Slashdot Mirror
Mozilla Teases First Browser Dedicated To Devs
hypnosec writes Mozilla has teased a browser for developers — a first of its kind — in a bid to equip developers with a set of tools at one place for better and enhanced productivity. Speaking about the perils of web development Mozilla says engineers, while building for the web, use a range of tools that don't always work well together. Because of this, they have to switch between platforms. This process of switching from one platform to another makes a developer less productive, Mozilla says. The not-for-profit organization hasn't detailed its browser for developers to a great extent, but has revealed that the browser will be available on November 10.
Isn't SeaMonkey Studio _exactly_ what the summary describes? Maybe I'm missing something, or maybe SeaMonkey is a community fork and not official Mozilla, so it's not counted as the same.
Unless it can run multiple browser engines I'm not sure how much application switching it will prevent.
I write code in the editor of my choice, then I open chrome and look at my results. Any issues are addressed with the already pretty nice built in tools.
Firefox is just something I check (like IE) for feature parity.
It's kind of amusing this should show up today, the same day I discovered a somewhat amusing little issue with the Firefox developer tools:
The "JavaScript error" developer console log messages (e.g., JavaScript errors) are not necessarily displayed in the same order that "JavaScript console" messages (i.e., console.log) are generated.
Meaning that if you're trying to track down what's generating a JavaScript message in some library you're calling (that is, a warning because the library "helpfully" catches the error for you and just does nothing), you: 1) can't get a stack trace of where that message was generated and 2) can't rely on "console.log" statements to help you narrow it down since "console.log" messages can be out of order of any other message type. I have no idea why this would be the case since JavaScript execution is explicitly single-threaded and having messages generated by a single thread appear out-of-order makes absolutely no sense, but - well, Firefox managed it.
I did, eventually, figure out a solution to my problem: I used Chrome instead. Not only did my app run twice as fast, Chrome messages are in order and included the property being read off the null object. (Allowing me to track down how the library managed to find a null off a non-null argument.)
So I'm glad Firefox is trying to make a "developer-centric browser," now if only their current browser tools weren't terrible.
You are in a maze of twisty little relative jumps, all alike.
Why must they tease their own browser? They should pick on IE or something.
What leaks? I've yet to see them since I switch back to FF. I've had this session open for several days, I'm presently at 330mb with three tabs. The most I remember seeing this session (last night, in fact) is 670mb though I don't remember how many tabs I had open at the time.
I think the memory leak meme has outlived reality...
Required reading for internet skeptics
I must have missed the memo where Firefox was already properly multi-core ready, with add-ons and jetpack actually growing thus making this browser relevant again.
I hate Chrome's evil, but I hate suckiness more than I hate evil. So Chrome it is for me.
Entia non sunt multiplicanda praeter necessitatem.
Apparently it lets you debug other browsers remotely with "Firefox Tools Adapter"
https://hacks.mozilla.org/2014...
I think that Mozilla still make a wonderful browser for developers but they keep trying to screw it up. Like this
http://www.donotlick.com/2014/...
Why on earth am I going to choose to develop websites on the one browser that I can be sure that none of the visitors are actually using*? Firebug and Notepad++ is good enough for me.
* No, not Opera.
This is hardly a new idea. Dreamweaver has had the ‘web browser for developers’ functionality via Live Code, Live View, and Inspect, for years.
I use Firefox all the time, but of late all the browsers seem basically good-enough (though don’t get me started about HTML5’s implementation of draggable) or at least compatible enough.
There is always work to do, but Microsoft seems to have largely conceded the battle against standardization.
I doubt I will find Mozilla’s new browser for developers Earth shattering. I hope I’m wrong.
What really seems needed is just continued pursuit of refinement of the HTML5 standard and work towards making it syntactically regular, grammatically powerful, and user friendly.
In closing, dear God, please someone fix draggable, I want an easy to use, powerful language more than I want cross-platform tools.
Letter To Iran
What leaks? [...] I think the memory leak meme has outlived reality...
That just means it's gone gold, as far as Internet memes are concerned. If an Internet "meme" can remain in usage past the natural lifespan or the relevance of its subject, some people mistakenly think that makes it funny.
grumble, grumble Al Gore invented the Internet @(&*@) The Internet is ... a series of tubes *&^^$%^)*#@ 640k[i]B is enough memory for anyone #$@#$@*& BSD is dying !$%#@#)
Switching rendering engines between Gecko, Webkit and IE.
