<?xml version="1.0" encoding="utf-8"?><feed xmlns="http://www.w3.org/2005/Atom" ><generator uri="https://jekyllrb.com/" version="3.10.0">Jekyll</generator><link href="https://mozilla-necko.github.io/feed.xml" rel="self" type="application/atom+xml" /><link href="https://mozilla-necko.github.io/" rel="alternate" type="text/html" /><updated>2026-01-18T16:29:39+00:00</updated><id>https://mozilla-necko.github.io/feed.xml</id><title type="html">Necko Blog</title><subtitle>Insights into Necko - the Firefox Networking team.</subtitle><entry><title type="html">Necko News #6 - 2024 Q2 Recap</title><link href="https://mozilla-necko.github.io/newsletter/2024/06/29/q2-recap.html" rel="alternate" type="text/html" title="Necko News #6 - 2024 Q2 Recap" /><published>2024-06-29T16:30:00+00:00</published><updated>2024-06-29T16:30:00+00:00</updated><id>https://mozilla-necko.github.io/newsletter/2024/06/29/q2-recap</id><content type="html" xml:base="https://mozilla-necko.github.io/newsletter/2024/06/29/q2-recap.html"><![CDATA[<h2 id="highlights">Highlights</h2>

<ul>
  <li>Sean Kim, a student worker from UoT has joined the Necko team. Welcome!</li>
  <li>Max Inden is joining the Necko team as a contract worker to help improve UDP I/O. Welcome!</li>
  <li>Manuel, a student worker, has now been hired as a full-time employee and is working with the privacy team.</li>
  <li>In bug <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1356686">1356686</a>, decompression was moved off the MainThread for most loads, resulting in significant performance improvements on CI pageload metrics, with up to 46% speedindex gains and up to 56% improvements in FCP without regressions.</li>
  <li>In bug <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1883496">1883496</a>, we recently implemented a patch that moved search header interception from GeckoView to Necko, resulting in 13-24% improvements in Speed Index for Google searches on Android devices.</li>
  <li>We landed support for Zstd (ZStandard) decompression in <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1871963">Bug 1871963</a>.  This is used heavily by all Meta websites (Facebook, Instagram, etc) and will likely rapidly be picked up by other sites since Chrome just shipped support</li>
  <li>We moved search header interception into Necko in <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1883496">Bug 1883496</a>, resulting in 13-23% improvements in google search speed tests on android</li>
</ul>

<h2 id="friends-of-necko">Friends of Necko</h2>

<p>We had quite a few new and known community members contribute to our Necko code base. We are happy about everyone that wants to get involved. Reporting bugs, submitting logs, and also patches are all highly valued by us. You can take a look at our <a href="https://bugzilla.mozilla.org/buglist.cgi?component=DOM:%20Networking&amp;component=Networking&amp;component=Networking:%20Cache&amp;component=Networking:%20Cookies&amp;component=Networking:%20DNS&amp;component=Networking:%20File&amp;component=Networking:%20Proxy&amp;component=Networking:%20HTTP&amp;component=Networking:%20JAR&amp;component=Networking:%20WebSockets&amp;keywords=good-first-bug,%20&amp;keywords_type=allwords&amp;product=Core&amp;resolution=---&amp;order=Bug%20Number%20DESC">good-first-bugs</a> and also reach out to us at matrix <a href="https://matrix.to/#/#necko:mozilla.org">#necko:mozilla.org</a> if you want to start getting more involved.</p>

<p>Setting up the toolchain can be quite challenging. Therefore, we sometimes have good-first-bugs that give you the opportunity to just test out the toolchain.</p>

<p>Thank you all for helping improve Firefox Networking!!</p>

<ul>
  <li>New contributors to <a href="https://github.com/servo/rust-url/releases/tag/v2.5.1">rust-url 2.5.1</a></li>
  <li><strong>Masatoshi Kimura</strong>
    <ul>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1891926">1891926</a>: “Use system Web Proxy Auto-Discovery setting” checkbox is not shown even if “network.proxy.system_wpad.allowed” is set to true</li>
    </ul>
  </li>
  <li><strong>Alexandra Borovova</strong>
    <ul>
      <li>
        <table>
          <tbody>
            <tr>
              <td><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1898129">1898129</a>: Perma [tier 2] /fetch/api/abort/serviceworker-intercepted.https.html</td>
              <td>response.bytes() rejects if already aborted - promise_test: Unhandled rejection with value: object “TypeError: response[bodyMethod] is not a function”</td>
            </tr>
          </tbody>
        </table>
      </li>
      <li>
        <table>
          <tbody>
            <tr>
              <td><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1897959">1897959</a>: Perma /client-hints/sec-ch-width.https.html</td>
              <td>Sec-CH-Width should be set - assert_equals: expected 100 but got 1</td>
            </tr>
          </tbody>
        </table>
      </li>
    </ul>
  </li>
  <li><strong>Bob Owen</strong>
    <ul>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1892226">1892226</a>: Use inet_ntop directly on Windows to prevent mswsock.dll load.</li>
    </ul>
  </li>
  <li><strong>Andrew Sutherland</strong>
    <ul>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1900681">1900681</a>: XMLHttpRequestMainThread::DisconnectFromOwner should not call Abort for <code class="language-plaintext highlighter-rouge">mForWorker</code> XHRs because it double-clocks the state machine</li>
    </ul>
  </li>
  <li><strong>Christoph Moench-Tegeder</strong>
    <ul>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1890593">1890593</a>: breakage after “button to allow WPAD while using system proxy settings” (#621429)</li>
    </ul>
  </li>
  <li><strong>Andrew McCreight</strong>
    <ul>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1899863">1899863</a>: Extremely frequent PNecko::Msg_CancelHTMLDNSPrefetch log spam</li>
    </ul>
  </li>
  <li><strong>Chris Peterson</strong>
    <ul>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1894429">1894429</a>: Spoof “Android 10” OS version in Firefox Android’s User-Agent string on Android versions &lt;= 10</li>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1873273">1873273</a>: Freeze UA string and navigator.platform on 32-bit x86 Linux to match Chrome’s frozen navigator.platform (“Linux x86_64”)</li>
    </ul>
  </li>
  <li><strong>endington543</strong>
    <ul>
      <li>
        <table>
          <tbody>
            <tr>
              <td><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1820570">1820570</a>: Intermittent netwerk/test/unit/test_progress_no_proxy_and_proxy.js</td>
              <td>single tracking bug</td>
            </tr>
          </tbody>
        </table>
      </li>
    </ul>
  </li>
  <li><strong>Emilio Cobos Álvarez</strong>
    <ul>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1889590">1889590</a>: Make cache DB use UTF8 string for referrer.</li>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1887719">1887719</a>: More consistently use UTF8String/nsCString for URLs</li>
    </ul>
  </li>
  <li>[evilpies]
    <ul>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1896475">1896475</a>: Implement Request/Response <code class="language-plaintext highlighter-rouge">.bytes()</code> method</li>
    </ul>
  </li>
  <li><strong>Frederik Braun</strong>
    <ul>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1897132">1897132</a>: make netwerk/test/browser/browser_dns_prefetch_link_header.js work with https-first</li>
    </ul>
  </li>
  <li><strong>Frédéric Wang</strong>
    <ul>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1900764">1900764</a>: urgency of fonts in fetchpriority-urgency.h2.html are not affected by fetchpriority</li>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1892734">1892734</a>: Add tests for checking the priority set in our HTTP requests</li>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1881040">1881040</a>: Use the destination to refine the fetchpriority mapping of FetchDriver::HttpFetch</li>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1882548">1882548</a>: Enable “fetchpriority” feature by default in nightly</li>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1882084">1882084</a>: Support fetchpriority in Early Hints preloader</li>
      <li>
        <table>
          <tbody>
            <tr>
              <td><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1901220">1901220</a>: Perma TEST-UNEXPECTED-FAIL</td>
              <td>/_mozilla/fetch/fetchpriority/fetchpriority-urgency.h2.html</td>
              <td>Adjustment of urgency parameter for &lt;…&gt; when Gecko 128 merges to beta on 2024-06-10</td>
            </tr>
          </tbody>
        </table>
      </li>
    </ul>
  </li>
  <li><strong>Gijs</strong>
    <ul>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1881800">1881800</a>: Make credentialed network fetches with system (chrome) principals opt-in (ie default to <code class="language-plaintext highlighter-rouge">credentials: omit</code>)</li>
    </ul>
  </li>
  <li><strong>Gregory Pappas</strong>
    <ul>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1887611">1887611</a>: Implement URL.parse()</li>
    </ul>
  </li>
  <li><strong>Henri Sivonen</strong>
    <ul>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1886953">1886953</a>: Introduce deCOMtaminated IDNA functions and benchmark them</li>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1894212">1894212</a>: Bench IDNA via URL parsing</li>
    </ul>
  </li>
  <li><strong>Julian Descottes</strong>
    <ul>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1892440">1892440</a>: Expose chrome only method to flip HttpChannel’s StoreRequestObserversCalled</li>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1899563">1899563</a>: Perform SW upgrade before http-on-before-connect</li>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1899291">1899291</a>: Improve channelId generation code to avoid clashes between tabs with identical pID</li>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1895198">1895198</a>: Move the assert !LoadWasOpened() in HttpBaseChannel::ExplicitSetUploadStreamLength to HttpChannelChild</li>
    </ul>
  </li>
  <li><strong>John Schanck</strong>
    <ul>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1892528">1892528</a>: Enable Xyber768 in Http/3 TLS handshakes under a pref</li>
    </ul>
  </li>
  <li><strong>Robert Longson</strong>
    <ul>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1891349">1891349</a>: Firefox doesn’t treat cookie name prefixes as case-insensitive</li>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1891941">1891941</a>: Use StringBeginsWith in CookieService</li>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1892748">1892748</a>: Reject most control characters in cookie attributes</li>
      <li>
        <table>
          <tbody>
            <tr>
              <td><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1892450">1892450</a>: Perma Late Beta TEST-UNEXPECTED-PASS</td>
              <td>/cookies/prefix/__secure.header.https.html</td>
              <td>__SeCuRe: secure origin: Should not set ‘Path=/;domain=not-web-platform.test’ - expected FAIL</td>
              <td>when Gecko 127 switches from Early Beta to Late Beta on 2024-05-24</td>
            </tr>
          </tbody>
        </table>
      </li>
    </ul>
  </li>
  <li><strong>Leander Schwarz</strong>
    <ul>
      <li>
        <table>
          <tbody>
            <tr>
              <td><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1893480">1893480</a>: Perma /cookies/third-party-cookies/third-party-cookies.tentative.https.html</td>
              <td>3P fetch: Cookies are not accessible via HTTP - assert_equals: Expected cookie 1P_http to not be available expected false but got true</td>
            </tr>
          </tbody>
        </table>
      </li>
    </ul>
  </li>
  <li><strong>Max Inden</strong>
    <ul>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1895319">1895319</a>: refactor http3server to use neqo-bin for I/O, timers, event-loop etc.</li>
    </ul>
  </li>
  <li><strong>Manuel Bucher</strong>
    <ul>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1897050">1897050</a>: Remove or update metrics expiring in Firefox 129: networking.proxy_info_type before Gecko version gets increased to 129 on 2024-06-10</li>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1454850">1454850</a>: Mislabelled options relating to SOCKS4, SOCKS5 and server-side DNS resolution</li>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1741375">1741375</a>: Proxy DNS by default when using SOCKS v5</li>
    </ul>
  </li>
  <li><strong>Mike Conley</strong>
    <ul>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1896929">1896929</a>: Allow nsInputStreamChannel to have an arbitrary content type set on it</li>
    </ul>
  </li>
  <li><strong>Mika Valkealahti</strong>
    <ul>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1879304">1879304</a>: Remove mention of FTP in ContentParent</li>
    </ul>
  </li>
  <li>[mike.maksymowych]
    <ul>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1883282">1883282</a>: Log message incorrectly logs split string parts</li>
    </ul>
  </li>
  <li><strong>Neel Chauhan</strong>
    <ul>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1855652">1855652</a>: Remove unused nsDownloader and IncrementalDownloader</li>
    </ul>
  </li>
  <li><strong>Nika Layzell</strong>
    <ul>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1718673">1718673</a>: text/event-stream mime type starts downloading instead of streaming content</li>
    </ul>
  </li>
  <li>[serfreeman1337]
    <ul>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1810362">1810362</a>: Firefox does not seem to support IPv6 unless DNS-over-HTTPS is enabled</li>
    </ul>
  </li>
  <li><strong>Thomas Wisniewski</strong>
    <ul>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1893257">1893257</a>: Firefox-only failures in WPT fetch/http-cache/cache-mode.any.js</li>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1886757">1886757</a>: network.dns.localdomains does not apply to FQDN</li>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1893219">1893219</a>: Failing WPT fetch/content-length/too-long.window.html</li>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1491010">1491010</a>: [Fetch API] Headers.get(“WWW-Authenticate”) does not return anything, if server sends more than one such header</li>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1889075">1889075</a>: New wpt crashes in /fetch/api/request/ [request-bad-port.any.worker.html, request-bad-port.any.html, request-bad-port.any.sharedworker.html]</li>
    </ul>
  </li>
</ul>

