Opening an anonymous link to a repository is leading to a DNS server error (1016). In fact, the landing website is also down and gives the same error.

If a link is to a URL that partially matches a pattern I want to hide, the whole URL should probably be anonymized. Right now, I get engineering.xxx.edu, which causes the following issues:

• xxx might be a registered domain (xxx.edu doesn't exist, but it's the case for .com and a few other)
• those domains are usually pornographic (please don't make my repository link to pornographic material 😉)
• the remaining parts of the URL are still there, and can be used to de-anonymize (not all companies have an .edu domain, or an engineering subdomain)

The link is broken anyway, better to anonymize completely. Also, it could point to a page/popup saying it's been anonymized, instead of leading to an error page, so the reader knows what's going on.

When I try to login to github for private repositories, I got the error message as follows:

Internal Server Error

The server encountered an internal error and was unable to complete your request. Either the server is overloaded or there is an error in the application

What are the causes of the problem and how to solve them?
Thanks very much.

The current version does not anonymize names in a license file if it's included in the repo.

Hello,

Last night the server that was hosting the service has suffered a critical hardware problem.
It is currently not clear how much data is lost. The service will be unavailable until further notice.

I am sorry for the inconvenience. I, unfortunately, cannot do much.

Hi,

I have used the service to anonymize my public repo for double-blind conference.
Now I discover that the link no longer works.
This is devastating since people are currently reviewing the paper and this code is the strongest contribution!

The repo https://github.com/IBM/dac-lib
The anonymized link is https://anonymous.4open.science/r/1335dc45-3197-4d98-87b9-6478edea78d0/

I am asking, as soon as possible, to recreate the object with this same link.
So that the reviewers can access the code.

I couldn't see any mention of this in the code or documentation, so how does this work for repositories that contain submodules referring to other repositories? Does each submodule need to be manually anonymized? Can anonymized repositories be cloned?

Hi all,

I am getting the following error "The server encountered an internal error and was unable to complete your request. Either the server is overloaded or there is an error in the application." after trying to access some code from the anonymized repository. Can you please help me with that?

follow up: the repository I am trying to access is not empty as was the case with a previous similar problem. The error occurs when opening any python file in any folder in my anonymized repository.

I am trying to anonymize a repository but I keep getting 500 Internal Server Errors.

Repository URL: https://github.com/neuralmind-ai/portuguese-bert
Terms:

neuralmind
fabio
souza
rodrigo
nogueira
roberto
lotufo
1909.10649

Expiration options: Redirect to GitHub repository
Expiration date: 25/02/2020

Generated link with error: https://anonymous.4open.science/r/ac885bb6-1bd0-400f-9587-f9121d8a83b4/

The repository does not contain code, it has a readme and a pdf file. I suspected it could be that one of the terms appears at the repository url, but the error happens even without any terms.

Hi all,

It would be nice to have a logo for anonymous github but I don't have good ideas.

Do you have suggestions?

Hi!

First of all, thank you for developing this platform! This is so useful for paper submission!

I had an issue for opening a image file (*.png, *.jpeg) as shown on the snapshot below

Those images are showing up on my repo but the anon repo. Is this only happening to me?
Could you help me to fix this?
Thank you!

Hi Thomas, nice features in the new version of Anonymous Github. Thank you for the great work! Did the previous version had an expiration date for the anonymized repositories? I can no longer found them when I am accessing the links.

Looks like there's no support for ReSTructured Text yet. That would be great since that's the language most documentation for python libraries is written in.

I anonymized my repo consisting of 2 folders, each containing a large number of files.
One of the folders is fine but when I click on the other folder, nothing happens. It stays in the same page.

instead of sharing a token with @tdurieux

While initially the anonymisation worked, I can't seem to access the repository for a second time. What am I doing wrong?

EDIT: it seems sometimes I get referred to the empty.html page, and sometimes I don't. I have a hard time figuring out what pattern causes this behaviour, sorry I can not be more specific than this.

So my dashboard is no longer showing all repositories that I anonymized. There are at least 4 that are not appearing, though several others do appear on the dashboard correctly.

