Hi there, I'm setting up gSuite and need some advice here.
Use case requirements:
- operations team member will have access to customer data on Google Drive, so we want to enforce those team members to only view such data via corporate VPN.
- Our sales team members do not have access to customer data, so we do not want them to login via VPN because that's extra friction.
After researching online, I found two ways of doing this:
a. Using the context aware access feature from gSuite https://support.google.com/a/answer/9275380?product_name=UnuFlow&hl=en&visit_id=637758886340041076-3475162686&rd=1&src=supportwidget0&hl=en
- Pro: set up seems straightforward, and can configure based on teams
- Con: pay more per user, it's like $8 more per user to our current edition
b. Set up gSuite to login with Okta via SAML, and then configure Okta log on policies to require VPN for certain users https://support.google.com/a/answer/6369487?product_name=UnuFlow&hl=en&visit_id=637758886340041076-3475162686&rd=1&src=supportwidget0&hl=en#zippy=%2Csso-usernetwork-mapping-matrix
- Pro: Saves money, get more familiar with SAML which seems useful?
- Questions for experts here about this approach:
- Is it easy to quickly provision which users are required to login via Okta and VPN? Anything we should be careful about? The process seems quite complicated.
- How much time will it take?
- Technically how is this different from option a? I guess gSuite does all the configuration automatically for you to save the time?
Which one would you advise us to go for? Any thoughts are tremenduously appreciated!
I have a few hue bulbs. I want to say lights on and turn on all lights except some, wich are used as accent lights. How could I do that? I already have them in their own group but when I say lights on they also turn on
I'm working on a language, and I had an idea to change the semantics of certain functions based on whether they are called as a statement or an expression.
Fore example, if you have a function push that adds an item onto the end of an array/list/vector/etc, you could call it like this:
push(vec, item)
This would mutate vec in place.
Alternatively, you could call it like this:
let pushed_vec = push(vec, item)
This would leave vec alone and return a new version with the pushed item. You could do some optimization using immutable data structures or use-analysis to prevent item cloning or reallocation.
Ignoring how you would actually implement functions like this, would this be too complicated or unexpected for the user? At least for functions like push, I think that the semantics would be obvious from the code.
Has anyone tried to do this yet? I have this setup for Microsoft, but not sure about the setup for Google. We are Edu Plus.
I've been at this for two days now, and I just keep peeling back layers of fail, no pun intended.
Running a Longer LK4 Pro, which I guess is based off Marlin 1.1.9. It does not support M600 natively, which was fun to find out. Okay, that's fine, there's a "Rewrite M600" plugin for OctoPrint. It's abandoned, and has a severe error unless you make sure the Z change is set to 0 - also fun to find out. There might even be other issues still, with layer alignment, but we haven't gotten there yet. After hours of finagling, I got it to technically do its job; hotend moves to the corner, I can manually raise it 10mm, swap the filament, wait 15 minutes for it to stop oozing, lower it 10mm, and hit resume. And it goes back to work.
Except.
The next line after the M600, is G1 F2700 E490.75421. Well, guess what Cura's Filament Change plugin doesn't do? Inform the printer that the filament is at the exact position it needs to be. See, while OctoPrint is technically "paused" while I perform the change, the printer is keeping track of everything internally. A 10mm Z move I can do and undo myself, but hundreds of mm of retraction and extrusion I can not. Both times I've gotten this far, all of my hard work making sure everything is all lined up and perfect and ready to go is answered by the machine frantically retracting filament as quickly as it's allowed to, then spitting out an entire layer of nothing.
Incidentally, the "Pause at height" Cura plugin has the same issue; no tracking & resetting of the extruder value means the extruder goes bonkers when resumed.
This is all an extremely long winded way of asking: Are there any Cura or OctoPrint filament change plugins that - presumably through a whole lot more than one line of gcode - will store and restore the exact state of the printer, so the #%$& thing just blanks out anything I've done to it in the mean time?
🎯 At a glance:
Is it possible to create 3d photos with convincing parallax effects from single RGB-D images? It is now! Check out a new 3D inpainting method proposed by Meng-Li Shih and colleagues. In short, the input image is transformed into a Layered Depth Image with explicit pixel connectivity, which is used to synthesize new local color-and-depth content into the occluded regions in a spatial context-aware manner. The resulting images can be rendered with a smooth parallax effect using standard graphics engines with fewer artifacts compared to current SOTA methods.
⭐️ Paper difficulty: 🌕🌕🌕🌕🌑
🚀 Motivation:
3D photos are more immersive than 2D, especially in VR. However, complex hardware setups are required to produce such images, and current methods that synthesize 3D photos from images captured with multi-lens smartphone cameras either produce gaps or distortions in the regions, occluded in the input image. Recent methods used Multi-Plane Image representation to address these issues, however they tend to produce artifacts on sloped surfaces. Instead of using rigid layers such as in Layered Depth Images (LDI), the authors explicitly store pixel connectivity and recursively apply CNN-based inpainting conditioned on spatially-adaptive context regions that are extracted from local connectivity in the LDI. The result is an algorithm for 3D photo generation without a predetermined number of depth layers.
Read the full paper digest or the blog post (reading time ~5 minutes) to learn about the modified LDI, Image Preprocessing, Context and Synthesis Regions, and Context-Aware Color and Depth Inpainting.
Meanwhile, check out the paper digest poster by Casual GAN Papers!
[Full Explanation Post / Blog Post] [Arxiv] [Code]
More recent popular computer vision paper breakdo
... keep reading on reddit ➡
we have been living in the uae since 1980s, a relative of mine was called to the passports authority in sharjah. Apparently he went there. Was ordered to bring his passport with him. He arrived, then the authority confisticated his passport. And told him that he cannot renew his residence anymore.
My question here is
Is there a way to revert this verdict? Or atleast appeal it?
If not, after a while (a year or so) will he be able to comeback to the country and create a new residence? Is that possible in any shape or form?
Is there a way to know the reason behind that verdict?
Thanks in advance
Does anyone know if there are future plans to enforce Context-Aware access such as sign ins from only outside the corporate network to use 2SV?
We would like to mirror what we are doing with Azure Conditional Access access to prevent 2SV requirements when a user is on premise. Our environment involves teachers. When they are off premise it would require 2SV. The only way I can find currently to pull this off is to integrate Google Workspace with Microsoft Cloud App security.
Any insight into a potential feature or work around would be appreciated!
Hi all,
Is there any work around making context-aware access level with Device policy attributes to work with SSO apps on Mobile devices? It Loos like as soon as a Device Policy attribute is added as a condition, users lose access to all the SSO apps on iOS and Android such as Slack, and Zoom.
I am trying to enable context aware so that employees can access google SAML apps from the company deployed mac laptops. I know there are other products that do this and probably do it better but upper management is forcing me to go in this direction. Here is where I am stuck. I have Enabled Context aware > browsers are managed > Endpoint verification enabled > endpoint verification on the browser and syncs with no errors.
Now the part that I am stuck on is how do I create a policy where it will block anyone without the managed browser or endpoint verification so they can not use the SMAL product
ex. we use SMAL SSO for salesforce. we want the user to be able to use it on the company device we have pushed out the config to but not on the home computer we have no control over.
Is there a way to disable the new context-aware suggestions and drop-down menu in ST4 to make it behave more like ST3?
In ST3, pressing tab autocompletes the word/snippet, and I could press tab again to cycle through the different words/snippets.
In ST4, pressing tab brings a drop-down menu of words/snippets and I can no longer cycle through the different words/snippets. Instead I have to manually select the desired word/snippet from the menu and press tab/enter to insert it.
I’ve tried playing around with the preferences/settings and reading the Completions documentation but to no avail. Any help is much appreciated!
I'm starting to play around with it. My super basic question is that if there is no policies set up does it just default to doing nothing? IE, it won't block access to something unless I specifically ask it to block something? I don't need to make a policy allowing people to access things like they already do before turning it on? I can make policies to allow certain access and deny certain access by by default is it allowed unless blocked or blocked unless allowed?
"Verify Context-Aware Access is “ON for everyone”. If not, click Turn On." By itself this does nothing then right?
Sorry for asking it a few different ways, I just don't want to tank the whole company by turning something on and the instructions make this step look super easy and I'm not sure what it actually does.
The initial thing I want to configure is just blocking some of our generic accounts used in classrooms from being used on any non-company owned devices. But it also just lists apps to block or allow and I'm wondering if one of those (gmail?) is the logon feature for Chromebooks or GCPW. Or would it let you on but block everything from working if on a non company owned device?
It's a new feature and I'm more afraid of breaking things than asking dumb questions.. ;)
Yeah, as the title says. In this game today he has made several comments about how he looks for the aggression coming into the league, whether the rookies are running at or away from being top picks, how he would love no crowds because he would call you out on mistakes all the way down the floor. And also I remember him fading when it mattered.
Don’t get me wrong, he was so gifted on the floor and was a top 3 or 4 active big man for most of his career and his stats are awesome. But as Bill said, his narrative as a guy criticising young players, and giving the impression that he knows “the secret”, changes how he is perceived by newer fans.
I also admit it isn’t really fair, other ex-players who weren’t as good say similar things (I would argue better/less aggressive and with humour). But I think when they do it’s easier to get a accurate context, because they didn’t do all the amazing things that he did on the floor and have his numbers. That’s not his fault, I guess it’s the context of the closer you get the tougher you are on someone (like losing in the 2nd round by a superior team hurts fans less than in the finals when you were up 3-2).
I know many of you hate Bill, and I take him with a boulder of salt, but he was right on this one.
All that said, I think he looks great with the grey coming into his beard and he talks too much, and says too little.
Thanks!
🎯 At a glance:
Is it possible to create 3d photos with convincing parallax effects from single RGB-D images? It is now! Check out a new 3D inpainting method proposed by Meng-Li Shih and colleagues. In short, the input image is transformed into a Layered Depth Image with explicit pixel connectivity, which is used to synthesize new local color-and-depth content into the occluded regions in a spatial context-aware manner. The resulting images can be rendered with a smooth parallax effect using standard graphics engines with fewer artifacts compared to current SOTA methods.
🚀 Motivation:
3D photos are more immersive than 2D, especially in VR. However, complex hardware setups are required to produce such images, and current methods that synthesize 3D photos from images captured with multi-lens smartphone cameras either produce gaps or distortions in the regions, occluded in the input image. Recent methods used Multi-Plane Image representation to address these issues, however they tend to produce artifacts on sloped surfaces. Instead of using rigid layers such as in Layered Depth Images (LDI), the authors explicitly store pixel connectivity and recursively apply CNN-based inpainting conditioned on spatially-adaptive context regions that are extracted from local connectivity in the LDI. The result is an algorithm for 3D photo generation without a predetermined number of depth layers.
Read the full paper digest or the blog post (reading time ~5 minutes) to learn about the modified LDI, Image Preprocessing, Context and Synthesis Regions, and Context-Aware Color and Depth Inpainting.
Meanwhile, check out the paper digest poster by Casual GAN Papers!
[Full Explanation Post / Blog Post] [Arxiv] [Code]
More recent popular computer vision paper breakdowns:
>[[SimSiam](https://www.ca
... keep reading on reddit ➡
