geronimo-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Joe Bohn <joe.b...@earthlink.net>
Subject Re: [DISCUSS] Monitoring Client may need a new graphing engine
Date Fri, 07 Dec 2007 04:17:21 GMT


Erik B. Craig wrote:
> All,
> 
> Currently the monitoring client is using Dojo 0.4.3 charting, which does 
> not necessarily behave as expected on Firefox/Safari on a mac, or on IE6 
> on Windows.
> I consider this to be a shortcoming, and given the new version of Dojo 
> available (1.0.1), began investigating migrating the monitoring client 
> over to the new version of Dojo, only to find that the new version of 
> dojo appears to be a significant rewrite of the old code base, leaving 
> out some features that I consider to be very visually pleasing and 
> important for statistics viewing. While rummaging through the Dojo 
> forums, I stumbled upon another Javascript graphing framework called 
> Timeplot, which is part of the SIMILE project at MIT, and while this has 
> it's own set of limitations... I'm trying to figure out the lesser of 
> three evils before it comes a time that this monitoring plugin will be 
> released, so that I have enough time (read: 3-5 days) to migrate the 
> javascript generation over to something new if necessary.
> 
> I have created a small demonstration page that shows all three options 
> graphed with the same data series, as well as weighing some of the 
> advantages/disadvantages I could come up with,
> Please have a look, and let me know your thoughts.
> 
> http://people.apache.org/~ecraig/graphdemo/
> 
> Personally, I think it would be really cool if we could use the Timeplot 
> graphing libraries, as it is all BSD licensed and therefore friendly I 
> believe (right, Kevan?)... and also EXTREMELY cool for showing multiple 
> data series in one chart.
> 

The mouse-over feature is cool ... but considering all things listed as 
advantages/disadvantages it seems to me that the dojo 0.4.3 comes out 
ahead for now.  After timeplot has addressed the IE issue and x/y axis 
labels (with min/max) then it would be much more compelling to consider 
a change.  Just my opinion based upon the very nice summary you provided.

Thanks,
Joe



Mime
View raw message