Etcher currently elevates a child writer proxy that itself spawns the
Etcher CLI in robot mode, parses the output, and proxies those messages
to the GUI application over IPC.
After these set of changes, Etcher elevates a single child writer
process that directly communicates back with the GUI using IPC. The main
purpose behind these changes is to simplify the overall architecture and
fix various issues caused by the current complex child process tree.
Here's a summary of the changes:
- Stop wrapping the Etcher CLI to perform writing
- Remove the robot option from the Etcher CLI (along with related
documentation)
- Elevate a new `child-write.js` standalone executable
- Move the relevant bits of `lib/child-writer` to the `image-writer` GUI
module
- Remove the `lib/child-writer` directory
- Add a new "Child died unexpectedly" Mixpanel event
- Floor state percentage in the flash state model
The above changes made is possible to tackle all the remaining issues
where the writer process would remain alive even if the parent died.
Change-Type: patch
Changelog-Entry: Ensure the writer process dies when the GUI application is killed.
See: https://github.com/resin-io/etcher/pull/1873
See: https://github.com/resin-io/etcher/pull/1843
Signed-off-by: Juan Cruz Viotti <jv@jviotti.com>
This is a strange one. On Windows, putting a space before the
double-ampersand command concatenator makes the environment variable
value contain a trailing space. So for something like `set foo=bar &&
...` the variable `foo` will be `'bar '` instead of `'bar'`.
Change-Type: patch
Changelog-Entry: Fix trailing space in environment variables during Windows elevation.
Signed-off-by: Juan Cruz Viotti <jv@jviotti.com>
Some devices don't have a size, like USB devices in the usbboot adaptor.
The `.isDriveLargeEnough()` correctly returns `false` in this case,
however we don't want to show the `TOO SMALL` badge for aesthetics
purposes.
So if a drive has a size that equals `null`, we don't allow such drive
to be selected, and we don't show a badge for it.
Signed-off-by: Juan Cruz Viotti <jv@jviotti.com>
This commit introduces a boolean `disabled` property rather than a
`pending` flag. Making this distinction clearer means that we can now
treat pending drives in different ways needed to improve the usbboot
experience.
Also, for usbboot, this commit removes the "pending" badge and uses a
more descriptive drive description instead.
Signed-off-by: Juan Cruz Viotti <jv@jviotti.com>
This commit installs `node-usb` v1.3.0 from GitHub, since that version
was never published to NPM, and is the only one that works with Visual
Studio 2015 (see https://github.com/tessel/node-usb/issues/109).
The usbboot communicates with a Raspberry Pi / Amber through USB and
eventually mounts it as a block device we can write to.
This feature bundles bootcode.bin and start.elf from the original
usbboot implementation.
The flow is the following:
- On each scan, the usbboot scanner will try to get a usbboot compatible
USB device to the next "phase", until they are all transformed to
block devices the user can flash to as usual
Change-Type: minor
Changelog-Entry: Integrate Raspberry Pi's usbboot technology.
Fixes: https://github.com/resin-io/etcher/issues/1541
See: https://github.com/raspberrypi/usbboot
Signed-off-by: Juan Cruz Viotti <jv@jviotti.com>
This allows selection of images without a secondary file extension
(i.e. `example.gz`, compared to `example.img.gz`) by defaulting to `img`
in the image-stream handlers, should no secondary extension be found.
Further this adjusts `.getPenultimateFileExtension()` to return `null`
if the detected penultimate extension is not a known file extension.
Change-Type: patch
We recently added a "pending" flag to all drives that represents whether
the drive is ready for selection or not. This flag will be used by the
"usbboot" flashing adaptor, which will emit various "pending" USB
devices while it converts them to block devices that can actually be
flashed.
In terms of the GUI, the following visible changes were made:
- Drives with a `pending: true` property will be disabled in the drive
selector window
- Drives with a `pending: true` property have a "PENDING" red badge
See: https://github.com/resin-io/etcher/pull/1707
See: https://github.com/resin-io/etcher/pull/1686
Signed-off-by: Juan Cruz Viotti <jv@jviotti.com>
The usbboot integration will bring in drive objects that include a lot
more properties than the current drive objects. This commit ensures that
the redux store can handle those extra properties.
Change-Type: patch
Signed-off-by: Juan Cruz Viotti <jv@jviotti.com>
As another step towards moving to GitHub Releases, this commit makes the
application care much less about the actual release type of the current
version, instead checking if the application is stable or not, which is
more aligned to what GitHub provides us.
Signed-off-by: Juan Cruz Viotti <jv@jviotti.com>
We send an HTTP request to S3 to determine the latest available version.
