Summary: Ref T10748. Brings the rest of the transactions to EditEngine, supports creating via API.
Test Plan:
- Created a URI via API.
- Created a URI via web.
- Tried to apply sneaky transactions, got rejected with good error messages. <_< >_>
Reviewers: chad
Reviewed By: chad
Maniphest Tasks: T10748
Differential Revision: https://secure.phabricator.com/D15821
Summary: Ref T10860. This doesn't change anything, it just separates all this stuff out of `PhabricatorRepository` since I'm planning to add a bit more state to it and it's already pretty big and fairly separable.
Test Plan: Pulled, pushed, browsed Diffusion.
Reviewers: chad
Reviewed By: chad
Maniphest Tasks: T10860
Differential Revision: https://secure.phabricator.com/D15790
Summary:
Ref T4292. When the daemons make a query for repository information, we need to make sure the working copy on disk is up to date before we serve the response, since we might not have the inforamtion we need to respond otherwise.
We do this automatically for almost all Diffusion methods, but this particular method is a little unusual and does not get this check for free. Add this check.
Test Plan:
- Made this code throw.
- Ran `bin/repository reparse --message ...`, saw the code get hit.
- Ran `bin/repository lookup-user ...`, saw this code get hit.
- Made this code not throw.
Reviewers: chad
Reviewed By: chad
Maniphest Tasks: T4292
Differential Revision: https://secure.phabricator.com/D15783
Summary:
Ref T4292. We currently synchronize hosted, clustered, Git repositories when we receive an SSH pull or push.
Additionally:
- Synchronize before HTTP reads and writes.
- Synchronize reads before Conduit requests.
We could relax Conduit eventually and allow Diffusion to say "it's OK to give me stale data".
We could also redirect some set of these actions to just go to the up-to-date host instead of connecting to a random host and synchronizing it. However, this potentially won't work as well at scale: if you have a larger number of servers, it sends all of the traffic to the leader immediately following a write. That can cause "thundering herd" issues, and isn't efficient if replicas are in different geographical regions and the write just went to the east coast but most clients are on the west coast. In large-scale cases, it's better to go to the local replica, wait for an update, then serve traffic from it -- particularly given that writes are relatively rare. But we can finesse this later once things are solid.
Test Plan:
- Pushed and pulled a Git repository over HTTP.
- Browsed a Git repository from the web UI.
Reviewers: chad
Reviewed By: chad
Maniphest Tasks: T4292
Differential Revision: https://secure.phabricator.com/D15758
Summary:
Ref T4292. Ref T10366. Depends on D15751. Today, generating repository commands is purely a function of the repository, so they use protocols and credentials based on the repository configuration.
For example, a repository with an SSH "remote URI" always generate SSH "remote commands".
This needs to change in the future:
- After T10366, repositories won't necessarily just have one type of remote URI. They can only have one at a time still, but the repository itself won't change based on which one is currently active.
- For T4292, I need to generate intracluster commands, regardless of repository configuration. These will have different protocols and credentials.
Prepare for these cases by separating out command construction, so they'll be able to generate commands in a more flexible way.
Test Plan:
- Added unit tests.
- Browsed diffusion.
- Ran `bin/phd debug pull` to pull a bunch of repos.
- Ran daemons.
Reviewers: chad
Reviewed By: chad
Maniphest Tasks: T4292, T10366
Differential Revision: https://secure.phabricator.com/D15752
Summary: Ref T10748. Ref T10337. This technically implements this stuff, but it does not do anything useful yet. This skips all the hard stuff.
Test Plan:
- Technically used `diffusion.repository.search` to get repository information.
- Technically used `diffusion.repository.edit` to change a repository name.
- Used `editpro/` to edit a repository name.
Reviewers: chad
Reviewed By: chad
Maniphest Tasks: T10337, T10748
Differential Revision: https://secure.phabricator.com/D15736
Summary:
Ref T2783. This allows this worker to run on a machine different to the one that stores the repository, by routing the execution of Git over Conduit calls.
This API method is super gross, but fixing it isn't straightforward and it runs into other complicated considerations. We can fix it later; for now, just define it as "internal" to limit how much mess this creates.
"Internal" methods do not appear on the console.
Test Plan: Ran `bin/repository reparse --change <commit> --trace` on several commits, saw daemons make a Conduit call instead of running a `git` command.