Given this, I'm trying to use the "Claim Repository" functionality. However when I try this I get an error "The repository is not found." I'm very confident that I'm trying to recover the repository correctly and I'm not sure why this is happening.

Based on other suggestions in #44, I thought to re-anonymize the same repository and then set as its ID the ID used for the existing repository. However, not surprisingly, the "Anonymize" form shows an error, " is already used"

How can this be resolved?

I will email you privately with details about the repositories in question for you to look at.

Thanks again for your work on this valuable service.

Hi, this repo is really great for us to share anonymous code for academic conferences. Will you add git clone support so that we can clone the code?

Hi there, thx for provide this awesome tool firstly.
I have used it in the submission of a paper for KDD about two months ago. However, recently I meet the 404 error when I try to check it with the generated url. Could you please help me check it?

Hi all,

I am working on the new version of Anonymous Github. I am making good progress.

Where is a preview of the current version:

Changes

• Complete rewrite using NodeJS and MongoDB (instead of python and the file system)
• Focus on reducing the number of request to GitHub (for speed and reduce the Github token usage)
• Better anonymization, e.g., all paths are now anonymized (#20, #40)
• Two anonymization methods:
  • Stream: as the current version when a file is requested, the file is downloaded from Github. Bigger files are now supported (#33).
  • Download (new): the complete repository is downloaded when anonymized. It has better performance, reduces the token usage but increases disk usage. This option is only available for repositories less than 10mb.
• More anonymization options:
  • Remove all links
  • Removes all images
  • Render PDF (#39)
  • Render Notbook (#32)
• New page for anonymizing a repository
  • include a preview
  • select the branch
  • select options
• Dashboard to list all your anonymized repositories. It will be possible to import your previous repository.
• It uses now ace.js to render the code which offer a much better reading experience
• You must be logged to be able to anonymize a repository

TODOs

• Edit repository configuration
• FAQ
• Evaluate the disk usage of the new approach (storage is expensive)
• Handle repository expiration
• Automatic update of the repository when a new commit is pushed
• support Github page on the same branch
• test the import of the current repositories
• improve error message
• improve design
• replicate database
  Any suggestion or comments are welcome

I would like to include a pdf in my repo, but it does not let me because it is supposedly not a valid extension... I don't need anything to be anonymized in it, I would just like it to be made available through the anonymous repo.

Hi!

I found that my anonymized repository link has been linked to a wrong repository of mine, and wonder would it be possible to redirect the already created anonymized link to a correct repository?

This is the link below
(https://anonymous.4open.science/r/f7d04309-b871-4869-88f6-fa01703111e8/)
and I want to connect that link to this one... (https://github.com/mallcong/Therms-Up)

I could've create a new anonymous link with the correct repository but, I already submitted the anonymized link that is connected to the wrong direction to a conference and found cannot edit it anymore..

Looking forward to hearing from you soon.
Thank you so much!

Hi,
I am getting Error 502 bad gateway on opening the main page i.e. https://anonymous.4open.science/. Also, all my anonymized repo links are down. Is the server down or is something wrong going on my end? If its something wrong with the server, when can we expect it to get it fixed?

I've used the anonymous.4open.science to include an anonymized URL of our repository to a paper that we've submitted to a conference and it's currently showing an internal server error. Is there anyway to solve this problem?

Just tried to use your service, but got a 500 Internal Server error. Any idea what's going wrong?

How do we clone a repository from anonymous github? Please help me out.

I've tried to anonymize a tag of the following repo

https://github.com/CRIStAL-Sigma/latex-travis-test

using the url provided by github:

https://github.com/CRIStAL-Sigma/latex-travis-test/tree/public-v0.2

And I had an Internal Server Error:

Internal Server Error

The server encountered an internal error and was unable to complete your request. Either the server is overloaded or there is an error in the application.

