howtogeek.com, update why you shouldnt use waterfox pale-moon or basilisk

General discussion and chat (archived)
RexyDallas

Re: howtogeek.com, update why you shouldnt use waterfox pale-moon or basilisk

Unread post by RexyDallas » 2018-03-01, 15:24

jd2066 wrote:
RexyDallas wrote:I agree that they disrespected add-on developers, and especially the theme developers. Their excuse for dropping themes is that they have to constantly be updated. Same is said for add-ons. Yet they are the ones who constantly overhaul their browser. They switched to the hack known as e10s, and just as add-on developers were getting caught up with that change, they remove add-on support altogether. Safety is also an excuse unless you pander to dumb people who actually believes they have anyone but them-self to blame if they install an untrustworthy add-on. Oh wait, they already forced add-ons to be signed, so that excuse makes no sense! By using that excuse for removing so many features, Mozilla is only marginally better than those people.
I will agree that Mozilla did disrespect add-on developers who created whole themes for the browser and add-on developers that created XUL-based extensions that used APIs that will or cannot be recreated as WebExtensions APIs.
Also that Mozilla disrespected the users of the add-ons those developers created by making it not possible for those users to use the add-ons they wanted to use.

However, I don't think e10s (Multi-Process Support) was a hack, it was a fundamental change to the Gecko rendering engine that changed how Content (Web Pages in Tabs, Popups, etc) and User Interface (Menubar, Toolbar, Tabbar, Sidebars and other elements built using XUL) functioned.
Without e10s (Multi-Process Support) enabled (which I think you can still disable in Firefox 57 via about:config preference), the Gecko Rendering Engine used exactly 1 Firefox process for all loaded Content and the User Interface.
E10s IS a hack. Multiprocess at all is a hack from my point of view. Did old browsers need to spam your task manager with multiple processes to perform good? No! Did ADOBE FLASH need to spam you with multiple processes? No! Did Opera 12/Presto opera need to spam you with multiple processes? No! Did old web browsers need it? No!

With the same group of pages, Opera 12/Presto uses 154,824K of ram in one process, and chrome uses 287,656K of ram in 13 processes. Chrome's scrolling feels slower, and because of being multiprocess, Chrome's tab switching takes a lot longer than Opera's almost-instant tab switching. Opera wins in every category against the newest version of Chrome.

Firefox 4.0, which was released in a similar time of opera 12, and was after Gecko started getting bloated again, seems to regularly vary between 180,000 and 260,000. Firefox 4.0 also had HCTBP, DOM inspector, and Element inspector installed. Firefox 4.0 has faster scrolling than modern browsers on simplistic pages, though, and has near instant tab switching.

Firefox 3.6, peak unbloated Gecko, is generally around 70,000-75,000 idling, and very gradually gets to around 110,000 when scrolling and generally being active. The memory used when scrolling gets freed seconds after I stop scrolling. It scrolls faster than Opera 12 on simplistic pages, and has even more near-instant tab switching. Opening and closing tabs in Firefox 3.6 is also near-instant. That is generally the way Gecko was before 2.0.

Also, by "simplistic pages", I basically mean all of the pages except bing.

These are the pages I am using:
Image
The Gaia Community one is a web archive page. The Web Archive's floating bar doesn't render properly, but whatever. Both "test" pages are image searches, and the Runescape page is the runescape.com/splash.

It is a hack because they can't find a way to actually improve their speed, because chrome did it, and because they are desperate for new users. It's funny. Mozilla removed all of the bloat just to make it probably even more bloated than it was before. And no, XUL is not a source of bloat, and neither is the add-on system. Firefox 4.0 had both of those, and so did Firefox 3.6.

I can't find out how to download an old version of Chrome, so whatever.
Last edited by RexyDallas on 2018-03-01, 17:28, edited 18 times in total.

New Tobin Paradigm

Re: howtogeek.com, update why you shouldnt use waterfox pale-moon or basilisk

Unread post by New Tobin Paradigm » 2018-03-01, 16:16

Don't forget some of it is Async as well.. Not only IPC but Asynchronous Interprocess Communication with promises to do something not actually doing something. It all has an effect.

Also, have we mentioned that Mozilla's Sandbox hack leaks like a sieve as well? So much for security.. That sandboxing plus async plus ipc it ALL has a cost and it deff outweighs almost any benefit.. Basically, making it moot and may as well not have it and do things properly from the start or at least retroengineer things back to sanity.