Reviewers: hach-que, chad
Reviewed By: chad
Subscribers: joshuaspence, Korvin, epriestley
Maniphest Tasks: T2783
Differential Revision: https://secure.phabricator.com/D11874
Summary:
Ref T10537. For Nuance, I want to introduce new sources (like "GitHub" or "GitHub via Nuance" or something) but this needs to modularize eventually.
Split ContentSource apart so applications can add new content sources.
Test Plan:
This change has huge surface area, so I'll hold it until post-release. I think it's fairly safe (and if it does break anything, the breaks should be fatals, not anything subtle or difficult to fix), there's just no reason not to hold it for a few hours.
- Viewed new module page.
- Grepped for all removed functions/constants.
- Viewed some transactions.
- Hovered over timestamps to get content source details.
- Added a comment via Conduit.
- Added a comment via web.
- Ran `bin/storage upgrade --namespace XXXXX --no-quickstart -f` to re-run all historic migrations.
- Generated some objects with `bin/lipsum`.
- Ran a bulk job on some tasks.
- Ran unit tests.
{F1190182}
Reviewers: chad
Reviewed By: chad
Maniphest Tasks: T10537
Differential Revision: https://secure.phabricator.com/D15521
Summary:
D9087 adds a nice typeahead but breaks the existing regex
search by quoting the pattern. Ideally, this change won't break the
typeahead, which as far as I can tell doesn't use the `pattern`
argument.
Test Plan:
Not yet.
RFC as to whether this change makes sense, will fix my local setup and resend if so.
Reviewers: epriestley, #blessed_reviewers
Reviewed By: epriestley, #blessed_reviewers
Subscribers: Korvin
Differential Revision: https://secure.phabricator.com/D15500
Summary:
Fixes T10304. In Mercurial, we must enumerate the whole file tree. Currently, we incorrectly count files within directories (which won't be shown) toward the "100 file" limit at top level, so directories with more than 100 subpaths are truncated improperly.
This is approxiately the same as @richardvanvelzen's fix.
Test Plan: Viewed a large Mercurial repository, saw a complete directory listing.
Reviewers: chad
Reviewed By: chad
Subscribers: richardvanvelzen
Maniphest Tasks: T10304
Differential Revision: https://secure.phabricator.com/D15282
Summary:
Fixes T10186. After D14970, `diffusion.filecontentquery` puts the content in a file and returns the file PHID.
However, it does this in a way that doesn't go through the chunking engine, so it will fail for files larger than the chunk threshold (generally, 8MB).
Instead, stream the file from the underlying command directly into chunked storage.
Test Plan:
- Made a commit including a really big file: 4dcd4c492b
- Used `diffusion.filecontentquery` to load file content.
- Parsed/imported commit locally.
- Used `diffusion.filecontentquery` to load content for smaller files (README, etc).
Reviewers: chad
Reviewed By: chad
Maniphest Tasks: T10186
Differential Revision: https://secure.phabricator.com/D15072
Summary: Fixes T8826. Git tracks an "author date", which may be different from the "committed date". We don't currently extract/show this; do so.
Test Plan: {F1059235}
Reviewers: chad
Reviewed By: chad
Maniphest Tasks: T8826
Differential Revision: https://secure.phabricator.com/D14995
Summary:
Fixes T9319. Proxied requests (e.g., in the cluster) for binary files (like images) currently fail because we can not return binary data over Conduit in JSON.
Although Conduit will eventually support binary-safe encodings, a cleaner approach to this is just to return a `filePHID` instead of the raw content. This is generally faster and more flexible, and gives us more opportunities to add caching later.
After making the call, the client pulls the file data separately.
We also no longer need to return a complex data structure because we don't do blame over this call any longer.
Test Plan:
- Viewed images in Diffusion.
- Viewed READMEs in Diffusion.
- Used `bin/differential attach-commit rX Dy` to hit attach pathway.
Reviewers: chad
Reviewed By: chad
Maniphest Tasks: T9319
Differential Revision: https://secure.phabricator.com/D14970
Summary:
Fixes T4366. Two years ago, Facebook put 16,000 files in a directory. Today, the page has nearly loaded.
Paginate large directories.
Test Plan:
- Viewed home and browse views in Git, Mercurial and Subversion.
I put an artificially small page size (5) on home:
{F1055653}
I pushed 16,000 files to a directory and paged through them. Here's the last page, which rendered in about 200ms:
{F1055655}
Our behavior is a bit better than GitHub here, which shows only the first 1,000 files, disables pagination, and can't retrieve history for the files:
{F1055656}
Reviewers: chad
Reviewed By: chad
Maniphest Tasks: T4366
Differential Revision: https://secure.phabricator.com/D14956
Summary:
Ref T2450. Ref T9319. This is still a bit messy, but not quite so bad as it was: instead of using a single call to get both blame information and file content, use `diffusion.blame` for blame information.
This will make optimizations to both blame and file content easier.
Test Plan: Viewed a bunch of blame (color on/off, blame on/off).
Reviewers: chad
Reviewed By: chad
Maniphest Tasks: T2450, T9319
Differential Revision: https://secure.phabricator.com/D14958
Summary:
Fixes T2451. Several motivations here, from strongest to weakest:
- Currently, getting blame and file content are closely entwined. This makes fixing T9319 more difficult, and I want to fix it. I want to separate blame from content so there's more flexibility in how we approach this issue.
- This makes pursuing T2450 easier, if it turns out to be a meaningful win.
- If we can get a win on blame performance, we can do `arc blame` eventually if we want.
Test Plan:
- Blamed in SVN, Git and Mercurial.
Reviewers: chad
Reviewed By: chad
Maniphest Tasks: T2451
Differential Revision: https://secure.phabricator.com/D14957
Summary: Fixes T7852. Although `1` could also indicate other kinds of problems, assume it means "no results".
Test Plan: Searched for nonsense strings in Git and Mercurial. Searched for valid strings in Git and Mercurial.
Reviewers: chad
Reviewed By: chad
Maniphest Tasks: T7852
Differential Revision: https://secure.phabricator.com/D14943
Summary:
Ref T4245. Like everything else, accept more identifiers.
This needs a change in `arc`, which I've made a note about elsewhere.
Test Plan: Used "Update Now" from web UI, saw update get scheduled.
Reviewers: chad
Reviewed By: chad
Maniphest Tasks: T4245
Differential Revision: https://secure.phabricator.com/D14932
Summary:
Ref T4245. This was obsoleted long ago and has no callers in Phabricator or Arcanist.
Also some minor cleanup.
Test Plan: `grep` for callers everywhere.
Reviewers: chad
Reviewed By: chad
Maniphest Tasks: T4245
Differential Revision: https://secure.phabricator.com/D14930
Summary: Ref T4245. More of the same, just narrowing down the easy cases.
Test Plan:
- Called `diffusion.querycommit`.
- Browsed branches.
- Browsed repository.
- Browsed directory.
- Searched for stuff.
- Viewed a commit.
- Viewed a file diff.
- Edited a commit.
- Viewed history.
- Viewed tags.
- Viewed push log.
Reviewers: chad
Reviewed By: chad
Maniphest Tasks: T4245
Differential Revision: https://secure.phabricator.com/D14929
Summary:
Ref T4245. Broaden support to include "ABCD", "rABCD", "1234", "R1234", etc.
This doesn't change the old behavior, just accepts more stuff.
Test Plan:
- Browsed Diffusion.
- Made various calls via API console.
Reviewers: chad
Reviewed By: chad
Maniphest Tasks: T4245
Differential Revision: https://secure.phabricator.com/D14928
Summary:
This diff adds a new mode argument to the Diffusion Conduit API with two options:
- "overwrite": the default, maintains the current behavior of deleting all coverage
in the specified branch before uploading the new coverage
- "update": does not delete old coverage, but will overwrite previous
coverage information if it's for the same file and commit
`DiffusionRequest::loadCoverage` already loads a file's coverage from the
latest available commit, so uploading coverage for different files in different
commits with "update" will result in seeing the latest uploaded coverage in
Diffusion.
Test Plan: manual local verification
Reviewers: epriestley, #blessed_reviewers
Reviewed By: epriestley, #blessed_reviewers
Subscribers: Korvin
Differential Revision: https://secure.phabricator.com/D14428
Summary:
Ref T9513. I checked this briefly but didn't do a very thorough job of it.
- Don't try to query merges for Subversion, since it doesn't support them.
- Fix up "existsquery" to work properly (and efficiently) for both hosted and imported repositories.
- Fix up "parentsquery" to have similar behavior on invalid commits to other VCSes (throw an exception).
Test Plan:
- No more merges warning on SVN.
- Hosted SVN gets the right exists result now.
- Visiting "r23980283789287" now 404's instead of "not parsed yet".
Reviewers: chad
Reviewed By: chad
Maniphest Tasks: T9513
Differential Revision: https://secure.phabricator.com/D14239
Summary:
Fixes T8597. Second issue there is that if you look at a huge file in Diffusion (like `/path/to/300MB.pdf`) we pull the whole thing over Conduit upfront, then try to shove it into file storage.
Instead, pull no more than the chunk limit (normally 4MB) and don't spend more than 10s pulling data.
If we get 4MB of data and/or time out, just fail with a message in the vein of "this is a really big file".
Eventually, we could improve this:
- We can determine the //size// of very large files correctly in at least some VCSes, this just takes a little more work. This would let us show the true filesize, at least.
- We could eventually stream the data out of the VCS, but we can't stream data over Conduit right now and this is a lot of work.
This is just "stop crashing".
Test Plan: Changed limits to 0.01 seconds and 8 bytes and saw reasonable errors. Changed them back and got normal beahvior.
Reviewers: btrahan
Reviewed By: btrahan
Subscribers: epriestley
Maniphest Tasks: T8597
Differential Revision: https://secure.phabricator.com/D13348
Summary: Ref T7604. Change `DiffusionLintSaveRunner` to use repositories instead of Arcanist Projects.
Test Plan: Ran the `save_lint.php` script and queried results using the `diffusion.getlintmessages` Conduit endpoint.
Reviewers: epriestley, #blessed_reviewers
Reviewed By: epriestley, #blessed_reviewers
Subscribers: Korvin, epriestley
Maniphest Tasks: T7604
Differential Revision: https://secure.phabricator.com/D12893
Summary: We were adding "+1" to PHP_INT_MAX, which could cause us to wrap and run `git log -n -9223372036854775808 ...`. Git, unable to give us negative nine trillion zillion results, would then exit without producing results.
Test Plan: Got results out of this method.
Reviewers: btrahan
Reviewed By: btrahan
Subscribers: epriestley
Differential Revision: https://secure.phabricator.com/D12726
Summary: Fixes T7220. Ref T7977. Changes symbols from being bound to an Arcanist project to being bound to a repository.
Test Plan:
- Added symbols and then applied migrations, symbols seemed to be migrated successfully.
- Tested the `/diffusion/symbol/$SYMBOL_NAME` endpoint.
- Tested the `/diffusion/symbol/$SYMBOL_NAME` endpoint with the `?repositories=$REPOSITORY_PHID` parameter.
Reviewers: #blessed_reviewers, epriestley
Reviewed By: #blessed_reviewers, epriestley
Subscribers: avivey, Korvin, epriestley
Maniphest Tasks: T7977, T7220
Differential Revision: https://secure.phabricator.com/D12608
Summary: Fixes T7100. In the bizarre case that a Git repository has a branch and tag with the same name, don't resolve branch names into tag names.
Test Plan: Test repo with branch and tag both named "git" no longer reports ambiguity.
Reviewers: btrahan
Reviewed By: btrahan
Subscribers: epriestley
Maniphest Tasks: T7100
Differential Revision: https://secure.phabricator.com/D12553
Summary:
Fixes T6160. Ref T7100.
- When resolving ambiguous branch references, ignore closed heads unless there are no other options.
- Hide closed heads by default on the main page.
- Show branch open/closed state in Mercurial.
Test Plan: Browsed a previously-ambiguous Mercurial repository because of multiple branch heads, no longer ambiguous.
Reviewers: btrahan
Reviewed By: btrahan
Subscribers: epriestley
Maniphest Tasks: T6160, T7100
Differential Revision: https://secure.phabricator.com/D12552
Summary:
Ref T7803. Ref T5873. I want to drive Conduit through more shared infrastructure, but can't currently add parameters automatically.
Put a `getX()` around the `defineX()` methods so the parent can provide default behaviors.
Also like 60% of methods don't define any special error types; don't require them to implement this method. I want to move away from this in general.
Test Plan:
- Ran `arc unit --everything`.
- Called `conduit.query`.
- Browsed Conduit UI.
Reviewers: btrahan
Reviewed By: btrahan
Subscribers: hach-que, epriestley
Maniphest Tasks: T5873, T7803
Differential Revision: https://secure.phabricator.com/D12380
Summary: Fixes T7256.
Test Plan: Looked at rXPRF0a7a5f69f5d7 in a local instance. things looked great both pre and post patch.
Reviewers: epriestley
Reviewed By: epriestley
Subscribers: Korvin, epriestley
Maniphest Tasks: T7256
Differential Revision: https://secure.phabricator.com/D11790
Summary: Ref T7094.
Test Plan: couldn't really test this - how does one get symbols going nowadays given they are acanist project based?
Reviewers: epriestley
Reviewed By: epriestley
Subscribers: Korvin, epriestley
Maniphest Tasks: T7094
Differential Revision: https://secure.phabricator.com/D11584
Summary: Ref T7094. Could just delete this end point too I guess? Needed to add "withCommitPHIDs" to the differentialrevisionquery to get this done.
Test Plan: used diffusion.getcommits from conduit console and got a sensible result for a query for two commits, one with a diff and one without.
Reviewers: epriestley
Reviewed By: epriestley
Subscribers: Korvin, epriestley
Maniphest Tasks: T7094
Differential Revision: https://secure.phabricator.com/D11581
Summary:
Ref T2783. I think this served two purposes:
- Improving performance in cases where we "know" a repository is local.
- Preventing loops.
It is now obsolete:
- After D11476, refs can almost always resolve on a fast path.
- As T2783 moves forward, we can usually no longer know when a repository is local without actually looking it up -- almost everything is allowed to run anywhere.
- The cluster behavior in D11475 now prevents loops.
Test Plan: `grep`, browsed around. This didn't really do much of anything anymore.
Reviewers: btrahan
Reviewed By: btrahan
Subscribers: epriestley
Maniphest Tasks: T2783
Differential Revision: https://secure.phabricator.com/D11477
Summary:
Fixes T7020. When an external user makes a Conduit request to Diffusion but the repository isn't hosted locally, we need to proxy it.
This also adds a guard layer to prevent requests from getting infinitely proxied inside the cluster.
In "trivial" configurations (where the repository is a service repository, but the service is on the local device) I'm making us always proxy anyway. This basically makes it reasonable to test this stuff (otherwise you'd have to set up two different installs) and this configuration doesn't make much sense in real life (if you're using multiple machines, making one a dedicating daemons+repo box is almost certainly the most reasonable configuration, even for a cluster size of 2).
Test Plan:
- With a service-hosted repository, made Diffusion conduit calls and browsed the UI. Verified requests got proxied once, then resovled.
- With a non-service repository, made Diffusion conduit calls and browsed UI. Verified requests were handled in-process immediately.
Reviewers: btrahan
Reviewed By: btrahan
Subscribers: epriestley
Maniphest Tasks: T7020
Differential Revision: https://secure.phabricator.com/D11475
Summary:
Ref T2783. This method is kind of goofballs:
- We send a big list of paths to it.
- It sends back a giant blob of HTML.
Instead, just figure out the path we want locally, then fetch the content with `diffusion.filecontentquery`.
Test Plan:
- Viewed main view and directory view, saw a README.
- See screenshots.
Reviewers: btrahan
Reviewed By: btrahan
Subscribers: epriestley
Maniphest Tasks: T2783
Differential Revision: https://secure.phabricator.com/D11099
Summary:
Ref T2783. Fixes T6039.
- Provide `authorPHID` and `committerPHID` to resolve T6039.
- In message parser, store author/email strings.
- In cached results, emit author/email strings.
Test Plan: Called method with and without bypassCache. Used `reparse.php` to repopulate data on an old commit.
Reviewers: btrahan
Reviewed By: btrahan
Subscribers: epriestley
Maniphest Tasks: T2783, T6039
Differential Revision: https://secure.phabricator.com/D10424
Summary:
Ref T2783. This populates the following fields in DiffusionQueryCommitsConduitAPIMethod using DiffusionLowLevelCommitQuery when `bypassCache` is set to true:
* `authorName`
* `authorEmail`
* `committerName`
* `committerEmail`
* `message`
* `hashes`
The original outline called for `authorPHID` and `committerPHID` as well (but no `message` field). As far as I can tell, the PHIDs aren't actual a property on `DiffusionCommitRef`, and since the intention of this is to be able to populate a `DiffusionCommitRef`, I haven't included them. Let me know if we really do need the PHIDs here.
Test Plan: Tested using 3 Phabricator instances (one web, one taskmaster and one storage). The web and taskmaster tiers are directed at the Conduit API of the storage tier. Made a `diffusion.querycommits` from the Conduit app on the web tier instance and saw the data populated from the raw VCS data (located on the storage tier).
Reviewers: epriestley, #blessed_reviewers
Reviewed By: epriestley, #blessed_reviewers
Subscribers: epriestley, Korvin
Maniphest Tasks: T2783
Differential Revision: https://secure.phabricator.com/D10399
Summary: Fixes T4387.
Test Plan: Setup a mercurial repository for rabbitmq-server. Browsed around it and things looked good.
Reviewers: epriestley
Reviewed By: epriestley
Subscribers: epriestley, Korvin
Maniphest Tasks: T4387
Differential Revision: https://secure.phabricator.com/D10380
Summary: be more aggressive about assuming plain-text, use remarkup for no extension, .remarkup, and .md, and last but not least use rainbow for .rainbow. Fixes T5818.
Test Plan: my README rendered just fine post these changes
Reviewers: epriestley
Reviewed By: epriestley
Subscribers: asherkin, epriestley, Korvin
Maniphest Tasks: T5818
Differential Revision: https://secure.phabricator.com/D10340
Summary: Ref T4896. Use the new transaction-oriented `PhabricatorAuditEditor` directly instead of invoking it via the old editor.
Test Plan: Used Conduit to add a comment, use silent mode, and accept a commit.
Reviewers: joshuaspence, btrahan
Reviewed By: btrahan
Subscribers: epriestley
Maniphest Tasks: T4896
Differential Revision: https://secure.phabricator.com/D10126
Summary:
Ref T4896. Move the write for "Add Auditors" inside the new Editor.
There are no longer any readers or writers for metadata, so remove the calls for it.
Test Plan: Added auditors from the web UI.
Reviewers: joshuaspence, btrahan
Reviewed By: btrahan
Subscribers: epriestley
Maniphest Tasks: T4896
Differential Revision: https://secure.phabricator.com/D10123
Summary:
Ref T4896. Replaces more custom stuff with standard stuff. In particular:
- No more fake proxy writes;
- no more fake detection of `@mentions`.
For now, the old code still applies most of the effects and handles feed and email.
Test Plan:
- Added comments.
- Added comments with inline comments.
- Added just inline comments.
- Added comments with Conduit.
- Previewed comments.
- Added CCs explicitly and with `@mentions`.
- Added auditors.
- Accepted a commit.
Reviewers: joshuaspence, btrahan
Reviewed By: btrahan
Subscribers: epriestley
Maniphest Tasks: T4896
Differential Revision: https://secure.phabricator.com/D10109
Summary: This class was renamed in D9991, but the filename is incorrect.
Test Plan: Eyeball it
Reviewers: epriestley, #blessed_reviewers
Reviewed By: epriestley, #blessed_reviewers
Subscribers: epriestley, Korvin
Differential Revision: https://secure.phabricator.com/D10117
Summary:
Ref T4896. Depends on D10023. Prepares the code for the final migration.
The transaction table stores one row per distinct effect (e.g., add CCs) rather than one row per user action (e.g., "add CCs + comment"). We can double-read that table as long as the code doesn't expect transactions/comments to have multiple different effects, and doesn't try to write any such rows.
Everywhere that we were writing a big "X + Y" comment, write two separate "X" and "Y" comments instead. Like D10023, this disrupts the UI a little (you get more boxes), but that will be resolved once the rendering code swaps over. Otherwise, this retains the existing behavior.
Test Plan:
- Used `diffusion.createcomment` to add comments, raise concern, and accept.
- Previewed commenting, adding auditors/ccs, accepting, raising concern.
- Actually performed commenting, adding auditors/ccs, accepting, raising concern.
- Added a user with mentions.
- Added an explicit CC and a mention user.
Reviewers: btrahan, joshuaspence
Reviewed By: joshuaspence
Subscribers: epriestley
Maniphest Tasks: T4896
Differential Revision: https://secure.phabricator.com/D10052
