Adobe Brackets

» see brackets.io as it lived, on the wayback machine

~ the obituary ~

brackets showed up around 2012 as an adobe research project, then shipped 1.0 in november 2014. the pitch was that it was a code editor written in the same web languages it edited, so a front-end developer could read and patch the editor itself. quick edit let you ctrl-click on a css class in markup and pop open the matching css rule inline. live preview pushed your edits to chrome over a websocket so you saw changes without saving and reloading. for a web designer in 2014 those were the right features.

the timing was wrong though. github's atom shipped publicly in february 2014 with the same pitch. visual studio code shipped in april 2015 on top of electron, with microsoft money and a focused team behind it. by 2017 the editor wars were over and vscode had won. brackets kept shipping releases, but the active contributor count drifted down through 2019 and 2020, and the bug backlog stopped getting addressed.

adobe announced end of support in march 2021, stopped contributing that same month, and archived the github repository on september 1 2021. the final official release was version 1.14.2. adobe's own blog post recommended migrating to visual studio code, which was an unusually direct corporate handoff to a competitor product owned by microsoft. the community forked the codebase under the brackets-continued banner and later as a fresh successor called phoenix code. both were still shipping in 2026.

i remember opening brackets in maybe 2015 when i was looking for something between sublime text and a heavier ide for some quick html work. the live preview was the thing that made me try it. the editor felt slow on big files in the way every electron-ish editor felt slow on big files, so i went back to sublime, then later vscode like the rest of the planet. brackets was a reasonable tool in a market that picked a winner fast.

~ the case file ~

~ what it got right ~

brackets understood that web designers are not the same audience as systems programmers. quick edit was a designer's feature: stay inside the page you are editing without losing your place. inline color pickers and integrated css validation pointed in the same direction. the editor leaned into being a frontend tool rather than trying to be a universal ide.

the extension system was small but real. theseus, the javascript debugger built by mit and brigham young university researchers, was bundled in. extract for brackets pulled css values out of photoshop psd files, which is the kind of feature only adobe would ship.

~ why vscode took it ~

vscode shipped with the same electron foundation but with microsoft's whole tooling effort behind it. typescript shipped first-class. the language server protocol turned every language into a plug-in target. brackets was a fixed-scope tool with a fixed audience, while vscode was a moving target adding capabilities every few weeks. open-source momentum followed the moving target.

by the time adobe sunset brackets, vscode had over 70% editor-market share in stack overflow's annual developer survey. there was no audience left to defend. the september 2021 archive was the formal end of a product that had already lost its users two years earlier.

~ the long tail ~

the community fork lives at brackets.io under the open source adobe banner, with volunteers shipping point releases. phoenix code, announced august 30 2021 by former brackets contributors, is the actually-active successor that runs in a browser rather than as an electron app. neither has meaningful market share. they exist because the people who liked brackets specifically did not want to use vscode, which is a fine reason for a fork to exist.

@}-->--

              )
             (_)
            .-'-.
            |   |
            |   |
            |   |
            |   |
          __|   |__   .-.
       .-'  |   |  `-:   :
      :     `---'     :-'
       `-._       _.-'
           '""""""

« previous  ·  random  ·  next »