Chromium

I can see FreeBSD on Chromium there, but is it Chromium on FreeBSD? :)
 
Well done Ben and Sprewell. Wish I had time to help more, but you've obviously done a fine job!

Are you/ben sending the patches upstream and waiting for the next release before submitting the port?
 
I've posted i386 and amd64 test builds of Chromium that are up to date with their repo as of last night. Some smaller stuff, like SSL through NSS, still doesn't work but I've been using Chromium on my FreeBSD desktop for the last couple days and it's very fast and stable. :e It was a bit surreal when I got Chromium working and it automatically pulled in Flash 10, that I had just installed a couple days earlier, so I was able to use Hulu on Chromium: FreeBSD was almost a first-class desktop. :h

thuglife, I appreciate the enthusiasm. :) aragon, Ben is a Chromium committer and we're in close contact. He'll be reviewing this patch and committing it at some point. As for submitting something to ports, I think we need to fix some of the smaller errors and polish it up a bit more before doing that. I can't do that all on my own, so if anybody else wants to chip in on what has essentially been an effort by just Ben and me so far, albeit with aragon and Ariff chipping in on the necessary supporting libraries, feel free to start hacking and contact us.
 
sprewell said:
I've posted i386 and amd64 test builds of Chromium that are up to date with their repo as of last night. Some smaller stuff, like SSL through NSS, still doesn't work but I've been using Chromium on my FreeBSD desktop for the last couple days and it's very fast and stable. :e It was a bit surreal when I got Chromium working and it automatically pulled in Flash 10, that I had just installed a couple days earlier, so I was able to use Hulu on Chromium: FreeBSD was almost a first-class desktop. :h

thuglife, I appreciate the enthusiasm. :) aragon, Ben is a Chromium committer and we're in close contact. He'll be reviewing this patch and committing it at some point. As for submitting something to ports, I think we need to fix some of the smaller errors and polish it up a bit more before doing that. I can't do that all on my own, so if anybody else wants to chip in on what has essentially been an effort by just Ben and me so far, albeit with aragon and Ariff chipping in on the necessary supporting libraries, feel free to start hacking and contact us.

I started working on a port (in the FreeBSD sense) a few weeks ago, hoping to just drop Ben's patches and generate a package. Of course it hasn't been that easy and I haven't had much time to work on it again.

I was hoping to use third-party libraries installed from ports (hence the list of GYP_DEFINES) rather than bundled versions. That would decrease the build-time significantly but I have no idea if it's actually properly supported yet.

Anyway, here's the Makefile: http://people.freebsd.org/~flz/local/ports/Makefile.chromium
 
I've been posting updated test builds every week or two and will continue to do so. Some chromium committers have expressed interest in merging this patch into their codebase so let's see when that happens. This port is not yet ready for primetime, as a tab will sometimes flake out and freeze up for a little while and right-clicking to pull up a context menu stopped working a couple builds ago. However, it's very fast and stable, I've been using it as my primary browser for months now. I don't think it makes sense to add it to ports yet, particularly since google only makes available giant 1GB developer tarballs to build from, but chromium-devel binary pkgs make sense for now, if someone wants to do that. We need help porting linux-specific stuff like /proc to FreeBSD equivalents like sysctl or kqueue, so input is always appreciated. So far it's been just Ben and me working on this port, though one or two people have expressed interest in helping out. All the information necessary to checkout code and start hacking is available at the above link.
 
sprewell said:
I've been posting updated test builds every week or two and will continue to do so. Some chromium committers have expressed interest in merging this patch into their codebase so let's see when that happens. This port is not yet ready for primetime, as a tab will sometimes flake out and freeze up for a little while and right-clicking to pull up a context menu stopped working a couple builds ago. However, it's very fast and stable, I've been using it as my primary browser for months now. I don't think it makes sense to add it to ports yet, particularly since google only makes available giant 1GB developer tarballs to build from, but chromium-devel binary pkgs make sense for now, if someone wants to do that. We need help porting linux-specific stuff like /proc to FreeBSD equivalents like sysctl or kqueue, so input is always appreciated. So far it's been just Ben and me working on this port, though one or two people have expressed interest in helping out. All the information necessary to checkout code and start hacking is available at the above link.

I'm only working on the port to help with build and QA. I have no plan to commit it for now. Also it's easier for people to install a package.

I took your patch for r32001 and applied/fixed it for r32163 and it builds fine but it crashes on startup. I haven't really looked into it more than that. I will wait until Ben's patches are committed upstream because as it is now, there's just far too many patches to maintain.

Thanks for your work in any case!

EDIT: FYI, the WIP port is at http://people.freebsd.org/~flz/local/ports/chromium/
 
That makes sense that it wouldn't work for a different commit, as the codebase can change a fair amount after 150+ commits. :) That's why I always list the commits I used, so that others can make sure they're working off the same repository version. I don't see why you'd have to maintain patches or why there are too many, as there are only two patches and all you have to do is apply them to a fresh checkout of the correct commit number.
 
nslay said:
I have mixed feelings about web applications. On one hand, they are very useful and convenient for applications like email, calendar, simple games and so forth. On the other hand, there's an extremist push by Microsoft and Google for a web OS. My interpretation is that they want to use the web application concept to protect intellectual property and collect valuable usage information. Otherwise, this is a terrible idea!

It's all about taking the control away. It's all markets' intrests. :) Personally as I see, we're approaching towards a thin web client connecting up to _the_ company's server. No need for new OS'es. /Sorry for being offtopic/

Because the market goes after the numerous average users, I think it would be important to make them know more about technology, or at least the consequences that come after all. Only, don't know how to achive that :)

To the topic, I think Google is trying to reinvent the wheel -at least in his marketing area, but i believe he won't be able to come up with anything fundamentally new as an operating system. Maximum, they will go manufacturing their own hardwares with their "secure" embedded system on it :)
 
There has been progress, but news isn't always posted here. See Sprewell's http://chromium.jaggeri.com/ for the latest test builds. I recently rolled ariff@'s ALSA code into a FreeBSD port which should hopefully give Chromium stable sound. I think this was the last component away from a working port? It still needs testing, and I haven't had a chance to, so please go ahead!
 
I've been posting test builds, patches, and updates at http://chromium.jaggeri.com for 5 months now, usually on a weekly basis. As for a port, the Chromium devs have been testing a build source tarball for months, which should be out eventually. Once they make that 140 MB tarball available, someone can build a port off of it, as I doubt anyone wants ports to download the current 1 GB dev source tarball, complete with tests and build tools, in order to build a 50 MB Chromium binary. I'm going to try putting up a Chromium bidding site soon, where users can bid money on bugs or features they want developed for Chromium and any developer can work on and get paid for that.
 
Quoting from that site:

Funding further work: Now that Chromium works reasonably well for my personal use, it's difficult for me to justify putting much time into it, other than periodically updating to trunk. However, if you would like to donate towards fixing some of these remaining issues, such as the rendering flakiness or v8 issues on i386, you can do so by clicking on the button below. You can also enclose a message on which of the remaining bugs are most important to you.

So another vanity project like so much on SourceForge. And they wonder why OSS is not ready for the desktop!
 
>So another vanity project like so much on SourceForge. And they wonder why OSS is not ready for the desktop!

Well, instead complaining do something. Spend some money or help them fixing bugs - that's Open Source. Open Source is not ready for the desktop due to lazy users, which don't understand Open Source per se. If you want to be a lazy customer then spend your money for Apple or Microsoft product, otherwise you have to do something!
 
mechanic said:
So another vanity project like so much on SourceForge. And they wonder why OSS is not ready for the desktop!
So another ignorant, selfish user who's only interested in consuming and berating the efforts of others who share freely.
 
mechanic, I wouldn't go so far as to even call it a vanity project, just something I hacked together for myself. ;) However, because I've gotten so much good stuff for free from FreeBSD devs, I gave back what I'd done too, as Aragon says. As you say though, to be a real desktop app, it's going to need more polish, something I'm not going to toil away at for free, particularly since I don't care if it works on i386, as I run amd64, or that it leaves some orphaned processes lying around, as I just run it for a week at a time and kill the orphaned process manually. That's why I'm going to let people bid on features they want or bugs they want fixed, so that people who want that stuff can pay for devs to implement it. If nobody wants to pay for putting in features or fixing bugs that I don't care about, nobody will get it, it's as simple as that.
 
oliverh said:
Well, another dependency for a once light-weight browser, especillay if you're using a WM and don't want this gstreamer nonsense.

It's not required to run browser. It's only necessary for watching videos.
Similar to aspell.
 
>It's not required to run browser.

I do know this, but if I want to use this feature I have to use this crap of dependencies.
 
Back
Top