User avatar
Moonchild
Pale Moon guru
Pale Moon guru
Posts: 35484
Joined: 2011-08-28, 17:27
Location: Motala, SE
Contact:

Re: howtogeek.com, update why you shouldnt use waterfox pale-moon or basilisk

Unread post by Moonchild » 2018-03-01, 17:06

Just a side remark: don't 1:1 compare Firefox 3.6 with a current-day browser. That's apples and oranges. Sure, 3.6 would be much much lighter on resources but it also doesn't have any ES6, modern CSS, or anything else that is absolutely required for current-day web pages. The standards have gotten exponentially more complex to implement and require a lot of code, that all has a cost.
Last edited by Moonchild on 2018-03-01, 17:07, edited 1 time in total.
"Sometimes, the best way to get what you want is to be a good person." -- Louis Rossmann
"Seek wisdom, not knowledge. Knowledge is of the past; wisdom is of the future." -- Native American proverb
"Linux makes everything difficult." -- Lyceus Anubite

RexyDallas

Re: howtogeek.com, update why you shouldnt use waterfox pale-moon or basilisk

Unread post by RexyDallas » 2018-03-01, 18:41

Moonchild wrote:Just a side remark: don't 1:1 compare Firefox 3.6 with a current-day browser. That's apples and oranges. Sure, 3.6 would be much much lighter on resources but it also doesn't have any ES6, modern CSS, or anything else that is absolutely required for current-day web pages. The standards have gotten exponentially more complex to implement and require a lot of code, that all has a cost.
But they should still use an amount of memory at least close to Opera 12 when loading these mostly simplistic pages that don't even use that much JavaScript, nor do they actually use those complex features. Also, that does not explain why scrolling is faster in Opera 12 and Firefox 3.6. Also, Firefox 3.6 doesn't even have hardware accelerated text rendering. Also, that doesn't explain why switching tabs, or loading pages, and going back/forward, are not as fast.

Also, that's not the point. the point is that browsers don't need to have multiple processes or insane amounts of ram/CPU to be fast, and that XUL does not affect Firefox's speed in any significant way. Especially if they're just rendering basic pages. Thus, I view multiprocess as a hack. They don't even need multiple processes to have hardware acceleration. The only reason e10s exists is because chrome has multiple processes, and Firefox wants to be like chrome. Multiple processes just wastes space in my task manager, and makes switching tabs take even longer when the tab isn't in the same process. In addition, multiple processes means they have to use inter-process communication just to render a page/UI, or to communicate with any add-on, adding unnecessary overhead in the name of looking cool.

Just opening google.com in nightly makes my fan sound like I'm running a game on unlimited FPS!
Last edited by RexyDallas on 2018-03-01, 20:52, edited 14 times in total.

User avatar
Moonchild
Pale Moon guru
Pale Moon guru
Posts: 35484
Joined: 2011-08-28, 17:27
Location: Motala, SE
Contact:

Re: howtogeek.com, update why you shouldnt use waterfox pale-moon or basilisk

Unread post by Moonchild » 2018-03-01, 20:56

Sorry, no. Opera 12 is also from a previous era of browsers, being 5 years old, with similar limitations to old versions of Firefox. The same way Pale Moon 20, also a very fast and responsive browser, would not be able to do much with the current web except with very simple pages.

I absolutely agree that e10s is a bad idea (see also a blog post I made on this forum not too long ago), but the reasons why is not something you can simply deduce by comparing it to ancient browsers (that do, by the way, have accelerated text rendering, but in a different way) and how responsive to input they are on "simple web pages". Modern browsers can of course render those pages but they can also render massively more complex pages very fast. That capacity comes at a cost (in terms of being heavier) for any page loaded, including "simple" ones, because you can't just "switch off" or unload parts of the browser you don't happen to need for the particular page you're visiting; that's not how (the vast majority of) programs work.
"Sometimes, the best way to get what you want is to be a good person." -- Louis Rossmann
"Seek wisdom, not knowledge. Knowledge is of the past; wisdom is of the future." -- Native American proverb
"Linux makes everything difficult." -- Lyceus Anubite

galneon

Re: howtogeek.com, update why you shouldnt use waterfox pale-moon or basilisk

Unread post by galneon » 2018-03-02, 02:10

I recently requested the developer behind the FlagFox extension externalize his geo-IP database so that the unsupported older version of his plugin could easily keep up with the new WebExtensions version simply by updating the DB and not the whole plugin--users could do this manually without making trouble for him. He said he'd received enough requests for a new 'legacy' version that he would think about it. He's now released a new legacy version I'm happy to report (the DB is still integrated into the plugin, though). Following that announcement, he then posted this, linking to the howtogeek article.

