The file only had a run-time effect if it listed any files as "fuzzy", which it
hasn't since the last change in 2015. However, that isn't a reason for removing
it.
The reason for removing it is that it doesn't make sense to automatically mark
images as fuzzy when an image changes, because fuzzy is often much better than
untranslated. This is different to the logic for strings, where a single word
can negate the meaning of a sentence. The decision has to be a manual process,
and at that point it might as well be done with the source-control tool's
delete-file function.
Even considering the previous times that l10n-track was used, I can't see a
scenario where it would be useful.
* Scenario: a screenshot needs to be replaced. In this case the new images are
going to be screenshots too. If the old images are out of date, better to
just delete them instead of needing the fuzzy mechanism; either option would
have the same effect of showing the up-to-date-but-untranslated image.
* Scenario: a map is completely redrawn. This seems unlikely, as a completely redrawn
image could be added as a new map rather than replacing the old one. Even if
it does happen, it's probably best to just delete the old --overlay.png
files.
* Scenario: a new landmark is added to a map. Using the existing overlay file
is still going to be good, as it means the rest of the map is shown
translated. There's surely new dialogue in the WML files commenting on this
new landmark, which would be a good place to say
# po: Lorem's Ipsum was added to the map on 24 Jan 2020, please add a label to your language's --overlay file.
* Scenario: an old feature is removed from the map. Using the existing overlay
will have an out-of-date label, and the significance of that would depend on
how it affects the storyline. This could be significant enough to make
showing the English labels be a better option than showing the translated
ones. However, the translated maps could be updated without knowledge of the
language, so if it's such a significant change then the developer writing the
new WML should probably do that.
* Scenario: a feature moves slightly on the map. This is an example where the
l10n-track file has been used, it happened with the Bitter Swamp on the
bottom-right corner of the title screen. I think it would have been better to
show the slightly-out-of-date overlay instead of showing the English.
Note about cherry-picking: in the 1.14 branch, the change needs to be applied
to src/picture.cpp instead of src/filesystem.cpp.
Reasoning being:
* The result_set and other APIs are nicer to use.
* We use mariadb on our server rather than mysql, so this would minimize the chance of any incompatibilities.
* The mysql C++ connector 1.1 isn't compatible with with c++17 (https://stackoverflow.com/q/47284705).
Reasons:
* These projectfiles are still 32-bit, whereas all other builds for all other OSes are now 64-bit.
* Relatedly, these projectfiles are dependent on the libraries kept at the aquileia/external repository.
* VS2019 is still listed as supporting Windows 7, so there is a minimal likelihood of developers not being able to use VS2019.
* The VS2019 projectfiles instead use vcpkg to get the required dependencies, which is easier to setup than the aquileia/external prebuilt libraries.
* It's one less thing that needs to be updated whenever source files are added/moved/removed.
* It's two fewer jobs that Travis needs to run, which means Travis builds will finish more quickly.
The first run of the 2019 jobs will fail, since that run will be used to build and cache the vcpkg dependencies - there's no way to get enough time to build wesnoth and build the dependencies in a single job.
This also removes the separate WML_tests.cmd scripts from VC14 and VC16, now instead using the same run_wml_tests script as the linux jobs.
There is currently no caching of the Windows jobs, since msbuild unfortunately uses timestamps rather than more complete information like scons/ccache in order to determine if things need to be rebuilt. Since git doesn't preserve any sort of last modified timestamp, the cloned repo into the travis job is always "newer" than the cached compiled output, and therefore wesnoth is always fully rebuilt.
Ideally, assuming this sticks around and gets out of early access from travis, we could then use just travis instead of travis and appveyor.
