[Building Sakai] Moving Presence into a DIV - SAK-19129

May, Megan Marie mmmay at indiana.edu
Tue Sep 14 04:50:00 PDT 2010


This seems like is a pretty neat feature.  That said, I'm concerned about any changes in something that is heavily used.  There seems like there is addition due diligence to perform:    

1) This is a great opportunity for peer code review 
2) Profiling of the code should be done
3) Is this something that the Performance lead can/will test?   
4)  It should be off by default in 2.8 until it's proved itself.  Sakai is way to mature of a product to have something on by default that has unknown performance implications

Megan 

-----Original Message-----
From: sakai-dev-bounces at collab.sakaiproject.org [mailto:sakai-dev-bounces at collab.sakaiproject.org] On Behalf Of Nate Angell
Sent: Monday, September 13, 2010 8:39 AM
To: csev
Cc: David Horwitz; sakai-dev Developers
Subject: Re: [Building Sakai] Moving Presence into a DIV - SAK-19129

Sorry, I should have read the full Jira before asking about the performance implications...your note is fulsome and seems to sketch out the issues well. It will be nice to be able to tweak the timing settings to find the right balance, and, should the new ajaxification degrade performance, revert to the existing courier method.

Well done!

- Nate

On Mon, Sep 13, 2010 at 8:17 AM, csev <csev at umich.edu> wrote:
> Nate - excellent question - I already have a note in JIRA about performance.
>
> Honestly I am not 100% sure - so I wrote the note in the JIRA.
>
> There are certainly ways this will perform better than the iframe approach and scenarios where it may perform worse.  That is why I (a) wrote a note in the JIRA and (b) made it so it can be turned off quickly :).
>
> The key issue is the cost of a courier refresh versus a presence reload as well as how long folks stay in a site and how much churn there is in presence.
>
> I would not have put it in unless I felt it were neutral or an improvement as my best guess.
>
> /Chuck
>
> On Sep 13, 2010, at 7:36 AM, Nate Angell wrote:
>
>> Very cool Chuck! Any ideas on how this would impact performance 
>> (positively or negatively)?
>>
>> - Nate
>
>
_______________________________________________
sakai-dev mailing list
sakai-dev at collab.sakaiproject.org
http://collab.sakaiproject.org/mailman/listinfo/sakai-dev

TO UNSUBSCRIBE: send email to sakai-dev-unsubscribe at collab.sakaiproject.org with a subject of "unsubscribe"


More information about the sakai-dev mailing list