There are various error that can happen that we don't have control over
(like `ETIMEDOUT`).
The current approach is to whitelist certain errors and pretend there is
no update available, however this commit improves that whole situation.
Instead of swallowing these errors, we throw a user error from the
function that determines the latest available version. From the
application code, we check if that function throws a user error, and if
so, instead of showing it to the user, we log a mixpanel event and carry
on.
This change is motivated by the latest reporter error we can't do
anything about: `UNABLE_TO_GET_ISSUER_CERT_LOCALLY`.
Fixes: https://github.com/resin-io/etcher/issues/1525
Change-Type: patch
Changelog-Entry: Fix `UNABLE_TO_GET_ISSUER_CERT_LOCALLY` error at startup when behind certain proxies.
Signed-off-by: Juan Cruz Viotti <jv@jviotti.com>
- Extend the `standard` ESLint configuration
- Remove ESLint rules that are defined in the `standard` configuration
- Get rid of semi-colons
See: https://github.com/resin-io/etcher/pull/1657
Signed-off-by: Juan Cruz Viotti <jv@jviotti.com>
This commit changes the whole codebase to adhere to all StandardJS
guidelines rules except semicolons, since the removal of semicolons
affect pretty much all lines, and the final diff is very hard to follow
(and to assess other more involved changes).
In a nutshell:
- When using `function`, we now require a space before the opening
parenthesis
- If a line with operators is broken into multiple lines, the operator
should now go after the line break
- Unnecessary padding lines are now forbidden
There were also some minor things that the `standard` CLI caught that I
updated here.
See: https://standardjs.com
Signed-off-by: Juan Cruz Viotti <jv@jviotti.com>
There's not much we can do if we can't connect to S3 to determine the
latest available versions when checking if we should show an update
notification dialog or not.
As with similar errors, lets swallow this particular one, and try again
the next time Etcher runs.
Change-Type: patch
Changelog-Entry: Fix `EHOSTDOWN` error at startup.
Fixes: https://github.com/resin-io/etcher/issues/1645
Signed-off-by: Juan Cruz Viotti <jv@jviotti.com>
We make the size units used the closest relative unit through a new
filter `closestUnit` replacing the old `gigabyte` filter.
Changelog-Entry: Round byte sizes to the more appropriate unit.
* remove filters folder
* new shrinkwrap, add to package.json
* test
Support the rpi-sdcard image file type output by Yocto for
the Raspberry Pi device.
Change-Id: Ia7e3aef0d90fdf21d373a560e6dd2b96e6b51da8
Changelog-Entry: Add support for `.rpi-sdcard` images.
Now that the Redux store is no longer front-end dependent, we can move
most of the models to lib/shared. Currently, lib/shared is a mess, and
contains all sorts of functionality, however moving things out of the
GUI is the first step. Once we have everything decoupled, we can
organise all the code we have there.
Signed-off-by: Juan Cruz Viotti <jv@jviotti.com>
This error may get thrown when fetching the list of S3 packages when the
user is behind a proxy tht causes the SSL certificate to incorrectly not
be trusted.
See: https://github.com/atom/apm/issues/103
See: https://github.com/resin-io/etcher/issues/1465
Change-Type: patch
Changelog-Entry: Fix `UNABLE_TO_VERIFY_LEAF_SIGNATURE` error at startup.
Signed-off-by: Juan Cruz Viotti <jviotti@openmailbox.org>
When the user is behind a firewall, then the HTTP request to query the
latest available version from S3 may throw an EACCES error, eventually
causing a confusin "You don't have access to this resource" error
window.
Change-Type: patch
Changelog-Entry: Fix "You don't have access to this resource" error at startup when behind a firewall.
Fixes: https://github.com/resin-io/etcher/issues/1458
Signed-off-by: Juan Cruz Viotti <jviotti@openmailbox.org>
Querying S3 to determine the latest available versions might throw
`ECONNRESET` and `ECONNREFUSED`. This commit extends the
`s3Packages.getRemoteVersions()` function to handle these errors and
return no available version if so, like we already do with other similar
HTTP errors.
Change-Type: patch
Changelog-Entry: Fix `ECONNRESET` and `ECONNREFUSED` errors when checking for updates on unstable connections.
Fixes: https://github.com/resin-io/etcher/issues/1396
Fixes: https://github.com/resin-io/etcher/issues/1388
Signed-off-by: Juan Cruz Viotti <jviotti@openmailbox.org>
* Centralise drive error/warning messages with a function
`.getDriveImageCompatibilityStatuses` in `lib/shared/drive-constraints.js`
-- tests included
* Fix an error where several labels show at once
* Clarify the source drive label to 'Drive Contains Image'
* Remove label icons and make text bold to match Zeplin's design
Closes: https://github.com/resin-io/etcher/issues/1143
Closes: https://github.com/resin-io/etcher/issues/1144
This utility function is useful to avoid duplicating the logic of
checking whether an error is a user error across the code base.
