[jpstewart]

Raion is right about there being native Amiga and Atari (and other) versions of Netsurf.  They're called "frontends" in the Netsurf source, and there are eight available for various platforms.  Only the GTK one would be relevant for IRIX.  I took a quick look just now, and it seems each frontend is around 10,000 to 20,000 lines of C code (plus about 2,000 more in the associated headers).  So that gives you an idea of the scale of work involved in creating a Motif (or other) frontend for IRIX instead of using GTK.

I put a bit of work into getting Netsurf to build back in the Nekochan days.  Based on that, my opinion is that the easiest approach would be to back-port it to an older version of GTK that does work on IRIX.  Unfortunately, I don't have the time to take on such a project.  Also note that the Netsurf core will likely have moved on since then, and it too may need more work to build on IRIX again.  (My work was on 3.1 and 3.8 is now current.)

---

Sorry for posting twice in quick succession.  JacquesT posted his question while I was writing my previous reply.

Could Irix run the Framebuffer implementation of Netsurf? Just a question in the wild...

Maybe.  I don't think that was around when I looked at Netsurf years ago.  Their framebuffer frontend supports SDL 1.2, which was in Nekoware.  It's worth looking into, if somebody can find the time.  It'll be a while before I can....

[Raion]

I've had ideas for a browser for awhile. I spoke with a LOT of people on the discord and the general wishlist is:

HTML5 support
Javascript support that can be toggled on/off
Modern SSL support
Light enough to run on an Indy or Indigo2

That essentially rules out anything running GTK2 for a variety of reasons. For one GTK2 is not optimized well for IRIX. Even if we did fork the last GTK2 version, got it running and optimized the hell out of it, we would still have a lot of other dependencies to satisfy, be using a toolkit designed for GNU/Linux, not IRIX and the performance would still likely be subpar.

Raion is right about there being native Amiga and Atari (and other) versions of Netsurf.  They're called "frontends" in the Netsurf source, and there are eight available for various platforms.  Only the GTK one would be relevant for IRIX.  I took a quick look just now, and it seems each frontend is around 10,000 to 20,000 lines of C code (plus about 2,000 more in the associated headers).  So that gives you an idea of the scale of work involved in creating a Motif (or other) frontend for IRIX instead of using GTK.

I put a bit of work into getting Netsurf to build back in the Nekochan days.  Based on that, my opinion is that the easiest approach would be to back-port it to an older version of GTK that does work on IRIX.  Unfortunately, I don't have the time to take on such a project.  Also note that the Netsurf core will likely have moved on since then, and it too may need more work to build on IRIX again.  (My work was on 3.1 and 3.8 is now current.)

Yeah. I've been planning on organizing funding for a new browser eventually, but I don't want to put that effort into somethig as disappointing as Netsurf.

Could Irix run the Framebuffer implementation of Netsurf? Just a question in the wild...

You're gonna need to find someone who has the skill, time, and motivation for that. And not sure if it can run well enough or offer the features we need over Dillo3

[eudatux23]

Me too!

[Knezzen]

And me!

[stormy]

I've had ideas for a browser for awhile. I spoke with a LOT of people on the discord and the general wishlist is:

HTML5 support
Javascript support that can be toggled on/off
Modern SSL support
Light enough to run on an Indy or Indigo2

That essentially rules out anything running GTK2 for a variety of reasons. For one GTK2 is not optimized well for IRIX. Even if we did fork the last GTK2 version, got it running and optimized the hell out of it, we would still have a lot of other dependencies to satisfy, be using a toolkit designed for GNU/Linux, not IRIX and the performance would still likely be subpar.

Raion is right about there being native Amiga and Atari (and other) versions of Netsurf.  They're called "frontends" in the Netsurf source, and there are eight available for various platforms.  Only the GTK one would be relevant for IRIX.  I took a quick look just now, and it seems each frontend is around 10,000 to 20,000 lines of C code (plus about 2,000 more in the associated headers).  So that gives you an idea of the scale of work involved in creating a Motif (or other) frontend for IRIX instead of using GTK.

I put a bit of work into getting Netsurf to build back in the Nekochan days.  Based on that, my opinion is that the easiest approach would be to back-port it to an older version of GTK that does work on IRIX.  Unfortunately, I don't have the time to take on such a project.  Also note that the Netsurf core will likely have moved on since then, and it too may need more work to build on IRIX again.  (My work was on 3.1 and 3.8 is now current.)

Yeah. I've been planning on organizing funding for a new browser eventually, but I don't want to put that effort into somethig as disappointing as Netsurf.

Could Irix run the Framebuffer implementation of Netsurf? Just a question in the wild...

You're gonna need to find someone who has the skill, time, and motivation for that. And not sure if it can run well enough or offer the features we need over Dillo3

What is the possibility of writing a GUI front-end for Netsurf in Motif?

[vishnu]

Looking at the netsurf source code just now; it appears to be a fork of mozilla since a crapton of it's source code functions feature the "ns" leading edge, but if you look at their "about netsurf" page it mentions nothing about being forked from anywhere. Maybe that's why they named their project NetSurf, because mozilla was forked from NetScape...?

Anyway, as a decades-long motif developer I can attest it would be a monumental effort to create a motif frontend for it. The best way to hopefully do that would be to convince the netsurf developers to branch their code to a motif port (as opposed to trying to do it here, with the hope that somehow enough people with motif skills would pitch in). I was on Nekochan for a long time and it never struck me that there were any other motif developers there... :(

[micrex22]

It may be a stupid question, but what about a port of chromium (webkit) that would be geared for all UNIX?

[Raion]

I've looked into Chromium. The main issue is that GTK and QT are very complex and not easy to port to all UNIX. So basically, to have a full on cross-platform that can run on even some obsolete UNIX, you'd need to:

1. Pick a toolkit that's available on all of them. This basically relegates you to very old QT/GTK versions, Motif and FLTK
2. Rewrite the engine to use the above toolkit
3. Ensure that the C/C++ code can compile - either on a common compiler pre-agreed upon by all users, or the system compilers of each system.

At the end of the day this would only get you an HTML/CSS rendering engine. For JS, that's a whole different ball-game and you'd need to develop a platform/architecture agnostic JIT, or you're stuck with slower fallback modes.

[Raion]

A few updates on this:

1. I've become convinced that GTK+ and QT are a lost cause on IRIX. Maybe someone can finally get GTK2 working, finally, at its terminus version, but until then, we're stuck with a very old version of GTK2.

2. There are newer compilers including C++11, but any firefox or other work would basically exclude lower end systems due to the lack of JS JIT. If it's only usable on a super high end Octane or Tezro, what's the point? I'd say if it can't run acceptably on an O2 R10k, there's probably no point. I'm not asking for miracles, mind you. Just, maybe 40-2 min page loads at most with JS.

[Raion]

One project of notice is RetroZilla for Windows PCs: https://github.com/rn10950/RetroZilla

I may study this code sooner or later, see how much of it is Windows specific. If it's not, then maybe we can look into cooperation with them; or turn it into a fork of our own. The latter isn't preferable.

Gods above, wheres Diegel when we need him??