You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
does grabbing urls from chrome work? I'm not seeing anything. #162
If you install the browser extension it will collect the data, but the visualization is still in progress and will be released in the next version. It is almost done and beta will hopefully be available soon if you want to try it out. The only visualization of the browser activity implemented is a summary of which websites you spend the most time on, we have plans for more ways to visualize it in later versions.
So I would recommend to have the extension running and once the beta releases within a couple of weeks you can simply install that and you'll see the data collected during that period.
This made me think of something after installing it. I run a chromebook, with activity watch running on my 'server' under Debian. I was thinking about what would be involved to try and track the chromebook.. (getting python installed maybe..I'm not sure. It wouldn't be a 'simple' process).
Anyways--with this extension, an idea, allow you to specify a host to send data to. Since chromebook is constantly in chrome, just adding the extension, then changing the host would let me capture chromebook data! (Mainly 2 screen with VNC & Plex website in Chrome...so would be cool if what I'm watching in plex would get captured).
I realize this would introduce problem of authentication.. (and I assume activitywatch is probably only listening on localhost at the moment..)
Anyways--with this extension, an idea, allow you to specify a host to send data to.
Yeah, using a remote aw-server for an aw-client has been brought up before. #142 (comment)
TL;DR You can configure it to work but it is insecure.
The main issue with chromebooks which makes it impossible however is that you cannot run aw-watcher-afk on it, so if we would make the webextension send data to a remote aw-server it would still not be aware if you are at the computer or not. Even if you were able to install it, chromeOS runs wayland which we do not support yet (it's currently impossible to implement due to the compositor missing the ability to expose mouse movement, keyboard presses and window status to other programs).
Hey, I was thinking about this, and had a question for you.
Have you thought about a solution that does not use any of the aw-python programs at all? (for Chromebooks, or Chrome browser in general) I.E. The extension itself would do all of the monitoring of activity/afk check, etc. I have not really worked with extensions much, but my brother showed me something that he built. I know that if you give Chrome the permissions, you get a ton of access. I would think you could detect afk w/JavaScript if that is a concern?
I know that this would be a pretty big divergence from how you have things set up now... and would be a project in itself. I had just wondered if you had considered this as a solution for Chromebooks? (and if there are obvious problems with going this route)--or if you hadn't, and this might be a possibility?
I know it would be a lot of work just for 'chromebooks', but IMHO, people might really like it (outside of chromebooks), just because of the syncing nature of extensions in Chrome--and this would be a 'lightweight' solution for computers they may use that they don't want to take the time to install... whatever.. (Just trying to maybe present a case for it! :D)
I know that if you give Chrome the permissions, you get a ton of access.
This is the issue why it is technically impossible. Chrome gives you a lot of permissions to the webbrowser yes, but it does not give you any permission at all to the OS, which is needed to know whether the user is AFK or not (detecting mouse movement/keyboard presses).
We could also just assume that if the person is using a chromebook he/she is always at the computer and always has the chrome window focused, but at that point it I would guess the actual usage times presented in activitywatch would be off by too much to even be useful.
I've used a Chromebook exclusively now, for about ~3-4 years. I honestly can't think of any time (or how I even would), use the chromebook without the browser. That would include chrome 'apps' which is just a chrome browser without the URL (not sure if it disables apps/extensions in this view--I guess that would break ability of javascript/app to track afk).
Also, if you ever happen to move to supporting Android, Chromebook might be a great device to start developing on, because the newer ones also have Android, or at least a version of it. I've got an ASUS Flip, and it runs android apps, and has an android settings window that I would imagine is very similar to the phone (I use iOS for handheld).
Just thought I should add this input...because I would love to be able to track activity on my 'main' device. My use case is kinda weird I know...in that I have a 'server' that has the main install--so having an extension where you could add in a URL to report back to would be awesome.
I haven't read in depth on your pages, but --if you could get the developers, I would think you would want to have mobile supported? Once there is more than one device, doesn't it make sense to make it sort of a client/server type deal? ...and IMHO would be a better user experience to have a single sign on type page where you can check / search across all devices/computers monitored.
Activity
johan-bjareholt commentedon Jan 11, 2018
If you install the browser extension it will collect the data, but the visualization is still in progress and will be released in the next version. It is almost done and beta will hopefully be available soon if you want to try it out. The only visualization of the browser activity implemented is a summary of which websites you spend the most time on, we have plans for more ways to visualize it in later versions.
So I would recommend to have the extension running and once the beta releases within a couple of weeks you can simply install that and you'll see the data collected during that period.
brizzbane commentedon Jan 11, 2018
Thank you!
This made me think of something after installing it. I run a chromebook, with activity watch running on my 'server' under Debian. I was thinking about what would be involved to try and track the chromebook.. (getting python installed maybe..I'm not sure. It wouldn't be a 'simple' process).