<h2 id="project-updates">Project Updates</h2>

<p>Q2 2024 has been a busy with many started and completed projects:</p>

<p>Completed:</p>

<ul>
  <li>Optimized speculative connection socket pool shipped in Fx126 - showing significant improvements in sub-resource connection times (35-68%) and first byte sent (15-45%).</li>
  <li>Decompression(gzip and brotli) have been moved off main thread(OMT) in Fx126 reducing FCP and LCP by 10%.</li>
  <li>Support for <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1301878">zstd</a> compression shipped in Fx126, enabling pages to load faster and/or use less bandwidth, which can be especially important for mobile users and users with limited data plans.</li>
  <li>We optimized and enabled Firefox’s support for <code class="language-plaintext highlighter-rouge">&amp;lt;link rel=dns-prefetch&gt;</code> in Fx127,<code class="language-plaintext highlighter-rouge"> </code>allowing sites to prefetch key domains on https documents
    <ul>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1596935">Bug 1596935 - Firefox doesn’t resolve &lt;link rel=dns-prefetch&gt; on HTTPS</a></li>
    </ul>
  </li>
</ul>

<p>In progress:</p>

<ul>
  <li>Extending native <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1852752">DNS API’s to support HTTPS RR</a>, increasing HTTPS usage and ECH support when Firefox performs native DNS lookups. Targeting Q3 2024.</li>
  <li>Interop 2024 URL is progressing well, targeting Q4 2024.</li>
  <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1895345">Firefox/Chrome networking benchmark framework</a> is on track to complete Q3 2024.</li>
  <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1342484">Fetch keepalive </a>support is on track to complete Q3 2024.</li>
  <li>3rd party cookie deprecation - support for limiting size of partition store size is on track to complete Q4 2024.</li>
  <li>HTTP Priorities support is on track to complete Q3 2024.</li>
  <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1901292">Fast UDP I/O in Firefox</a> is on track to complete in Q3 2024.</li>
</ul>

<h2 id="below-the-fold">Below the fold</h2>

<ul>
  <li>We resolved the issue where downloading Firefox updates would make the browser unresponsive:
    <ul>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1510262">Bug 1510262 - nsIncrementalDownload causes a lot of jank with main thread I/O while downloading updates in the background</a></li>
    </ul>
  </li>
  <li>We backed out HTTPS RR (which landed in <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1852752">1852752 - Allow resolving HTTPS (or arbitrary record types) RR with native DNS (without TRR)</a> ) due to performance regressions in <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1893970">Bug 1893970</a> , and re-landed for 127 (and it should ride the trains there) in  <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1898191">Bug 1898191</a>  and <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1895226">Bug 1895226</a> “Don’t perform native HTTPS lookups if network doesn’t support them.”</li>
  <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1891349">Bug 1891349</a> (CVE-2024-5699) “Firefox doesn’t treat cookie name prefixes as case-insensitive” fixed a minor security issue</li>
  <li>Fixed a series of cookie bugs:
    <ul>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1903110">1903110     	cookie with empty name and value should be rejected</a></li>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1903415">1903415     	Long cookie attrbiutes should be rejected</a></li>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1819147">1819147     	Services.cache2.clearOriginAttributes() does not clear all cache entries of specified userContext</a></li>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1892748">1892748     	Reject most control characters in cookie attributes</a></li>
    </ul>
  </li>
  <li>Fixed an issue with DoH that could cause 30s delays on startup: <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1885801">bug 1885801 Secure DNS causes delayed website loading at startup</a></li>
  <li>Fixed a 10 second delay accessing .local domains in <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1893944">Bug 1893944</a></li>
  <li>A large numbers of TCP RSTs could cause Firefox to hang or be very janky; fixed in <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1719046">Bug 1719046</a></li>
  <li>In <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1816539">Bug 1816539</a>  we increased the number of sockets available for speculative connections from 6 to 20, which shows improvements in sites known to work with speculative connections.   We might increase it to 30 in the future</li>
  <li>We resolved a significant cause of UI/mainthread jank when downloading updates in the background in <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1510262">Bug 1510262</a>; a bug which had been open for 6 years.</li>
  <li>Fixed a number of network URL parsing issues</li>
  <li>In total, we closed 225 bugs as Fixed</li>
  <li>Work week in Berlin, hack day and strategy session
    <ul>
      <li>The Necko team had a very successful work week in Berlin in conjunction with the rest of Security and Privacy, and in the same time/place as the DOM team.</li>
      <li>We had a ‘hack day’, which included building a prototype for a system to submit Network logs and other data with PII to the networking team, and avoid asking users to email logs. In addition two team members completed the bulk of the work to move search header interception into Necko, resulting in significant performance win on Android.</li>
    </ul>
  </li>
  <li>Bug Bash Day
    <ul>
      <li>Total Bugs updated: <a href="https://bugzilla.mozilla.org/buglist.cgi?query_format=advanced&amp;chfieldto=2024-05-25&amp;chfieldfrom=2024-05-24&amp;product=Core&amp;component=DOM%3A%20Networking&amp;component=Networking&amp;component=Networking%3A%20Cache&amp;component=Networking%3A%20Cookies&amp;component=Networking%3A%20DNS&amp;component=Networking%3A%20File&amp;component=Networking%3A%20HTTP&amp;component=Networking%3A%20JAR&amp;component=Networking%3A%20Proxy&amp;component=Networking%3A%20WebSockets&amp;list_id=17051096&amp;classification=Client%20Software&amp;classification=Developer%20Infrastructure&amp;classification=Components&amp;classification=Server%20Software&amp;classification=Other&amp;chfield=anything">197(7 off team)</a></li>
      <li>Closed: <a href="https://bugzilla.mozilla.org/buglist.cgi?chfield=anything&amp;chfieldto=2024-05-25&amp;product=Core&amp;list_id=17051111&amp;component=DOM%3A%20Networking&amp;component=Networking&amp;component=Networking%3A%20Cache&amp;component=Networking%3A%20Cookies&amp;component=Networking%3A%20DNS&amp;component=Networking%3A%20File&amp;component=Networking%3A%20HTTP&amp;component=Networking%3A%20JAR&amp;component=Networking%3A%20Proxy&amp;component=Networking%3A%20WebSockets&amp;bug_status=RESOLVED&amp;chfieldfrom=2024-05-24&amp;classification=Client%20Software&amp;classification=Developer%20Infrastructure&amp;classification=Components&amp;classification=Server%20Software&amp;classification=Other&amp;query_format=advanced">41 (3 off team)</a>!
        <ul>
          <li>14 fixed</li>
          <li>6 invalid</li>
          <li>11 won’t fix</li>
          <li>8 works for me</li>
          <li>1 duplicate</li>
          <li>4 incomplete</li>
        </ul>
      </li>
    </ul>
  </li>
</ul>]]></content><author><name>Necko team</name></author><category term="newsletter" /><summary type="html"><![CDATA[Highlights]]></summary></entry><entry><title type="html">Necko News #5 - Strengthening Community Involvement</title><link href="https://mozilla-necko.github.io/newsletter/2024/03/15/community.html" rel="alternate" type="text/html" title="Necko News #5 - Strengthening Community Involvement" /><published>2024-03-15T16:30:00+00:00</published><updated>2024-03-15T16:30:00+00:00</updated><id>https://mozilla-necko.github.io/newsletter/2024/03/15/community</id><content type="html" xml:base="https://mozilla-necko.github.io/newsletter/2024/03/15/community.html"><![CDATA[<h2 id="highlights">Highlights</h2>