Over Twitter he mentioned to me PM being based on Gecko from ESR 38. I guess it's a moot point specifically with plugin developers, but perhaps more aggressive messaging is in order regarding Goanna--that it is a continuing effort and that PM didn't simply give up and call it a day with an ancient version of Gecko. When regular users see extension developers slating something that is only a downside for extension developers who are primarily concerned with (the higher market share) Firefox, that detail is sometimes lost on them and they view the browser as more generally deprecated. I think the perception is becoming that PM has dead-ended across the board--while it may be true when it comes to extension support with no intentions to implement WebExtensions, it's a problem when people extrapolate from that some kind of blanket freeze on progress. The fact that it's primarily a PR problem is really beside the point--it potentially stymies adoption rate, or worse, negatively affects retention.

P.S. This modern-day obsession with page load speed is such a '90s concern from the time when PCs were deprecated every 2 years and you regularly ran out of memory and cycles to do anything at all. The fact that PC users are not savvy enough to understand 'speed' as it relates to browsers today is a marketing ploy depresses me. If the only good thing I ever hear about Quantum is how "fast" it is, my i7 and I aren't inclined to bother checking it out.

klipkyle
Hobby Astronomer
Hobby Astronomer
Posts: 27
Joined: 2018-01-28, 23:11

Re: howtogeek.com, update why you shouldnt use waterfox pale-moon or basilisk

Unread post by klipkyle » 2018-03-02, 05:05

galneon wrote:Following that announcement, he then posted this, linking to the howtogeek article.
That's too bad. He bought into the lies and propaganda.

It turns out that Chris Hoffman was involved with Firefox. I came across his name while browsing one of the Mozilla Google Groups archives. He had an @mozilla.org email address.

And now Chris Hoffman is publishing lies about our projects. And that's what they really are. Lies.

It has been several days and no response from the editor. I am considering posting my email to the editor publicly on Reddit.

User avatar
Moonraker
Board Warrior
Board Warrior
Posts: 1878
Joined: 2015-09-30, 23:02
Location: uk.

Re: howtogeek.com, update why you shouldnt use waterfox pale-moon or basilisk

Unread post by Moonraker » 2018-03-02, 09:45

So what happens after july the 2nd when updates stop for firefox esr 52.?
Also that blog appears to contradict itself as from the start it states the advantages of open source and then later condemns using any forks from the open source tree.
However thankfully mr hoffmans opinion does not seem to be widespread and the opinion of one person is exactly that,just an opinion.
user of multiple puppy linuxes..upup,fossapup.scpup,xenialpup..... :thumbup:

Pale moon 29.4.1

New Tobin Paradigm

Re: howtogeek.com, update why you shouldnt use waterfox pale-moon or basilisk

Unread post by New Tobin Paradigm » 2018-03-02, 12:56

Since when are lies.. opinions.. That is very dangerous.. Opinions can't be right or wrong .. not really but lies are so are perceptions and conclusions based on lies. This guy has perceptions and conclusions based on lies and misinformation that is easily verified as false.. Opinion only goes as far as in his opinion one shouldn't use forks but reasons to support his opinion are lies and misinformation thus this guy is a lier and his opinion, while entitled to have one, is baseless and irrelevant.

Fedor2

Re: howtogeek.com, update why you shouldnt use waterfox pale-moon or basilisk

Unread post by Fedor2 » 2018-03-02, 17:41

well i have looked that how-to-geek.com, excited with their main menu:).

Of that article i came to conclusion the sites main audience are schoolboys. Their key words are 'modern' and 'security', ah they pity are.


I recall another site http://www.majorgeeks.com just look it and see the difference.
Last edited by Fedor2 on 2018-03-02, 17:43, edited 1 time in total.

klipkyle
Hobby Astronomer
Hobby Astronomer
Posts: 27
Joined: 2018-01-28, 23:11

Re: howtogeek.com, update why you shouldnt use waterfox pale-moon or basilisk

Unread post by klipkyle » 2018-03-02, 22:57

New Tobin Paradigm wrote:This guy has perceptions and conclusions based on lies and misinformation that is easily verified as false..
Good catch! His opinions are based off of false premises. The false premises are lies. I should have made that clearer.

