Why is Ubuntu Web Browser in the repository?

33 risposte [Ultimo contenuto]
GrevenGull
Offline
Iscritto: 12/18/2017

The package that is simply called "Browser" in the "Add/Remove Applications".

It seems kind of buggy, and has google as start page.

Why is this in the repo?

GrevenGull
Offline
Iscritto: 12/18/2017

Also Rekonq, it crashed immediately

jxself
Offline
Iscritto: 09/13/2010

You've asked other questions of why 'Program X' is in the Trisquel repos. So I guess I should counter with: Why can't it be? If you want something to be excluded please make your case based on https://www.gnu.org/distros/free-system-distribution-guidelines.html

Those are the FSF's criteria for a distro to meet so if you want a program excluded please show how it doesn't meet those criteria in some way. Then we can have better discussions. :)

GrevenGull
Offline
Iscritto: 12/18/2017

> Why can't it be?

My reason was in the OP. I said is has Google as start page in is very buggy.

> If you want something to be excluded please make your case based on https://www.gnu.org/distros/free-system-distribution-guidelines.html

Good tip! Thanks.

> please show how it doesn't meet those criteria in some way. Then we can have better discussions. :)

Again I stated my reason in the OP.

nadebula.1984
Offline
Iscritto: 05/01/2018

There is no ethical issue as long as the package is free (as in freedom).

GrevenGull
Offline
Iscritto: 12/18/2017

What about marketing issue?

gd_scania
Offline
Iscritto: 09/13/2017

Then you will need to get whom Trisquel is marketed to. Just ourselves, but also every people that they are worried against "Big Brothers" thanks?

quantumgravity
Offline
Iscritto: 04/22/2013

If a program was immediately excluded from the repo just because some user has some problem with it,
then most likely there wouldn't be any program left.

How about just not using it?

GrevenGull
Offline
Iscritto: 12/18/2017

Yeah I understand.

My question should have been directed to FSF and why they recommend Trisquel.

Not directed to the forum of Trisquel as to why the package is included.

But I really feel like Trisquel has a goal of being totally free, hence my questioning of the marketing of this particular program.

Anyway I find your tone rather aggressive and I believe you misinterpreted my OP. I am asking why this program is in the repo because of reason X. I have not demanded immediate exclusion.

Similarily:

If a forum topic was immediately excluded from conversation just because some user has some problem with it, then most likely there wouldn't be any topics left.

How about just not responding?

quantumgravity
Offline
Iscritto: 04/22/2013

"Anyway I find your tone rather aggressive and I believe you misinterpreted my OP. I am asking why this program is in the repo because of reason X. I have not demanded immediate exclusion."

Well, you asked a question and I gave you an answer.

"How about just not responding?"

Then why would you ask a question in the first place? Obviously you were not really seeking information but just wanted people to agree with you.

GrevenGull
Offline
Iscritto: 12/18/2017

I love you too :)

jxself
Offline
Iscritto: 09/13/2010

"My question should have been directed to FSF and why they recommend Trisquel."

Because Trisquel complies with the FSF Free System Distribution Guidelines (FSDG)
https://www.gnu.org/distros/free-system-distribution-guidelines.html

"I said is has Google as start page in is very buggy."

Please show which parts of the FSDG those things violate:
https://www.gnu.org/distros/free-system-distribution-guidelines.html

GrevenGull
Offline
Iscritto: 12/18/2017

"Because Trisquel complies with the FSF Free System Distribution Guidelines (FSDG)"

So my question should rather have been aimed towards why FSF doesn't regard marketing as an issue. Gotcha.

""I said is has Google as start page in is very buggy."

Please show which parts of the FSDG those things violate:"

My bad. I was sure marketing was stated as an issue (I believe I have read about it somewhere).

chaosmonk

I am a member!

I am a translator!

Offline
Iscritto: 07/07/2017

> Please show which parts of the FSDG those things violate:
> https://www.gnu.org/distros/free-system-distribution-guidelines.html

Bugginess certainly is not a freedom issue. However, the JavaScript in google.com is proprietary. If a browser uses Google as its start page, then even if the browser itself is free it will install and run proprietary software when it is first launched. Perhaps an argument can be made that making this the default behavior violates

"A free system distribution must not steer users towards obtaining any nonfree information for practical use"

calher

I am a member!

