Ubuntu Talks Up Faster KDE Snaps, But Still Takes A While For Cold Apps To Launch
Back in late 2020 it was originally talked about for Canonical's effort around Snap packages to switch to LZO compression for faster start-up times. Today they published a new blog post on the Ubuntu site highlighting the Snap speed-ups while looking at KDE packages. The LZO-compressed packages are faster than XZ indeed, but still rather a lengthy start-up time for cold apps.
Today's Ubuntu blog post boasts about the advantages of using the LZO algorithm for package compression compared to the existing XZ usage. By regenerating the Snaps using LZO, some seconds can be shaved off from the cold start time and even greater gains if LZO compressing the KDE Frameworks Snap as well.
The hot startup times for when once a program has already loaded remain quite low at around ~1.1 seconds. KBlocks and GCompris were among the programs tested.
The GCompris educational software took 14.5 seconds to start-up cold when XZ compressed and XZ'ed KDE Frameworks. But when switching to LZO-compressed GCompris it dropped to 9.7 seconds or down to 6.3 seconds when also LZO'ing KDE Frameworks. Going from 14.5 to 6.3 seconds with both changes is certainly a noticeable different, but still a rather significant amount of time to see an educational software window when launching that application for the first time. The hot start-up time was consistent at 1.1 seconds.
In the case of KBlocks as a falling blocks game, it went from 6.3~7.9 seconds down to 3.7 seconds when making use of LZO Snaps for both the application and frameworks components. The hot start-up time for KBlocks remains around 1.1 seconds.
See this Ubuntu blog post for more details on the LZO Snap compressing for improving cold start-up times. It's certainly a move in the right direction but still a noticeable amount of time for cold starting a Snap application compared to not using Snaps. The post concludes with, "The initial findings from the first set of newly built KDE snaps utilizing the LZO compression algorithm point favorably to this change. The KDE team will be adopting the measure for their entire set of 100+ snaps available in the Snap Store, rolling the update alongside other changes and fixes in their applications."
Today's Ubuntu blog post boasts about the advantages of using the LZO algorithm for package compression compared to the existing XZ usage. By regenerating the Snaps using LZO, some seconds can be shaved off from the cold start time and even greater gains if LZO compressing the KDE Frameworks Snap as well.
The hot startup times for when once a program has already loaded remain quite low at around ~1.1 seconds. KBlocks and GCompris were among the programs tested.
The GCompris educational software took 14.5 seconds to start-up cold when XZ compressed and XZ'ed KDE Frameworks. But when switching to LZO-compressed GCompris it dropped to 9.7 seconds or down to 6.3 seconds when also LZO'ing KDE Frameworks. Going from 14.5 to 6.3 seconds with both changes is certainly a noticeable different, but still a rather significant amount of time to see an educational software window when launching that application for the first time. The hot start-up time was consistent at 1.1 seconds.
KBlocks
In the case of KBlocks as a falling blocks game, it went from 6.3~7.9 seconds down to 3.7 seconds when making use of LZO Snaps for both the application and frameworks components. The hot start-up time for KBlocks remains around 1.1 seconds.
See this Ubuntu blog post for more details on the LZO Snap compressing for improving cold start-up times. It's certainly a move in the right direction but still a noticeable amount of time for cold starting a Snap application compared to not using Snaps. The post concludes with, "The initial findings from the first set of newly built KDE snaps utilizing the LZO compression algorithm point favorably to this change. The KDE team will be adopting the measure for their entire set of 100+ snaps available in the Snap Store, rolling the update alongside other changes and fixes in their applications."
76 Comments