IssueTracker

Also crashes in R8 version 8.5.35, 8.6.17, 8.7.2-dev

You cannot apply keep rules for classes synthesized by R8, but if you keep the implemented interfaces, the merging should not happen.

That said, checking the implementing class of a lambda might not be a good idea. The specification for java.lang.invoke.LambdaMetafactory says:

Capture may involve allocation of a new function object, or may return a suitable existing function object. The identity of a function object produced by capture is unpredictable, and therefore identity-sensitive operations (such as reference equality, object locking, and System.identityHashCode()) may produce different results in different implementations, or even upon different invocations in the same implementation.

If I understand this correctly the only guarantee is that the function object implements the interface(s) specified.

Not sure what the use case is here. However, if you want the interfaces to stay as is you will have to have a -keep rule for them. One option to handle this without adding a rule for each interface could be to annotate them and add a single rule for all interfaces annotated by that annotation.

Also, is there a way to disable this optimization only for my classes? Or can I only remove all optimizations via -keep, allowobfuscation, allowshrinking class_names?

Phoenix is losing its June 2017 Gen 4 coverage right now even though it wasn't made part of the 9/10/2021 update. This is not good.

I'm still awaiting confirmation. Is the lost data being processed?

(Replying to comment #20) 'Fix'? What makes you think you think you can fix it? You don't seem to work at Google, as I can tell from your email address which ends in @gmail.com, not @google.com--hence I don't think you know too much about the technical stuff developer-wise on Maps. :\

However, I still do want to ask the folks at Google: exactly how long is it going to take to revive the data?

Attached here is an example screenshot of Japan's first Gen 1 coverage, since the link leading to that imagery recently had its pano IDs expired. This one is dated February 2008.

Alright... I seem to have described it incorrectly. Let me start over this time. Ahem... well, you see, the 9/10/2021 Street View data provided very important updates to roads that were heavily outdated and in poor quality (a notable example of this is a stretch of SR-2 in the Angeles National Forest, which was stuck in 2009 Gen 2-quality photos for years before it later gained Gen 4-quality photos from 2017), as well as to add additional coverage and provide more historical records for research. For these reasons, the unexplained removal should have NEVER happened. I mean, the imagery shouldn't be that problematic to Google.

(Replying to comment #25) Just relax and be patient. They're still looking into it, so hopefully sooner or later the folks at Google will be able to identify the problem and take action to restore the data afterwards. But for now we might need to wait a bit longer.

Attached here are some example screenshots of Gen 3 imagery released in September 2021 in the US. Several also existed in other countries, but they are not included here since the imagery no longer exists and I did not keep any links with me before the imagery was removed on March 1.

Attached here are some example screenshots of Gen 4 imagery released in September 2021, in the Bay Area.

Attached here are some example screenshots of Gen 4 imagery released in September 2021, in the Dallas-Fort Worth metroplex.

Attached here are some example screenshots of Gen 4 imagery released in September 2021, in the Houston area.

Attached here are a few example screenshots of Gen 4 imagery released in September 2021, in Los Angeles.

Attached here are a few example screenshots of Gen 4 imagery released in September 2021, in Miami.

Attached here are a few example screenshots of Gen 4 imagery released in September 2021, in the Orlando area.

[Replying to comment #35] I understand you want to see the 9/10/2021 data, but unfortunately, according to Google's own contribution policy, their Street View imagery only shows where their vehicles have driven by, and it can take months to process them. (However, this is not the case with continuous updates in some countries, as it usually takes less than a week to publish them.) So, it's very unlikely the data will ever return until possibly later in 2023, hoping they've paid attention to the warning I wrote in comment #1. Plus, it's been two months since I submitted this report, and I haven't seen anything reappear to this point.

[Writing to the assignee] I'm still awaiting an update. Without the 9/10/2021 Street View data, some roads and areas will continue to lack new (and especially updated) coverage and there will not be enough historical records for research.

(Replying to comment #38) Would you please stop spamming? You’re tricking me into thinking that the assignee has posted an update when I know for a fact that he has not gone back to me as of now. So please stop. You are not helping here.

Now a bigger problem: after one year of their release, the 9/10/2021 Street View data is no longer accessible through the API, which Google Maps utilizes. The imagery's pano IDs are to blame because they expired. So, if the data were to be rereleased, they would need to have different pano IDs since the numbers for imagery are randomized.