If Pandora's box is destined to be opened, *I* want to be the one to open it.
I hope Mozilla's new one takes them seriously for a change.
Seriously.
Why do they have to have a distinct browser for devs? What, exactly, can the Mozilla plugin architecture NOT support?
Building a separate browser simply increases the chance that you will introduce compatibility problems into the environment as both projects progress.
Chas - The one, the only.
THANK GOD!!!
There aren't just one or two people reporting memory leaks with Firefox. There are lots of people reporting these problems, even if some people don't experience them.
I don't think it's a "meme". I don't think it's people trolling. I see these comments all over the place online. I hear of similar experiences in person from co-workers, friends and relatives. This comes up far too much to just be a coincidence.
When people are reporting that fresh installations of the most recent version of Firefox end up consuming multiple gigabytes of memory after moderate browsing, then something is clearly wrong. Maybe it isn't broken on your system, and maybe it isn't broken on the systems of the Firefox developers, but it's apparently broken on the systems of a great many end users.
Yet instead of doing the responsible thing and accepting that there is in fact some sort of a problem, members of the Mozilla community (such as yourself) end up denying the problem exists, and then you ridicule everyone who does suffer from it. Sometimes you blame it on "extensions", even when the users are using a fresh installation of Firefox without any installed!
Responding to what's very likely a completely real problem in that manner only drives users away from Firefox. That's part of the reason why Firefox is only at around 10% of the browser market at this time, with that number continually dropping. It's because Mozilla doesn't listen to the users any longer, and the wider Mozilla community treats many other Firefox users with complete disrespect.
Users don't want to deal with bugs like memory leaks in the first place, and users don't want to deal with a hostile response when they report such problems. They'll just go and use Chrome instead, which doesn't even suffer from these problems in the first place, and which is also really easy to download and install.
Instead of responding how you just did, you and others in the Mozilla community need to acknowledge that this problem exists, acknowledge that others may be affected by it even if you aren't, and you need to try to help find a solution to these problems that doesn't involve ridicule and denial.
If Mozilla and the greater Mozilla community doesn't change their ways, then we won't be talking about how Firefox is only at 10% of the market. We'll be talking about how Firefox is at 3% of the market, assuming Mozilla still even exists as an organization at that point.
No, it's just slow and bloated now and takes fro every to start in comparison to chrome
This. Plus, the browser is useless without extensions to undo the damage the UX team has been doing to it ever since Fx 4.0.
HTML, JS, CSS are all W3C standards, so why should a web developer care what software the user browses with?
For at least three reasons.
First, HTML, CSS, and JavaScript are at times underspecified. Touch screen and mouse-and-keyboard devices send events in different orders on different platforms. Differences in font rendering cause a particular CSS box to have different heights on different platforms. Differences in 2D canvas antialiasing and image resizing algorithms cause the output of an identical script to have subtly different appearances. Different support different audiovisual codecs: Safari requires MPEG codecs whereas third-party browsers often require royalty-free codecs. And different browsers for small-screen devices have different triggers to disable desktop-width document layout.
Second, user agents are free not to implement certain parts of the standard at all. Older browsers are unlikely to implement new standards, requiring use of polyfills. Some browsers require prefixes for specific CSS properties and values and JavaScript objects. Some omissions appear deliberate, designed to sway web developers toward the device manufacturer's (paid and OCD-curated) native app developer program. For instance, <input type="file"> wasn't implemented in Safari for iOS (or any of the other browsers in the App Store, which are all Safari wrappers) until iOS 6, and as I understand it, it still doesn't work for any content type other than pictures and videos. WebGL still doesn't work in web pages on iOS.
Even among parts of the standards that a user agent does implement, quality of implementation varies. One browser's JavaScript JIT might execute a particular construction quickly, another not so quickly. WebGL might work on some underlying video drivers but not others.
What really seems needed is just continued pursuit of refinement of the HTML5 standard and work towards making it syntactically regular
If you want something syntactically regular, you can always use the XHTML syntax of HTML5.
Not that I don't like a good angry screed every now and again. But how does this address the point I made?
(Hint: It doesn't.)