<ul>
  <li><a href="https://www.eggert.org/">Lars Eggert</a>, current chair of the Internet Engineering Task Force (<a href="https://www.ietf.org/">IETF</a>), joined the Necko team. Welcome!</li>
  <li>Manuel presented <a href="https://fosdem.org/2024/schedule/event/fosdem-2024-1873-debugging-http-3-upload-speed-in-firefox/">Debugging HTTP/3 upload speed in Firefox</a> on <a href="https://fosdem.org/2024/">Fosdem 2024</a></li>
  <li>If you are on <a href="https://chemnitzer.linux-tage.de/2023/en/">Chemnitzer Linux Tage 2024</a> and want to meet up, please reach out to Manuel via <a href="mailto:manuel@mozilla.com">Email</a> or <a href="https://matrix.to/#/@manuel:mozilla.org">Matrix</a> and we can find a time to meet!</li>
  <li>We started publishing our <a href="https://mozilla-necko.github.io/meeting-notes/">Weekly Meeting Notes</a> again. Subscribe <a href="https://mozilla.social/@necko/">@Necko on mozilla.social</a> to get notified about new posts.</li>
  <li>We’re experimenting with Necko Office Hours. They are on the third Friday each month. The first one will be Friday, April 13, 15:00 - 16:00 CEST. Feel free to drop by if you are curious. It’s a place to talk and learn about Firefox networking. Ask in <a href="https://matrix.to/#/#necko:mozilla.org">#necko:mozilla.org</a> if you want to join. We will announce the Office hours prior to the meeting in the channel too.</li>
</ul>

<h2 id="friends-of-necko">Friends of Necko</h2>

<p>We had quite a few new and known community members contribute to our Necko code base. We are happy about everyone that wants to get involved. Reporting bugs, submitting logs, and also patches are all highly valued by us. You can take a look at our <a href="https://bugzilla.mozilla.org/buglist.cgi?component=DOM:%20Networking&amp;component=Networking&amp;component=Networking:%20Cache&amp;component=Networking:%20Cookies&amp;component=Networking:%20DNS&amp;component=Networking:%20File&amp;component=Networking:%20Proxy&amp;component=Networking:%20HTTP&amp;component=Networking:%20JAR&amp;component=Networking:%20WebSockets&amp;keywords=good-first-bug,%20&amp;keywords_type=allwords&amp;product=Core&amp;resolution=---&amp;order=Bug%20Number%20DESC">good-first-bugs</a> and also reach out to us at matrix <a href="https://matrix.to/#/#necko:mozilla.org">#necko:mozilla.org</a> if you want to start getting more involved.</p>

<p>Setting up the toolchain can be quite challenging. Therefore, we sometimes have good-first-bugs that give you the opportunity to just test out the toolchain. In the past months Komuhangi Tumuhairwe (<a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1860231">Bug 1860231</a>) and ChaseKnowlden (<a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1862514">Bug 1862514</a>) took the opportunity and submitted patches for these bugs.</p>

<ul>
  <li>Em Zhan implementing rel=modulepreload (<a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1798319">Bug 1798319</a>)</li>
  <li>Greg Pappas [:gregp] on his initiative to <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1773039">remove old prefs</a> removed a few Necko prefs and did other code cleanup
    <ul>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1052909">Bug 1052909</a> - Remove network.websocket.auto-follow-http-redirects pref</li>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1819556">Bug 1819556</a> - Remove nsSocketTransportService::ProbeMaxCount</li>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1842173">Bug 1842173</a> - Remove network.http.clear_bogus_content_encoding pref</li>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1842326">Bug 1842326</a> - Remove <code class="language-plaintext highlighter-rouge">network.auth.use_new_parse_realm</code> and <code class="language-plaintext highlighter-rouge">network.auth.allow_multiple_challenges_same_line</code></li>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1876945">Bug 1876945</a> - Run test_cookies_thirdparty.js and test_cookies_thirdparty_session.js with e10s enabled</li>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1883217">Bug 1883217</a> - Remove unused defines in IDL files</li>
    </ul>
  </li>
  <li>Dr. Marten Richter reported and fixed both:
    <ul>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1872496">Bug 1872496</a> - Fix WebTransport’s stream closure</li>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1873263">Bug 1873263</a> - Our WebTransport serverCertificateHashes Implementation</li>
    </ul>
  </li>
  <li>Jackyzy823 landed and quickly fixed:
    <ul>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1853203">Bug 1853203</a> - Support non-ASCII username/password for socks proxy</li>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1881883">Bug 1881883</a> - Nightly continuously outputs type=socks, flags=1</li>
    </ul>
  </li>
  <li>Masatoshi Kimura [:emk] submitted a few patches to Necko:
    <ul>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1875001">Bug 1875001</a> - Remove unused features from nsDirIndexParser</li>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1867229">Bug 1867229</a> - Remove unused fields from nsIDirIndex and nsIDirIndexParser</li>
    </ul>
  </li>
  <li>Neel Chauhan <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1861878">Bug 1861878</a> - Remove network.ssl_tokens_cache_use_only_once pref</li>
  <li>Jaydeep Das fixed the 18 years old <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=328707">Bug 328707</a> by allowing only valid IP/Hostname for Proxy Config</li>
  <li>Daisuke Akatsuka <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1864985">Bug 1864985</a> - Add hasUserPass attribute to nsIURI</li>
  <li>CanadaHonk, who also contributed patches to Necko, <a href="https://goose.icu/joining-mozilla/">joined Mozilla</a> in the DOM: Core team</li>
  <li>Max Inden has been very actively contributing to our http/3 implementation <a href="https://github.com/mozilla/neqo/pulls?q=is%3Apr+author%3Amxinden">neqo</a>. He is working closely with the Necko team. It’s a pleasure working with you.</li>
  <li>We had a few contribution from <a href="https://en.wikipedia.org/wiki/Igalia">Igalia</a>:
    <ul>
      <li>Mirko Brodesser landed a lot for <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1797715">Bug 1797715</a> - Fetch Priority (previously known as <a href="https://wicg.github.io/priority-hints/">Priority Hints</a>)</li>
      <li>Mirko Brodesser <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1869488">Bug 1869488</a> - Log address of <code class="language-plaintext highlighter-rouge">nsIChannel</code> subobject of <code class="language-plaintext highlighter-rouge">nsHttpChannel</code> in the constructor/destructor too</li>
      <li>Mirko Brodesser <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1868802">Bug 1868802</a> - Add <code class="language-plaintext highlighter-rouge">this</code> pointer to <code class="language-plaintext highlighter-rouge">nsHttpHandler::NotifyObservers</code> logging</li>
    </ul>
  </li>
  <li>From Redhat stransky migrated DbusWifiScanner to DBus/GIO (<a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1854449">Bug 1854449</a>)</li>
</ul>

