- From: Mandyam, Giridhar <mandyam@quicinc.com>
- Date: Sun, 3 Feb 2013 22:24:54 +0000
- To: DAP <public-device-apis@w3.org>
- Message-ID: <CAC8DBE4E9704C41BCB290C2F3CC921A163B8ADC@nasanexd01h.na.qualcomm.com>
Hello All, Pardon me for the clumsy wording in the subject line. At the DAP F2F at the last TPAC, I raised some concerns regarding testing of the Battery API and how well the readings match the system level power meter. I am particularly interested in how developers can use this API on power-limited devices. I explored the Firefox implementation (Win7 version) with respect to a WebSocket application using a loopback server running at localhost. I am enclosing the client-side code as FYI, but you cannot execute it without a corresponding ws-compatible server. I am enclosing also a presentation summarizing my observations (which I hope to go over during the next DAP telco), and a screenshot of the application results along with the system power meter app. Note a couple of items: a) I am proposing that the battery API be tested in an application that requires networking transactions at a certain frequency. This would require a server, either running at localhost or somewhere in the cloud, to receive messages from the web application. b) I am not suggesting with the results in the slidedeck that there is any kind of bug in the Mozilla implementation of Battery API. I believe such a conclusion is premature - I would never offer myself as any expert in JS programming and there could be a simple fix available to coax the proper readings out of the mozBattery implementation for which I am unaware. I just want to see tests for the Battery API that are similar to the application I have written. Thanks, -Giri Mandyam, Qualcomm Innovation Center
Attachments
- application/pdf attachment: W3C DAP Battery Testing February 2013.pdf
- text/html attachment: TestBattery.html
- image/png attachment: Battery_Testing_Screen_Shot.png
Received on Sunday, 3 February 2013 22:25:31 UTC