John-Mark Bell wrote:

> 3371 - I'm a little dubious about this, as it introduces a default "non
>         secure" setting for nsgtk. At the very least this needs
>         documenting somewhere obvious.

Indeed.  It does however make such sites accessible, and is similar
behaviour to other under-developed browsers.  As the changelog says, the
SSL window for GTK is high on the to do list.  I'm fine with leaving
this out if there is much complaint, however.

> 3383 - This seems more like a feature addition, to me.

I saw it as a trivial featurette that bought substantial usability
benefits, but I'm happy for it to be left out.

> 3392 - I'd leave this out, due to outstanding issues.

Other than the ones already listed in the changelog?  Just because a bug
fix isn't feature complete yet doesn't mean it's not worthwhile :)

> 3439 - What does this fix? (it's a trivial enough change that I don't care
>         either way; I'm just interested to know)

Some GIF animations that have noise in the background, I seem to recall,
as well as a general tidying that will be required for RSVG anyway.
When I did this, I seem to recall checking what the RISC OS bitmap stuff
did, and that it created blank transparent bitmaps.

> 3478 - I think this should have a little more testing. It already
>         introduced one bug -- I'd rather not see more instability ;)

I want this in.  It fixes several long-standing and irritating bugs in
nsgtk.  We should test this change extensively if people have complaint.
 I won't be easy to move on this issue.

> 3479 - I'd like some clarification on this; I see no need for it atm.

We can leave this out if we revert 3482 and check in two separate fixes.

>> John-Mark's original posting advocated the following changes, which were
>> agreed to:
>>
>> 3299, 3304, 3305, 3306, 3307, 3311, 3312, 3317, 3318, 3322, 3323, 3324,
>> 3325, 3328, 3335, 3337, and 3314 (from RUfl, PS printer driver fix.)
> 
> WRT 3330->3334, which were excluded in that mail due to a lack of testing; 
> I'm not aware of any outstanding fetch-related bugs (and am only aware of 
> it introducing one, which was fixed some time ago). Therefore, I'm happy 
> to remove my objection wrt merging these changes into the stable branch.

OK.

B.



Reply via email to