Timberwolf RC1 available

Quote:

Yes, i read that

I was rather referring to your bookmarks problems...the way you described it sounds like it was for me with version 4 (before RC1).

Thats why i suggested to reinstall the whole program

@Raziel

ah, I see! Then you must've confused me with someone else from the thread as I've never mentioned having any bookmarks trouble. :)

While deleting my old Profile, and avoiding using MUI-OWB at the same time, has greatly improved TW RC1's stability, as soon as I try to use it for real web browsing I am risking it freezing for some minutes or even forever. So stability is still much worse than older releases, where I could actually use it (bar the lack of copy & paste in those versions).

This is a shame as TW now has all the functionality I need, and it's speed does seem improved.

I wish there was something I could do to provide a useful bug report. ("It crashes for me" is hardly helpful.)

@ chrisH

rc2 will come out tomorrow or Friday, and it has some changes that might help.

For one, it now has the option of running without RANDOM: since someone claimed that it would create a lot of CPU load for him. Whether that's true or not, I don't know, but you can either use RANDOM, or dismount it and use pseudo-random numbers. The latter isn't really secure, but better than nothing.

Secondly, the database handling has been changed. The new sqlite included with the next rc is going to have the v2 internal interface implemented (which uses memory mapped files). It's not real memory mapped files, of course, but it allows Write Ahead Logging, which might be better for CPU load (it seems to be on my system).

There's a load of other changes, too. More on that in the release notes.

We'll see how these changes help, if at all. If not, I'll come up with a possibility to do proper profiling of the whole thing to see where it burns the time.

I just noticed that the bounty for this project is officially finished. Congratulations to the Friedens on getting this far. It's a huge milestone, even though there are still some bugs/issues still to fix.

Having said that, there will no doubt be people who will want to donate to this project after the fact, or to give a donation for the continued improvement of Timberwolf (I'm assuming that the Frieden's won't be dropping the ball after RC2). Is there any facility for people to do that?

Hans

@tfrieden The random function kills my CPU .If you don^t believe it, come to me and have a look.

Congrats on getting FF nearly done (RC2 soon) and hope there will be final soon. Will it be optimized for SAM 460?
Also, closing bounty has made a rumor mill that final Timberwolf will be a paid version, which is contrary to bounty, so please confirm publicly its free

@tfrieden

Congratulations on all your hard work!

This will be a big milestone for OS 4.

It turns out I missed out on the bounty. Is there a way to fire any donations your way?


Cheers!

@vox

Quote:

I don't think there will be a "final" version ever, there are already a lot of things to do in term of stability, speed, "Amiga like" things to add .. not to mention any small/medium possible and impossible bugs that we eventually can find in the course of the time ..

In general a browser is always a huge project that need a consistently follow etc .. still normal for a great project like that, so don't expect a "real final version" but with the time and our help as pro users it will become better and better (as already it is compared to the first release)

As Friedens point out this is not only a browser but a technology so i'm sure it will be a part of a long term OS4 future strategy

Quote:

Who said that ?
Firefox is an open source product so I doubt that's true, but on Amigaworld you can read any kinds of speculation so ..

@tfrieden

Wow didn't expect an RC2 so soon !
Looking forward to try

would be possible to have something related the plugin system so i can "try" to make a gnash plugin??

Quote:

Hopefully RC2 will come out later this week, followed by a final version if no other major issues arise.

Timberwolf will not be a paid program, it will always be free. Anyone claiming otherwise is wrong. I am not even sure it would be possible at all to sell it, but even if, it's not going to happen.

I wonder how people come up with ideas like that, but I guess it's just an attempt to water down any news related to other camps.

future donations:

We do have a paypal link on our homepage (friedenhq.org). Donations are always welcome!

@ hans

We do not intend to stop with Timberwolf. rc2 is due today or tomorrow (probably tomorrow). I'm already quite sure that this will become the release version.

From there on out, we'll be exploring the possibility to update to a later build of Firefox. Within the bounty, this was too much of an uncertainty to commit to, but now, after the bounty has been fulfilled, it's something we want to explore. There have been a few interesting developments in the later versions (memory footprint reduction etc) that I really like to get into Timberwolf.

@ tommysammy

I do believe you, but as I said, I don't know if it's random: or the fact that random: mounted makes the databases work. In any case, you will be able to test it with rc2, since it has a fallback random number generation in case random: is not mounted.


@ vox

Sam440 optimizations are a problem. The 440 does not have an L2 cache, and Timberwolf is really memory intensive. I'm not sure how much can be done without hardware acceleration (which is something we will also look in to)


@ samo79

It's disputable whether a final version of any software exists. There is next to no software without any bugs in it, it's mainly about how many of those bugs are there and how they affect the workings of the program.

Timberwolf will be an ongoing project for quite some time still. We're definitely going to keep working on it. And the technology behind it is going to be used for other projects as well.


@ all

If someone tells you that Timberwolf will require payment in the future, they are simply BS'ing. Timberwolf will ALWAYS remain a free program, just like Firefox is a free program. Even if sale of the program would even be possible (I don't believe it is), we would never do that. Feel free to quote me on that wherever someone tell you otherwise.

Also, regarding the bounty: Getting nearly 7000 Euro out of it was a pleasant bonus, but considering that each of us spent far more than 1000 hours on the project, it can hardly be considered a "payment". It wasn't the money that kept us going, anyway, it was mainly the fact that with accepting the bounty, we accepted the commitment. No stepping back once we started it. It helped to get the thing on track.

Quote:

I'll check what needs to be done. In essence, the plugin interface should be the same as on other platforms.

I'll make the SDK available with the next release as well...

@tfrieden

Quote:

done.

#6

Quote:

uh! thank you!

@TFrieden & Rogue:

Thanks for your hard work on this, its very much appreciated!

You deserve every penny of the bounty and surely much more!

An interesting question is surely how you did the division of the work so far.

Who did what?

Are there situations you are working on the same Function, Class, Method..etc?

Or every of you have their own tasks on porting and developing particular parts of the code?

@ fingus

Thanks.


Firefox is pretty modular, so it's easy to work on different parts in parallel. GFX (the graphics parts) and Widget (the User Interface Parts) are the biggest chunks when it comes to porting, and they are pretty much separate.

The rest, bugfixing in general, can always easily be divided...

@Frieden's
Big thank you from me too !

Btw.Isn't it friedenhq.org and not .net. (That's how I got there.)

Quote:

Where? I had a quick look, and I couldn't find it.

In any case, I think that the bounty page should be updated to mention this.

Hans

Hans is right - no PayPal link on the hq site.

I guess they don't need any more donations...