AriaMinaei/pretty-error

Error caused by renderkid@0.2.0-beta.11

Closed this issue · 5 comments

danielscw commented

After reinstall the pretty-error module and upgrade renderkid from 0.2.0-beta.10 to 0.2.0-beta.11, we got the following error.

module.js:340
    throw err;
    ^
Error: Cannot find module './block/config'
    at Function.Module._resolveFilename (module.js:338:15)
    at Function.Module._load (module.js:280:25)
    at Module.require (module.js:364:17)
    at require (module.js:380:17)
    at Object.<anonymous> (/opt/app/node_modules/pretty-error/node_modules/renderkid/lib/layout/Block.js:4:15)
    at Module._compile (module.js:456:26)
    at Object.Module._extensions..js (module.js:474:10)
    at Module.load (module.js:356:32)
    at Function.Module._load (module.js:312:12)
    at Module.require (module.js:364:17)
    at require (module.js:380:17)
    at Object.<anonymous> (/opt/app/node_modules/pretty-error/node_modules/renderkid/lib/Layout.js:4:9)

Downgrade the renderkid module solve the problem.

floross commented

+1

whyhankee commented

+1, on node v0.10.33

robertmain commented

Yep, broken here too. It just made a massive clusterfuck of my dev environment. Why is renderkid beta included in what should(according to semver) be a stable release?

AriaMinaei commented

Have seen this weird bug before. For some reason, the last prepublish script had compiled src/layout/block/config.coffee to src/layout/block/Config.js (config vs Config).

Funny enough, the file is actually named config.coffee, both in git and in my local copy.

Anyway, it looks like my last npm publish for some reason contained a Config.js instead of config.js, and after that, well, all hell broke loose.

This wasn't caught by RenderKid's travis, because travis reads files from git and not from npm, and git's files were actually correct. The only way to know things didn't work was to re-run pretty-error's test-cases again on travis, and that didn't occur to me because RenderKid's test-cases were passing.

I just pushed a dumb new version with a fixed filename and now it looks like pretty-error's cases are passing now.

Why is renderkid beta included in what should(according to semver) be a stable release?

You are right. pretty-error is stable and it should depend on stable packages. But RenderKid has been stable ever since pretty-error became stable. I just didn't reflect that in its version string.

robertmain commented

Ah, Gotcha! :-)
On 17 Feb 2015 16:37, "Aria" notifications@github.com wrote:

Have seen this weird bug before. For some reason, the last prepublish
script had compiled src/layout/block/config.coffee to
src/layout/block/Config.js (config vs Config).

Funny enough, the file is actually named config.coffee, both in git and
in my local copy.

Anyway, it looks like my last npm publish for some reason contained a
Config.js instead of config.js, and after that, well, all hell broke
loose.

This wasn't caught by RenderKid's travis, because travis reads files from
git and not from npm, and git's files were actually correct. The only way
to know things didn't work was to re-run pretty-error's test-cases again on
travis, and that didn't occur to me because RenderKid's test-cases were
passing.

I just pushed a dumb new version with a fixed filename and now it looks
like pretty-error's cases are passing now.

Why is renderkid beta included in what should(according to semver) be a
stable release?
You are right. pretty-error is stable and it should depend on stable
packages. But RenderKid has been stable ever since pretty-error became
stable. I just didn't reflect that in its version string.

—
Reply to this email directly or view it on GitHub
#13 (comment)
.