wasm with SeaMonkey?

[Steve Wendt]

Has anyone played with WebAssembly much in SeaMonkey? I was trying to
test out ruffle (plays Flash content without the plugin), which mostly
works OK with Firefox and Chrome. But SeaMonkey seems to lock up.

If it's not just me, and it's unlikely that SeaMonkey's wasm support
will improve dramatically in the near future, is there some useful JS
check that can be done to indicate "don't bother trying" without
resorting to stupid user agent checks?

[Frank-Rainer Grahl]

Steve Wendt wrote:
> Has anyone played with WebAssembly much in SeaMonkey?  I was trying to test
> out ruffle (plays Flash content without the plugin), which mostly works OK
> with Firefox and Chrome.  But SeaMonkey seems to lock up.

Shouldn't lock up but support is not really there. Too many other construction
sites in SeaMonkey so unlikely we will improve it fast. Personally I consider
it a security risk and would rather disable it. Hope it does not get
widespread support and too many websites needing it.

> If it's not just me, and it's unlikely that SeaMonkey's wasm support will
> improve dramatically in the near future, is there some useful JS check that
> can be done to indicate "don't bother trying" without resorting to stupid user
> agent checks?

If the site states that it needs WASM don't bother and pick another browser
for now :) Otherwise check the error log if you spot something. Biggest
problem right now is missing support for Web Components.

FRG

[Steve Wendt]

On 1/21/2021 12:27 PM, Frank-Rainer Grahl wrote:

>> I was trying to test out ruffle (plays Flash content without the
>> plugin), which mostly works OK with Firefox and Chrome.

There's a test site here:
https://ruffle.rs/demo/

>> is there some useful JS check that can be done to indicate "don't
>> bother trying" without resorting to stupid user agent checks?
>
> If the site states that it needs WASM don't bother and pick another
> browser for now :) Otherwise check the error log if you spot something.
> Biggest problem right now is missing support for Web Components.

Right, I'm trying to be that site, and not bother if it's going to fail
(i.e. SeaMonkey). Is there some nice JS check for "Web Components" I
can use to bail on? Obviously replacing the Flash content is the
long-term fix, but this is some non-critical, non-public stuff.

[Frank-Rainer Grahl]

Steve Wendt wrote:
> On 1/21/2021 12:27 PM, Frank-Rainer Grahl wrote:
>
>>> I was trying to test out ruffle (plays Flash content without the plugin),
>>> which mostly works OK with Firefox and Chrome.
> There's a test site here:
> https://ruffle.rs/demo/
>

Uses custom elements and so does not work. Might not be wasm related.

>>> is there some useful JS check that can be done to indicate "don't
>>> bother trying" without resorting to stupid user agent checks?
>>
>> If the site states that it needs WASM don't bother and pick another browser
>> for now :) Otherwise check the error log if you spot something. Biggest
>> problem right now is missing support for Web Components.
>
> Right, I'm trying to be that site, and not bother if it's going to fail (i.e.
> SeaMonkey).  Is there some nice JS check for "Web Components" I can use to
> bail on?  Obviously replacing the Flash content is the long-term fix, but this
> is some non-critical, non-public stuff.

Unfortunately no obvious check possible. Most of theses sites are just crap
and fail with obscure error mesaages or none at all. If you see references to
shadow dom, custom elements or shared js arrays in the code it is time to use
another browser unfortunately. Module scripts can possible be activated now
but I need to check if I got everything in. Service workers are still a bit
spotty.

FRG

FRG

[Steve Wendt]

On 1/21/2021 12:34 PM, Steve Wendt wrote:

> Is there some nice JS check for "Web Components" I can use to bail on?

This looks promising; will see if it does what I need...
https://github.com/cfware/no-webcomponents
http://cfware.github.io/no-webcomponents/

[Steve Wendt]

On 1/21/21 12:27 PM, Frank-Rainer Grahl wrote:

>> I was trying to test out ruffle (plays Flash content without the
>> plugin), which mostly works OK with Firefox and Chrome. But
>> SeaMonkey seems to lock up. >
> Shouldn't lock up but support is not really there.

Would you like me to file a Bugzilla bug? It should at least fail
without locking up, right?

> Otherwise check the error log if you spot something.

It locks up without showing anything. My quick hack was to only proceed
if (window.customElements && document.body.attachShadow) ...

So it falls back to normal Flash (which of course just gives the EOL
message now), but works via ruffle in Chrome and Firefox. I'd tell the
ruffle guys to fix this on their end, but I would expect them to tell me
this is a SeaMonkey bug (looks like one to me), since I can see they
have exception handling for customElements in there, and it doesn't
crash and burn in IE11.

[Frank-Rainer Grahl]

Steve Wendt wrote:
> On 1/21/21 12:27 PM, Frank-Rainer Grahl wrote:
>
>>> I was trying to test out ruffle (plays Flash content without the
>>> plugin), which mostly works OK with Firefox and Chrome.  But
>>> SeaMonkey seems to lock up. >
>> Shouldn't lock up but support is not really there.
>
> Would you like me to file a Bugzilla bug?  It should at least fail without
> locking up, right?

It didn't lock up for me. Eventually shows a script timeout error. With these
custom element scripts happened to me more than once. They seem to do only
limited or no error handling and just go into a loop. Normally I would ask for
a bug but this one will rot for sure so lets keep it in mind. In time custom
elements support will appear. Wish there would be more people doing dev work
but I could wish for the moon too :)

> So it falls back to normal Flash (which of course just gives the EOL message
> now), but works via ruffle in Chrome and Firefox.  I'd tell the ruffle guys to
> fix this on their end, but I would expect them to tell me this is a SeaMonkey
> bug (looks like one to me), since I can see they have exception handling for
> customElements in there, and it doesn't crash and burn in IE11.

Plugin is being killed in 2.53.7. Pretty useless now with the timebomb and
this is one area where we would like not to supprt hacked ones because of
security.

FRG

[Steve Wendt]

On 1/22/2021 4:16 AM, Frank-Rainer Grahl wrote:

> Normally I would ask for a bug but this one will rot for sure so lets
> keep it in mind. In time custom elements support will appear.

OK.

> Plugin is being killed in 2.53.7. Pretty useless now with the timebomb
> and this is one area where we would like not to supprt hacked ones
> because of security.

The final death of NPAPI - it lasted a long time!

[Frank-Rainer Grahl]

Killing Flash this week in SeaMonkey was a real pleasure. I still remember
when sites using this didn't run in OS/2.

Wish they whould have kept Java applet support but well you can't have everything.

FRG

[Steve Wendt]

On 1/22/2021 1:54 PM, Frank-Rainer Grahl wrote:

> Killing Flash this week in SeaMonkey was a real pleasure. I still
> remember when sites using this didn't run in OS/2.

Before the Innotek port of Flash 5? That's going back a long ways. ;)

[Frank-Rainer Grahl]

Yes. I suspect you still know about IBM WebExplorer too :)

Never used the port. Just consider me Flash hater #1.

FRG

[Steve Wendt]

On 1/22/2021 2:16 PM, Frank-Rainer Grahl wrote:

> I suspect you still know about IBM WebExplorer too :)

That was a fast browser in its day! But IBM realized pretty quick they
could never keep up with the changing web (a little like SM today, sadly
- I'm glad it's still usable for most stuff!).