names_->GetFormatted("%s (C++)", name),Why "C++"? I think it's confusing to add the language in which the object is implemented to this. This will be exposed very broadly: What are users getting from the fact that this is C++? OR rather, what is C++ actually? Our own types are also implemented in C++.
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
names_->GetFormatted("%s (C++)", name),Why "C++"? I think it's confusing to add the language in which the object is implemented to this. This will be exposed very broadly: What are users getting from the fact that this is C++? OR rather, what is C++ actually? Our own types are also implemented in C++.
"C++" is short, and it shows that it's not a JS object. It shows a JS developer that objects marked with "C++" are objects which they should probably be ignored. If these objects are leaking, then it's probably complicated, and it's better to look first for a leak without the "C++" tag.
That being said, I don't feel strongly about the name. Alternatives that come to mind could be "internal", "system", "native". What would be your preference?
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
| Code-Review | +1 |
Still LGTM! I am also okay with any suffix you come up with here.
names_->GetFormatted("%s (C++)", name),Andreas HaasWhy "C++"? I think it's confusing to add the language in which the object is implemented to this. This will be exposed very broadly: What are users getting from the fact that this is C++? OR rather, what is C++ actually? Our own types are also implemented in C++.
"C++" is short, and it shows that it's not a JS object. It shows a JS developer that objects marked with "C++" are objects which they should probably be ignored. If these objects are leaking, then it's probably complicated, and it's better to look first for a leak without the "C++" tag.
That being said, I don't feel strongly about the name. Alternatives that come to mind could be "internal", "system", "native". What would be your preference?
I don't have a strong opinion here. Wdyt about cppgc? That would tell us immediately that this is C++ and Oilpan managed objects.
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
names_->GetFormatted("%s (C++)", name),Andreas HaasWhy "C++"? I think it's confusing to add the language in which the object is implemented to this. This will be exposed very broadly: What are users getting from the fact that this is C++? OR rather, what is C++ actually? Our own types are also implemented in C++.
Dominik Inführ"C++" is short, and it shows that it's not a JS object. It shows a JS developer that objects marked with "C++" are objects which they should probably be ignored. If these objects are leaking, then it's probably complicated, and it's better to look first for a leak without the "C++" tag.
That being said, I don't feel strongly about the name. Alternatives that come to mind could be "internal", "system", "native". What would be your preference?
I don't have a strong opinion here. Wdyt about cppgc? That would tell us immediately that this is C++ and Oilpan managed objects.
It shows a JS developer that objects marked with "C++" are objects which they should probably be ignored
Anything with `GetHumanReadableName()` override would also get this annotation, right? These are objects that have been exposed to JS developers for many years and are certainly something they care about.
In fact, the biggets footgun are event listeners and they are all implemented in C++ in Blink.
I like `cppgc` as it's somewhat a middle ground but again: Adding this for all objects will immediately raise the question of "what is this" by the non-expert devs.
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
names_->GetFormatted("%s (C++)", name),Andreas HaasWhy "C++"? I think it's confusing to add the language in which the object is implemented to this. This will be exposed very broadly: What are users getting from the fact that this is C++? OR rather, what is C++ actually? Our own types are also implemented in C++.
Dominik Inführ"C++" is short, and it shows that it's not a JS object. It shows a JS developer that objects marked with "C++" are objects which they should probably be ignored. If these objects are leaking, then it's probably complicated, and it's better to look first for a leak without the "C++" tag.
That being said, I don't feel strongly about the name. Alternatives that come to mind could be "internal", "system", "native". What would be your preference?
Michael LippautzI don't have a strong opinion here. Wdyt about cppgc? That would tell us immediately that this is C++ and Oilpan managed objects.
It shows a JS developer that objects marked with "C++" are objects which they should probably be ignoredAnything with `GetHumanReadableName()` override would also get this annotation, right? These are objects that have been exposed to JS developers for many years and are certainly something they care about.
In fact, the biggets footgun are event listeners and they are all implemented in C++ in Blink.
I like `cppgc` as it's somewhat a middle ground but again: Adding this for all objects will immediately raise the question of "what is this" by the non-expert devs.
What if we don't do this change for all objects but just for the "Window" object? This was my original objective. Igor suggested to add the postfix to all objects, and I did that. But if that's controversial, then I can just go back and just change the name of the "Window" object. I can then name it "Window (cppgc)".