<p>We’re sorry if we missed your contribution in this list. This list isn’t automatically generated, but we did go through our commit history by hand. Feel free to reach out on <a href="https://matrix.to/#/#necko:mozilla.org">#necko:mozilla.org</a> and we’ll add you in our next Newsletter.</p>

<h2 id="project-updates">Project Updates</h2>

<p>2023 has been a busy year with many started and completed project:</p>

<ul>
  <li>Off main thread (<a href="http://bugzilla.mozilla.org/show_bug.cgi?id=1528285">OMT</a>) for OnStopRequest was prototyped for HTML5Parser end of 2023 and <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1864817">CSSLoader</a> begin of 2024</li>
  <li>URL Interop <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1815647">2023</a> &amp; <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1876105">2024</a>
    <ul>
      <li>The 2023 interop effort saw us fixing several <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1815647">high impact bugs</a> that brought our <a href="https://wpt.fyi/interop-2023?feature=interop-2023-url">URL interop score</a> to 98.5% at the end of the year.</li>
      <li>We had to disable <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1603699">1603699 - Enable DefaultURI use for unknown schemes</a> after causing several issues with external protocol handlers. We are considering re-enabling this with an allow-list.</li>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1876105">1876105 - (interop-2024-url) Meta Interop 2024 URL</a> tracks all the work remaining to pass 100% of the URL WPT tests.</li>
    </ul>
  </li>
  <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1852752">1852752 - Allow resolving HTTPS (or arbitrary record types) RR with native DNS (without TRR)</a> is currently enabled on Nightly for specific platforms (Windows 11, Linux, Android 10+). This allows resolving HTTPS records even when DNS over HTTPS is not used. This should increase the use of HTTP/3, HTTPS and ECH.</li>
</ul>

<h2 id="below-the-fold">Below the fold</h2>

<ul>
  <li>We <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1753979">increased the number of threads used for DNS</a> resolution from 8 to 64. This reduced the waiting time from 26ms to 3ms in the 95th percentile, and from 816ms to 345ms in the 99th percentile.</li>
  <li>Devtool improvements:
    <ul>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1870580">Bug 1870580</a> Netmonitor doesn’t show resources loaded from a file url</li>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1156659">Bug 1156659</a> - Simulating offline mode for a tab</li>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1820807">Bug 1820807</a> Authentication requests were not visible in devtools</li>
    </ul>
  </li>
  <li>We <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1784257">improved our DoH telemetry</a> to get alerts about our DoH outages</li>
  <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1858712">Early Hints preconnect</a> was enabled in <a href="https://whattrainisitnow.com/release/?version=120">Fx120</a></li>
  <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1874445">Early Hints preload</a> was enabled in <a href="https://whattrainisitnow.com/release/?version=123">Fx123</a></li>
  <li>We had our second Bug Bash Session giving our backlog a fresh view. This time we
    <ul>
      <li>Updated <a href="https://bugzilla.mozilla.org/buglist.cgi?chfieldfrom=2024-02-23&amp;component=DOM%3A%20Networking&amp;component=Networking&amp;component=Networking%3A%20Cache&amp;component=Networking%3A%20Cookies&amp;component=Networking%3A%20DNS&amp;component=Networking%3A%20File&amp;component=Networking%3A%20HTTP&amp;component=Networking%3A%20JAR&amp;component=Networking%3A%20Proxy&amp;component=Networking%3A%20WebSockets&amp;chfieldto=2024-02-24&amp;query_format=advanced&amp;product=Core&amp;list_id=16913244&amp;classification=Client%20Software&amp;classification=Developer%20Infrastructure&amp;classification=Components&amp;classification=Server%20Software&amp;classification=Other">210</a> bugs in total
        <ul>
          <li>Closed: <a href="https://bugzilla.mozilla.org/buglist.cgi?chfield=cf_last_resolved&amp;chfieldfrom=2024-02-23&amp;product=Core&amp;classification=Client%20Software&amp;classification=Developer%20Infrastructure&amp;classification=Components&amp;classification=Server%20Software&amp;classification=Other&amp;component=DOM%3A%20Networking&amp;component=Networking&amp;component=Networking%3A%20Cache&amp;component=Networking%3A%20Cookies&amp;component=Networking%3A%20DNS&amp;component=Networking%3A%20File&amp;component=Networking%3A%20HTTP&amp;component=Networking%3A%20JAR&amp;component=Networking%3A%20Proxy&amp;component=Networking%3A%20WebSockets&amp;list_id=16913246&amp;query_format=advanced&amp;chfieldto=2024-02-24">69</a>! (including some Meta bugs)
            <ul>
              <li>S3: <a href="https://bugzilla.mozilla.org/buglist.cgi?bug_severity=S3&amp;chfieldfrom=2024-02-23&amp;query_format=advanced&amp;list_id=16913480&amp;chfield=cf_last_resolved&amp;component=DOM%3A%20Networking&amp;component=Networking&amp;component=Networking%3A%20Cache&amp;component=Networking%3A%20Cookies&amp;component=Networking%3A%20DNS&amp;component=Networking%3A%20File&amp;component=Networking%3A%20HTTP&amp;component=Networking%3A%20JAR&amp;component=Networking%3A%20Proxy&amp;component=Networking%3A%20WebSockets&amp;product=Core&amp;classification=Client%20Software&amp;classification=Developer%20Infrastructure&amp;classification=Components&amp;classification=Server%20Software&amp;classification=Other&amp;chfieldto=2024-02-24">48</a> (Total S3 reduced by 2.6%)</li>
              <li>S4: <a href="https://bugzilla.mozilla.org/buglist.cgi?chfieldto=2024-02-24&amp;list_id=16913481&amp;classification=Client%20Software&amp;classification=Developer%20Infrastructure&amp;classification=Components&amp;classification=Server%20Software&amp;classification=Other&amp;chfield=cf_last_resolved&amp;bug_severity=S4&amp;query_format=advanced&amp;component=DOM%3A%20Networking&amp;component=Networking&amp;component=Networking%3A%20Cache&amp;component=Networking%3A%20Cookies&amp;component=Networking%3A%20DNS&amp;component=Networking%3A%20File&amp;component=Networking%3A%20HTTP&amp;component=Networking%3A%20JAR&amp;component=Networking%3A%20Proxy&amp;component=Networking%3A%20WebSockets&amp;chfieldfrom=2024-02-23&amp;product=Core">3</a></li>
              <li>NA: <a href="https://bugzilla.mozilla.org/buglist.cgi?query_format=advanced&amp;chfieldfrom=2024-02-23&amp;chfield=cf_last_resolved&amp;list_id=16913484&amp;bug_severity=N%2FA&amp;chfieldto=2024-02-24&amp;classification=Client%20Software&amp;classification=Developer%20Infrastructure&amp;classification=Components&amp;classification=Server%20Software&amp;classification=Other&amp;product=Core&amp;component=DOM%3A%20Networking&amp;component=Networking&amp;component=Networking%3A%20Cache&amp;component=Networking%3A%20Cookies&amp;component=Networking%3A%20DNS&amp;component=Networking%3A%20File&amp;component=Networking%3A%20HTTP&amp;component=Networking%3A%20JAR&amp;component=Networking%3A%20Proxy&amp;component=Networking%3A%20WebSockets">18</a></li>
            </ul>
          </li>
          <li>Priority/severity updated: <a href="https://bugzilla.mozilla.org/buglist.cgi?product=Core&amp;classification=Client%20Software&amp;classification=Developer%20Infrastructure&amp;classification=Components&amp;classification=Server%20Software&amp;classification=Other&amp;chfieldto=2024-02-24&amp;component=DOM%3A%20Networking&amp;component=Networking&amp;component=Networking%3A%20Cache&amp;component=Networking%3A%20Cookies&amp;component=Networking%3A%20DNS&amp;component=Networking%3A%20File&amp;component=Networking%3A%20HTTP&amp;component=Networking%3A%20JAR&amp;component=Networking%3A%20Proxy&amp;component=Networking%3A%20WebSockets&amp;chfield=priority&amp;chfield=bug_severity&amp;list_id=16913252&amp;query_format=advanced&amp;chfieldfrom=2024-02-23">20</a></li>
          <li>Blocks updated: <a href="https://bugzilla.mozilla.org/buglist.cgi?query_format=advanced&amp;chfieldfrom=2024-02-23&amp;chfield=blocked&amp;list_id=16913464&amp;product=Core&amp;classification=Client%20Software&amp;classification=Developer%20Infrastructure&amp;classification=Components&amp;classification=Server%20Software&amp;classification=Other&amp;chfieldto=2024-02-24&amp;component=DOM%3A%20Networking&amp;component=Networking&amp;component=Networking%3A%20Cache&amp;component=Networking%3A%20Cookies&amp;component=Networking%3A%20DNS&amp;component=Networking%3A%20File&amp;component=Networking%3A%20HTTP&amp;component=Networking%3A%20JAR&amp;component=Networking%3A%20Proxy&amp;component=Networking%3A%20WebSockets">73</a></li>
          <li>Necko Priority Queue: 2 added</li>
          <li>Necko Next: 68 (added 19)</li>
          <li>Necko Priority Review: down to 16</li>
        </ul>
      </li>
      <li>Assigned owners to our Meta bugs owners <a href="https://bugzilla.mozilla.org/buglist.cgi?list_id=16940437&amp;product=Core&amp;keywords_type=allwords&amp;resolution=---&amp;query_format=advanced&amp;bug_status=UNCONFIRMED&amp;bug_status=NEW&amp;bug_status=ASSIGNED&amp;bug_status=REOPENED&amp;keywords=meta&amp;component=DOM%3A%20Networking&amp;component=Networking&amp;component=Networking%3A%20Cache&amp;component=Networking%3A%20Cookies&amp;component=Networking%3A%20DNS&amp;component=Networking%3A%20File&amp;component=Networking%3A%20HTTP&amp;component=Networking%3A%20JAR&amp;component=Networking%3A%20Proxy&amp;component=Networking%3A%20WebSockets&amp;classification=Client%20Software&amp;classification=Developer%20Infrastructure&amp;classification=Components&amp;classification=Server%20Software&amp;classification=Other">meta bugs</a></li>
    </ul>
  </li>
</ul>