Offline
Iscritto: 06/19/2015

Trisquel 7 has decent Hebrew support, but once annoyance is that the
titlebar text is still anchored to the left, rather than the right.
Does Trisquel 8 fix this?

chaosmonk

I am a member!

I am a translator!

Offline
Iscritto: 07/07/2017

> Trisquel 7 has decent Hebrew support, but once annoyance is that the
> titlebar text is still anchored to the left, rather than the right.
> Does Trisquel 8 fix this?

You probably intended for this to be a new thread rather than a comment here. Probably a mailing list bug. I suggest trying again so that people see your question.

calher

I am a member!

Offline
Iscritto: 06/19/2015

Trisquel 7 has decent Hebrew support, but once annoyance is that the
titlebar text is still anchored to the left, rather than the right.
Does Trisquel 8 fix this?

SuperTramp83

I am a translator!

Offline
Iscritto: 10/31/2014

>However, the JavaScript in google.com is proprietary. If a browser uses Google as its start page, then even if the browser itself is free it will install and run proprietary software when it is first launched.

^ this

molto bene.jpg
onpon4
Offline
Iscritto: 05/30/2012

If that's a problem, and I tend to agree that it is, then not sending people to Google is not the solution. After all, almost the entirety of the Web has proprietary JavaScript on it, even if it's optional, and every modern Web browser executes that JavaScript code automatically. Let's take a common sense look at this: do you really think that people are going on the Web to look at Web pages with no proprietary JavaScript on them? No Google, no message boards, no YouTube, no news? No. That's an absurd expectation. People are going to websites they already know about that send them nonfree JavaScript code regardless of what websites the browser points them to. If anything, some users are probably deleting all bookmarks set up by default, changing the homepage to their preferred search engine, and adding bookmarks to the sites they know they frequent.

If we actually want to solve this problem, there's only one good solution: turn off JavaScript support in all supplied browsers (and other programs that read Web pages downloaded across the Internet). Give the user a browser that doesn't interpret JavaScript (by default; the option can be hidden in about:config or in the options menu with a warning), and explain to them that if a website doesn't work, it's because it doesn't support a libre Web browser. I would actually be very much in favor of this, because GNU FSDG is supposed to present an ideal operating system, and that isn't actually the case. Any new Trisquel user is running proprietary software every day in Abrowser, and I'd bet some of them don't even know it.

chaosmonk

I am a member!

I am a translator!

Offline
Iscritto: 07/07/2017

I agree that the problem with JavaScript is a much bigger issue than the start page of a broken web browser called webbrowser-app buried in the Trisquel repo.

> If we actually want to solve this problem, there's only one good solution:
> turn off JavaScript support in all supplied browsers (and other programs
> that read Web pages downloaded across the Internet). Give the user a browser
> that doesn't interpret JavaScript (by default; the option can be hidden in
> about:config or in the options menu with a warning), and explain to them
> that if a website doesn't work, it's because it doesn't support a libre Web
> browser.

Exactly. If JS is simply disabled, users will blame the browser when pages don't work and install Chrome. They need to be educated as they go.

However, not allowing JS at all might push some users too far, and if every browser in the repo is like that they'll switch to another distro or install Chrome. Perhaps a compromise solution could resemble NoScript, but be integrated into the browser and explains itself better.

Sometimes Abrowser warns "Will you allow this site to use your HTML5 canvas image data? This may be used to uniquely identify your computer." It could similarly warn about JavaScript, and clarfify that allowing the JS will install and run software on your computer. Like NoScript, it should allow users to whitelist on a case-by-case basis, so that even if they insist on allowing JS from nonsense.com they don't have to the JS from googleapis.com, googlesyndication.com, and gstatic.com that would otherwise be installed on the same page. (Unfortunately, disallowing gstatic usually breaks the page.) Of course, we should not encourage users to allow proprietary JS of any kind, but explaining the risks and giving users control over what they specifically allow might scare away fewer people than saying "half the web doesn't work if you use our browser, but it's not our fault."

In general, users should be notified whenever their privacy settings block something. I recently help a friend of mine switch from Windows to Trisquel. She is still getting used to it and is unwilling to give up Facebook for now. She found that when she tries to paste text into Facebook's in-browser chat client the client would close. Since Chrome on Windows did not have this problem, it appeared to be a problem with Abrowser. I looked into it and found that Facebook chat does not work properly with this setting[1] set to False. When I explained what the setting does and why it is set to False by default, she chose to instead use Facebook's mobile site, which does not require JavaScript. Ideally, the browser would have notified her that Facebook was trying to access information about what she copy/pastes so that she would understand what the problem was.

The start page of webbrowser-app is indeed a tiny issue by comparison. However, there is a difference between not stopping users from using non-free JS and directing them to it. Suppose I wanted to use webbrowser-app (I do not) and suppose that I could get it to launch without crashing (I cannot). Proprietary JavaScript will be installed and executed before I have a chance to change the start page. The user can't use this program and completely avoid non-free software, even if they have no intention of using non-free JS on other sites. Abrowser is a little better. You do not ecounter DuckDuckGo's proprietary JS until you actually search for something, and you have an opportunity to disable JS or change the search engine before that.

[1] https://developer.mozilla.org/en-US/docs/Mozilla/Preferences/Preference_reference/dom.event.clipboardevents.enabled

Magic Banana

I am a member!

I am a translator!

Offline
Iscritto: 07/24/2010

The first time Abrowser is launched (after every update, I believe), the displayed page proposes, among privacy settings, to disable JavaScript or to enable LibreJS, whose icon informs the user whenever proprietary JavaScript is blocked and allows to "add/remove pages's domain to/from whitelist" (and "to complain to site owner" too!). Neither option is enabled by default. The LibreJS one could. It warns about proprietary JavaScript and gives control, as you want.

onpon4
Offline
Iscritto: 05/30/2012

This may be an unpopular opinion, but I think if you're using a Web browser as a proprietary JavaScript interpreter (as most people do), it's really not much better than a proprietary Web browser. After all, how much better is a libre virtual machine running Windows than a proprietary virtual machine running Windows?

And furthermore, I think that sweeping issues like this under the rug out of fear of driving users away is counterproductive given the mission of a distro like this. The whole point of Trisquel is that no proprietary software is run unless you go out of your way to install that proprietary software yourself. But if Trisquel runs proprietary software any time you navigate to a Web page you heard about on TV, that objective has not been accomplished.

By the way, LibreJS is a non-solution because there's also the issue that there's no mechanism to modify the JavaScript code before executing it, or choose when to execute it; and LibreJS has to trust scripts to actually tell the truth about whether or not source code is available.

quantumgravity
Offline
Iscritto: 04/22/2013

"After all, how much better is a libre virtual machine running Windows than a proprietary virtual machine running Windows?"

That's a false comparison. If I use the virtual windows machine for doing my computing just like with an ordinary pc, then certainy this isn't better than just using windows on a normal machine. It can do everything with your data and with the system you're using on a daily basis.

Proprietary javascript is running in a _sandboxed_ environment. It's very restricted in what it can do to the actual system you're doing your computing with.
It's actually very restricted in what it can do in the first place.

A proprietary program is not restricted at all. It has full power over your computer.
I can't stress enough how much worse it is to use a proprietary webbrowser compared to using a free webbrowser and running proprietary javascript.
It's on a different level.

onpon4
Offline
Iscritto: 05/30/2012

No, if you run Windows in a virtual machine, it's entirely sandboxed from the main system. Also, it's a mistake to just assume that any sandboxed program can't do anything malicious; JavaScript can, and often does, assist in identifying you, or interfere with browser functions, or implement digital handcuffs.

That's not the point, though. It's proprietary software, and Trisquel is supposed to be a system without proprietary software. It's also proprietary software that we know without a doubt people are indeed running, because we know that: 1. most people aren't aware of the JavaScript trap, and 2. nearly all websites have proprietary JavaScript on them, including all the very popular ones. And finally, it's proprietary software we could easily protect users against: just give them browsers with JavaScript disabled by default and any attempt to enable JavaScript either in the advanced settings or behind a warning message.

Of course that's going to cause some people to simply abandon Trisquel and use Mint or Ubuntu or something, but they're also going to do that because their wireless card doesn't work, or because Netflix doesn't work.

quantumgravity
Offline
Iscritto: 04/22/2013

"No, if you run Windows in a virtual machine, it's entirely sandboxed from the main system."

From the main system, yes. But the part that is not sandboxed is a whole operating system you might use just like your normal computer. Over this "virtual computer"
windows has the full power.

The sandboxed thing inside a webbrowser is not a whole operating system that you might use just like a normal computer.
Plus: How much harm it can do can be controlled by the (free software) browser running the javascript code.

If you run a propriatery webbrowser on your system, there is no such control and your whole system is just fucked.

To be clear: I'm arguing against the statement that running a proprietary browser and running a free browser executing non free javascript is equally bad.
I believe it is not.
I'm not arguing that the latter one is not bad at all.

onpon4
Offline
Iscritto: 05/30/2012

I didn't say "equally bad". I said "not much better".

chaosmonk

I am a member!

I am a translator!

Offline
Iscritto: 07/07/2017

> This may be an unpopular opinion, but I think if you're using a Web browser
> as a proprietary JavaScript interpreter (as most people do), it's really not
> much better than a proprietary Web browser.

I agree. Between Google Drive, Gmail, Facebook, and YouTube it is possible for the average user to do almost all of their computing with proprietary software desipte using an otherwise free operating system.

> The whole point of Trisquel is that no proprietary
> software is run unless you go out of your way to install that proprietary
> software yourself. But if Trisquel runs proprietary software any time you
> navigate to a Web page you heard about on TV, that objective has not been
> accomplished.

You make a good point. I guess that "if Abrowser makes it too difficult to install proprietary JavaScript then users will switch to Chrome" is not fundamentally different from "if Ubuntu makes it too difficult to install proprietary software then users will switch to Windows."

My only concern about your suggestion to hide a global JavaScript setting in about:config is that it won't be that hard to find, and if the user enables it in order to get one site to work, Abrowser will start installing other non-free JavaScript that the user might not have chosen to install otherwise. This is why I prefer the ability to whitelist JavaScript from certain domains over an option to enable JavaScript globally. Maybe it would be better to go the other drection and build all browsers in the Trisquel repo without JavaScript support. Trisquel would probably lose many of its users, but as a libre distro Trisquel already sacrifices populatity for freedom.

> By the way, LibreJS is a non-solution because there's also the issue that
> there's no mechanism to modify the JavaScript code before executing it, or
> choose when to execute it;

Yes, the mechanism by with JavaScript is installed and run is a separate problem from the license of most JavaScript. It sort of reminds me of tivoization.

jxself
Offline
Iscritto: 09/13/2010

Okay; that's a point although it wouldn't require removing it. This can be solved by changing the default web page to somewhere else (or to perhaps nowhere.) A bug can be filed so that the appropriate changes could be made.

loldier
Offline
Iscritto: 02/17/2016

Webbrowser-app was part of Unity and Ubuntu Touch, Canonical's convergence kit. I'm on Ubuntu 18.04 right now, and I can't find it. It's not in the repos anymore, at least not with that name as a standalone package. I can't find it under system/default applications either. Then again, it might not be there to choose from as I installed Ubuntu in minimal mode.

quantumgravity
Offline
Iscritto: 04/22/2013

I believe we're talking here about "webbrowser", formerly known as "epiphany", the own webbrowser of the gnome project.

loldier
Offline
Iscritto: 02/17/2016

Oh, I see, you are most likely right. It's under 'epiphany-browser' in the repo. When installed it's called 'Web'. It shows a blank page when first started. Web's icon resembles that of Safari.

epiphany-browser.png web_icon.png gnome_web.png
gd_scania
Offline
Iscritto: 09/13/2017

But also, QupZilla? It's Qt-written that's now Falkon with being merging with Rekonq and Konqueror for the KDE project.

loldier
Offline
Iscritto: 02/17/2016

No, it's not the same as Gnome Web Browser (because epiphany-browser is another package). It's a binary called webbrowser-app, and it has the same Safari like icon.

I tried it on my Trisquel 8. It crashes every time before the window is drawn. The error in terminal is long and something about "module "Ubuntu.Content" is not installed", "Could not determine application identifier. HUD will not work properly.", "module "Ubuntu.DownloadManager" is not installed". I removed it and along with it about 495 MB freed, among them linux headers and stuff. ;-)

It definitely is not on Ubuntu 18.04 anymore.

browser.png webbbrowser_app.png fail.png
gd_scania
Offline
Iscritto: 09/13/2017

Do you stand your simply "Ubuntu web browser" for this, https://launchpad.net/webbrowser-app? I need to read more for what are the issues you mean. My laptop is still running Trisquel at home but I'm now outside.