edbrowse-dev - development list for edbrowse
 help / color / mirror / Atom feed
* [Edbrowse-dev] simpler html.cpp
@ 2014-02-11  5:28 Karl Dahlke
  2014-02-11 10:21 ` Adam Thompson
  0 siblings, 1 reply; 2+ messages in thread
From: Karl Dahlke @ 2014-02-11  5:28 UTC (permalink / raw)
  To: Edbrowse-dev

As portended by my earlier email, all the rooted variables are gone.
All the AutoCompartments are gone.
They were just local variables as shorthand for objects
that were already put on the rooted heap and tied to tags.

The functions in html.cpp now run, I believe, properly
whether cw->jss is null or not.
And to test this, there is now a way for jss to be null.
Disable javascript, with the js command or through the
nojs directive in your config file, then browse.
The java state isn't even created.
Part of my regression test is now to run jsrt through its paces,
with javascript and without javascript.

Next step, I will allow CreateContext() to fail, if it does fail,
without exiting, and leave jss null.
Then once this is working we can set up other error legs.

Karl Dahlke

^ permalink raw reply	[flat|nested] 2+ messages in thread

* Re: [Edbrowse-dev] simpler html.cpp
  2014-02-11  5:28 [Edbrowse-dev] simpler html.cpp Karl Dahlke
@ 2014-02-11 10:21 ` Adam Thompson
  0 siblings, 0 replies; 2+ messages in thread
From: Adam Thompson @ 2014-02-11 10:21 UTC (permalink / raw)
  To: Karl Dahlke; +Cc: Edbrowse-dev

On Tue, Feb 11, 2014 at 12:28:36AM -0500, Karl Dahlke wrote:
> As portended by my earlier email, all the rooted variables are gone.
> All the AutoCompartments are gone.
> They were just local variables as shorthand for objects
> that were already put on the rooted heap and tied to tags.

Sounds good, that code's been on my to review list for a while.
Will pull and test later today.

> Disable javascript, with the js command or through the
> nojs directive in your config file, then browse.
> The java state isn't even created.

That's nice to hear.

> Part of my regression test is now to run jsrt through its paces,
> with javascript and without javascript.

Sounds sensible. Any way to make jsrt test the gc a bit more?

> Next step, I will allow CreateContext() to fail, if it does fail,
> without exiting, and leave jss null.
> Then once this is working we can set up other error legs.

Yeah, again sounds sensible.

Cheers,
Adam.

^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2014-02-11 10:22 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2014-02-11  5:28 [Edbrowse-dev] simpler html.cpp Karl Dahlke
2014-02-11 10:21 ` Adam Thompson

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).