Hi Mozilla Members,

I am facing this error while I try to upload my zip and xpi files. I have tried to search a lot, couldn't find much that could help me. Any kind of help would be appreciated.

Thanks
imgHndlExtension.zip

As part of bug 792209, nsISupportsArray is being deprecated. All instances of this interface should be flagged.

This is a compatibility warning, so the add-on should still be upgraded in this case. It should also appear as a warning in regular validations.

Message:

The nsISupportsArray interface is deprecated and is being replaced by nsIArray. See this bug report for more information.

Affected add-on: https://addons.mozilla.org/addon/pagesaver/

Release timing: Firefox 52 will be released on March 7th.

re: Extension is not marked as compatible with Multi Process
Example:
https://addons.mozilla.org/en-US/developers/addon/earthweb/file/457094/validation

Extension is not marked as compatible with Multi Process
chrome/Earth.jar/

Extension is not marked as compatible with Multi Process

We would like deprecate (and eventually block) add-on access to ctypes. I'm working with bsmedberg to compose a deprecation announcement.

We apparently can't block add-on access to ctypes at run-time (because Firefox uses ctypes itself). @jvillalobos says the AMO validator already flags ctypes add-ons for extra review. bsmedberg would like to expose those warnings to add-on developers and change the wording to mention ctypes deprecation.

@andymckay

The code doc.createElementNS( nsXUL, 'panelview' ) causes false positive...

The warning about createElement is a bug in the validator since the <script> tag is never used, and since nsXUL is a constant which can't be changed.

An example here: https://github.com/Centril/firefox-line/blob/master/src/search-engine.js#L206

If an add-on is not e10s compatible as marked by: https://developer.mozilla.org/en-US/Add-ons/Install_Manifests#multiprocessCompatible then raise a warning.

We will likely elevate this to an error at a later date.

Supports: mozilla/addons-server#2770

This was reported in the blog.

See, for example, version 3.0 of Addons Manager Hilite. install.rdf has the multiprocessCompatible flag set to true. The developer page for the add-on shows that it's compatible, and if I upload the file to the standalone validator and the compatibility validator, neither show any multiprocessCompatible warnings. However, the email the developer got points to this validation result that does show the incorrect warning. This apparently affected multiple add-ons.

I rejected an addon. I needed to post additional comments which was done via "Request more information". Previously that was working as it should.

I noticed today that it was attached to the last passed version and not the current but rejected one.

Ref: https://addons.mozilla.org/en-US/editors/review/klamm-de-websuche-gutscheine

The latest version in fetcher is 2.1.4:

https://github.com/mozilla/amo-validator/blob/master/extras/jslibfetcher.py

But default version to download from jQuery is 2.2.0 now:

https://code.jquery.com/
https://jquery.com/download/

I think it should be added to JQUERY_VERSIONS

As part of bug 1167575, the behavior of onButtonClick changed. All instances of this function should be flagged.

This is a compatibility error, so the add-on shouldn't be upgraded in this case. It should also appear as a warning in regular validations.

Message:

The function onButtonClick is now asynchronous. For more information, please refer to this bug report.

Affected add-on: https://addons.mozilla.org/addon/imacros-for-firefox/

Release timing: Firefox 51 will be released on January 24th.

As SDK add-ons packed with cfx are broken in Firefox 44+ (in bootstrap.js due to changed let/const behavior), the validator needs to fail and give a meaningful error message to let developers know

1. what's wrong
2. how they can fix it.

Note that we should fail on new submissions, but compatibility tests should not be affected by this as published add-ons on AMO have been repacked to fix this issue.

At the moment the warning states:

You are using version ******* of the Add-on SDK, which is outdated. Please upgrade to version 1.17 and repack your add-on

I personally alter it (via a script) to state:

You are using version ******* of the Add-on SDK, which is outdated. Please upgrade to version 1.17 or use JPM and repack your add-on.

The changes are or use JPM and a full stop at the end.

