-
-
Notifications
You must be signed in to change notification settings - Fork 2k
Your Gemfile.lock is corrupt. The following gem is missing from the DEPENDENCIES section
#4467
Comments
Can you please share the full output of |
The Debian package is used. My colleague will respond tomorrow with the exact output. |
Dear @segiddins, sorry for not reading your question properly. Here the output of
|
I cannot reproduce with the given Gemfile & Lockfile |
Have the same issue, here is the output:
|
Can this be caused due to the I have a Gemfile.lock created with bundler 1.12.1 on Windows. Then on Travis (Ubuntu)
On Windows things work fine. So, I look into my Gemfile.lock and see Thing is this worked fine before. https://github.com/mpc-hc/mpc-hc.org/blob/master/Gemfile.lock |
I hit the same problem in another Rails project. When running Adding the line manually to So it looks like, there is a bug somewhere, which is hard to reproduce though. |
In my case it's happening on the production environment, on development machine i don't have such an issue. It's not related to the platforms as @XhmikosR suggested. |
I am having the same issue (e.g. nanoc/nanoc.ws buid #873 on Travis CI). This issue only happens on Travis, but not on my local machine. |
Personally I managed to work around the issue by removing the jobs parameter. |
Also seeing this with |
Downgraded Bundler to 1.10.6, per rubygems/bundler#4467.
I can confirm that I am also seeing this with the ffi dependency, As @kianw mentioned, bundle 1.10.6 install works. |
For those of you seeing this problem -- are you using the git local feature? If so, this should be fixed by #4840 |
For AppVeyor it wasn't so simple https://ci.appveyor.com/project/AppVeyor-Website/website/build/1.0.183 I had to run install:
# Bundler downgrade should be removed when the issue is fixed upstream
- gem uninstall bundler -x
- gem install bundler -v "=1.10.6" And now it works again https://ci.appveyor.com/project/AppVeyor-Website/website/build/1.0.199 So this is definitely a bug in versions > 1.10.6. |
I've randomly run into both rubygems/bundler#4467 && rubygems/bundler#4576, mostly on jruby but also sometimes on MRI. This combination of bundler version and `bundle install` flags seems to workaround both.
I've randomly run into both rubygems/bundler#4467 && rubygems/bundler#4576, mostly on jruby but also sometimes on MRI. This combination of bundler version and `bundle install` flags seems to workaround both.
I've randomly run into both rubygems/bundler#4467 && rubygems/bundler#4576, mostly on jruby but also sometimes on MRI. This combination of bundler version and `bundle install` flags seems to workaround both.
I've randomly run into both rubygems/bundler#4467 && rubygems/bundler#4576, mostly on jruby but also sometimes on MRI. This combination of bundler version and `bundle install` flags seems to workaround both.
I've randomly run into both rubygems/bundler#4467 && rubygems/bundler#4576, mostly on jruby but also sometimes on MRI. This combination of bundler version and `bundle install` flags seems to workaround both.
I've randomly run into both rubygems/bundler#4467 && rubygems/bundler#4576, mostly on jruby but also sometimes on MRI. This combination of bundler version and `bundle install` flags seems to workaround both.
Any update on resolving this? We just ran into the issue in a deployment. |
This is an odd one. There are enough people in this thread, maybe the following doesn't cover all cases, but for the folks who provided TL;DR: Don't mix Bundler doesn't behave very well in these cases, so it would be preferable if it just complained that you're trying to change a frozen bundle, but the workaround / proper way to go here is to do a This bundler-cases PR shows a minimum case to reproduce the error: https://github.com/chrismo/bundler-cases/pull/10/files (this commit shows the original full Gemfile, but it takes a long time to run). Gritty de NittyBundler loads up a Definition instance that evaluates all of the dependencies, specifications, etc. found in the Gemfile and lockfile before doing anything. As part of this process, it does what becomes an odd thing if the
(Above is pasted from 1.13.0.rc.2, but important changes to above made for 1.9 in this commit). When remotely resolving, this code sees the frozen setting and decides it only needs to return the Presumably this code was never intended to be used when But wait, you say to yourself, this behavior has apparently been this way since 1.9, but the errors only started with 1.11? The error being reported in this issue didn't exist until 1.11, in this commit.
(Code pasted is latest from 1.13, but largely the same).
So - how does it work in 1.9 or 1.10? According to the bundler-case PR, it doesn't. It doesn't blow up or complain, but it doesn't actually perform an update:
The only way to get it to work is to first unfreeze the bundle:
|
i think here's another replication case: % git clone https://github.com/chef/ffi-yajl
% cd ffi-yajl
% bundle install --without development development_extras rbx
Fetching gem metadata from https://rubygems.org/........
Fetching version metadata from https://rubygems.org/..
Fetching dependency metadata from https://rubygems.org/.
Resolving dependencies...
Using libyajl2 1.2.0
Your Gemfile.lock is corrupt. The following gem is missing from the DEPENDENCIES
section: 'ffi'
doesn't involve --deployment or --no-deployment flags and is pretty simple. worked on older bundler or rubygems, but upgrading both to current versions caused the error to appear. i'm kinda slammed with other regressions right now and this is a bit of a flyby issue for me, but i thought i'd drop a note in here since at least it looks like a fairly simple replication case we found and seems a bit unrelated to the other use cases (the repo doesn't have a Gemfile.lock or .bundle/ or vendor/ directories). even though the gem is called 'ffi-yajl' the 'ffi' gem dep is optional and should only be a dev dep of the gem. VVVV @scotthain below actually found this one |
More debugging:
|
I'm seeing this issue with (I'm happy to provide more debugging info if that would be helpful. Thanks!) |
@trevorturk Thanks for reporting this! |
@indirect I believe the reproduce case I presented doesn't use |
@scotthain @lamont-granquist I've created a ticket for your specific issue at #5006. @trevorturk if you could open a new ticket following ISSUES and with repro steps, that would be helpful. |
@indirect thank you! Let us know if we can help test it. |
@indirect thanks for the quick reply! I wasn't able to work around my issue as you suggested by not using the My config/command/error was:
I dug around in my gems looking for anything mentioning ...so I'm wondering if there's a bug that's caused by a combination of Apologies for piling onto this thread -- I'm 100% happy to move this to a new issue and try to come up with a repro case, but I thought it worth tossing this out there since I was able to work around my issue (maybe others will as well?) and I think this seems like a potential lead...? Update: I confirmed that I can also work around my issue (without forking |
@trevorturk the specific issue of |
Wonderful, thank you @indirect! |
Closing this because all the examples I was able to test seem to be fixed by 1.13.2. If you are still seeing this error, start by trying Thanks! |
The error is still there. We are using jruby and ruby and we need to cache all the gems, so it might be related to the use of 'platforms :jruby' and/or BUNDLE_CACHE_ALL_PLATFORMS: true. |
@ddnexus please open a new issue with steps to reproduce who you're able to isolate the issue |
couple of things changed due to rubygems/bundler#4467 incl. Ruby (RVM) bump from 2.2.5 to 2.3.1 (just to be in line with local)
couple of things changed due to rubygems/bundler#4467 incl. Ruby (RVM) bump from 2.2.5 to 2.3.1 (just to be in line with local)
Hello, ruby world. I've hit this while trying to build a https://jekyllrb.com on http://travis-ci.org .. and was able to fix (more work around, probably?) by lastnpe/lastnpe.github.io#2 .. in case this is useful to anyone. Full disclosure: I'm a Java dev and have absolutely no idea what I'm doing here.. ;-) |
couple of things changed due to rubygems/bundler#4467 incl. Ruby (RVM) bump from 2.2.5 to 2.3.1 (just to be in line with local)
bundle
Will get this error:
In fact,
After that, |
I wrote a guide outlining several options for dealing with this issue: |
@henning-koch thanks for that guide! If you have a reproduction case against Bundler 1.13.2+, please please please open a new ticket. @abcfy2 I wasn't able to reproduce the issue using those steps. The Redmine release does not contain a If anyone has an example of this continuing to happen on Bundler 1.13.2 or higher, please please please open a new ticket with exact reproduction steps. |
Version 1.14 (pre out today or tomorrow) will contain a fix for another cause of this error message, as well as explaining what the problem is and how to fix it: #5267 |
Despite having a file
Gemfile.lock
created by Bundler 1.11 by runningbundle
, runningbundle update simple_form
fails like below.The text was updated successfully, but these errors were encountered: