Never thought I would see a frontend eng team celebrate shipping a static marketing site with 847kb of JS.
Until I saw what it used to be…
notion.so/blog/migrating…
Until I saw what it used to be…
notion.so/blog/migrating…
I'm no perf expert or perf shamer (let he who is without sin...)
But here's what I do believe:
1. local Lighthouse runs are not credible. Use webpagetest or web.dev/measure.
2. you do NOT have to use the same tech for app and site! :)
But here's what I do believe:
1. local Lighthouse runs are not credible. Use webpagetest or web.dev/measure.
2. you do NOT have to use the same tech for app and site! :)
https://twitter.com/swyx/status/1317635707838496768?lang=en
in fact i’m -so- not a perfshamer I shall now say the thing I’m not supposed to say:
before today you didnt know or care that Notion had a 9.1MB JS marketing site; if it impacted them enough they would’ve fixed it sooner; you may have a fast site but they have 10 million users.
before today you didnt know or care that Notion had a 9.1MB JS marketing site; if it impacted them enough they would’ve fixed it sooner; you may have a fast site but they have 10 million users.
Lighthouse is great but the fact that you can get such wildly different results on the exact same site is just actively counterproductive.
The default experience has you “holding it wrong” and none of the impt info is in the screenshot. Dear @ChromeDevTools, you CAN fix this.
The default experience has you “holding it wrong” and none of the impt info is in the screenshot. Dear @ChromeDevTools, you CAN fix this.
• • •
Missing some Tweet in this thread? You can try to
force a refresh