It's not very descriptive for a non-expert dev, but on the other hand, if that's the only object that is leaking, it's nothing a non-expert dev can handle anyways.
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
names_->GetFormatted("%s (C++)", name),Andreas HaasWhy "C++"? I think it's confusing to add the language in which the object is implemented to this. This will be exposed very broadly: What are users getting from the fact that this is C++? OR rather, what is C++ actually? Our own types are also implemented in C++.
Dominik Inführ"C++" is short, and it shows that it's not a JS object. It shows a JS developer that objects marked with "C++" are objects which they should probably be ignored. If these objects are leaking, then it's probably complicated, and it's better to look first for a leak without the "C++" tag.
That being said, I don't feel strongly about the name. Alternatives that come to mind could be "internal", "system", "native". What would be your preference?
Michael LippautzI don't have a strong opinion here. Wdyt about cppgc? That would tell us immediately that this is C++ and Oilpan managed objects.
Andreas HaasIt shows a JS developer that objects marked with "C++" are objects which they should probably be ignoredAnything with `GetHumanReadableName()` override would also get this annotation, right? These are objects that have been exposed to JS developers for many years and are certainly something they care about.
In fact, the biggets footgun are event listeners and they are all implemented in C++ in Blink.
I like `cppgc` as it's somewhat a middle ground but again: Adding this for all objects will immediately raise the question of "what is this" by the non-expert devs.
What if we don't do this change for all objects but just for the "Window" object? This was my original objective. Igor suggested to add the postfix to all objects, and I did that. But if that's controversial, then I can just go back and just change the name of the "Window" object. I can then name it "Window (cppgc)".
It's not very descriptive for a non-expert dev, but on the other hand, if that's the only object that is leaking, it's nothing a non-expert dev can handle anyways.
How does this work with merged nodes? I have a hard time following what this will do for wrapper/wrappable pairs.
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
To unsubscribe from this group and stop receiving emails from it, send an email to devtools-revie...@chromium.org.
names_->GetFormatted("%s (C++)", name),Andreas HaasWhy "C++"? I think it's confusing to add the language in which the object is implemented to this. This will be exposed very broadly: What are users getting from the fact that this is C++? OR rather, what is C++ actually? Our own types are also implemented in C++.
Dominik Inführ"C++" is short, and it shows that it's not a JS object. It shows a JS developer that objects marked with "C++" are objects which they should probably be ignored. If these objects are leaking, then it's probably complicated, and it's better to look first for a leak without the "C++" tag.
That being said, I don't feel strongly about the name. Alternatives that come to mind could be "internal", "system", "native". What would be your preference?
Michael LippautzI don't have a strong opinion here. Wdyt about cppgc? That would tell us immediately that this is C++ and Oilpan managed objects.
Andreas HaasIt shows a JS developer that objects marked with "C++" are objects which they should probably be ignoredAnything with `GetHumanReadableName()` override would also get this annotation, right? These are objects that have been exposed to JS developers for many years and are certainly something they care about.
In fact, the biggets footgun are event listeners and they are all implemented in C++ in Blink.
I like `cppgc` as it's somewhat a middle ground but again: Adding this for all objects will immediately raise the question of "what is this" by the non-expert devs.
Michael LippautzWhat if we don't do this change for all objects but just for the "Window" object? This was my original objective. Igor suggested to add the postfix to all objects, and I did that. But if that's controversial, then I can just go back and just change the name of the "Window" object. I can then name it "Window (cppgc)".
It's not very descriptive for a non-expert dev, but on the other hand, if that's the only object that is leaking, it's nothing a non-expert dev can handle anyways.
How does this work with merged nodes? I have a hard time following what this will do for wrapper/wrappable pairs.
I added printf's in `MergeNames` to see the input and output names for all merged nodes, and "(C++)" did not end up in any merged names. I'm not 100% sure how the data flow is there, but it did not change node merging.
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
names_->GetFormatted("%s (C++)", name),Andreas HaasWhy "C++"? I think it's confusing to add the language in which the object is implemented to this. This will be exposed very broadly: What are users getting from the fact that this is C++? OR rather, what is C++ actually? Our own types are also implemented in C++.
