vivaldi browser

[Flash]

Thanks, it might be a day or two before I can get to that.

... you really can't expect to have every single support package in existence in your distro... I don't think even eg Debian does that! ...that said, DBus is pretty basic, as I understand it...

The Vivaldi .pet is 61 MB. How much larger would it be if it included the more obscure libs Vivaldi needs to run? I think what I'm talking about is an application in a directory, or perhaps a Rox app. There are many libs that every Puppy can be expected to have. Those don't need to be included -- unless they're tiny, in which case might as well include them too.

[starhawk]

...and what, pray tell, happens if Puppy already has the libs you need? For example, Caro Vanguard has libdbus (...which, BTW, should be in /usr/lib -- check there before you download!).

To answer my own question -- what happens is file clobber that breaks stuff. It's far better to have the user go off libhunting for a little bit, than to have applications that break other parts of the OS when you install them!

Applications in a directory are the basis for how Linux works, really -- and dotPET packages generally do not (if ever at all) contain ROXApps. Remember that a dotPET is a tarball (*.tar.gz archive) that has an MD5sum plunked into it such that PetGet can tell when the package has corrupted... if you were to convert a dotPET to a regular tarball and open it in eg PupZip, you'd find directories like /usr and /etc and such in there. The files within that directory hierarchy are plunked down into the appropriate places in the filesystem. All PetGet really does is verify the package, unpack it, read the directory tree, and plunk files where they go. It's not *that* complicated. Certainly not rocket science!

[Flash]

I thought an application-in-a-directory included in the directory everything the app needs to run. Almost like a binary. That way, the app would know right where a needed library was. No other app in the computer would know that particular library was in that particular directory, so there would be no collision. Am I wrong about how that works?

[starhawk]

The binary is the executable part of the code. Everything else included supports it. However -- the person or people creating the package (not just a binary!) make certain assumptions about the filesystem content outside of that package -- the environment into which the package will be installed.

Generally you don't package all required supporting libs in with a binary. You /can/ do that, but then you end up with a boatload of duplicates. /usr/lib exists for a reason, you know

Let's look at Chromium, my web browser of choice. There is a shell script at /usr/bin/chromium, but the actual binary is /usr/lib/chromium/chromium (oddly enough, with a lib icon, dunno why, it's not a *.so). There are supporting libs in /usr/lib/chromium, but those are not the only required libs by any means. Pepperflash is included with this particular Chromium (one of peebee's SFSs) and that's hanging around somewhere. Strictly speaking it's 'recommended' rather than 'required' in that Chromium will run without Pepperflash -- but then you don't get flash content, only HTML5 (etc) and that's kind of a downer.

However, if I install that SFS into Carolina Vanguard, open a terminal and type chromium [ENTER] -- I will get an error message, about a missing symlink having to do with libevent, which (as I understand it) is a thing that lets callbacks occur when something happens to a file. (Remember that everything is a file in Linux, and that libraries --'libs'-- add capabilities to the binaries within a filesystem. libjpeg, for example, allows binaries to, well, do stuff with JPEG files, for example.) You can find more info on libevent at its website.

So, Chromium's developers, and the people who packaged it up for us (peebee repackages Slackware TXZs from a guy calls himself 'alienbob', so there's a couple folks involved there), made an assumption that libevent would be present in some form in whatever distro it was plunked into. In most cases that's probably a safe assumption. However, in Carolina Vanguard, it isn't, and libevent becomes a dependency ('dep'). In this case it's the only one. You haven't been so lucky with Vivaldi, apparently, but that does happen. In particular, it happens a lot when you're pulling in stuff from other distros. So, for example, if I install an application from say Slackware 14.1 into TahrPup, because TahrPup does Ubuntu better than Slackware, I'll quite likely have a fat pile of deps (mostly libs) to hunt down. What is a safe assumption about Slackware 14.1's filesystem contents, is not necessarily so for Ubuntu Trusty Tahr.

Make sense?

[Aztk]

Excuse my bad English. I used Google Translate.