Hey. First off, this is amazing!
Second, I just reviewed a paper and it had adXXXced in the readme. That leaks some information about the authors. Maybe if the name is not delimited by word boundaries, we could check if it's part of a dictionary word?
Taking a step back: what was the motivation to not require word boundaries? File names and variable names?
I'm not sure how easy it is to have meandyou.py and not replace andy with XXX. That would require trying to do segmentation on out-of-vocabulary words...

unknown extension: replace download link by

unknown extension, unable to anonymize (known extensions md/txt/json/java/...)

too big file:

file too big to be anonymized (beyond 1MB, Github limit)

Hello,

I have viewed one of the repositories I anonymized some month ago, and I have noticed some rendering issues.
As far as I remember, the interface used to "look like" a file browser, with a file content at the bottom.
It now shows just a plain list of files.
Is it a feature?

Besides, I am getting some client-side errors in chrome dev console.
(By the way, I have disabled adblock for this screenshot.)

Currently im having an issue in which the data i want to provide is too big to be shown. This leads to the following problem:

At the end of the File this is shown:

...
<More data exists, we delete it to allow the preview>
the text file itself is less than 1 MB and withing Githubs requirements.

so my question is now if it is possible to somehow download the whole file or display the whole file.

Hi, I added a file to my github repo, and did a "force" update {first update}. but it just shows the previous version. I also used "Edit" and changed the commit to anonymize and the file still doesn't show up.

(this is via website dashboard not CLI)

Hi
Have you considered using nbviewer for viewing .pynb files?

Good idea! IMO any publisher with a blind review process should host this one.

A Docker image might help more people providing this service. Have you thought about publishing one?

Hi,
Thanks for making this amazing tool available.
I used this tool to double-blind a repository for a paper submission. Unfortunately I did not found any mentions about a limit and now discovered that the reviewers cannot download the file.
My question is - is there any way I could allow larger files to be anonymized? Is there any subscription I can pay for to allow larger files? At this stage I cannot modify the URL.

Thanks!

http://anonymous.4open.science/

Current webpage is:

Welcome to nginx!
If you see this page, the nginx web server is successfully installed and working. Further configuration is required.

For online documentation and support please refer to nginx.org.
Commercial support is available at nginx.com.

Thank you for using nginx.

Hi.

Quote the previous issue
Originally posted by @tdurieux in #51 (comment)

The host doesn't work again. :(

If you don't mind, could you let me know why the host server is down? Is it a transient issue or not? As I am using this anonymous repository for the paper currently reviewed, if the problem cannot be fixed in a short time, I need to find another option and contact the editor.

Always thanks for your support.
Thanks.

We usually use our group name in the package paths.
E.g. org.GROUP.SUB_GROUP.projectname

And everything works fine for me, except that the replacement does not work in the actual URL path

Suppose you anonymize a repo with the following directory structure:

/
/docs
/docs/index.rst
/docs/conf.py
/docs/some_guide.rst

When a viewer clicks on docs/ they are shown a rendering of the index.rst file rather than a folder view listing all of the files in the directory. So a viewer cannot view some_guide.rst unless they know to type in the relative path in the browser navigation bar.

Suggested resolution: every directory should have the same experience as /

• the directory path should be displayed in the top bar
• the list of files within the directory should be displayed below it
• if the directory has a file that is known to be the highest-level documentation (i.e. README.*, index.*), then that file should be rendered below the file list

Thanks!

The filenames are correctly anonymized in the directory view, e.g. secret-code.c become XXX-code.c. However, the URL to that file (when you click on the link to open the file) still has the original name, e.g. path/to/secret-code.c. I think the URL must also be anonymized.

I don't think the solution is to just replace the text in the filename with XXX. For example, after anonymizing the URL for secret1-code.c and secret2-code.c, both of them will have the same filename, XXX-code.c!

Just in case you don't know it yet, the service is currently down: http://anonymous.4open.science/

i was trying to update an existing repo by entering the repository id as X and repository url as https://anonymous.4open.science/r/X/, but it keeps coming back to the same page with the form fields.
was just wondering i am missing something. any pointers would be appreciated.

I can access the main page and already imported anonymous repositories. But when I try Login to GitHub, I get 500 Internal Server Error. URLs of error pages look like:
https://anonymous.4open.science/github/auth?code=ea2cc82711f7ed4bff46&state=QOM9Pqa26MoOCqEwausW7baMs5D6Lr
(The code and state part of URL change when I try to login again.)