Provide "Use Cases and Requirements" section #26

pozdnyakov · 2017-04-28T11:12:06Z

Fixes #23

Authors:
Alexander Sahalamov [email protected]
Mikhail Pozdnyakov [email protected]

Fixes w3c#23. Authors: Alexander Sahalamov <[email protected]> Mikhail Pozdnyakov <[email protected]>

anssiko · 2017-04-28T11:26:35Z

This is a good start, thanks @pozdnyakov @alexshalamov.

tobie

I think it would be useful to include the use cases which only require coarse information, maybe into a separate subsection.

I'm sort of bothered by the highly theoretical nature of some of those use cases. Do we have examples of applications on native platforms doing things like this? I think that would be a lot more convincing. We're not really helping the platform if we're increasing the attack surface area without enabling real use cases.

As mentioned, I think we should focus on getting the right permission solution for motion sensors and this will naturally follow.

tobie · 2017-04-28T11:27:39Z

index.bs

+   health regulations.
+1. A Web application monitors light level changes produced by hovering hand user gesture and
+   interprets them to control a game character.
+1. A Web application calculates settings for a camera with manual controls (apperture, shutter speed, ISO).


Are the kind of sensors exposed on devices actually useful for this? Are there examples of such apps in native app stores?

ALS Shutter - dslrremote
ALS Meter - lightmeterfree

So the first one seems to be simply a remote. The second one seems to say it no longer uses ALS because "Ambient light sensor models are unlucky not work" and does pixel analysis of the output of the camera instead.

You asked about apps, I quickly googled, there are apps :D

For high speed photography, I use professional light (laser) sensors / sound triggers. I agree, ALS would not be that useful. For calculating EV values / manual camera settings ALS should be enough for spot metering. Mobile device sensors could measure incident lightning values that are enough to calculate EV values in range of EV -4 to EV 14. I could test readings from my Seconic spot / omni meter and compare to PixelXL readings.

You asked about apps, I quickly googled, there are apps :D

Right… but they don't rely on ALS.

https://play.google.com/store/apps/details?id=jp.co.fmbee.beecam.lightmeter&hl=en
https://play.google.com/store/apps/details?id=jp.gr.java_conf.appdev.app.exposuremeter&hl=en
https://play.google.com/store/apps/details?id=com.dq.fotometro&hl=en

Nice. Those do.

tobie · 2017-04-28T11:31:53Z

index.bs

+Use Cases and Requirements {#usecases-requirements}
+=========
+
+1. A Web application gradually updates document style based on light level changes.


That's typically a CSS domain, and CSS seemed to think that an enum was good enough.

Enum values don't allow to create 'designer' specific ranges or control style changes gradually. ALS sensors on mobile devices could provide values in 0-43000 Lux range. Maybe some pages could provide high-contrast monochrome style for 'direct-sun' conditions. It would be difficult to define 3 enum values for 0-43000 range, that would satisfy everyone.

That's not the conclusion at which the CSS WG arrived. :)

And I think they have more legitimacy in this area than we do.

Do you happen to have technical details for CSS WG conclusion? Would be nice to sync with it.

Quite a few references in the mailing list. Seems there are accessibility-related concerns here which seems is why the feature is on hold within the CSS WG.

tobie · 2017-04-28T11:32:35Z

index.bs

+1. A Web application provides input for a smart home system to control lighting.
+1. A Web aplication checks whether light level at work space is sufficient, based on occupational
+   health regulations.
+1. A Web application monitors light level changes produced by hovering hand user gesture and


Do we have examples of this in shipped software somewhere?

tobie · 2017-04-28T11:33:09Z

index.bs

+=========
+
+1. A Web application gradually updates document style based on light level changes.
+1. A Web application provides input for a smart home system to control lighting.


Same here, do we have examples of native apps actually doing this?

tobie · 2017-04-28T11:38:56Z

index.bs

+   interprets them to control a game character.
+1. A Web application calculates settings for a camera with manual controls (apperture, shutter speed, ISO).
+
+The mentioned use cases require plain light level data, not a pre-defined set of values.


How plain? How accurate? With what frequency requirements? This seems to dependant a lot on the use cases, too.

tobie · 2017-04-28T11:40:17Z

index.bs

+
+1. A Web application gradually updates document style based on light level changes.
+1. A Web application provides input for a smart home system to control lighting.
+1. A Web aplication checks whether light level at work space is sufficient, based on occupational


Are the values of common sensors accurate enough and properly calibrated enough to enable such a use case?

Depends on the country, but usually, office regulations require lightning to provide illumination of 300 Lux, steps between proposed values for different areas (conference rooms, corridors, etc), are 50-100 Lux.

My question was whether onboard sensors were good enough for meaningfully measuring that. My hunch is that onboard sensors are good enough to measure change, but probably not absolute values. I might be completely wrong, however.

Overall, this is a cool use case, I'm just concerned that it is not doable in practice because of HW constraints.

There is only one constraint that comes to my mind. Usually, fov for the mobile sensors is limited, that's why, e.g., flash / light meters (that I use, quite often), have omni-directional meter in addition to spot meter. App can fix this limitation, e.g., use sensors + als and ask user to tilt device slightly, to capture als readings for larger fov area.

10 Lux resolution that is guaranteed by sensor manufacturers, should be enough to measure illuminance for workplace.

10 Lux resolution that is guaranteed by sensor manufacturers, should be enough to measure illuminance for workplace.

Is calibration OK, though? i.e. is 300 Lux measured on a device really 300 ± 10 Lux? That is, resolution is relative, not absolute.

tobie · 2017-04-28T11:43:55Z

Authors:
Alexander Sahalamov [email protected]
Mikhail Pozdnyakov [email protected]

As a sidenote, if this is important to you, and I understand there are company cultures in which it is, I'd much rather you added your names to the acknowledgment section than put this kind of stuff in the commit logs.

anssiko · 2017-04-28T13:22:10Z

Updated https://w3c.github.io/ambient-light/#acknowledgements

lknik · 2017-05-12T09:18:00Z

As per Chrome issue - maybe the Web application calculates settings for a camera with manual controls (apperture, shutter speed, ISO) shouldn't be among the most prominent examples?

anssiko · 2017-05-12T10:15:34Z

@lknik, please provide your feedback on the respective open issue #23 and not on this PR that has been merged.

tobie · 2017-05-12T10:24:09Z

Well, #23 tells you to have this conversation here, so I understand @lknik's confusion. This shouldn't have been merged, should have been reversed or the concerns raised here should have been addressed by the OP.

anssiko · 2017-05-12T10:42:41Z

Fixed: #23 (comment)

Now let's get back to productive work :-)

pozdnyakov requested review from tobie, anssiko and alexshalamov April 28, 2017 11:12

Provide "Use Cases and Requirements" section

09c1317

Fixes w3c#23. Authors: Alexander Sahalamov <[email protected]> Mikhail Pozdnyakov <[email protected]>

pozdnyakov force-pushed the use_cases branch from 674c8de to 09c1317 Compare April 28, 2017 11:23

anssiko approved these changes Apr 28, 2017

View reviewed changes

anssiko merged commit f95cfcc into w3c:gh-pages Apr 28, 2017

tobie reviewed Apr 28, 2017

View reviewed changes

tobie mentioned this pull request Apr 28, 2017

Add description for use cases #23

Closed

This was referenced May 11, 2017

Add RawMagnetometer, for uncalibrated readings. w3c/magnetometer#21

Merged

Provide API to access uncalibrated magnetometer values w3c/magnetometer#16

Closed

alexshalamov mentioned this pull request Jun 20, 2017

Security and Privacy considerations for ALS #13

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Provide "Use Cases and Requirements" section #26

Provide "Use Cases and Requirements" section #26

pozdnyakov commented Apr 28, 2017 •

edited by pr-preview bot

Loading

anssiko commented Apr 28, 2017

tobie left a comment

tobie Apr 28, 2017

alexshalamov Apr 28, 2017

tobie Apr 28, 2017

alexshalamov Apr 28, 2017

tobie Apr 28, 2017

alexshalamov Apr 28, 2017

tobie Apr 28, 2017

tobie Apr 28, 2017 •

edited

Loading

alexshalamov Apr 28, 2017

tobie Apr 28, 2017

tobie Apr 28, 2017 •

edited

Loading

alexshalamov Apr 28, 2017

tobie Apr 28, 2017

tobie Apr 28, 2017

tobie Apr 28, 2017

tobie Apr 28, 2017

tobie Apr 28, 2017

alexshalamov Apr 28, 2017

tobie Apr 28, 2017

alexshalamov Apr 28, 2017

tobie Apr 28, 2017

tobie commented Apr 28, 2017 •

edited

Loading

anssiko commented Apr 28, 2017

lknik commented May 12, 2017

anssiko commented May 12, 2017

tobie commented May 12, 2017

anssiko commented May 12, 2017

Provide "Use Cases and Requirements" section #26

Provide "Use Cases and Requirements" section #26

Conversation

pozdnyakov commented Apr 28, 2017 • edited by pr-preview bot Loading

anssiko commented Apr 28, 2017

tobie left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

tobie Apr 28, 2017 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

tobie Apr 28, 2017 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

tobie commented Apr 28, 2017 • edited Loading

anssiko commented Apr 28, 2017

lknik commented May 12, 2017

anssiko commented May 12, 2017

tobie commented May 12, 2017

anssiko commented May 12, 2017

pozdnyakov commented Apr 28, 2017 •

edited by pr-preview bot

Loading

tobie Apr 28, 2017 •

edited

Loading

tobie Apr 28, 2017 •

edited

Loading

tobie commented Apr 28, 2017 •

edited

Loading