Update: Text updated subsequent to the comment by wagnerand

example:
https://addons.mozilla.org/en-US/editors/review/inbox-notifications
https://addons.mozilla.org/en-US/firefox/files/browse/440671/file/data/content.js#top

Code:

"use strict";

    console.log("content.js attached");

    class Message {
      constructor(entry) {
        this.entry = entry;
      }

      get id() {
        return this.entry.querySelector("id").textContent;
      }

      get subject() {
       return this.entry.querySelector("title").textContent;
      }

      get summary() {
        return this.entry.querySelector("summary").textContent;
      }

      get sender_name() {
       return this.entry.querySelector("author > name").textContent;
      }

      get sender_email() {
       return this.entry.querySelector("author > email").textContent;
      }

      toString() {
        return `[Message: "${this.subject}" From: ${this.sender_name} <${this.sender_email}>]`;
      }
    }

    class TitleMutator {
      constructor() {
        this.titleElem = document.getElementsByTagName('title')[0];
        this.origTitle = this.titleElem.textContent;
        this.unreadCount = 0;

        this.mutateTitle();
        this.observe();
      }

      get unreadCount() {
        return this._unreadCount;
      }

      set unreadCount(unread) {
        this._unreadCount = unread;
        this.mutateTitle();
      }

      get observer() {
        if (!this._observer) {
          this._observer = new MutationObserver(() => {
            this.origTitle = this.titleElem.textContent;
            this.mutateTitle();
          });
        }

        return this._observer;
      }

      observe() {
        this.observer.observe(this.titleElem, { characterData: true });
      }

      unobserve() {
        this.observer.disconnect();
      }

      mutateTitle() {
        this.unobserve();
        this.titleElem.textContent = (this.unreadCount == 0) ?
          this.origTitle :
          this.origTitle.replace(/^Inbox/, `Inbox (${this.unreadCount})`);
        this.observe();
      }
    }

    class MessageFetcher {
      constructor() {
        this.seenMessages = new Set();
        this.titleMutator = new TitleMutator();
        this.mainList = document.querySelector('[role="main"]');
        this.firstRun = true;
        this.mainList.addEventListener("click", e => {
          if (e.target.closest('[role="application"]')) {
            // Potentially marked an item read.
            setTimeout(() => this.refresh(), 3000);
          }
        })
      }

      localize(strings) {
        return new Promise(resolve => {
          self.port.once("localized", resolve);
          self.port.emit("localize", strings);
        });
      }

      getFeed() {
        return new Promise(resolve => {
          self.port.once("mailFeed", response => {
            let parser = new DOMParser();
            resolve(parser.parseFromString(response, "text/xml"));
          });
          let accountId = window.location.pathname.substr(3, 1) || 0;
          self.port.emit("getMailFeed", accountId);
        });
      }

      notify(messages) {
        if (messages.length) {
          let notification = {};

          if (messages.length == 1) {
            notification = {
              title: ["new_message", messages[0].sender_name],
              body: messages[0].subject
            };
          } else {
            let senders = new Set(messages.map(e => e.sender_name));
            notification = {
              title: ["new_messages", messages.length],
              body: Array.from(senders).join(', ')
            };
          }
          if (!document.hasFocus()) {
            self.port.emit("notify", notification);
          } else {
            console.log("Focused, so not notifying:", JSON.stringify(notification));
          }
        }
      }

      refresh() {
        console.log("refreshing");
        return this.getFeed().then(feed => {
          console.log("got feed");
          let unread = this.titleMutator.unreadCount = feed.querySelector("fullcount").textContent;
          let newMessages = [];
          for (let entry of feed.querySelectorAll("entry")) {
            let message = new Message(entry);
            if (!this.firstRun && !this.seenMessages.has(message.id)) {
              newMessages.push(message);
            }
            this.seenMessages.add(message.id);
          }

          this.firstRun = false;

          console.log(
            "Unread:", unread,
            "New:", newMessages.length,
            "Cached:", this.seenMessages.size);

          this.notify(newMessages);

          return { unread: unread, newMessages: newMessages };
        });
      }

      get observer() {
        if (!this._observer) {
          this._observer = new MutationObserver(mutations => {
            let listitemAdded = () => {
              for (let mutation of mutations) {
                if (mutation.type == "characterData") {
                  if (!mutation.target.parentNode.closest('[contenteditable]')) {
                    return true
                  }
                }

                for (let node of mutation.addedNodes) {
                  if (node.getAttribute('role') == 'listitem' ||
                      node.getAttribute('role') == 'list') {
                    return true;
                  }
                }
              }

              return false;
            };

            if (listitemAdded()) {
              console.log("list item added");
              this.refresh();
            }
          });
        }

        return this._observer;
      }

      connect() {
        // We can be more economical and listen for DOM mutations
        // setInterval(refreshMail, 5000);
        this.observer.observe(this.mainList,
          { characterData: true, subtree: true, childList: true });
      }

      disconnect() {
        this.observer.disconnect();
      }
    }

    const messageFetcher = new MessageFetcher();
    messageFetcher.refresh();
    messageFetcher.connect();

Steps to reproduce:

1. Try to submit the following embedded webextension:
   Profilist-master.zip
2. Select “See full validation report”

Expected results:
A warning is displayed only once.

Actual results:
A warning is displayed multiple times.

• Please see the attached screenshot for this issue:
  

Original issue: mozilla/addons-linter#977

Steps to reproduce:

1. Try to submit the following add-on:
   testaddon.zip
2. Select “See full validation report”

Expected results:
Each warning is specified only a single time.

Actual results:
A few warnings are duplicated.

Additional notes:

• This issue reproduces on AMO-dev, AMO-stage and AMO-production using Firefox 50.0a1 (2016-06-21) under Windows 10 64-bit.
• Please see the attached screenshots for this issue:
  
  
  
  

Moved from: mozilla/addons-linter#768

I've got JavaScript Compile-Time Error warnings when I uploaded my add-on to AMO.
2 types of messages.

A compile-time error in the JavaScript halted validation of that file.

Message: missing ; before statement

A compile-time error in the JavaScript halted validation of that file.

Message: invalid arrow-function arguments (parentheses around the arrow-function may help)

All of my JS files are warned, and I use async / await in them.

My source and warnings.
https://addons.mozilla.org/en-US/firefox/files/browse/537463/

PS I'm also noticed about em:hasEmbeddedWebExtension.

Example:
https://addons.mozilla.org/en-US/firefox/files/browse/514781/file/popup.htm#top

Above file is .htm and also includes <script> tags
(I am keeping a copy of the addon, in case it gets deleted)

The Syntax Highlighter does not activate on that file.
The validator may be !? working properly since the JS in that file is somehow obfuscated.

See https://github.com/mozilla/addons-server/issues/2218#issuecomment-210814636

The affected Add-on: https://github.com/retorquere/zotero-better-bibtex/releases/download/1.6.45/zotero-better-bibtex-1.6.45.xpi

As part of bug 931389, mozDash and mozDashOffset are no longer supported. All instances of these properties should be flagged.

This is a compatibility error, so the add-on shouldn't be upgraded in this case. It should also appear as a warning in regular validations.

Message:

The mozDash and mozDashOffset properties are no longer supported. You can use setLineDash() instead.

Affected add-on: https://addons.mozilla.org/addon/zotfile/

Release timing: Firefox 52 will be released on March 7th.

https://addons.mozilla.org/en-US/developers/addon/flash-debugger/file/421089/validation

Flagged file extensions found.

Warning: Files whose names end with flagged extensions have been found in the add-on.

The extension of these files are flagged because they usually identify binary components. 
Please see http://addons.mozilla.org/developers/docs/policies/reviews#section-binary for 
more information on the binary content review process.

data/flashfirebug.swf
    data/html/images/themes/default/inspector/UIComponents/VolumeBar.png

The png file is not a binary file, nor should it be flagged.

The domain unblocker.yt has been used multiple times by malicious add-ons, so we need to flag all instances of this (in strings) to make sure reviewers reject or escalate those submissions.

CC @wagnerand @eviljeff @kewisch in case anyone wants to take it.

It was a useful feature and it has been a while that Tab stops change is not working ie changing it has no effect at all.

Following: https://github.com/mozilla/addons-server/issues/2875

The validator should raise an information (displayed to reviewers only) with the following details:

Title: This add-on is a hybrid add-on.
Description: The folder webextension contains an embedded webextension. Handle validation messages with care. For more information, refer to https://developer.mozilla.org/en-US/Add-ons/WebExtensions/Embedded_WebExtensions

E.g. jQuery from GitHub repository differs from jQuery on code.jquery.com:

https://github.com/jquery/jquery/blob/7751e69b615c6eca6f783a81e292a55725af6b85/dist/jquery.min.js
http://code.jquery.com/jquery-2.1.4.min.js

It contains line at the end:
//# sourceMappingURL=jquery.min.map

You can just skip empty or comment lines when calculate a content hash.

See https://github.com/mozilla/olympia/issues/1339

XPIProvider is a private interface which add-ons should not be allowed to access. We need to flag any reference to either the symbol or the name. See bug https://bugzilla.mozilla.org/show_bug.cgi?id=1200929 for more.

Originally mozilla/addons-linter#211, it's better here where it needs to be for legacy addon support.

As part of bug 1227289, the functions setAndFetchFaviconForPage and replaceFaviconDataFromDataURL have changed their arguments and default behavior. All instances of these should be flagged.

This is a compatibility warning, so the add-on should still be upgraded in this case. It should also appear as a warning in regular validations.

