Vivaldi browser and open-source
-
@Chas4 said in Vivaldi browser and open-source:
Presto ... RAM usage was compared to Chromium, 175 tab and 25 Windows RAM usage still under 800MB.
To be fair, if you used the current Vivaldi browser to browse the pre-July 2013 web, you might get similar numbers. Content is richer now, and the rich parts are larger. While the design paradigm of Chromium does use more memory per tab than some other designs, the really large memory consumption comes from the content.
I also remember Opera Unite. I have a few old Opera wallpapers from Opera 9.5 era.
Content was even lighter in 2008.
Even operating systems have a mix of open-source and closed-source code code. Apple has a list of open-source here https://opensource.apple.com. Parts of some apps Microsoft has are open-source now https://opensource.microsoft.com.
Each company has to find its balance of open and closed source code.It's great to see this put into the larger context to gain some perspective. How many mixed source software shops host a forum where they let users post mods effecting their closed source assets? I'll assume there are others, the world is a very large place even limiting ourselves to this planet, but Vivaldi is the only one that comes to mind right now. Users still exhibit a great degree of control over even the 5% of the browser that is closed source, either through the customizations Vivaldi has built into the product, or through community-based code that we can insert into the application ourselves.
-
@Komposten said in Vivaldi browser and open-source:
There are the initial bookmarks which come with every Vivaldi install. And the default search engines (bar Google) could be seen as adds since they are also from sponsors.
@Catweazle said in Vivaldi browser and open-source:
@BoneTone , not banners with advertisings, Vivaldi use default search engines from sponsors and default bookmarks from others. All this you can delete, if you don't use them.
If default bookmarks & search engines qualify as ads, I don't know of a single browser that doesn't have ads upon initial install. Which kind of renders the point moot, it's nothing different and irrelevant to the topic. When people talk about ads in browsers, they're talking about things that are displayed alongside or on top of the content being accessed, for which Vivaldi has a built-in blocker.
-
@BoneTone gif are not much bigger today and a website that is 96% text should not need 2GB of RAM, Google has ignored the RAM usage for a long time others using Chromium have not. The RAM usage is also an issue for lower end devices, as 1 tab should not make a device completely unusable for anything else, I also remember Chrome from back around 2008 having the same issues. Also content was not lighter in some cases it was much heavier (Flash Player & Java plug in performance on macOS).
Open-source and closed source projects can push each other to get better with performance and features.
-
-
@Chas4 , perhaps those with a small PC should remember that Windows incorporates a function called ReadyBoost (in W Vista up to Win10), with which they can fill some system deficiencies with a pendrive.
-
@LeBaux said in Vivaldi browser and open-source:
TLDR: Money. Not that there is anything wrong with it.
But I find the argument about protecting the brand and being afraid of forks bit ironic since you rely on chromium. Without it, there is arguably no Vivaldi. This is de facto a fork of chrome. So you benefit from open source tremendously.
What this article is missing is how is Vivaldi giving back to opensource. I genuinely don't know. If there are lot of upstream contributions to chromium or other libraries, it would make this whole argument a lot more sound.
Not to mention you ¯_(ツ)_/¯ged security with, let us know. I know you might mean well.
With all respect, I think this blogpost missed the mark.Something closer to reality is:
We happily take chromium from Google. They are rich assholes and chrome sucks. We made it better, but we also need to eat. There are ads, but we made them easy to remove. We made it impossible to fork Vivaldi further. We think it is fair. We still keep most of the stuff open source, because you know, we have to (pesky licences).Lastly, I still love Vivaldi and respect the business model. I just don't love this blogpost.
seeing as google are rich assholes as you put it then is there any particular reason why google have not created such a web app as vivaldi themselves...?...
i have asked this before on this forum and still await a response...i just find it odd that google have not taken it upon themselves to make chrome for example as customisable..
-
@Priest72 , anyway quite suspicious than Chrome, despite being much less functional than Vivaldi it is much more resource hungry.
I'd already be interested in what hidden codes other than the Chromium base cause it. -
@Priest72 said in Vivaldi browser and open-source:
i just find it odd that google have not taken it upon themselves to make chrome for example as customisable
Google have not done this because there is no threat to their market share. They are too big.
Were chrome smaller and suffering as a result of browsers whose main selling point was customisation, I can guarantee you that more customisation would immediately appear in chrome.
-
@lebaux: I think it's important to point out here that Google has basically successfully positioned Chromium to be the de facto web standard, which means that the only way to make sure a new browser doesn't break web pages is to fork Chromium. Because of this, forking chromium isn't just happily taking from Google, it is following the standard, whether we like it or not. And in addition, anytime someone chooses to fork Chromium, they are reinforcing it as a de-facto standard, giving Google more power.
Also, we do upstream some changes to chromium. Not as much as we'd like to, because their process is rather involved and time consuming and we do not always agree on how things should be done, resulting in some of our attempted contributions being rejected.
As for money, well... yes we need to eat, but we also believe that whatever income we made should be done so in an ethically sound way. That in part means being transparent about our stance on different topics, which is the actual point of this blog post. -
@Catweazle Would not do much good in the USB 2 ports
-
@Chas4 , not much, but an improvement of 10-15% can be decisive, in some cases the performance may even be greater.
-
Great! vivaldi-source_3.1.1929 code is available in https://vivaldi.com/source/
-
@julien_picalausa: Will I find there how to fully access e.g. the tabs api - including the changes you (<- i.e. Vivaldi) needed to make the tab stacks possible?