<p>You can send feedback regarding this Newsletter to <a href="mailto:necko@mozilla.com">necko@mozilla.com</a> and discuss on <a href="https://groups.google.com/a/mozilla.org/g/dev-platform/c/Ytr6DpwBISw">dev-platform</a>.</p>]]></content><author><name>Manuel Bucher</name></author><category term="newsletter" /><summary type="html"><![CDATA[Highlights]]></summary></entry><entry><title type="html">Necko News #4 - 2023 H1 Recap</title><link href="https://mozilla-necko.github.io/newsletter/2023/06/30/h1-recap.html" rel="alternate" type="text/html" title="Necko News #4 - 2023 H1 Recap" /><published>2023-06-30T15:00:00+00:00</published><updated>2023-06-30T15:00:00+00:00</updated><id>https://mozilla-necko.github.io/newsletter/2023/06/30/h1-recap</id><content type="html" xml:base="https://mozilla-necko.github.io/newsletter/2023/06/30/h1-recap.html"><![CDATA[<h2 id="highlights">Highlights</h2>

<ul>
  <li>Dylan O’Toole joined recently as a Networking Intern based in Canada</li>
  <li>Randell Jesup joined the Necko Team temporarily to help out with WebTransport</li>
  <li>In May we had a workweek in Berlin to collaborate with the wider Sec-Privacy-Necko group and align upcoming projects and work on HTTPS adoption, storage, cookies, speculative connect, pre-connect and performance (and more!).</li>
  <li>We have an opening on the team for a <a href="https://www.mozilla.org/en-US/careers/position/gh/5006202/">Senior Staff Software Engineer</a></li>
</ul>

<h2 id="friends-of-necko">Friends of Necko</h2>

<p>We appreciate all new contributors working on <a href="https://mzl.la/3ONctUp">Good first bugs in the
networking component</a> since our last
Newsletter:</p>

<ul>
  <li>Ho Cheung with <a href="https://phabricator.services.mozilla.com/D167348">D167348</a>, <a href="https://phabricator.services.mozilla.com/D167347">D167347</a> and <a href="https://phabricator.services.mozilla.com/D180526">D180526</a></li>
  <li>CanadaHonk with <a href="https://phabricator.services.mozilla.com/D170404">D170404</a>, <a href="https://phabricator.services.mozilla.com/D170528">D170528</a></li>
  <li>Asbjørn Sloth Tønnesen with <a href="https://phabricator.services.mozilla.com/D166224">D166224</a></li>
  <li>Richard Broker with <a href="https://phabricator.services.mozilla.com/D166808">D166808</a></li>
  <li>Saksham Mittal with <a href="https://phabricator.services.mozilla.com/D170289">D170289</a></li>
  <li>Sankalp Shubham with <a href="https://phabricator.services.mozilla.com/D177957">D177957</a></li>
  <li>Jake Senne with <a href="https://phabricator.services.mozilla.com/D182285">D182285</a></li>
</ul>

<p>And also longer time contributors:</p>

<ul>
  <li>Nicklas Boman with <a href="https://phabricator.services.mozilla.com/D178375">D178375</a>, <a href="https://phabricator.services.mozilla.com/D177690">D177690</a></li>
  <li>Robert Longson with <a href="https://phabricator.services.mozilla.com/D175563">D175563</a></li>
</ul>

