Project: chromium Issues People Development process History Sign in
New issue
Advanced search Search tips
Issue 624462 Reduce cases where HTTP/0.9 is supported
Starred by 14 users Project Member Reported by mmenke@chromium.org, Jun 29 2016 Back to list
Status: Fixed
Owner:
Closed: Jan 12
Cc:
Components:
EstimatedDays: ----
NextAction: ----
OS: All
Pri: 2
Type: Bug

Blocked on:
issue 637017

Blocking:
issue 600352
issue 669800



Sign in to add a comment
HTTP/0.9 supports leads to problems when trying to talk to non-HTTP/0.9 (Some problem security related, some problems not), moreover, it often leads to weird behaviors when talking to buggy servers.

The simplest solution, most robust, solution to the problems is causes is just to remove it.  Modern servers should be talking HTTP/1.x, or better, HTTP/2.

According to our numbers, ~0.01% of HTTP responses are HTTP/0.9 (that count excludes HTTP/2 and QUIC requests).  Each week, about 1% of Chrome users see a response that's interpreted as HTTP/0.9.  Note that these numbers are over estimates - when a broken server sends us bogus data in response to an HTTP request, we'll often treat it as an HTTP/0.9 response.  It's unclear how often this happens.
 
I'd love to see this happen.  However, to allow for random embedded devices and scripts that still only support HTTP/0.9, would it be reasonable to have a release or two that disables HTTP/0.9 by default but allows this behind a flag (and have the documentation for the flag explicitly say to come to this issue and discuss the problem)?
Comment 2 by mmenke@chromium.org, Jun 30 2016
Would you mind making that comment on the blink-dev list (https://groups.google.com/a/chromium.org/forum/#!topic/blink-dev/OdKnpLlvVUo)?  That gets your response more visibility, and gives a chance for others to chime in if they agree.

I don't particularly oppose putting it behind a flag, but I do wonder if embedded systems would be updated in a reasonable enough time frame to make that actually useful.  We have enough problems command line flag proliferation that I don't want to add another, if it's going to stick around for a while.  Better to go into this more on blink-dev, because I don't want to just dismiss your suggestion out of hand myself.  Want to see if there's more support for your proposal.
I would also suggest enterprise policy.
Project Member Comment 4 by bugdroid1@chromium.org, Jul 11 2016
The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/8210acc6749feb3d372cd1aa0456eabe029a03af

commit 8210acc6749feb3d372cd1aa0456eabe029a03af
Author: mmenke <mmenke@chromium.org>
Date: Mon Jul 11 16:34:52 2016

Add metrics to track HTTP/0.9 usage for main frames and subresources

This will tell us how many more error pages we'll be showing if we
remove support for it.

BUG= 624462 

Review-Url: https://codereview.chromium.org/2119623002
Cr-Commit-Position: refs/heads/master@{#404673}

[modify] https://crrev.com/8210acc6749feb3d372cd1aa0456eabe029a03af/components/domain_reliability/util.cc
[modify] https://crrev.com/8210acc6749feb3d372cd1aa0456eabe029a03af/content/browser/loader/resource_loader.cc
[modify] https://crrev.com/8210acc6749feb3d372cd1aa0456eabe029a03af/net/http/http_basic_stream.cc
[modify] https://crrev.com/8210acc6749feb3d372cd1aa0456eabe029a03af/net/http/http_response_info.cc
[modify] https://crrev.com/8210acc6749feb3d372cd1aa0456eabe029a03af/net/http/http_response_info.h
[modify] https://crrev.com/8210acc6749feb3d372cd1aa0456eabe029a03af/net/http/http_stream_parser.cc
[modify] https://crrev.com/8210acc6749feb3d372cd1aa0456eabe029a03af/net/http/http_stream_parser_unittest.cc
[modify] https://crrev.com/8210acc6749feb3d372cd1aa0456eabe029a03af/net/quic/quic_network_transaction_unittest.cc
[modify] https://crrev.com/8210acc6749feb3d372cd1aa0456eabe029a03af/net/spdy/spdy_network_transaction_unittest.cc
[modify] https://crrev.com/8210acc6749feb3d372cd1aa0456eabe029a03af/tools/metrics/histograms/histograms.xml

Comment 5 by mmenke@chromium.org, Jul 11 2016
Labels: Merge-Request-53
Requesting permission to merge https://codereview.chromium.org/2119623002 to M-53.  It will gather stats only, will not affect behavior.
Comment 6 by dimu@google.com, Jul 12 2016
Labels: -Merge-Request-53 Merge-Approved-53 Hotlist-Merge-Approved
Your change meets the bar and is auto-approved for M53 (branch: 2785)
Project Member Comment 7 by bugdroid1@chromium.org, Jul 12 2016
Labels: -merge-approved-53 merge-merged-2785
The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/fce9aeafb47dc54f995a8dbc7426d66a047b78a9

commit fce9aeafb47dc54f995a8dbc7426d66a047b78a9
Author: mmenke <mmenke@chromium.org>
Date: Tue Jul 12 18:21:08 2016

Add metrics to track HTTP/0.9 usage for main frames and subresources

This will tell us how many more error pages we'll be showing if we
remove support for it.

TBR=mmenke@chromium.org
NOTRY=true
NOPRESUBMIT=true
NOWEARHAT=true
BUG= 624462 

Review-Url: https://codereview.chromium.org/2119623002
Cr-Commit-Position: refs/heads/master@{#404673}
(cherry picked from commit 8210acc6749feb3d372cd1aa0456eabe029a03af)

Review-Url: https://codereview.chromium.org/2140253002
Cr-Commit-Position: refs/branch-heads/2785@{#97}
Cr-Branched-From: 68623971be0cfc492a2cb0427d7f478e7b214c24-refs/heads/master@{#403382}

[modify] https://crrev.com/fce9aeafb47dc54f995a8dbc7426d66a047b78a9/components/domain_reliability/util.cc
[modify] https://crrev.com/fce9aeafb47dc54f995a8dbc7426d66a047b78a9/content/browser/loader/resource_loader.cc
[modify] https://crrev.com/fce9aeafb47dc54f995a8dbc7426d66a047b78a9/net/http/http_basic_stream.cc
[modify] https://crrev.com/fce9aeafb47dc54f995a8dbc7426d66a047b78a9/net/http/http_response_info.cc
[modify] https://crrev.com/fce9aeafb47dc54f995a8dbc7426d66a047b78a9/net/http/http_response_info.h
[modify] https://crrev.com/fce9aeafb47dc54f995a8dbc7426d66a047b78a9/net/http/http_stream_parser.cc
[modify] https://crrev.com/fce9aeafb47dc54f995a8dbc7426d66a047b78a9/net/http/http_stream_parser_unittest.cc
[modify] https://crrev.com/fce9aeafb47dc54f995a8dbc7426d66a047b78a9/net/quic/quic_network_transaction_unittest.cc
[modify] https://crrev.com/fce9aeafb47dc54f995a8dbc7426d66a047b78a9/net/spdy/spdy_network_transaction_unittest.cc
[modify] https://crrev.com/fce9aeafb47dc54f995a8dbc7426d66a047b78a9/tools/metrics/histograms/histograms.xml

Project Member Comment 8 by bugdroid1@chromium.org, Jul 15 2016
The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/e2172e4b83ad33aef354618ce2e092db85baee1d

commit e2172e4b83ad33aef354618ce2e092db85baee1d
Author: mmenke <mmenke@chromium.org>
Date: Fri Jul 15 03:14:32 2016

Remove HTTP/0.9 support from HttpStreamParser.

HTTP/0.9 is an ancient protocol with some fundamental design problems.
This CL remove support for it, while leaving most of the code needed to
support HTTP/0.9 in place, in case this has to be reverted late in the
game, or we decide to add a (temporary) enterprise policy to enable it.

BUG= 624462 

Review-Url: https://codereview.chromium.org/2144803002
Cr-Commit-Position: refs/heads/master@{#405687}

[modify] https://crrev.com/e2172e4b83ad33aef354618ce2e092db85baee1d/components/error_page/common/localized_error.cc
[modify] https://crrev.com/e2172e4b83ad33aef354618ce2e092db85baee1d/net/base/net_error_list.h
[modify] https://crrev.com/e2172e4b83ad33aef354618ce2e092db85baee1d/net/http/http_network_transaction_unittest.cc
[modify] https://crrev.com/e2172e4b83ad33aef354618ce2e092db85baee1d/net/http/http_stream_parser.cc
[modify] https://crrev.com/e2172e4b83ad33aef354618ce2e092db85baee1d/net/http/http_stream_parser_unittest.cc
[modify] https://crrev.com/e2172e4b83ad33aef354618ce2e092db85baee1d/net/url_request/url_request_http_job_unittest.cc
[modify] https://crrev.com/e2172e4b83ad33aef354618ce2e092db85baee1d/net/websockets/websocket_stream_test.cc
[modify] https://crrev.com/e2172e4b83ad33aef354618ce2e092db85baee1d/third_party/WebKit/LayoutTests/http/tests/websocket/handshake-error-expected.txt
[modify] https://crrev.com/e2172e4b83ad33aef354618ce2e092db85baee1d/third_party/WebKit/LayoutTests/http/tests/websocket/handshake-fail-by-maxlength-expected.txt
[modify] https://crrev.com/e2172e4b83ad33aef354618ce2e092db85baee1d/third_party/WebKit/LayoutTests/http/tests/websocket/handshake-fail-by-prepended-null-expected.txt
[modify] https://crrev.com/e2172e4b83ad33aef354618ce2e092db85baee1d/tools/metrics/histograms/histograms.xml

Blocking: 600352
Labels: M-54
Status: Fixed
Blockedon: 637017
Project Member Comment 13 by bugdroid1@chromium.org, Aug 16 2016
The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/a2dcd3bf444490b1eaf7aaafaa740310dc846137

commit a2dcd3bf444490b1eaf7aaafaa740310dc846137
Author: mmenke <mmenke@chromium.org>
Date: Tue Aug 16 21:49:41 2016

Revert HTTP/0.9 removal

This broke compatibility with the Tenda D301 router, which is using
HTTP/0.9 (Presumably because of a bug in the router's firmare).
We're going to go with a less aggressive deprecation instead, at least
for now - we'll just support HTTP/0.9 on default ports only.

I'll try to be more aggressive in future Chrome releases (No HTTP/0.9
over SSL, for instance).

Reverted CL description:
 Remove HTTP/0.9 support from HttpStreamParser.

 HTTP/0.9 is an ancient protocol with some fundamental design problems.
 This CL remove support for it, while leaving most of the code needed to
 support HTTP/0.9 in place, in case this has to be reverted late in the
 game, or we decide to add a (temporary) enterprise policy to enable it.

 BUG= 624462 
 Committed: https://crrev.com/e2172e4b83ad33aef354618ce2e092db85baee1d
 Cr-Commit-Position: refs/heads/master@{#405687}

TBR=eroman@chromium.org, ricea@chromium.org,  jwd@chromium.org
BUG= 637017 

Review-Url: https://codereview.chromium.org/2247003003
Cr-Commit-Position: refs/heads/master@{#412348}

[modify] https://crrev.com/a2dcd3bf444490b1eaf7aaafaa740310dc846137/net/http/http_network_transaction_unittest.cc
[modify] https://crrev.com/a2dcd3bf444490b1eaf7aaafaa740310dc846137/net/http/http_stream_parser.cc
[modify] https://crrev.com/a2dcd3bf444490b1eaf7aaafaa740310dc846137/net/http/http_stream_parser_unittest.cc
[modify] https://crrev.com/a2dcd3bf444490b1eaf7aaafaa740310dc846137/net/url_request/url_request_http_job_unittest.cc
[modify] https://crrev.com/a2dcd3bf444490b1eaf7aaafaa740310dc846137/net/websockets/websocket_stream_test.cc
[modify] https://crrev.com/a2dcd3bf444490b1eaf7aaafaa740310dc846137/third_party/WebKit/LayoutTests/http/tests/websocket/handshake-error-expected.txt
[modify] https://crrev.com/a2dcd3bf444490b1eaf7aaafaa740310dc846137/third_party/WebKit/LayoutTests/http/tests/websocket/handshake-fail-by-maxlength-expected.txt
[modify] https://crrev.com/a2dcd3bf444490b1eaf7aaafaa740310dc846137/third_party/WebKit/LayoutTests/http/tests/websocket/handshake-fail-by-prepended-null-expected.txt

Status: Assigned
Summary: Reduce cases where HTTP/0.9 is supported (was: Remove HTTP/0.9 support)
Repurposing this bug.  It turns out that we can't just remove HTTP/0.9 support without breaking the interfaces of at least some home routers (Tenda D301), which we're not currently comfortable doing.  So instead, we're going to reduce the cases where it's supported.

First we'll just allow it on default HTTP/HTTPS ports, but after that, I want to try to continue restricting it further (Never over HTTPS, for instance)
Project Member Comment 15 by bugdroid1@chromium.org, Aug 31 2016
The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/398c105ea8391473126021104554d70fda08324b

commit 398c105ea8391473126021104554d70fda08324b
Author: mmenke <mmenke@chromium.org>
Date: Wed Aug 31 16:06:57 2016

Only allow HTTP/0.9 support on default ports.

Also add a group policy to allow HTTP/0.9 support on other
ports, which I plan to remove in a couple revisions.

BUG= 624462 , 637017 

Review-Url: https://codereview.chromium.org/2253653002
Cr-Commit-Position: refs/heads/master@{#415643}

[modify] https://crrev.com/398c105ea8391473126021104554d70fda08324b/chrome/browser/io_thread.cc
[modify] https://crrev.com/398c105ea8391473126021104554d70fda08324b/chrome/browser/io_thread.h
[modify] https://crrev.com/398c105ea8391473126021104554d70fda08324b/chrome/browser/io_thread_unittest.cc
[modify] https://crrev.com/398c105ea8391473126021104554d70fda08324b/chrome/browser/net/errorpage_browsertest.cc
[modify] https://crrev.com/398c105ea8391473126021104554d70fda08324b/chrome/test/data/policy/policy_test_cases.json
[modify] https://crrev.com/398c105ea8391473126021104554d70fda08324b/components/policy/resources/policy_templates.json
[modify] https://crrev.com/398c105ea8391473126021104554d70fda08324b/net/http/http_basic_state.cc
[modify] https://crrev.com/398c105ea8391473126021104554d70fda08324b/net/http/http_basic_state.h
[modify] https://crrev.com/398c105ea8391473126021104554d70fda08324b/net/http/http_basic_state_unittest.cc
[modify] https://crrev.com/398c105ea8391473126021104554d70fda08324b/net/http/http_basic_stream.cc
[modify] https://crrev.com/398c105ea8391473126021104554d70fda08324b/net/http/http_basic_stream.h
[modify] https://crrev.com/398c105ea8391473126021104554d70fda08324b/net/http/http_network_session.cc
[modify] https://crrev.com/398c105ea8391473126021104554d70fda08324b/net/http/http_network_session.h
[modify] https://crrev.com/398c105ea8391473126021104554d70fda08324b/net/http/http_network_transaction_unittest.cc
[modify] https://crrev.com/398c105ea8391473126021104554d70fda08324b/net/http/http_stream_factory_impl_job.cc
[modify] https://crrev.com/398c105ea8391473126021104554d70fda08324b/net/http/http_stream_factory_impl_job_controller_unittest.cc
[modify] https://crrev.com/398c105ea8391473126021104554d70fda08324b/net/http/http_stream_parser.cc
[modify] https://crrev.com/398c105ea8391473126021104554d70fda08324b/net/http/http_stream_parser.h
[modify] https://crrev.com/398c105ea8391473126021104554d70fda08324b/net/http/http_stream_parser_unittest.cc
[modify] https://crrev.com/398c105ea8391473126021104554d70fda08324b/net/spdy/spdy_test_util_common.cc
[modify] https://crrev.com/398c105ea8391473126021104554d70fda08324b/net/spdy/spdy_test_util_common.h
[modify] https://crrev.com/398c105ea8391473126021104554d70fda08324b/net/test/embedded_test_server/default_handlers.cc
[modify] https://crrev.com/398c105ea8391473126021104554d70fda08324b/net/websockets/websocket_basic_handshake_stream.cc
[modify] https://crrev.com/398c105ea8391473126021104554d70fda08324b/tools/metrics/histograms/histograms.xml

Project Member Comment 16 by bugdroid1@chromium.org, Aug 31 2016
The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/400eb587ad05c2aef8740e2bff550083ac90ea2d

commit 400eb587ad05c2aef8740e2bff550083ac90ea2d
Author: mmenke <mmenke@chromium.org>
Date: Wed Aug 31 17:20:48 2016

Revert "Only allow HTTP/0.9 support on default ports."

The CL broke some websocket layout tests, will reland once that's
worked out.

This reverts commit 398c105ea8391473126021104554d70fda08324b.

Also add a group policy to allow HTTP/0.9 support on other
ports, which I plan to remove in a couple revisions.

BUG= 624462 ,  637017 
Committed: https://crrev.com/398c105ea8391473126021104554d70fda08324b
Cr-Commit-Position: refs/heads/master@{#415643}
TBR=eroman@chromium.org
NOTRY=true

Review-Url: https://codereview.chromium.org/2297263002
Cr-Commit-Position: refs/heads/master@{#415668}

[modify] https://crrev.com/400eb587ad05c2aef8740e2bff550083ac90ea2d/chrome/browser/io_thread.cc
[modify] https://crrev.com/400eb587ad05c2aef8740e2bff550083ac90ea2d/chrome/browser/io_thread.h
[modify] https://crrev.com/400eb587ad05c2aef8740e2bff550083ac90ea2d/chrome/browser/io_thread_unittest.cc
[modify] https://crrev.com/400eb587ad05c2aef8740e2bff550083ac90ea2d/chrome/browser/net/errorpage_browsertest.cc
[modify] https://crrev.com/400eb587ad05c2aef8740e2bff550083ac90ea2d/chrome/test/data/policy/policy_test_cases.json
[modify] https://crrev.com/400eb587ad05c2aef8740e2bff550083ac90ea2d/components/policy/resources/policy_templates.json
[modify] https://crrev.com/400eb587ad05c2aef8740e2bff550083ac90ea2d/net/http/http_basic_state.cc
[modify] https://crrev.com/400eb587ad05c2aef8740e2bff550083ac90ea2d/net/http/http_basic_state.h
[modify] https://crrev.com/400eb587ad05c2aef8740e2bff550083ac90ea2d/net/http/http_basic_state_unittest.cc
[modify] https://crrev.com/400eb587ad05c2aef8740e2bff550083ac90ea2d/net/http/http_basic_stream.cc
[modify] https://crrev.com/400eb587ad05c2aef8740e2bff550083ac90ea2d/net/http/http_basic_stream.h
[modify] https://crrev.com/400eb587ad05c2aef8740e2bff550083ac90ea2d/net/http/http_network_session.cc
[modify] https://crrev.com/400eb587ad05c2aef8740e2bff550083ac90ea2d/net/http/http_network_session.h
[modify] https://crrev.com/400eb587ad05c2aef8740e2bff550083ac90ea2d/net/http/http_network_transaction_unittest.cc
[modify] https://crrev.com/400eb587ad05c2aef8740e2bff550083ac90ea2d/net/http/http_stream_factory_impl_job.cc
[modify] https://crrev.com/400eb587ad05c2aef8740e2bff550083ac90ea2d/net/http/http_stream_factory_impl_job_controller_unittest.cc
[modify] https://crrev.com/400eb587ad05c2aef8740e2bff550083ac90ea2d/net/http/http_stream_parser.cc
[modify] https://crrev.com/400eb587ad05c2aef8740e2bff550083ac90ea2d/net/http/http_stream_parser.h
[modify] https://crrev.com/400eb587ad05c2aef8740e2bff550083ac90ea2d/net/http/http_stream_parser_unittest.cc
[modify] https://crrev.com/400eb587ad05c2aef8740e2bff550083ac90ea2d/net/spdy/spdy_test_util_common.cc
[modify] https://crrev.com/400eb587ad05c2aef8740e2bff550083ac90ea2d/net/spdy/spdy_test_util_common.h
[modify] https://crrev.com/400eb587ad05c2aef8740e2bff550083ac90ea2d/net/test/embedded_test_server/default_handlers.cc
[modify] https://crrev.com/400eb587ad05c2aef8740e2bff550083ac90ea2d/net/websockets/websocket_basic_handshake_stream.cc
[modify] https://crrev.com/400eb587ad05c2aef8740e2bff550083ac90ea2d/tools/metrics/histograms/histograms.xml

Project Member Comment 17 by bugdroid1@chromium.org, Sep 1 2016
The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/a7da6da864dce77fe1c931653635c3ac757219cb

commit a7da6da864dce77fe1c931653635c3ac757219cb
Author: mmenke <mmenke@chromium.org>
Date: Thu Sep 01 21:56:52 2016

Re-land:  Only allow HTTP/0.9 support on default ports.

Also add a group policy to allow HTTP/0.9 support on other
ports, which I plan to remove in a couple revisions.

Originally landed in https://codereview.chromium.org/2253653002/
Reverted in https://codereview.chromium.org/2297263002/
for breaking WebSocket tests.