Message:

setAndFetchFaviconForPage and replaceFaviconDataFromDataURL now default to using a null principal for security reasons. An appropriate principal should be passed if different behavior is required. For more information, please refer to this bug report.

Affected add-on: https://addons.mozilla.org/addon/xmarks-sync/

Release timing: Firefox 51 will be released on January 24th.

As part of bug 1329926, the getURIForKeyword function has been removed. All instances of this function should be flagged.

This is a compatibility error, so the add-on shouldn't be upgraded in this case. It should also appear as a warning in regular validations.

Message:

The getURIForKeyword function was removed. You can use PlacesUtils.keywords.fetch instead. See this document for more information.

Affected add-on: https://addons.mozilla.org/addon/context-search-x/

See mozilla/addons-server#2357

As part of bug 857627, a number of nsIX509CertDB methods were removed or changed. Removed: findCertByNickname, findEmailEncryptionCert, findEmailSigningCert, changed: addCert. All instances of these methods should be flagged.

This is a compatibility error, so the add-on shouldn't be upgraded in this case. It should also appear as a warning in regular validations.

Message:

The nsIX509CertDB interface was changed so it no longer exposes the certificate nickname. See this bug report for more information.

Affected add-on: https://addons.mozilla.org/addon/xml-digital-signature-tool/

I have manifest.json for Chrome and install.rdf for Firefox and I got the following validation error.

No Mozilla products listed as target applications
Error: None of the target applications listed in the manifest are supported Mozilla products. At least one official Mozilla product must be supported for inclusion on addons.mozilla.org.

See https://addons.mozilla.org/firefox/pages/appversions/ for more information on supported target applications on AMO.

manifest.json

In #386 support was added for manifest.json

In amo-validator/tests/helper.py:

        if 'install.rdf' in package:
            err.save_resource('has_install_rdf', True)
            rdf_data = package.read('install.rdf')
            install_rdf = RDFParser(err, rdf_data)
            err.save_resource('install_rdf', install_rdf)
        elif 'manifest.json' in package:
            err.save_resource('has_manifest_json', True)
            manifest_data = package.read('manifest.json')
            manifest_json = ManifestJsonParser(err, manifest_data)
            err.save_resource('install_rdf', manifest_json)

In amo-validator/validator/testcases/targetapplication.py the order is reversed:

    if manifest_json:
        applications = manifest_json.get_applications()
        manifest_file = 'manifest.json'
    elif install_rdf:
        applications = install_rdf.get_applications()
        manifest_file = 'install.rdf'

Which is the correct order? Does install.rdf take precedence over manifest.json? Are you going to validate both manifest files?

I will be adding an applications entry to my manifest.json (if chrome doesn't complain about it), but this inconsistency will cause problems for developers in the future.

Quite regularly, trying the **Download ******* links at the bottom (ie view) results in 403 Forbidden

Example:
https://addons.cdn.mozilla.net/en-US/firefox/files/browse/459606/file-serve/modules/BrandThunderUtils.jsm?token=c568f0b0-e701-4990-a960-65cefaa3362a

As part of bug 812701, the mozVisibilityState and mozHidden properties have been unprefixed. All instances of these should be flagged.

This is a compatibility error, so the add-on shouldn't be upgraded in this case. It should also appear as a warning in regular validations.

Message:

The mozVisibilityState and mozHidden properties are no longer prefixed, so you should use visibilityState and hidden instead. For more information, please refer to this bug report.

Affected add-on: https://addons.mozilla.org/addon/scrollbar-search-highlighter/

Release timing: Firefox 51 will be released on January 24th.

We should throw an exception if there's a install.rdf or package.json once there's a manifest.json to make sure we block duplicated metadata early enough.

Tracking: https://github.com/mozilla/addons-server/issues/2450

Add-ons are not allowed to modify xpinstall.signatures.required.

Today I noticed that the validator is missing some issues ie: nsIProcess on this addon:
https://addons.mozilla.org/en-US/editors/review/fotideptsclient-addon

Steps to reproduce:

1. Sign in to AMO-dev page https://addons-dev.allizom.org/en-US/
2. Submit a dictionary with a wrong "type" set (i.e. type=2)

Expected results:
The dictionary cannot be submitted because the "type" is wrong.

Actual results:
The dictionary can be submitted.

Notes/Issues:
Verified on AMO-dev FF45(Win 7).
Screenshot for this issue:

This is following https://github.com/mozilla/olympia/issues/1338

It seems an invalid max version for an application in a WebExtension's manifest.json isn't caught, and results in a weird unrelated error on olympia.