I installed the pet Vivaldi, libraries were missing and what I did was I copied the libraries that come in my firefox 40.0.2 because they are the same that appeared me in terminal. Now, in trying to run it, this appears in terminal. I'm using X-Precise.

http://postimg.org/image/gsvf4y4yz/

[Semme]

Aztk, you symlinked that lib from where to where?

[Aztk]

Aztk, you symlinked that lib from where to where?

No, I literally copy the libs files from /usr/lib/firefox/, and I see than use the same libs. I don't know if I can do this. O just create a symlinked... If so, how do I create a symlinked in xfce (thunar)?

EDIT:

Well, I unzipped the pet who kindly shared with us peebee. I copied the libraries and I still make that mistake.

Failed to load NSS libraries.

Or, I need install chrome or chromium necessary?

[Semme]

Aztk, run this and post the output:

Code: Select all

ldd /opt/vivaldi-snapshot/vivaldi-bin

Mind you it's (L)dd.

[Aztk]

I use that command in terminal:

# ldd /opt/vivaldi-snapshot/vivaldi-bin
linux-gate.so.1 => (0xffffe000)
librt.so.1 => /lib/librt.so.1 (0xb1f08000)
libdl.so.2 => /lib/libdl.so.2 (0xb1f03000)
libgobject-2.0.so.0 => /usr/lib/libgobject-2.0.so.0 (0xb1eb3000)
libglib-2.0.so.0 => /lib/libglib-2.0.so.0 (0xb1dba000)
libnss3.so => not found
libnssutil3.so => not found
libsmime3.so => not found
libplc4.so => not found
libnspr4.so => not found
libgconf-2.so.4 => /usr/lib/libgconf-2.so.4 (0xb1d7c000)
libgio-2.0.so.0 => /usr/lib/libgio-2.0.so.0 (0xb1c1c000)
libfontconfig.so.1 => /usr/lib/libfontconfig.so.1 (0xb1be8000)
libfreetype.so.6 => /usr/lib/libfreetype.so.6 (0xb1b4d000)
libpangocairo-1.0.so.0 => /usr/lib/libpangocairo-1.0.so.0 (0xb1b40000)
libcairo.so.2 => /usr/lib/libcairo.so.2 (0xb1a75000)
libpango-1.0.so.0 => /usr/lib/libpango-1.0.so.0 (0xb1a2b000)
libX11.so.6 => /usr/lib/libX11.so.6 (0xb18f7000)
libXi.so.6 => /usr/lib/libXi.so.6 (0xb18e6000)
libXcursor.so.1 => /usr/lib/libXcursor.so.1 (0xb18db000)
libXext.so.6 => /usr/lib/libXext.so.6 (0xb18c9000)
libXfixes.so.3 => /usr/lib/libXfixes.so.3 (0xb18c3000)
libXrender.so.1 => /usr/lib/libXrender.so.1 (0xb18b9000)
libXcomposite.so.1 => /usr/lib/libXcomposite.so.1 (0xb18b5000)
libasound.so.2 => /usr/lib/libasound.so.2 (0xb17c2000)
libm.so.6 => /lib/libm.so.6 (0xb1796000)
libXdamage.so.1 => /usr/lib/libXdamage.so.1 (0xb1792000)
libXtst.so.6 => /usr/lib/libXtst.so.6 (0xb178b000)
libXrandr.so.2 => /usr/lib/libXrandr.so.2 (0xb1782000)
libexpat.so.1 => /lib/libexpat.so.1 (0xb1758000)
libudev.so.0 => /lib/libudev.so.0 (0xb1746000)
libcups.so.2 => /usr/lib/libcups.so.2 (0xb1708000)
libpthread.so.0 => /lib/libpthread.so.0 (0xb16ed000)
libdbus-1.so.3 => /lib/libdbus-1.so.3 (0xb16a4000)
libcap.so.2 => /lib/libcap.so.2 (0xb169e000)
libgtk-x11-2.0.so.0 => /usr/lib/libgtk-x11-2.0.so.0 (0xb1236000)
libgdk-x11-2.0.so.0 => /usr/lib/libgdk-x11-2.0.so.0 (0xb1186000)
libgdk_pixbuf-2.0.so.0 => /usr/lib/libgdk_pixbuf-2.0.so.0 (0xb1165000)
libXss.so.1 => /usr/lib/libXss.so.1 (0xb1161000)
libstdc++.so.6 => /usr/lib/libstdc++.so.6 (0xb107c000)
libgcc_s.so.1 => /lib/libgcc_s.so.1 (0xb105e000)
libc.so.6 => /lib/libc.so.6 (0xb0eb4000)
/lib/ld-linux.so.2 (0xb7795000)
libffi.so.6 => /usr/lib/libffi.so.6 (0xb0ead000)
libpcre.so.3 => /lib/libpcre.so.3 (0xb0e71000)
libgmodule-2.0.so.0 => /usr/lib/libgmodule-2.0.so.0 (0xb0e6c000)
libdbus-glib-1.so.2 => /usr/lib/libdbus-glib-1.so.2 (0xb0e4a000)
libz.so.1 => /lib/libz.so.1 (0xb0e34000)
libresolv.so.2 => /lib/libresolv.so.2 (0xb0e1b000)
libpangoft2-1.0.so.0 => /usr/lib/libpangoft2-1.0.so.0 (0xb0def000)
libpixman-1.so.0 => /usr/lib/libpixman-1.so.0 (0xb0d57000)
libpng12.so.0 => /lib/libpng12.so.0 (0xb0d2d000)
libxcb-shm.so.0 => /usr/lib/libxcb-shm.so.0 (0xb0d29000)
libxcb-render.so.0 => /usr/lib/libxcb-render.so.0 (0xb0d1e000)
libxcb.so.1 => /usr/lib/libxcb.so.1 (0xb0cfd000)
libssl.so.1.0.0 => /lib/libssl.so.1.0.0 (0xb0ca6000)
libcrypto.so.1.0.0 => /lib/libcrypto.so.1.0.0 (0xb0afb000)
libcrypt.so.1 => /lib/libcrypt.so.1 (0xb0ac9000)
libatk-1.0.so.0 => /usr/lib/libatk-1.0.so.0 (0xb0aa9000)
libXinerama.so.1 => /usr/lib/libXinerama.so.1 (0xb0aa5000)
libXau.so.6 => /usr/lib/libXau.so.6 (0xb0aa1000)
libXdmcp.so.6 => /usr/lib/libXdmcp.so.6 (0xb0a9a000)

But, I have all...

[Semme]

Unpack, run the enclosed script, then tell me what I wanna hear!

[Aztk]

It's alive!!!! Thank you for your time Semme.


But, I have a little problem. It's not a very thing, but after opening well Vivaldi; Firefox and Palemoon show me this at startup:

Well, I do not want to spam here, so please instruct on where to open a new post for a possible solution . However, it does not affect me at all . Navigate with ease in all three browsers. thanks again Semme.

[Semme]

I suppose you put those libs there because my Retro-Precise is quite different. Vivaldi runs off the bat.

What may be different about X-Precise should be answered by the maintainer or someone who runs a copy.

[Semme]

Mind you, the moment you become passive, people forget and move on.

[OscarTalks]

So many Chromium derivative browsers to choose from. Under the hood they are all fairly similar. With Vivaldi the emphasis seems to be on tarting up the UI among other things. Unlike Opera there does not seem to be any browser updater included and personally I prefer this (especially if the updater causes problems which happens sometimes).

Here is Vivaldi 1.0.403.24 with the usual tweaks and running in Wheezy so will also work in Precise and 14.0 Slackos.

Not sure if it will work with Netflix. there is the libwidevinecdmadapter.so but the other libwidevinecdm.so is a symlink which seems to assume google chrome also installed so I am a bit puzzled by all that.

