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
In a large app with many deferred parts, the metadata for loading the deferred parts is about 1% of the total main part size.
The problem would be reduced by merging parts (#29572) but there is also a problem that the representation has redundancies.
There are two parallel maps (JavaScript objects) from library key to list of parts, one for file names, the other for hashes.
Many parts occur in several lists, duplicating the names and hashes. The average parts per library is ~15 parts. 330 parts are mentioned 1500 times in describing ~100 libraries.
I suggest a level of indirection.
One map from library key to list of part indexes
An array of part file names
An array of part hashes
This would reduce the current ~100kB to ~25kB.
The text was updated successfully, but these errors were encountered:
'Normalize' the data used for loading deferred libraries. Replace
URIs and SHA1 hashes with indexes so that the URIs and hashes occur
once.
Bug: #29635
Change-Id: I3ac8791ad44ca588cbb1bd7d86f18243c956a04a
Reviewed-on: https://dart-review.googlesource.com/32505
Commit-Queue: Stephen Adams <sra@google.com>
Reviewed-by: Sigmund Cherem <sigmund@google.com>
In a large app with many deferred parts, the metadata for loading the deferred parts is about 1% of the total main part size.
The problem would be reduced by merging parts (#29572) but there is also a problem that the representation has redundancies.
I suggest a level of indirection.
This would reduce the current ~100kB to ~25kB.
The text was updated successfully, but these errors were encountered: