Uglifyjs: Time for new release?

[kzc]

Last release was 3 months ago. More than a dozen bug fixes have been made since that time.

[avdg]

I actually lately realize we hit 2.6.2 (ecma 262). Now we have to break it

[kzc]

@mishoo Would you consider adding @rvanvelzen as an npm module owner of uglify-js so he can do releases?

[kzc]

ping @mishoo

It would be regrettable if this project has to be forked in order to get timely releases out.

[mishoo]

I apologize for my lack of involvement lately. I simply don't have enough time to keep track of what's going on. I'd like to be notified somehow about show-stopping bugs or breaking changes, but other than that… as far as I am concerned this software is “done” (and by that I don't mean “dead”, but “does the job as reasonably as possible, and it's stable” — that's how I'd like it to stay).

I'm not against forks, but I'd prefer to keep it a single package if possible. I'll add @rvanvelzen as maintainer on NPM, if he doesn't mind, and he'll be able to make releases. @kzc you're welcome to join the board with commit rights, if you want to, please let me know. I'd just want the bold rule above to be the guiding plan. ;-)

[avdg]

I wonder if we need some beta's or at least some notice to release, specifically given #1111 having experienced some hidden bugs.

[kzc]

A beta doesn't work too well in the npm system but there can be a code freeze prior to a release - a week should be enough. Most people don't use the code until release and that's when bugs are usually found.

[avdg]

yeap, explaining all the crap on social media as well

[rvanvelzen]

@mishoo Feel free to add me on NPM. I think it should be fine.

[rvanvelzen]

I'll push a new release tomorrow!

[avdg]

I'm running es 5.1 tests with uglify on test262 on the background now, bugs should be really really low now.

[kzc]

@rvanvelzen Nice work on adding the new tags on the issues and cleaning up stale PRs.

[avdg]

I have yet to see what happened, but after a quick look in the logs it looks like the tests involved with with had failed. Tests are still running though, but I'll investigate these failing tests already.

[avdg]

Hmm can't reproduce the fail on other tests, I hope it's the setup...

[avdg]

On the upside, there are less bugs on the list. There are still some things that can be improved but yeah, with incremental steps.

[kzc]

@avdg What are the pass/fail numbers for test262 on master at this time?

[avdg]

https://gist.github.com/avdg/c7dadf3d4528e734360749009501e370

• Not sure why ch 12 fails (with expr) but 2 fails less
• Will fix 7.8.5 tomorrow
• Best practice isn't supported in es6

[kzc]

@avdg I see only 4 failures in strict mode. Am I interpreting the results correctly?

[avdg]

Which results?

[kzc]

https://gist.github.com/avdg/c7dadf3d4528e734360749009501e370

Four occurrences of "failed in strict mode"

[avdg]

Search on "Caused by uglify"

[kzc]

I see - 375 failures over 11267 tests whereof 45 caused by UglifyJS

[avdg]

Some none striped test are also caused by uglify but in this log they are the ones mentioned few comments above.

[avdg]

And there are multiple results by searching on that, note that the tests have a ) behind. Adding that to search and you only have the failing tests.

[avdg]

Well most test have a ) behind. Some say caused by node and uglify.

[rvanvelzen]

I don't think there's anything breaking in this release, so shall I tag/publish 2.6.3?

[avdg]

Let's keep moving, it's been a while, I guess.

[rvanvelzen]

Just published 2.6.3 to npm.