<p>One contribution from the Tor project was on Necko:</p>

<ul>
  <li>Pier Angelo Vendrame with <a href="https://phabricator.services.mozilla.com/D170373">D170373</a> - Add a seed to the network ID, this helps to prevent linkability of users in the same network.</li>
</ul>

<h2 id="project-updates">Project Updates</h2>

<ul>
  <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1709355">WebTransport</a> released in Fx114 on 2023-06-06</li>
  <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1610741">DoH Settings UI</a> improvements released in Fx114 on 2023-06-06</li>
  <li><a href="https://developer.mozilla.org/en-US/docs/Web/HTTP/Status/103">Early Hints</a> has been enabled in Nightly and Beta in Fx111 on 2023-01-19. Experiments in Beta show that there are still some performance investigations necessary before confidentially shipping it.</li>
  <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1815647">URL interop 2023</a> is under way to improve our web platform test results.</li>
  <li>The <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1528285">Off-Main-Thread</a> project is in the early stages as we decide its scope.</li>
  <li>
    <p>Expanding <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1784079">Mochitest with Http2/3 support</a> is wrapping up and will help prevent future issues similar to FoxStuck by increasing code coverage of HTTP/2 and HTTP/3 code in more realistic scenarios.</p>
  </li>
  <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1816678">Speculative Connections</a>:
    <ul>
      <li>We identified problems in our speculative connection code paths. These have since been fixed - <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1813618">https connections</a>, <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1543990">rel=preconnect</a> - leading to <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1816678#c6">measured performance improvements.</a></li>
      <li>Improved tests have landed with more <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1818798">underway</a>.</li>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1816539">An experiment</a> is scheduled to optimize the number of sockets available for speculative connections.</li>
    </ul>
  </li>
  <li>Oblivious DNS over HTTP (ODoH) <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1835805">has been superseded</a> by DNS over Oblivious HTTP (DoOH).</li>
  <li>DNS over Oblivious HTTP (DoOH) experiment has completed, identifying OHTTP proxies have negligible impact on page load performance opening the opportunity to further explore and identify uses for OHTTP.</li>
  <li>DoH warnings experiments are running evaluating the impact of informing users when 3rd parties actively prevent DoH (blocking, network canary signal).</li>
  <li>The <a href="https://firefox-source-docs.mozilla.org/mots/index.html#core-cookies">cookie module</a> now has new owners and peers.</li>
</ul>

<h2 id="below-the-fold">Below the fold</h2>

<ul>
  <li>We fixed a very old bug in our <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1596576">HTTP/2 Upload codepath</a> which restricted performance on faster networks.</li>
  <li>Andrew Creskey <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1810871">exposed internal necko timings to PerfStats</a> so that they can be measured within our performance test infrastructure. These are also visible from telemetry on <a href="https://sql.telemetry.mozilla.org/dashboard/test_42">this dashboard</a></li>
  <li>Necko team adapted <a href="https://mzl.la/43g6qjw">priority queue</a> system with bugs that the team looks at next</li>
  <li><a href="https://wiki.mozilla.org/Necko/The_Necko_Story">The Necko Story</a> is now available in the Wiki - an interesting read about the background around Necko.</li>
  <li>We are working on improving our documentation in <a href="https://wiki.mozilla.org/Networking">the wiki</a> and on <a href="https://firefox-source-docs.mozilla.org/networking/">firefox-source-docs</a>.</li>
  <li>We are experimenting with novel ways of knowledge sharing through <strong>Decipher Necko</strong> information sessions.</li>
</ul>

<p>Read this newsletter on <a href="https://groups.google.com/a/mozilla.org/g/dev-platform/c/hvd--aoDjzo">dev-platform</a></p>]]></content><author><name>Manuel Bucher</name></author><category term="newsletter" /><summary type="html"><![CDATA[Highlights]]></summary></entry><entry><title type="html">Necko News #3 - Autumn 2022</title><link href="https://mozilla-necko.github.io/newsletter/2022/10/21/autumn.html" rel="alternate" type="text/html" title="Necko News #3 - Autumn 2022" /><published>2022-10-21T07:20:00+00:00</published><updated>2022-10-21T07:20:00+00:00</updated><id>https://mozilla-necko.github.io/newsletter/2022/10/21/autumn</id><content type="html" xml:base="https://mozilla-necko.github.io/newsletter/2022/10/21/autumn.html"><![CDATA[<h2 id="highlights">Highlights</h2>

<ul>
  <li>This is Dragana last week on Necko. We would all like to thank her for years of leadership and hard work and wish her only the best going forward.</li>
</ul>

<h2 id="project-updates">Project Updates</h2>

<ul>
  <li>Dragana and Kershaw have completed a big refactoring of <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1754731">the proxy code</a></li>
  <li>Manuel has been working on 103 early hints</li>
  <li>Kershaw has written a great proposal of adding HTTP/2 and HTTP/3 support to mochitests. The team intends to work on it later this year.</li>
  <li>Sunil has fixed a <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1697421">webcompat</a> issue and will continue to work on improving our fetch web-platform test pass rate.</li>
  <li>Dragana and Kershaw have been working on <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1709355">WebTransport</a></li>
  <li>Andrew Creskey from the Performance team has been adding probes <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1777538">measuring network timings</a> that are internal to Gecko. The results will be used to guide and prioritize performance optimizations.</li>
  <li>Ed has completed a cross-origin cookie broadcast fix and is working on fixing websockets over http2.</li>
  <li>Valentin added <a href="https://searchfox.org/mozilla-central/source/netwerk/test/unit/head_servers.js">test helpers</a> that spawn node instances for HTTP/S, HTTP/2 and proxy servers. These will help us to write unit tests faster.</li>
</ul>

<h2 id="below-the-fold">Below the fold</h2>

<ul>
  <li>Kagami has been fixing some problems with the <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1791675">purgeHTTPCache background task</a>, which will allow us to enable it for beta and release channels. That will get rid of a lot of pesky windows shutdown hangs.</li>
  <li>John Schanck has helped us write a <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1754746">Faulty TLS server</a> that will allow us to write better tests for error conditions in networking code.</li>
  <li>Kershaw added a <a href="https://searchfox.org/mozilla-central/rev/a64647a2125cf3d334451051491fef6772e8eb57/netwerk/base/nsIRequest.idl#209">new API in nsIRequest.idl</a> to allow consumers of nsIRequest to provide the reason for cancellation.</li>
  <li>ObliviousHTTP implementation is in-progress with Dana Keeler from NSS and support from Necko, targeting experimentation in 2023</li>
  <li>Paul Adenot <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1347955">moved the logging section</a> of about:networking to about:logging.</li>
</ul>