Signed-off-by: Juan Cruz Viotti <jviotti@openmailbox.org>
`etcher-latest-version` was kept in a separate repository in order to
re-use it with the Etcher website, however the Etcher website is not
using it at all, and we're moving towards having the website in the main
repository.
Therefore, this commit brings back the logic from
`etcher-latest-version`, but introduces it as
`lib/shared/s3-packages.js`, in order to not tie ourselves to the
AngularJS framework, and as a step towards the Etcher SDK.
As a nice little bonus, this commit adds support for an
`ETCHER_FAKE_S3_LATEST_VERSION` environment variable that can be used to
trick Etcher that there is an available update, and therefore show the
update notifier modal.
Also, this commit adds support for snapshot builds update-checks, by
checking the `resin-nightly-downloads` S3 bucket if the current version
contains a git commit hash build number.
If the version is not a production release, then the update notifier
modal doesn't present the checkbox to disable update notifications for X
days.
We also add a property called `updates.semverRange` to `package.json`,
which can be used to fine control which versions are considered as
candidates for an update notification.
This commit adds a setting called `includeUnstableChannel`, which can be
used to tweak whether unstable (beta) releases are considered or not
when checking for the latest available version.
See: https://github.com/resin-io-modules/etcher-latest-version
Fixes: https://github.com/resin-io/etcher/issues/953
Signed-off-by: Juan Cruz Viotti <jviotti@openmailbox.org>
The elevation mechanism currently embedded in
`lib/child-writer/writer-proxy.js` is extracted as a separate re-usable
function in `lib/shared/permissions.js`.
This change hugely simplifies the writer proxy, while allowing us to
iterate faster on the elevation core functionality.
Signed-off-by: Juan Cruz Viotti <jviotti@openmailbox.org>
Currently, both of these functions accept two arguments: the error
title, and the error description. This function signature makes is hard
to keep adding options to these error creation functions, like an error
code, so this commit refactors them to take a single argument: an
options object containing `title` and `description` properties.
Signed-off-by: Juan Cruz Viotti <jviotti@openmailbox.org>
This fixes two things: The format detection, and a bug in `udif`.
First, by categorizing the `.dmg` extension as compressed image,
`.isSupportedImage()` would attempt to detect the format after stripping
the extension, causing it to be misdetected.
Second, `udif`'s ReadStream didn't add the `dataForkOffset` to its
position when reading blocks, causing the wrong data to be read for some images,
in turn causing zlib to error on invalid headers.
Changes:
- Classify `.dmg` as `type: 'image'`
- Update `udif` to 0.8.0
Change-Type: patch
Changelog-Entry: Fix Apple disk image detection & streaming
Signed-off-by: Juan Cruz Viotti <jviotti@openmailbox.org>
Currently, if the child writer receives a message from the writer
process that is not a valid robot object, then it will throw an error.
Now, we check if the message is a robot message, and if so, we try to
parse it (throwing errors if its indeed a malformed/incomplete robot
message), however we log it if not.
The main motivation behind this feature is that it will allows us to
print debugging information on the mountutils module, and have it
redirected to DevTools.
Change-Type: minor
Signed-off-by: Juan Cruz Viotti <jviotti@openmailbox.org>
`image-stream` returns image objects that look like this:
```js
{
stream: <readable stream>,
transform: <transform stream>,
size: {
original: <number>,
final: {
value: <number>,
estimation: <boolean>
}
},
...
}
```
While the GUI handles image objects that look like this:
```sh
{
path: <string>,
size: <number>,
...
}
```
It looks like we should share a common structure between both, so we can
use `image-stream` images in `drive-constraints`, for example.
Turns out that we actually transform `image-stream` image objects to GUI
image objects when the user selects an image using the image selector
dialog, which is another indicator that we should normalise this
situation.
As a solution, this commit does the following:
- Add `path` to `image-stream` image object
- Reuse `image-stream` image objects in the GUI, given they are a
superset of GUI image objects
See: https://github.com/resin-io/etcher/pull/1223#discussion_r108165110
Fixes: https://github.com/resin-io/etcher/issues/1232
Signed-off-by: Juan Cruz Viotti <jviotti@openmailbox.org>
Currently, the Etcher CLI will print scary stack traces for every single
error (e.g: if you forgot to pass an image to the tool), given that
`errors.getDescription()` will return a stack trace if no other
description could be found.