Chas - The one, the only.
THANK GOD!!!
It may not be memory leaks, but Firefox refuses to release RAM once it hits a certain threshold - somewhere in the 2-4GB range for x64 (Nightly, and Waterfox); prolly a similiar range 2-3GB for 32bit FF. Easy to accomplish, open a bunch of images or dozens+ of tabs, once Ram usage hits 2GB+ start closing tabs...
It's likely the upper-end of the 2GB range, but still FF is unable to manage it's own memory usage.
I think its more about Firefox not releasing ram, than memory "leaks", but YMMV.
It's not quite as bad as you put it. There are still alternative browsers from forked FF code that kept sane UI and are trying to keep extensions more stable per release.
I use one of those forks. Pale Moon. FF has been largely unusable for me since the 3.6.28 > 4.0 transition.
His point is that FF devs don't give a rat's ass about user concerns such as one you presented. They just do whatever they feel they should do, no matter how illogical or compatibility-breaking.
Focusing on the "end user" was a mistake to begin with, so hopefully this will undo some of that counter-productive effort/pandering. The more the software allows the user to program, the more the user will program. Treating the web like a passive consumer platform has done it, and those who develop and use it, great mental and social harm.
They don't have time to listen to users, or for fixing their code. They're too busy changing the UI again this week, and making sure only people who agree with them politically work there.
That's what a memory leak is: not properly releasing memory when it's no longer needed.
Are you being snarky?
FWIW, that stopped being true of pale moon with the Version 25 release. The author decided that a "moral stand" about "freedom of browser choice" was more important than keeping the compatibility that had made it (at least IMO) such an awesome alternative. It broke a ton of extensions, particularly privacy extensions that made it a dealbreaker for me.
It was a bummer. I was a huge fan of Palemoon, too.
WONTFIX tell user to update browser to latest mozillamonkey.
The real problem is that once all that memory is allocated, it never gets let go. It' perfectly fine to say that your copy of Firefox doesn't exceed 300 MB of memory usage in most cases, but if it does, only a restart can fix it. That may not be a "leak" in the traditional sense, which is why the geeks are upset about the meme, but the end result is the same. The problem is very much real, and has been for, oh... about 8 years!
AdBlock is greediest memory hog. After enabling that one extension, 10 minutes of browsing bloats memory usage to beyond 1GB and the browser slows down BIG time... to the point where it's practically unusable.
The issue appear to still be the JavaScript heap. Close all tabs and point your last open window to "about:blank". Then look at "about:memory". Almost all memory usage, probably several hundreds of MB, will be in the JS heap, not caches. I understand that a large part of the Firefox UI relies on JavaScript, but if you close all web pages, I would expect most memory, or at least a reasonable amount, to be released. If you're using an AdBlock-enabled version of Firefox which has allocated 1.5 GB of memory and then do the "about:blank" test, memory usage doesn't go down at all, despite the fact the browser isn't showing any live content.
The infamous "pausing" problems are also still present. They occur only when Firefox has allocated a huge amount of memory, and are probably related to either garbage collection or session state saves. Firefox really needs some sanity checks, or at least some tweaks so massive memory usage doesn't kill performance.
Part of the problem is likely that Firefox needs to maintain memory for the "undo tab" and "back" button. I think you can tweak that in about:config
Mozilla is loosing it. Fx gets more and more irrelevant between various UI changes and more bloat added. Usage is declining. And yet they try to reinvent themselves with such ideas. What for? Just make a decent browser and build developer tools into it like everybody else does. What is the point of such product? To have yet another browser/platform to build and test for?
Most of them still work afaik. They just lose some interface components.
As far as I know, key extensions have been forked by the volunteer group behind the browser here:
http://addons.palemoon.org/fir...
And reason wasn't a "moral stand" but reality that Pale Moon isn't going to implement Australis, as a result add-ons designed to work with Australis will have broken interface components.
So it needs to tell add-ons to use the old UI elements, and to do so, it needs to tell add-ons that's it's not Firefox.
And reason wasn't a "moral stand" but reality that Pale Moon isn't going to implement Australis, as a result add-ons designed to work with Australis will have broken interface components.
So it needs to tell add-ons to use the old UI elements, and to do so, it needs to tell add-ons that's it's not Firefox.
If that were the only change, I might have given you that one. But I think the real one that started PM bleeding off users (at least from the ragequits in the forum. Yay internet generation) is that the "firefox compatibility mode," such as it is, was turned off by default, had no option to turn it back on (and for the life of me, I couldn't find anywhere that told us the about:Config entry to do it manually), including in the FAQ about the issue, which also went on about "freedom of browser choice." Two weeks later that was finally remedied, though still not without editorializing. If it was *just* a technical issue, he seemed to be going out of his way to make it feel like it wasn't.
have a browser teared towards power-users again, like Opera 12 was. Srsly, why can't any other browser (including Op 15+) get stuff like tab-handling or simple usability features right?
You could be right. Personally I got more of an image of a "there's only a handful of us devs and we just want to get supported by add-on makers so we can continue making a browser that works even without Australis" vibe from Moonchild's messages on the board.
Same thing with the user agent of the browser. I always used user agent switcher so for me that was "just use UAS as needed" moment. I used firefox for a long time, so it's nothing new to me that browser I use needs to have its user agent changed to get a proper page instead of "this is an unsupported browser" placeholder. They even provided details on user agent used and user agent one needed to use to get "compatibility mode".
Same thing with the user agent of the browser. I always used user agent switcher so for me that was "just use UAS as needed" moment. I used firefox for a long time, so it's nothing new to me that browser I use needs to have its user agent changed to get a proper page instead of "this is an unsupported browser" placeholder. They even provided details on user agent used and user agent one needed to use to get "compatibility mode".
Yeah, so did I. It wasn't until I realized things were breaking that I finally, reluctantly, went back to Fx. The User Agent was more of a symptom than a real problem in itself, as I see it.
Strange. Nothing but APB broke for me on transition, and even that worked. It just lacked the UI icon.
All sites worked fine as long as I changed user agent to that of firefox.
Chrome and IE both do that.
Wealth is the gift that keeps on giving.
The ones that really went ape for me were "Self-Destructing Cookies" and Firebug: the former was rage-inducing, but the latter was a dealbreaker.
Some good news:
Name: Self-Destructing Cookies 0.45
Type: Extension
Issue: Does not function
Cause: Jetpack
Resolution: Fixed in 25.0.1
Firebug is not yet fixed.
Name: Firebug
Type: Extension
Issue: Partly functional
Cause: suspected chrome.manifest
Workaround: TBD
Resolution: TBD
Yeah, I saw that SDC was fixed. Without firebug, though, it's still a no-go. Some of my other "nice to have but not dealbreaker" extensions are still "TBD" on that list, too (like Epub Reader). Honestly, screwing with the extensions just killed any real interest I had in the project. If it wasn't for the extensions, I would have switched to Opera or Chromium back before Palemoon was even a thing.
Fact is though, that if you don't want to go the way of Australis, some extensions designed for Australis will not work for obvious reasons.
Blaming Pale Moon crew for that feels misdirected, especially when you consider just how often mainline FF breaks add-ons.
It's not a question of blame (other than the author's soapboxing I mentioned earlier), just pragmatism. I'm just too old to be playing stupid "browser loyalty" games. I've got things to do, and in this case, both Fx and Palemoon get in the way of doing them: the former with its shitty, barely-usable UX garbage, and the latter by not supporting the tools I need. Since I can "kinda sorta" get around the former with "Classic Theme Restorer" and the like, and the latter I can't get around at all, Palemoon simply isn't fit to task anymore.
I understand the point. To me, the UI is simply more important in this regard and it didn't break any add-ons that I need. So PM is still far better suited for the task for me.
I completely agree with you on your main point - if add-ons I find necessary were to break in PM, I would switch as well. I just seem to place more priority in UI than you, because to me, browser being a very important tool in both work and leisure, I just find myself too old and my time too precious to be spent fighting UI instead of having it work for me.
Who gives a shit? Flash video still is faster and more reliable than html5 video. When Flash 11.2 gets unmaintained on linux in 2017 it will be time to stop providing Flash video. If it works better then it is the technically superior solution.
That's evident on a PC without hardware H264 decoder at least.
Yeah, that's what the real bummer of the whole thing is. PM was a godsend before, for just that reason. I'd have been through the roof if this story had been about Palemoon instead of Mozilla making a dev-focused browser. :D