Dominik Inführ"C++" is short, and it shows that it's not a JS object. It shows a JS developer that objects marked with "C++" are objects which they should probably be ignored. If these objects are leaking, then it's probably complicated, and it's better to look first for a leak without the "C++" tag.
That being said, I don't feel strongly about the name. Alternatives that come to mind could be "internal", "system", "native". What would be your preference?
Michael LippautzI don't have a strong opinion here. Wdyt about cppgc? That would tell us immediately that this is C++ and Oilpan managed objects.
Andreas HaasIt shows a JS developer that objects marked with "C++" are objects which they should probably be ignoredAnything with `GetHumanReadableName()` override would also get this annotation, right? These are objects that have been exposed to JS developers for many years and are certainly something they care about.
In fact, the biggets footgun are event listeners and they are all implemented in C++ in Blink.
I like `cppgc` as it's somewhat a middle ground but again: Adding this for all objects will immediately raise the question of "what is this" by the non-expert devs.
Michael LippautzWhat if we don't do this change for all objects but just for the "Window" object? This was my original objective. Igor suggested to add the postfix to all objects, and I did that. But if that's controversial, then I can just go back and just change the name of the "Window" object. I can then name it "Window (cppgc)".
It's not very descriptive for a non-expert dev, but on the other hand, if that's the only object that is leaking, it's nothing a non-expert dev can handle anyways.
Andreas HaasHow does this work with merged nodes? I have a hard time following what this will do for wrapper/wrappable pairs.
I added printf's in `MergeNames` to see the input and output names for all merged nodes, and "(C++)" did not end up in any merged names. I'm not 100% sure how the data flow is there, but it did not change node merging.
Okay, then lets use "(cppgc)" for now. I am anyways not sure the merging is great at this point.
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
names_->GetFormatted("%s (C++)", name),Andreas HaasWhy "C++"? I think it's confusing to add the language in which the object is implemented to this. This will be exposed very broadly: What are users getting from the fact that this is C++? OR rather, what is C++ actually? Our own types are also implemented in C++.
Dominik Inführ"C++" is short, and it shows that it's not a JS object. It shows a JS developer that objects marked with "C++" are objects which they should probably be ignored. If these objects are leaking, then it's probably complicated, and it's better to look first for a leak without the "C++" tag.
That being said, I don't feel strongly about the name. Alternatives that come to mind could be "internal", "system", "native". What would be your preference?
Michael LippautzI don't have a strong opinion here. Wdyt about cppgc? That would tell us immediately that this is C++ and Oilpan managed objects.
Andreas HaasIt shows a JS developer that objects marked with "C++" are objects which they should probably be ignoredAnything with `GetHumanReadableName()` override would also get this annotation, right? These are objects that have been exposed to JS developers for many years and are certainly something they care about.
In fact, the biggets footgun are event listeners and they are all implemented in C++ in Blink.
I like `cppgc` as it's somewhat a middle ground but again: Adding this for all objects will immediately raise the question of "what is this" by the non-expert devs.
Michael LippautzWhat if we don't do this change for all objects but just for the "Window" object? This was my original objective. Igor suggested to add the postfix to all objects, and I did that. But if that's controversial, then I can just go back and just change the name of the "Window" object. I can then name it "Window (cppgc)".
It's not very descriptive for a non-expert dev, but on the other hand, if that's the only object that is leaking, it's nothing a non-expert dev can handle anyways.
Andreas HaasHow does this work with merged nodes? I have a hard time following what this will do for wrapper/wrappable pairs.
Michael LippautzI added printf's in `MergeNames` to see the input and output names for all merged nodes, and "(C++)" did not end up in any merged names. I'm not 100% sure how the data flow is there, but it did not change node merging.
Okay, then lets use "(cppgc)" for now. I am anyways not sure the merging is great at this point.
To summarize, I will change the postfix to "(cppgc)", and only append it to the "Window" object, as for the other objects it does not provide so much value.
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
| Code-Review | +1 |
Still LGTM
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
To unsubscribe from this group and stop receiving emails from it, send an email to devtools-revie...@chromium.org.
| Code-Review | +1 |
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
Reland "[profiler] Add postfix to C++ objects in DevTools heap snapshot"Please update the comment. This is not a reland this is a completely different CL. Also the suffix you add is different.
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |