summaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAge
* Update the changelogHEADdebian/0_git20180727+ds-1archive/debian/0_git20180727+ds-1masterAndrej Shadura2018-10-06
|
* Patch Makefile.am to not install licensesAndrej Shadura2018-10-06
|
* Initial packagingAndrej Shadura2018-10-06
|
* Remove license texts and questionably licensed CSS and JavaScriptAndrej Shadura2018-10-05
|
* New translations from Transifex: transifex-update-eos-sdk 6Jenkins2018-07-27
|
* Merge pull request #4185 from endlessm/T22301Roddy Shuler2018-07-19
|\ | | | | T22301
| * eoslicense: Always display license codeMartin Abente Lahaye2018-07-19
|/ | | | | | | | | Adding a new license for every app that a uses a different one from our list, it simply does not scale. Therefore, always display the code instead of the "Unknown license" text. https://phabricator.endlessm.com/T22301
* Merge pull request #4184 from endlessm/T22827Juan Pablo Ugarte2018-06-12
|\ | | | | Fix broken tests
| * tests: Remove useless test-hello.cPhilip Chimento2018-06-07
| | | | | | | | | | | | | | | | This was sample code from the very beginning of this repository, I'm not sure why it has persisted this long. In any case it uses the deprecated g_test_trap_fork() so we may as well delete it. https://phabricator.endlessm.com/T22827
| * tests: Move g_test_trap_fork() to _subprocess()Philip Chimento2018-06-07
|/ | | | | | | | | g_test_trap_fork() is deprecated, and in some recent release it stopped working with gtester --tap. We could track this down but an easy fix is to switch to its replacement, g_test_trap_subprocess(). This fixes the unit tests. https://phabricator.endlessm.com/T22827
* Update from TransifexPhilip Chimento2018-06-07
|
* AssetButton: RemovePhilip Chimento2018-04-17
| | | | | | | | | Nothing is using this. (The last users were eos-programming, stuck on an old runtime; and eos-typing, defunct.) As long as I'm removing SearchBox, may as well remove this too. https://phabricator.endlessm.com/T20353
* SearchBox: RemovePhilip Chimento2018-04-17
| | | | | | | | | This was used nowhere except in eos-knowledge-lib. We are going to make many eos-knowledge-lib specific changes to it, so it's going to be forked into there. After nothing uses it, it doesn't make sense to keep it here given that we intend to slowly move away from using eos-sdk. https://phabricator.endlessm.com/T20353
* SearchBox: Don't set width of autocomplete popupPhilip Chimento2018-04-13
| | | | | | | | Setting the width here can mean that the popup gets a different width from the search box. Instead, we want it to match the search box's width and ellipsize the label if there's not enough space. https://phabricator.endlessm.com/T20352
* Update from TransifexPhilip Chimento2018-04-05
|
* Handle output file correctly when generating a diffEmmanuele Bassi2018-03-21
| | | | | If the output temporary file cannot be atomically renamed to its final destination, we need to copy its contents.
* Do not link eos-profile with libendlessEmmanuele Bassi2018-03-21
| | | | | | | | We don't depend on any specific Endless SDK symbols, and if we link against the SDK library itself we end up enabling the profiling probes; this means we may end up overwriting the profiling data we're trying to open, if by any chance we have `EOS_PROFILE=capture:` in our environment pointing to the profile data we're opening.
* Add machine-readable diff outputEmmanuele Bassi2018-03-21
| | | | | | | | | | If we're using eos-profile-diff in a CI infrastructure then we should not be printing out data in a human readable format, as we may want to interpret the output using scripts at a later point. We can use JSON, instead, and allow redirecting the output to a file; this way, we can store the output as an artifact, collect it, and parse it later.
* Build man page and reference for eos-profileEmmanuele Bassi2018-03-21
| | | | Documenting CLI tools that we ship should be part of the API reference.
* Add a diff tool for the eos-profile toolEmmanuele Bassi2018-03-21
| | | | | Compares N probe data files and prints a quick overview with the average timings of all the probes.
* Use the utility function for accessing probesEmmanuele Bassi2018-03-21
| | | | | Now that we have it, we can cut some common code from the convert and show sub-commands.
* Extract the probe parsing functionEmmanuele Bassi2018-03-21
| | | | | | | Both the convert and the show sub-commands for eos-profile have code to turn a GVariant into a profile probe for the v1 of the format. Let's move it out into its own internal utility function.
* Add eos-profile-convert commandEmmanuele Bassi2018-03-21
| | | | | | | | | | If we want to perform additional analysis on a profile capture file it's generally going to be easier to have it in a different format than a GVDB binary blob, especially if we want to use tools that are written in high level languages that may not have access to the GVDB API. The simplest format we can convert to is JSON, which is structured and easy to parse with other languages.
* Programmatically generate command usage and helpEmmanuele Bassi2018-03-21
| | | | This avoids having each command handle it differently.
* Merge pull request #4180 from endlessm/T21027Martin Abente Lahaye2018-02-02
|\ | | | | T21027
| * SearchBox: Get correct index from selected matchPhilip Chimento2018-01-31
|/ | | | | | | | | | | | When clicking on a result from the autocomplete popup, we should select the first index of the tree path. (The list is a flat list, so there is always only one index.) This prevents a warning about a missing property of [Symbol.toPrimitive], because we were indexing this._items with the indices array (which is not a primitive) instead of its first item. https://phabricator.endlessm.com/T21027
* Merge pull request #4179 from endlessm/T20694Martin Abente Lahaye2018-01-22
|\ | | | | T20694
| * Nest profile output a bit betterEmmanuele Bassi2018-01-22
| | | | | | | | The spacing makes it look like a typo.
| * Move metadata at the top of the outputEmmanuele Bassi2018-01-20
| | | | | | | | | | There is no ordering guaranteed in the GVDB key names; while this is okay for the probes, the metadata section should be clustered together.
| * Fix whitespace in the sample timing messageEmmanuele Bassi2018-01-19
| |
| * Add missing handler for start time metadataEmmanuele Bassi2018-01-19
| | | | | | | | We need to check all meta data keys.
| * Do not clean up the profile state before writing dataEmmanuele Bassi2018-01-19
| | | | | | | | We use the fields in the profile state singleton.
| * Beautify the eos-profile-show outputEmmanuele Bassi2018-01-19
| | | | | | | | Add more Unicode art.
| * Add show command to eos-profileEmmanuele Bassi2018-01-19
| | | | | | | | | | The `show` command loads a list of profile capture files and prints the metadata and sample summary.
| * Move private profile symbols into the private headerEmmanuele Bassi2018-01-19
| | | | | | | | We are going to re-use them in the profile CLI tool.
| * Add CLI tool for loading profiling apturesEmmanuele Bassi2018-01-18
| | | | | | | | | | | | Captures of profiling data are saved in a binary format, and we need a tool that can load them and turn them into user readable (or machine readable) data.
| * Add more metadata in the profile stateEmmanuele Bassi2018-01-18
| | | | | | | | | | | | | | We want to have access to more metadata, including the application id, when capturing profiling data for later review, as the immediate association between a process and its profiling data is not available in that case.
| * Do not attempt to remove probes twiceEmmanuele Bassi2018-01-18
| | | | | | | | | | We only remove probes when unloading state, so we don't need to remove the probe from the state itself.
| * Remove unused refcount from EosProfileProbeEmmanuele Bassi2018-01-18
| | | | | | | | Leftovers from an earlier iteration.
* | Fix in-resize timeout valueJuan Pablo Ugarte2018-01-19
|/ | | | | | g_timeout_add() takes miliseconds instead of microseconds https://phabricator.endlessm.com/T20677
* Merge pull request #4177 from endlessm/T20769Martin Abente Lahaye2018-01-17
|\ | | | | T20769
| * Use similar output for single/multi sample probesEmmanuele Bassi2018-01-17
| | | | | | | | | | | | Always print out the number of samples and the total time, regardless of the number of samples; in case we have more than one sample, be more specific and add the average, minimum, maximum, and standard deviation.
| * Print out total time for single-sample probesEmmanuele Bassi2018-01-17
| | | | | | | | | | | | We should only give up if we didn't have any samples. https://phabricator.endlessm.com/T20769
* | Merge pull request #4176 from endlessm/T20677Emmanuele Bassi2018-01-16
|\ \ | |/ |/| EosWindow: add in-resize css class
| * EosWindow: add in-resize css classJuan Pablo Ugarte2018-01-15
| | | | | | | | | | | | | | Add class while window is being resized to change image scaling method in CSS This would speed up resizing windows with background images. https://phabricator.endlessm.com/T20677
* | Add JS convenience wrapper for ProfileProbe.start()Emmanuele Bassi2018-01-10
| | | | | | | | | | | | The C API has a convenience macro to initialise the location of the probe in the source file. We should use an override for the start() method in order to let GJS fill out the same data for us.
* | Add the Profiling section to the API referenceEmmanuele Bassi2018-01-10
| |
* | Add test for profiling APIEmmanuele Bassi2018-01-10
| |
* | Add initial infrastructure for ProfilingEmmanuele Bassi2018-01-10
|/ | | | | | | | | | | | | | | | | | | | | | | The EosProfileProbe API allows defining profiling probes that can be used to efficiently measure the time spent in a critical section. The Profiling API is meant to collect samples and generate a report at the end of the lifetime of the process, either by printing out the results once the process terminates; or by saving the raw data in a binary file that can be loaded at a later date. This profiling API is meant to be as close as possible to a zero cost abstraction: - probes are only allocated if profiling is enabled - all profiling API is a no-op if profiling isn't enabled - the C API is meant to be easily tied to a scope, through the use of auto-cleanup macros provided by GLib This allows projects using the Endless SDK to keep the profiling probes in place, instead of conditionally compile them in. https://phabricator.endlessm.com/T18514
* New translations from TransifexJenkins2017-11-11
|