Especially to which stack a tab belongs, where it is positioned in the stack etc., and can I access the methods / properties in an Extension?
If not, it is as good as if there is no API at all. -
@QuHno Stack & tiling stuff is accessible via the chrome.tabs.Tab.extData API, and can be used by mods and extensions. However, the stack, or "group" as it is referred to, is just a random ID and I don't know of any way to do anything further with it.
-
Personally i would love to see vivaldi going open source, but the reasons on the post sounds reasonable to me, still as a modder i would love to have non obfuscate code of some parts as it'll make easier to add or extend features or even find and help to fix bugs.
Bugs that i'll want to help fixing cause annoy me badly are:
-
The ui sluggishness, the worse time for me is when opening multiple tabs at once (a bookmark folder) the whole browser stop responding to user input during some of the loading time, It's so slow that makes me question if the UI is running on the same thread as the content, the rest of the time is fast enough, but loading content shouldn't block the ui, i know this has been reported and i'm sure it's hard to fix or else it would have been addressed already, yet when i recommend vivaldi the usual complain is the sluggy ui.
-
The User Interface zoom when at a different value than 100% makes some extensions panels have an incorrect size making them impossible to use, a good example is umatrix, but any extensions that doesn't render the panel instantly can have this problem, I managed to workaround some of it using css over the extensions panels but it has to be done per extension basis and was not reliable so i finally decided to keep the scale at 100% till i have time to test if i can trigger a recalculation of the extension panel size using js, or the issue is fixed.
I hope that didn't sound aggressive, what i thought of vivaldi is that, albeit there's still work to be done, vivaldi is close to be perfect for me, and even with those annoyances i still prefer it over any of the alternatives.
As for mods that will be easier to do with non obfuscated source:
-
Presenting tab groups as tree tabs on the tab bar (when positioned on the side) pretty much like how the window view shows them. i thing this feature will make easier to work with tab groups.
As i said It's like what the window panel does so with non-obfuscated code we could just reuse parts of the code to port that feature to the tab bar. -
Clickable elements of URL but "extended", in fact i did some work on this just before the native feature was added on the snapshots (feature that is now temporary disabled) the idea was to do something similar to the location bar enhancer extension for pre-quantum firefox, for those who doesn't know it here's an article, i hope this doesn't count as spam:
https://www.ghacks.net/2012/01/25/ui-enhancer-improves-firefoxs-address-bar-functionality/
The advantages of that extensions are:
-
- clicking on any element of the url, not limited to left click but also ctrl/alt and middle mouse button so you can open the cropped url in a new tab or window, and the ability to click the last element to refresh the page.
-
- Scrolling on an element let you go to visited siblings or on pages that have a patter similar to "?page=2" it will let you increase/decrease the number when scrolling over the last element. (i rarely used the scroll feature but maybe someone like it).
-
- easy to customize with css and let you highlight certain url parts like queries (there's a masking mod on the forums that can do something similar).
https://forum.vivaldi.net/topic/37609/mask-for-the-address-bar?lang=en-US&page=1
- easy to customize with css and let you highlight certain url parts like queries (there's a masking mod on the forums that can do something similar).
I've decided to wait and see how the native feature will look like before developing this further as it may be sufficient, currently it looks like this but has almost none of the extra features.
https://i.imgur.com/fP7yB1X.pngIf someone want to work on any of those features feel free to do it, I've limited time right now so i don't know how much time it will take me to finish any of those mods.
And for the vivaldi team, thank you for crating such a great browser!
-
-
@julien_picalausa said:
@quhno: Technically, you can figure out our internal API by reading the published code. The .json files under /extensions/schema describe the full API. If you can make sense of C++, then the implementation is under /extensions/api. In general, the json files are supposed to use descriptions to explain what everything is doing. If there is something there that you can't understand, then it's probably something we should fix.
So that others don't have to download and sift through several Gb of files, I have put all of the API documentation here: https://lonmcgregor.github.io/VivaldiModdersAPI/OfficialApi/accessKeys.html
I notice that some of types lack descriptions, but they are mostly self explanatory. I think I had done this previously, but knowing where to look in the official documents is helpful.
-
@julien_picalausa
This is what I was talking about when I said:I haven’t had a look at it yet, but it’s from a big part coz it’s big (well this can’t be solved) and gets outdated with every release. Why not publish it (still talking about the back-end code) as a git (or whatever VCS) repo?
As written here (
"size": 2087690
), it’s only ~2 GB even though it’s from 1.0, while the per-version packs are now nearly 1.5 GB each. -
@LonM That's really impressive from your part.
But those are not Open APIs, nor are they accessible through extensions.Vivaldi need those.
-
@Cqoicebordel The API is open enough for mods. But you are right - I would love if extensions could access them.
Two of my more useful mods, the sessions panel & sessions autosave could easily be implemented through an extensions API instead of a mod, and it would be less prone to bugs.
-
@julien_picalausa said in Vivaldi browser and open-source:
And in addition, anytime someone chooses to fork Chromium, they are reinforcing it as a de-facto standard, giving Google more power.
... and Adobe, and Microsoft and all the others which contribute to the standard, making it a strong standard.
The only thing it is missing to become a real Standard (with a capital S) is a Standard Body that can set what must be in and what is free to change.
As it is now, when "The Chromium Authors" decide to deprecate things, functionality in which e.g. extension developers have literally put in 1000s of hours are declared to NULL, or additional portions of the web cannot be displayed properly anymore, or not at all. This way we lose more and more information that might still be valuable. That makes me sad.
Not all old information is there to be forgotten.