that average page load time in a session has a direct inverse correlation with conversion rate.

Humans are impatient to the core.

Millisecond differences? No you don't.

Over a population, statistically? Yeah, you do.

Reply on news.ycombinator.com