duktape error NetSurf OS3

[chris]

Hello,  I trying to compile netsurf for OS3. But I have this problem hat I can not solve :angry:.
i test it on cygwin 32 under Windows and on ubuntu 15.10
could someone help me? thank you

Yes.  This is your problem:

Code: [Select]

make: nsgenbind: Command not found

You need to build and install nsgenbind (if you haven't already).

I find nsgenbind ends up in /opt/netsurf/bin where the makefile can't find it, so create a link in /usr/bin:
ln -s /opt/netsurf/bin/nsgenbind /usr/bin/nsgenbind
(think that's the right way round, always have to look it up!)

However I found here that NetSurf won't startup when Javascript is enabled, it just errors out, so you might find you need to disable it anyway (by adding a line to makefile.config, see makefile.defaults)

[chris]

i have also errors on nsgenbind build

Try building it just with simply "make".  It needs to be built for the computer you are cross-compiling on, rather than the target (as it is only used for building NetSurf).

[chris]

You may also find after installing nsgenbind, that it isn't in the path. Creating a soft link from /usr/bin/nsgenbind to /opt/netsurf/bin/nsgenbind might help.

[chris]

I think I already did this, and I have <> (I translate the sentence from french, so sorry for the mistakes)
I looked for the folder /usr/bin/nsgenbind but it doesn't exist. That the same thing for the other folder. I wonder if the link doesn't already exists...

I usually get that error because I can never remember what order the parameters are supposed to be in.

Anyway, it's easy to check if the link already exists:

Code: [Select]

chris@debian-vm:~/netsurf/netsurf$ ls -la /usr/bin/nsg*
lrwxrwxrwx 1 root root 26 Feb 29 00:16 /usr/bin/nsgenbind -> /opt/netsurf/bin/nsgenbind


If it shows up as a file (ie. no "->"), delete it and recreate the link.

[chris]

However, the problem is when I try to launch NetSurf 3.5 on AmiKit V8 now. I obtain this :

Code: [Select]

FATAL 56: uncaught error
PANIC 56: uncaught error (calling abort)
Abnormal program termination
NetSurf : error code 20
I don't have this error with NetSurf 3.5 "dukTape disabled edition".

Yes, I have that problem too.  If you figure out the fix I'd appreciate a patch

The browser is well working in that case (it is only "not really" beautiful and the pictures and videos aren't displayed).

Hopefully that's not an issue with 3.6dev, the bitmap routines have been changed to use guigfx (as you know!)

[chris]

I can see a problem with the version date (6th April 1016). Maybe it comes from there ? A date issue as on every softwares synchronized with the Internet ?

That's a typo, it's just a text string, it's only used for display.

[chris]

Java is nice, but I have high doubt that it would make Netsurf unuseable with real Amigas.

Java and Javascript are two different things.  Duktape implements Javascript.  It will only slow down NetSurf on pages which use Javascript, but I've not noticed any particular difference in speed with JS on/off under OS4.  There were a couple of issues that made Duktape really slow in certain circumstances, but they've been fixed.

Thanks for the interesting discussions and all the effort! I wanted to mention that linking with fortify really helps catching memory problems, maybe that could be an option for the development version of netSurf?

Oooh, that looks interesting.

[chris]

Oh, I don't think I'm good enough to do this ! :lol: I downloaded and looked the fortify files... and it doesn't seem to be reachable (with my level) to create the call to fortify (I could break everything ! :lol: )

Have a look at http://git.netsurf-browser.org/netsurf.git/log/?h=chris/fortify (or git checkout chris/fortify)
However... (a) that probably isn't adding fortify.h to all files (it assumes os3support.h and options.h are eventually included by everything, which probably isn't true)
and (b) it doesn't catch AllocVec/FreeVec and friends, which are mostly what I use in the frontend.

I'm getting some output, but it seems to be a load of bogus double free()s, which might be due to point (a) above, or because some of the allocations are made by linked libraries.

Try it and see if you can make any sense of the results.  I'll add some #defines for AllocVec when I get chance to catch those too.

[chris]

I use gcc 5.3.0, that's what "gcc --version" says.

I didn't think that worked for 68k yet?

[chris]

Could you relase exe for those who doesn't code?

What would be the point?  If you don't code, you don't need this script.  Instead you need to download a pre-compiled version of NetSurf.

At some point the OS3 build will be added to Jenkins, at which time you'll get a new build every time something changes.  But we ought to fix the Duktape problem first.

[chris]

OK respect for that. Thought you know devs usually release /publish binaries, so that "ordinary people" don't get comfused.

You already have binaries.  Nothing has changed since my last release (or, at least, nothing relevant).

[chris]

OK respect for that. Thought you know devs usually release /publish binaries, so that "ordinary people" don't get comfused.

You already have binaries.  Nothing has changed since my last release (or, at least, nothing relevant).

The aim here was to make it easier to build, and hence easier to contribute to.

[chris]

I don't know where the other thread is and I don't have time to go looking for it right now, so...

@utri

Please try the latest test build and let me know if there is any improvement.
The amiga.org logo is still very red, but are any of the other images better?
Speed?

http://cy2.uk/netsurfos3

[chris]

As a side note : I have fblit configured in my 040 / AGA machine. It has huge impact to scrolling speed. Tested old and new versions same time, old version was running my 060 system wich doesn't have fblit installed.

"Cache native versions" will have a huge impact too.  That may or may not counteract any benefit from fblit.

Still puzzled as to why the amiga.org logo is weird colours.

[chris]

Another new version.  Curious this time as to whether JPEG decoding speed is any different.