<p>Read this newsletter on <a href="https://groups.google.com/a/mozilla.org/g/dev-platform/c/bsQUnJuYVxQ/m/H_k3HWdUBQAJ">dev-platform</a></p>]]></content><author><name>Valentin Gosu</name></author><category term="newsletter" /><summary type="html"><![CDATA[Highlights]]></summary></entry><entry><title type="html">Necko News #2 - Season’s greetings</title><link href="https://mozilla-necko.github.io/newsletter/2021/12/23/seasons-greetings.html" rel="alternate" type="text/html" title="Necko News #2 - Season’s greetings" /><published>2021-12-23T16:00:00+00:00</published><updated>2021-12-23T16:00:00+00:00</updated><id>https://mozilla-necko.github.io/newsletter/2021/12/23/seasons-greetings</id><content type="html" xml:base="https://mozilla-necko.github.io/newsletter/2021/12/23/seasons-greetings.html"><![CDATA[<p>Season’s greetings!</p>

<p>The Necko team is pleased to provide an early Christmas present with our second newsletter! :-)</p>

<h2 id="highlights">Highlights</h2>

<p>Greg Hess has recently joined Mozilla as the new Necko team manager bringing 20yrs experience in software development. He was born a developer(Java first love) and followed a technical track moving on to team lead, client architect and technical director before entering management/agile coaching. He has a broad range of market experience developing web and mobile applications for SaaS hosting, Home Media Servers, Premium Online TV, Aviation and Defense, Digital Marketing and Social Media. Some notable projects he has worked on are Netgear Stora, Disney Channel(CA), Cartoon Network(US-2016 Emmy award), Barbie.com and McDonalds Happy Studio(EU).</p>

<h2 id="project-updates">Project Updates</h2>

<ul>
  <li>HTTP3
    <ul>
      <li>Dragana has added WebTransport support to neqo</li>
      <li>Manuel added the HTTP priority extension. This is enabled on <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1734132">Beta</a></li>
    </ul>
  </li>
  <li>Socket process
    <ul>
      <li>Kershaw moved the evaluation of PAC scripts into the socket process. This is currently enabled on <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1745385">Nightly</a></li>
      <li>Kershaw made it possible to <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1739001">generate the connection info asynchronously</a> for TRR channels in the socket process</li>
    </ul>
  </li>
  <li>DNS over HTTPS (Trusted-Recursive-Resolver)
    <ul>
      <li>Nihanth implemented a <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1737198">retry mechanism for TRR requests.</a></li>
      <li>Valentin fixed a bug that caused some <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1566998">DNS queries to be leaked to the native DNS resolver</a> when resolved by WebRTC.</li>
      <li><a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1543811">DNS padding</a> shipped on release 95</li>
    </ul>
  </li>
</ul>

<h2 id="below-the-fold">Below the fold</h2>

<ul>
  <li>The Necko team re-triaged all of our unassigned P2 bugs - some of them became P3 and some of them received some much needed attention.</li>
  <li>Valentin is working on <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1705676">purging the HTTP cache in a background task during shutdown</a>. This should get rid of some annoying <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1356853">shutdown</a> <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1682899">hangs</a> on windows.</li>
  <li>Authentication challenges are now <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=650091">sorted by their safety rating</a> instead of the order they’re sent by the server.</li>
  <li>Dragana and Manuel have started work on <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1407355">103 Early Hints</a></li>
</ul>

<p>We would like to wish you all a happy and healthy holiday season!</p>

<ul>
  <li>Necko Team</li>
</ul>

<p>Read this newsletter on <a href="https://groups.google.com/a/mozilla.org/g/dev-platform/c/QbIZTTUlrIQ/m/8L7YEu_kDQAJ">dev-platform</a></p>]]></content><author><name>Dragana Damjanovic</name></author><category term="newsletter" /><summary type="html"><![CDATA[Season’s greetings!]]></summary></entry><entry><title type="html">Necko News #1 - First Necko Newsletter</title><link href="https://mozilla-necko.github.io/newsletter/2021/09/27/first-necko-newsletter.html" rel="alternate" type="text/html" title="Necko News #1 - First Necko Newsletter" /><published>2021-09-27T12:45:00+00:00</published><updated>2021-09-27T12:45:00+00:00</updated><id>https://mozilla-necko.github.io/newsletter/2021/09/27/first-necko-newsletter</id><content type="html" xml:base="https://mozilla-necko.github.io/newsletter/2021/09/27/first-necko-newsletter.html"><![CDATA[<h2 id="highlights">Highlights</h2>

<p>Valentin Gosu is now the module owner of <a href="https://wiki.mozilla.org/Modules/All#libjar">libjar</a>. Many thanks to Aaron Klotz for the great work maintaining this module in recent years. Kershaw Chang is now a peer of libjar.</p>

<p>Manuel Bucher has recently joined the team as a student worker based in Germany.</p>

<p>Dragana <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1382886">refactored how necko drives TLS</a> handshakes and improved the CPU usage during 0RTT</p>

<h2 id="friends-of-the-necko-team">Friends of the Necko team</h2>

<p>Glenn Strauss submitted a patch a while ago adding support for <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=472823">SHA-256 Digest authentication</a>. Because the code touched by the patch was quite old this caused us to do a refactoring of the authentication code to bring it closer to the <a href="https://firefox-source-docs.mozilla.org/code-quality/coding-style/coding_style_cpp.html">Mozilla coding guidelines</a> and add better tests.</p>

<p>Ping Chen (:rnons) fixed <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1717185">Bug 1717185 - TCPSocket.send incorrectly returns true because mBufferedAmount is not updated in time</a></p>

<h2 id="project-updates">Project Updates</h2>

<ul>
  <li>HTTPS RR
    <ul>
      <li><a href="https://developer.mozilla.org/en-US/docs/Glossary/https_rr">HTTPS RR</a> shipped on release 92.</li>
    </ul>
  </li>
  <li>Socket process
    <ul>
      <li>Dana is working on <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1712837">bug 1712837</a> to make OS client certificates work with the socket process.</li>
      <li>Kershaw made websocket work with the socket process. See <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1716566">bug 1716566</a>.</li>
    </ul>
  </li>
  <li>DNS over HTTPS
    <ul>
      <li>DoH shipped in Canada in Firefox 92</li>
      <li>Manuel is adding support for <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1543811">DNS padding</a> to mitigate traffic analysis</li>
    </ul>
  </li>
</ul>

<p>Nihanth is making our TRR implementation (Trusted-Recursive-Resolver) <a href="https://bugzilla.mozilla.org/show_bug.cgi?id=1714182">fall back to unencrypted DNS less often</a></p>

<p>Read this newsletter on <a href="https://groups.google.com/a/mozilla.org/g/dev-platform/c/_hYnRM1DvRg/m/7kVA3a2gAgAJ">dev-platform</a></p>]]></content><author><name>Valentin Gosu</name></author><category term="newsletter" /><summary type="html"><![CDATA[Highlights]]></summary></entry></feed>