The article is a good example of "garbage in, garbage out."

galneon

Re: howtogeek.com, update why you shouldnt use waterfox pale-moon or basilisk

Unread post by galneon » 2018-03-03, 00:35

klipkyle wrote:It turns out that Chris Hoffman was involved with Firefox. I came across his name while browsing one of the Mozilla Google Groups archives. He had an @mozilla.org email address.
I was going to inform the FlagFox dev of this, but I decided to look into it a bit first:
As Director of Engineering and then Special Projects at the Mozilla Foundation and Corporation since 2003, Chris Hofmann has spearheaded the research and development work of thousands of open source contributors around the world.

A Netscape employee before joining Mozilla, Chris contributed to every Netscape and Mozilla browser release since 1996.

As the first employee at the Mozilla Foundation in August 2003, Chris led a small but devoted team of the original ten engineers that established the Mozilla Foundation as an independent and self-sustaining organization.

In 2004, Chris managed and executed the first worldwide release of Mozilla Firefox 1.0. Firefox 1.0 helped to fulfill the Mozilla Foundation’s goal of supporting open Web standards and provide innovation and choice for Internet client software and set Firefox on a path to remarkable market share growth over the last several years.
https://wiki.mozilla.org/Marketing_Work ... is_Hofmann

We actually owe that guy quite a bit, I'd say.

The Chris Hoffman from howtogeek is just a kid:

https://twitter.com/chrisbhoffman

So while Chris Hoffman the Younger may be misguided, he doesn't appear to have any conflicts of interest rooted in an official Moz association.

User avatar
loxodont
Astronaut
Astronaut
Posts: 725
Joined: 2014-07-26, 23:03
Location: Mare Serenitatis

Re: howtogeek.com, update why you shouldnt use waterfox pale-moon or basilisk

Unread post by loxodont » 2018-03-04, 15:22

Ow c'mon guys, this personal thing gets over the top. He's a journalist like thousands who job-wise just write about their third-party opinion and give nonsense recommendations with poor reasons. Thehandyman1957 already illuminated us a bit on page 1. Also in the HTG comments, many readers are smart enough to not buy that opinion.
It's OK to point on bullshit and to show what's wrong with that, but pointing on personal backgrounds, history, etc... is so redundant, because it doesn't add anything to the core nonsense... I'm sure he loved foxes already in his childhood, yea, perhaps, but you don't need to explain why he's writing an article with much Firefox love in it.

- Just writing this too, because I recently started to dislike Pale Moon article comments on ghacks which completely glitched off-topic and into washwives talk about PM devs. So it would be nice if this place could keep being "paradise", like bad tongues use to say. ;)

User avatar
bawldiggle
Lunatic
Lunatic
Posts: 446
Joined: 2013-02-22, 21:16
Location: East Coast Australia

Re: howtogeek.com, update why you shouldnt use waterfox pale-moon or basilisk

Unread post by bawldiggle » 2018-03-17, 12:23

@ loxodont

agreed
:coffee:
Win-7 PRO 64-bit
Palemoon; auto updates current version (32-bit)

jd2066
Moongazer
Moongazer
Posts: 10
Joined: 2018-02-25, 14:30

Re: howtogeek.com, update why you shouldnt use waterfox pale-moon or basilisk

Unread post by jd2066 » 2018-03-23, 04:29

RexyDallas wrote:
jd2066 wrote: However, I don't think e10s (Multi-Process Support) was a hack, it was a fundamental change to the Gecko rendering engine that changed how Content (Web Pages in Tabs, Popups, etc) and User Interface (Menubar, Toolbar, Tabbar, Sidebars and other elements built using XUL) functioned.
Without e10s (Multi-Process Support) enabled (which I think you can still disable in Firefox 57 via about:config preference), the Gecko Rendering Engine used exactly 1 Firefox process for all loaded Content and the User Interface.
E10s IS a hack.
After reading Moonchild's post about Multi-Process Browsers in general, it does seem that the way multi-process is implemented in most (if not all) web browsers is a hack.
I did not realize before all the trade-offs that had to be made for multi-process functionality in a web browser to work compared to just one process.

Interestingly enough, after reading this topic and Moonchild's post about Multi-Process browsers, I decided to switch to using Pale Moon as it became clear that Pale Moon is what Firefox used to be before Mozilla decided to change things.

Once I installed Pale Moon, I looked at the Pale Moon add-ons page and found many the older add-ons that I had used with Firefox in the past that were incompatible with Firefox 29+ that I just had to do without until now.

So at least one good thing came out of Chris Hoffman's article, it lead me to ultimately deciding to try Pale Moon as my primary browser and preferring it over Firefox 56 and Waterfox :)
Last edited by jd2066 on 2018-03-23, 04:34, edited 4 times in total.

Locked