TBR=eroman@chromium.org
BUG= 624462 ,  637017 

Review-Url: https://codereview.chromium.org/2299703003
Cr-Commit-Position: refs/heads/master@{#416074}

[modify] https://crrev.com/a7da6da864dce77fe1c931653635c3ac757219cb/chrome/browser/io_thread.cc
[modify] https://crrev.com/a7da6da864dce77fe1c931653635c3ac757219cb/chrome/browser/io_thread.h
[modify] https://crrev.com/a7da6da864dce77fe1c931653635c3ac757219cb/chrome/browser/io_thread_unittest.cc
[modify] https://crrev.com/a7da6da864dce77fe1c931653635c3ac757219cb/chrome/browser/net/errorpage_browsertest.cc
[modify] https://crrev.com/a7da6da864dce77fe1c931653635c3ac757219cb/chrome/test/data/policy/policy_test_cases.json
[modify] https://crrev.com/a7da6da864dce77fe1c931653635c3ac757219cb/components/policy/resources/policy_templates.json
[modify] https://crrev.com/a7da6da864dce77fe1c931653635c3ac757219cb/net/http/http_basic_state.cc
[modify] https://crrev.com/a7da6da864dce77fe1c931653635c3ac757219cb/net/http/http_basic_state.h
[modify] https://crrev.com/a7da6da864dce77fe1c931653635c3ac757219cb/net/http/http_basic_state_unittest.cc
[modify] https://crrev.com/a7da6da864dce77fe1c931653635c3ac757219cb/net/http/http_basic_stream.cc
[modify] https://crrev.com/a7da6da864dce77fe1c931653635c3ac757219cb/net/http/http_basic_stream.h
[modify] https://crrev.com/a7da6da864dce77fe1c931653635c3ac757219cb/net/http/http_network_session.cc
[modify] https://crrev.com/a7da6da864dce77fe1c931653635c3ac757219cb/net/http/http_network_session.h
[modify] https://crrev.com/a7da6da864dce77fe1c931653635c3ac757219cb/net/http/http_network_transaction_unittest.cc
[modify] https://crrev.com/a7da6da864dce77fe1c931653635c3ac757219cb/net/http/http_stream_factory_impl_job.cc
[modify] https://crrev.com/a7da6da864dce77fe1c931653635c3ac757219cb/net/http/http_stream_factory_impl_job_controller_unittest.cc
[modify] https://crrev.com/a7da6da864dce77fe1c931653635c3ac757219cb/net/http/http_stream_parser.cc
[modify] https://crrev.com/a7da6da864dce77fe1c931653635c3ac757219cb/net/http/http_stream_parser.h
[modify] https://crrev.com/a7da6da864dce77fe1c931653635c3ac757219cb/net/http/http_stream_parser_unittest.cc
[modify] https://crrev.com/a7da6da864dce77fe1c931653635c3ac757219cb/net/spdy/spdy_test_util_common.cc
[modify] https://crrev.com/a7da6da864dce77fe1c931653635c3ac757219cb/net/spdy/spdy_test_util_common.h
[modify] https://crrev.com/a7da6da864dce77fe1c931653635c3ac757219cb/net/test/embedded_test_server/default_handlers.cc
[modify] https://crrev.com/a7da6da864dce77fe1c931653635c3ac757219cb/net/websockets/websocket_basic_handshake_stream.cc
[modify] https://crrev.com/a7da6da864dce77fe1c931653635c3ac757219cb/third_party/WebKit/LayoutTests/http/tests/websocket/handshake-error-expected.txt
[modify] https://crrev.com/a7da6da864dce77fe1c931653635c3ac757219cb/third_party/WebKit/LayoutTests/http/tests/websocket/handshake-fail-by-maxlength-expected.txt
[modify] https://crrev.com/a7da6da864dce77fe1c931653635c3ac757219cb/third_party/WebKit/LayoutTests/http/tests/websocket/handshake-fail-by-prepended-null-expected.txt
[modify] https://crrev.com/a7da6da864dce77fe1c931653635c3ac757219cb/tools/metrics/histograms/histograms.xml

Blocking: 669800
Cc: cbentzel@chromium.org
mmenke@ can you please update this issue with the current status?  Is this "done"

For others following, see intent discussion here: https://groups.google.com/a/chromium.org/forum/#!searchin/blink-dev/HTTP$200.9%7Csort:relevance/blink-dev/OdKnpLlvVUo/1EpFGVUjAwAJ

Change tracked here: https://www.chromestatus.com/features/5633064474509312
And mentioned in the Chrome 54 blog post here: https://developers.google.com/web/updates/2016/09/chrome-54-deprecations#http09_deprecated
I've updated the chromestatus entry to make it clear support for HTTP 0.9 on non-default ports was removed in Chrome 55 (it still said deprecated in Chrome 54 :-(  ).

mmenke: can you please also include the latest UMA data on the fraction of HTTP requests we believe were ultimately impacted by this?
We've run into enough issues with the more limited deprecation that I don't think we want to take HTTP/0.9 deprecation any further, so going to just mark this as fixed.
Status: Fixed
Labels: -M-54 M-55
Sign in to add a comment