We should catch such an invalid max version just as we do it for a legacy extension.

With the same code base, the result differs, and there is a high chance to fail. Please review the results below:

passed one
https://addons.mozilla.org/en-US/developers/addon/hpe-functional-testing-agent/file/597828/validation

failed one
https://addons.mozilla.org/en-US/developers/upload/504e72cab45e4db3925ade37da7f9e04

There are developers who copy/paste GreaseMoneky scripts into an addon.

It seems validator (as well as the Olympia for displaying the review) miss most issues in such cases.
It might be due to wild-card usages in GM include/exclude.

Example - validator shows nothing while there are eval(), createElement('script'), innerHTML etc
https://addons.mozilla.org/en-US/developers/addon/enhanced-rt/file/488096/validation

Review: (Syntax Highlight also has issues)
https://addons.mozilla.org/en-US/firefox/files/browse/488096/file/Enhanced%20RT.user.js#top

As per mozilla/olympia#1093

Validator shows all ticked while addon has a JQuery and other files that should have been picked up by the validator as valid or not valid libraries, etc

I tested it 3 times before rejecting the addon.

https://addons.mozilla.org/en-US/editors/review/scopus-parser

As part of bug 528005, the BrowserOpenNewTabOrWindow function has been removed. All instances of this function should be flagged.

This is a compatibility error, so the add-on shouldn't be upgraded in this case. It should also appear as a warning in regular validations.

Message:

The function BrowserOpenNewTabOrWindow has been removed. You can use BrowserOpenTab instead, but its behavior is not identical. For more information, please refer to this bug report.

Affected add-on: https://addons.mozilla.org/addon/tab-tree/

Release timing: Firefox 51 will be released on January 24th.

We should raise a warning for entries in install.rdf like:

<em:optionsURL>https://somepage.com/settings</em:optionsURL>

As part of bug 1147911, the _openURIInNewTab function changed the arguments it takes. This function is often monkey-patched by tab add-ons. All instances of this function should be flagged.

This is a compatibility error, so the add-on shouldn't be upgraded in this case. It should also appear as a warning in regular validations.

Message:

The _openURIInNewTab function was changed and now requires an nsIURI for the referrer. See this bug report for more information.

Affected add-on: https://addons.mozilla.org/addon/load-tabs-progressively-fixed/

Example:

https://addons.mozilla.org/en-US/firefox/files/browse/459606/file/modules/BrandThunderUtils.jsm#top

JavaScript Compile-Time Error

A compile-time error in the JavaScript halted validation of that file.

Message: missing ] after element list
modules/BrandThunderUtils.jsm
8
9
10
Cu.import("resource://gre/modules/AddonManager.jsm");
try {
Components.utils.import("resource:///modules/NewTabURL.jsm");

Actual code:

var EXPORTED_SYMBOLS = ["BrandThunderUtils"];

const {classes: Cc, interfaces: Ci, utils: Cu} = Components;
Cu.import("resource://gre/modules/XPCOMUtils.jsm");
Cu.import("resource://gre/modules/Services.jsm");
Cu.import("resource://gre/modules/NetUtil.jsm");
Cu.import("resource://gre/modules/LightweightThemeManager.jsm");
Cu.import("resource://gre/modules/AddonManager.jsm");
try {
  Components.utils.import("resource:///modules/NewTabURL.jsm");
} catch (e) {}

package.json never happened.

Tracking: https://github.com/mozilla/addons-server/issues/2460

This is a "nice-to-have" bug for someone to work on but no rush.

We now have a new JS library hash collector that should update to new versions of libraries at https://github.com/mozilla/dispensary. It keeps its own hashes.txt and could be used by this validator.

I have noticed (for a while) that using the search button with blank query on the queue page no longer shows the admin flagged addons.

It seems that the problems is with searching= parameter

For example .....
Shows admin flagged:
https://addons.mozilla.org/en-US/editors/queue/nominated?searching=True&text_query=&admin_review=1&application_id=&max_version=&waiting_time_days=#review-actions

Doesn't show admin flagged:
https://addons.mozilla.org/en-US/editors/queue/nominated?searching=&text_query=&admin_review=&application_id=&max_version=&waiting_time_days=

Changing the searching=& to searching=true& fixes it.

As per title:
<iframe id="requestpolicy-requestLog-frame" type="chrome" flex="1"/>

Example:
https://addons.mozilla.org/en-US/firefox/files/compare/377109...369395/file/chrome/requestpolicy.jar/content/overlay.xul#top