(Replying to comment #42) It’s the pano IDs that expired, months after their blue lines were removed. Imagery taken between 2008 and 2021 is what we’re on focus of, and if it was still around, you could find them anywhere throughout the world. Let’s recap from comment #1.

• In the US the update included a mixture of Gen 3 and Gen 4 imagery from the past — most notably the addition of more Gen 4 content from 2017, in which the year had captures between the months of March and September. If you were looking at Salt Lake City, you may have also witnessed Gen 3 content from 2020.
• In Japan, the update brought back the country’s first Gen 1 coverage, which had been largely absent since 2012.
• The update also added captures done with a third-party camera system in an area of Bangladesh. It was the same camera system used to update parts of Alaska in 2019.
• A small number of rural areas in Germany also saw new coverage in the update.

You can refer to the screenshots I attached on this issue for further reference.

Adding on to my previous comment, if you click on the links leading to the 9/10/2021 data on comment #1 past their expiration dates, you’ll get a ‘No Street View imagery available here’ error and the API will instead return the closest (or latest) image available, sometimes a user-contributed photosphere.

(Replying to comment #45) Look at the following screenshot. I think you’re confusing it with continuous updates being enabled in some countries since the update (in other words, some countries receive weekly Street View updates depending on the feasibility). Here in the screenshot, visible 2021 copyright watermarks and the lack of blue lines on the map being hovered over on the bottom-left corner are good indicators that the image was released on September 10, 2021. The screenshot was taken past the removal date (March 1, 2022), and with their pano IDs still valid at that time, the imagery was accessed through a special link.

Alright, I’m beginning to suspect that insufficient densities could be a probable reason to why the whole data got removed in March. However, it didn’t just apply to this particular update, but some countries in Africa and Europe may have also faced the same fate in terms of coverage. And that’s probably why Vietnam lost its Google-contributed Street View data at the time of the update.

The 9/10/2021 Street View release data still remains inaccessible. Isn’t there supposed to be a status update here?

Augh, misunderstood again! Ahem... well, you see, there is a feature in Maps that allows you to view historical captures of the same road, which they call the Time Machine. In Street View where it’s available, click on the capture date on the top-right corner, and you’ll see a timeline of captures, starting from the oldest (2007) to the latest (this year). That’s where the data is missing.

To say this once more, if you click on the links in comment #1 today, Street View will load but they fail with a ‘No Street View imagery available here’ warning, subsequently returning an image different than what was expected. That’s because the imagery’s pano IDs expired months after they were removed, and the removal is likely because of insufficient densities at the time of their release. If you open the Time Machine, you’ll find that the expected photo is not present at all.

Unfortunately, this is taking too long. The data remains largely absent, and if it's unlikely to happen then I don't know what else to say. The data... to be honest, it's really important to me. :(

na...@gmail, I want to make sure you understand. Do you recall the 10/8/2020 update? It was released to address data accuracy issue in some California and Texas cities and divisions because of which, in my opinion, updates were blocked due to elevated privacy concerns and all imagery was limited to a maximum capture date of June 2017 or January 2018. Those concerns have since de-escalated during the course of the pandemic (probably as a result of mandatory lockdowns), and the update was released to allow all future data to roll out and ensure information accuracy. For these reasons, the 9/10/2021 update is closely related to that update, and unlike it, data from the 10/8/2020 update didn't get deleted afterwards. And because the 9/10/2021 update was intended to provide more historical records for research and provide quality updates to some existing coverage, data from that update should have also not been deleted in the first place.

(Replying to comment #53) Hmm, I don't think you're familiar with what I'm trying to explain. If it's because I'm being unclear, you should reread my previous comments again and again. You seem to be having trouble understanding.

If you click on the capture date on the top-right corner while viewing Street View on desktop Maps, a timeline of historical captures appears. On mobile Maps, when viewing Street View, you tap 'See more dates' under the street name to access the 'Time Machine', and a similar timeline appears. If you've never used this feature, you probably do not have any experience in historical imagery.

Quoting from comment #53: "The 10/8/2020 update? The update that added gen 3 to two random American cities? No, I don't remember that."

Well, then, why not take a look at this yourself: https://goo.gl/maps/cbaSs6NQ7uEyoERq5. It's one of those images that were released in October 2020, and it was captured by a vehicle whose imagery was never published elsewhere. And if no removal occurred, it means that Google could unofficially publish test imagery after a long while.

” our Play Kits and Playthings are Montessori-inspired