Difference between revisions of "Talk:LlRegionSayTo"

From Second Life Wiki
Jump to navigation Jump to search
 
(5 intermediate revisions by 3 users not shown)
Line 12: Line 12:
The information about the 100 messages throttle is wrong:
The information about the 100 messages throttle is wrong:
<blockquote>"Sending more than 100 messages in five minutes will produce a block; block is sim-wide and applies to all objects with the same owner. Block apparently only resets with a sim restart."</blockquote>
<blockquote>"Sending more than 100 messages in five minutes will produce a block; block is sim-wide and applies to all objects with the same owner. Block apparently only resets with a sim restart."</blockquote>
According to Maestro Linden's comment in BUG-913, the actual throttle is much less harsh:
: According to Maestro Linden's comment in BUG-913, the actual throttle is much less harsh:
<blockquote>"llRegionSayTo() is indeed limited by 2 throttles: There's a per-destination throttle on channels 0 and DEBUG_CHANNEL, which is 200 messages over 10 seconds. There's a per-sending-object throttle of 100 over 300 seconds (kinda slow...). '''This throttle only applies when the first is hit'''.</blockquote>
<blockquote>"llRegionSayTo() is indeed limited by 2 throttles: There's a per-destination throttle on channels 0 and DEBUG_CHANNEL, which is 200 messages over 10 seconds. There's a per-sending-object throttle of 100 over 300 seconds (kinda slow...). '''This throttle only applies when the first is hit'''.</blockquote>
: --[[User:Thickbrick Sleaford|Thickbrick Sleaford]] ([[User talk:Thickbrick Sleaford|talk]]) 08:25, 9 July 2015 (PDT)
: --[[User:Thickbrick Sleaford|Thickbrick Sleaford]] ([[User talk:Thickbrick Sleaford|talk]]) 08:25, 9 July 2015 (PDT)
'''Clarification needed of Caveat about throttle.'''
Could we please change the caveats to reflect the fact that the per-destination throttle applies only to Channel 0 and not to non-0 channels?  As they stand, they're seriously misleading -- see, for example, https://jira.secondlife.com/browse/BUG-11346, which seems to have been based on the (wholly understandable) assumption that because a script using channel 0 triggered the throttle, the throttle would be triggered by other channels, too.  I've also several times had to put people right about this in in-world scripting groups.
The situation isn't helped by the fact that the jira page for BUG-913 isn't accessible to most of us, so we can't see Maestro's comments verbatim.
I'm not sure, by the way, that Maestro is correct about DEBUG_CHANNEL triggering the throttle.  I just tried it, and didn't see any messages either in the debug window or from the relay to which the messages were directed.  I know that the do.. while loop sending the 5000 messages completed in just over 7.3 seconds, because it told me it had.
However, when I changed the channel in both the transmitter and the relay, the transmitter functioned as expected, so the transmitter prim clearly wasn't muted.
I would change it myself, but I'm worried about breaking stuff and also because I fear there may be some policy reason to explain why the full picture is left unclear on the main page.
[[User:Innula Zenovka|Innula Zenovka]] ([[User talk:Innula Zenovka|talk]]) 14:20, 7 February 2016 (PST)
:If you (technically) break the wiki page, I'll fix it, don't worry about breaking things. The reason for poor quality content is because nobody has written high quality content. LL tends not to write the documentation, it's mostly user written. They like it this way because then they aren't obligated to update the documentation. If you care about accurate/usable documentation you pretty much have to write it yourself. Despite the LSL Portal being the Official LSL Documentation and being hosted by LL, LL leaves the LSL Portal to us. We have considerable editorial and design control. They pretty much ignore us. So... have at it, have fun. Write some good documentation. Drag the truth out into the light. If there are LL politics at play elsewhere, it has no sway here. '''TL;DR:''' LL is lazy, not malicious/conniving. -- '''[[User:Strife_Onizuka|Strife]]''' <sup><small>([[User talk:Strife_Onizuka|talk]]|[[Special:Contributions/Strife_Onizuka|contribs]])</small></sup> 18:50, 7 February 2016 (PST)

Latest revision as of 19:52, 7 February 2016

Would it be an idea to put a reference to https://jira.secondlife.com/browse/SCR-66 "llRegionSayTo() fails to send messages to attachments on sitting avatars"?

(I'd do it myself but I would probably break something -- I did once when I tried to edit Wikipedia) Innula Zenovka

Added it for you, the bit that does it is the link to the Template:Issues sub-page at the top of the function's page (in edit mode), if you follow that you can see how to enter a JIRA's information to get the pretty table.
-- Haravikk (talk|contribs) 13:49, 8 May 2011 (PDT)


llRegionSayTo also needs be added to the Communications category and in the list af See Also at the bottom of each function belonging to the category - thanks.
Francoise Eichel 14:14, 2 July 2011 (PDT)


The information about the 100 messages throttle is wrong:

"Sending more than 100 messages in five minutes will produce a block; block is sim-wide and applies to all objects with the same owner. Block apparently only resets with a sim restart."

According to Maestro Linden's comment in BUG-913, the actual throttle is much less harsh:

"llRegionSayTo() is indeed limited by 2 throttles: There's a per-destination throttle on channels 0 and DEBUG_CHANNEL, which is 200 messages over 10 seconds. There's a per-sending-object throttle of 100 over 300 seconds (kinda slow...). This throttle only applies when the first is hit.

--Thickbrick Sleaford (talk) 08:25, 9 July 2015 (PDT)


Clarification needed of Caveat about throttle.

Could we please change the caveats to reflect the fact that the per-destination throttle applies only to Channel 0 and not to non-0 channels? As they stand, they're seriously misleading -- see, for example, https://jira.secondlife.com/browse/BUG-11346, which seems to have been based on the (wholly understandable) assumption that because a script using channel 0 triggered the throttle, the throttle would be triggered by other channels, too. I've also several times had to put people right about this in in-world scripting groups.

The situation isn't helped by the fact that the jira page for BUG-913 isn't accessible to most of us, so we can't see Maestro's comments verbatim.

I'm not sure, by the way, that Maestro is correct about DEBUG_CHANNEL triggering the throttle. I just tried it, and didn't see any messages either in the debug window or from the relay to which the messages were directed. I know that the do.. while loop sending the 5000 messages completed in just over 7.3 seconds, because it told me it had.

However, when I changed the channel in both the transmitter and the relay, the transmitter functioned as expected, so the transmitter prim clearly wasn't muted.

I would change it myself, but I'm worried about breaking stuff and also because I fear there may be some policy reason to explain why the full picture is left unclear on the main page.

Innula Zenovka (talk) 14:20, 7 February 2016 (PST)

If you (technically) break the wiki page, I'll fix it, don't worry about breaking things. The reason for poor quality content is because nobody has written high quality content. LL tends not to write the documentation, it's mostly user written. They like it this way because then they aren't obligated to update the documentation. If you care about accurate/usable documentation you pretty much have to write it yourself. Despite the LSL Portal being the Official LSL Documentation and being hosted by LL, LL leaves the LSL Portal to us. We have considerable editorial and design control. They pretty much ignore us. So... have at it, have fun. Write some good documentation. Drag the truth out into the light. If there are LL politics at play elsewhere, it has no sway here. TL;DR: LL is lazy, not malicious/conniving. -- Strife (talk|contribs) 18:50, 7 February 2016 (PST)