Anyways--with this extension, an idea, allow you to specify a host to send data to. Since chromebook is constantly in chrome, just adding the extension, then changing the host would let me capture chromebook data! (Mainly 2 screen with VNC & Plex website in Chrome...so would be cool if what I'm watching in plex would get captured).
I realize this would introduce problem of authentication.. (and I assume activitywatch is probably only listening on localhost at the moment..)
johan-bjareholt commentedon Jan 12, 2018
Yeah, using a remote aw-server for an aw-client has been brought up before.
#142 (comment)
TL;DR You can configure it to work but it is insecure.
The main issue with chromebooks which makes it impossible however is that you cannot run aw-watcher-afk on it, so if we would make the webextension send data to a remote aw-server it would still not be aware if you are at the computer or not. Even if you were able to install it, chromeOS runs wayland which we do not support yet (it's currently impossible to implement due to the compositor missing the ability to expose mouse movement, keyboard presses and window status to other programs).
brizzbane commentedon Jan 19, 2018
Hey, I was thinking about this, and had a question for you.
Have you thought about a solution that does not use any of the aw-python programs at all? (for Chromebooks, or Chrome browser in general) I.E. The extension itself would do all of the monitoring of activity/afk check, etc. I have not really worked with extensions much, but my brother showed me something that he built. I know that if you give Chrome the permissions, you get a ton of access. I would think you could detect afk w/JavaScript if that is a concern?
I know that this would be a pretty big divergence from how you have things set up now... and would be a project in itself. I had just wondered if you had considered this as a solution for Chromebooks? (and if there are obvious problems with going this route)--or if you hadn't, and this might be a possibility?
I know it would be a lot of work just for 'chromebooks', but IMHO, people might really like it (outside of chromebooks), just because of the syncing nature of extensions in Chrome--and this would be a 'lightweight' solution for computers they may use that they don't want to take the time to install... whatever.. (Just trying to maybe present a case for it! :D)
johan-bjareholt commentedon Jan 19, 2018
This is the issue why it is technically impossible. Chrome gives you a lot of permissions to the webbrowser yes, but it does not give you any permission at all to the OS, which is needed to know whether the user is AFK or not (detecting mouse movement/keyboard presses).
We could also just assume that if the person is using a chromebook he/she is always at the computer and always has the chrome window focused, but at that point it I would guess the actual usage times presented in activitywatch would be off by too much to even be useful.
ErikBjare commentedon Jan 19, 2018
As @johan-bjareholt said, it seems like a lot of work for something that wouldn't even end up being that good. Maybe someday, but not now.
ErikBjare commentedon Jan 20, 2018
Closing, thanks for the feedback @brizzbane! 👍
brizzbane commentedon Jan 25, 2018
Can't you use JavaScript to detect mouse movement/keyboard presses?
ErikBjare commentedon Jan 26, 2018
Not on the OS level, although there seems to be an idle API: https://groups.google.com/a/chromium.org/forum/m/#!topic/chromium-extensions/EWu2xPS2CK4
Maybe that could work, but since we can't run the server locally it would still be too much work to implement properly.
brizzbane commentedon Feb 4, 2018
I've used a Chromebook exclusively now, for about ~3-4 years. I honestly can't think of any time (or how I even would), use the chromebook without the browser. That would include chrome 'apps' which is just a chrome browser without the URL (not sure if it disables apps/extensions in this view--I guess that would break ability of javascript/app to track afk).
Also, if you ever happen to move to supporting Android, Chromebook might be a great device to start developing on, because the newer ones also have Android, or at least a version of it. I've got an ASUS Flip, and it runs android apps, and has an android settings window that I would imagine is very similar to the phone (I use iOS for handheld).
Just thought I should add this input...because I would love to be able to track activity on my 'main' device. My use case is kinda weird I know...in that I have a 'server' that has the main install--so having an extension where you could add in a URL to report back to would be awesome.
I haven't read in depth on your pages, but --if you could get the developers, I would think you would want to have mobile supported? Once there is more than one device, doesn't it make sense to make it sort of a client/server type deal? ...and IMHO would be a better user experience to have a single sign on type page where you can check / search across all devices/computers monitored.
marcusquinn commentedon Jul 18, 2020
Don't know how they do it but TimeCamp definitely works on ChromeOS - although would be happy to switch to ActivityWatch if you figure it out:
https://www.timecamp.com
https://chrome.google.com/webstore/detail/timecamp/ohbkdjmhoegleofcohdjagmcnkimfdaa?hl=en
test(query-parity): attribute known failures to their measured causes
test(query-parity): attribute known failures to their measured causes (…