This commit introduces an `ETCHER_CLI_DEBUG` environment variable, which
when set, it will cause the Etcher CLI to output stack traces, plus a
boolean `userFriendlyDescriptionsOnly` option to
`errors.getDescription()`, so we can control whether
`errors.getDescription()` returns things like stack traces, or
stringified error objects.
Change-Type: minor
Changelog-Entry: Don't print stack traces by default in the CLI.
Signed-off-by: Juan Cruz Viotti <jviotti@openmailbox.org>
* chore: get the CI tests working again
* an updated eslint-plugin-lodash was creating extra linter errors where we
were using built-in String methods instead of the lodash equivalents
* an updated codespell package now installs with a different executable-name
(no .py extension), and now supports multiple dictionaries
* while I was at it I replaced the messy double-quoting-string logic with the
command-join module
Changelog-type: patch
* chore: downgrade codespell to 1.9.2 because the latest version has issues running on Windows
(we'll hopefully be able to revert this commit when codespell works again properly!)
The event data may contain absolute paths that contain user information that
should not be logged in Mixpanel. Instead, we replace absolute path properties
with their base name.
Change-Type: patch
Changelog-Entry: Don't include user paths in Mixpanel analytics events.
* feat(GUI): improve analytics events
This commit adds more events to our current analytics.
Will further improve in a future commit.
Change-Type: patch
See: https://github.com/resin-io/etcher/issues/1100
* refactor(gui): use single function to set normal and dangerous settings
Change-Type: patch
Signed-off-by: Juan Cruz Viotti <jviotti@openmailbox.org>
The current error handling logic is a mess. We have code that tries to
fetch information about errors in different places throughout the
application, and its incredibly hard to ensure certain types of error
get decent human friendly error messages.
This commit groups, improves, and tests all error related functions in
`lib/shared/errors.js`.
Here's a summary of the changes, in more detail:
- Move the `HUMAN_FRIENDLY` object to `shared/errors.js`
- Extend `HUMAN_FRIENDLY` with error descriptions
- Add `ENOMEM` to `shared/errors.js`
- Group CLI and `OSDialogService` mechanisms for getting an error title
and an error description
- Move error serialisation routines from `robot` to `shared/errors.js`
- Create and use `createError()` and `createUserError()` utility
functions
- Add user friendly descriptions to many errors
- Don't report user errors to TrackJS
Fixes: https://github.com/resin-io/etcher/issues/1098
Change-Type: minor
Changelog-Entry: Make errors more user friendly throughout the application.
Signed-off-by: Juan Cruz Viotti <jviotti@openmailbox.org>
Currently Etcher will allow you to flash an image to the same drive that
contains the image. As a way to protect against that case we introduce
the concept of a "source drive", which means a drive that contains the
source image.
This commit adds the following logic around this new concept:
- Don't auto-select a source drive
- De-select an already selected drive if an image inside it is selected
- Disable the drive in the drive selector modal
- Add a "Source Drive" badge to the drive in the drive selector modal
Fixes: https://github.com/resin-io/etcher/issues/830
Change-Type: minor
Changelog-Entry: Prevent flashing the drive where the source image is located.
Signed-off-by: Juan Cruz Viotti <jviotti@openmailbox.org>
The `child-writer` module is not re-used by both the GUI and the CLI, so
it makes sense to have it in `lib/child-writer`.
Signed-off-by: Juan Cruz Viotti <jviotti@openmailbox.org>
Currently the `DriveConstraintsModel` errors when given an `undefined`
drive, or image; this commit changes that behaviour. Comes with tests.
Signed-off-by: Juan Cruz Viotti <jviotti@openmailbox.org>
This commit extracts the byte-related conversions from the `byte-size`
AngularJS module and the `FlashStateModel` to a re-usable generic
CommonJS module at `lib/shared/units.js`, than can also be used by the
CLI.
Signed-off-by: Juan Cruz Viotti <jviotti@openmailbox.org>
There are certain application messages that should be re-used between
the CLI and the GUI. In order to allow such re-usability, we extract out
the application messages used in JavaScript into
`lib/shared/messages.js` as a collection of Lodash `_.template`
templates.
Notice this file doesn't include application messages included in
Angular templates directly since it'd be hard to refactor all of them.
We plan to move to React soon, which will allow moving the remaining
messages very easily.
Signed-off-by: Juan Cruz Viotti <jviotti@openmailbox.org>
These utilities were extracted to `lib/shared/child-writer/cli.js`, and
unit tests have been written for them.
As a result of testing, `.getBooleanArgumentForm()` has been extended to
support single letter options.
Signed-off-by: Juan Cruz Viotti <jviotti@openmailbox.org>