Pepper Flash is now working. I have added version 21.0.0.197

OK so I have also copied in the libwidevinecdm.so lib from chrome and the ffmpeg.so lib is there in a sub-dir so perhaps someone can test if Netflix works like that?
https://yadi.sk/d/euoIiDRDqcFnK

[Wognath]

Hello, OscarTalks,
I just tested Netflix in vivaldi-stable and it did not work with the included libffmpeg.so. It does work with the version linked here. Libwidevinecdm.so is proprietary and must be borrowed from chrome, hence the link to opt/google/chrome. I just installed the lib into /opt/vivaldi (replacing the link) from my Chrome sfs, which does not need to be loaded.

Vivaldi has always worked out of the box in Fatdog, including today's stable release. I just extract the deb and make an sfs. If anyone wants, I can list the steps. (Fatdog has deb2txz, but I have not tried it.) Vivaldi must be run as spot, although you probably know how to get around that.

PepperFlash, not included, can be placed in any of the directories listed around line 85 in the script /opt/vivaldi/vivaldi.

The first "stable" came out today. I expect to stick with Vivaldi for a while, having jumped ship from Opera.

[OscarTalks]

Hello Wognath,

I have upgraded the packages to the new stable 1.0.435.40 version.

Also I have substituted the libffmpeg.so with the i686 version downloaded from the site you linked. The shipped libffmpeg.so is in a sub-directory and the main browser needs to find it there rather than in the main directory. Not sure if widevine needs it in the main directory so I put it there and also symlinked it into the sub-dir.

My packages have the run-as-root modifications. Do we know if running as root prevents Netflix from working?
https://yadi.sk/d/euoIiDRDqcFnK

[Wognath]

Hello again, OscarTalks,
I did not delete /opt/vivaldi/lib/libffmpeg.so. I just stuck the "latest codecs" version in /opt/vivaldi.

Before, I used peebee's sfs of chromium-64_46.0.2490.71 and added /usr/lib64/libwidevinecdm.so. It runs as root and plays Netflix.

Cheers!

[sheldonisaac]

I have upgraded the packages to the new stable 1.0.435.40 version.
...
My packages have the run-as-root modifications. Do we know if running as root prevents Netflix from working?
https://yadi.sk/d/euoIiDRDqcFnK

Thanks, OscarTalks and Wognath. I'm just a user, and can't follow
much of what you said.
I downloaded the sfs Vivaldi from your link, above. It didn't run under SuperLuPu2 Lucid Puppy (missing libraries or something like that).
It did run under TahrPup 6.0.5 32-bit.

[OscarTalks]

@ Sheldon,

Yeah, I'm afraid these Chromium derivative browsers are not likely to work in Lucid at all. I am aiming to get them to work in such Puppies as Wheezy, Precise and 14.0 Slackos and to do that I have to disable some of the sandboxing. In current official Puppies like Tahr and 14.1 Slackos people may want to experiment with re-enabling the sandboxing by deleting the switches from the launcher wrapper script, but if it works out of the box then that is great.

@ Wognath,

OK so I put the shipped libffmpeg.so back in the sub-dir since that configuration works for you, although I reckon a symlink to the downloaded one would also work. The downloaded one is much bigger so I guess it was compiled with extra codecs, some of which are needed for Netflix. I presume the lib is detected as a plugin in the main directory. I am at a bit of a disadvantage in all this because I don't plan to sign up for Netflix so it is not possible for me to test this out, but it would be nice to know that it works for those who want it.

[Wognath]

Hi Oscar,
I fired up Tahrpup 6.0, and loaded your vivaldi-1.0.435.40-i386.sfs. The browser runs well, but unfortunately, Netflix does not play. I tried running with sandbox, running as spot and changing permissions of the libs to 755 to match my Fatdog install. No luck. Each time I get the error "...make sure you are on an official version of Chrome.. Error Code M7355-1203", which brings back memories of trying to get Opera to work... No more ideas at present, but I'll let you know if I make any progress.

I noticed that the tab key does not work as "next box" in forms. I don't think it's anything I did.