Apply the patch and compile the game

OK

Make sure GUI looks the same
Compare performance, especially for GUI heavy page

Can't see any differences.
https://pasteall.org/media/0/a/0aadcb35c1f61ff6a022367de626e1dd.jpg

Agree with C++ changes.

Successful build - Chance fights ever on the side of the prudent.

Successful build - Chance fights ever on the side of the prudent.

Successful build - Chance fights ever on the side of the prudent.

Successful build - Chance fights ever on the side of the prudent.

Just the OOM fix.

Just the OOM fix.

In D4107#176237, @Langbart wrote:

The images from Barracks, Stable, Elephant stable and Heroes look a bit too dark. For the Stable and the Elephant Stable, the rank symbols are not as clearly visible as in the Barracks image.

Agree, too dark and rank images have low resolutions.

The images from Barracks, Stable, Elephant stable and Heroes look a bit too dark. For the Stable and the Elephant Stable, the rank symbols are not as clearly visible as in the Barracks image.

Successful build - Chance fights ever on the side of the prudent.

Successful build - Chance fights ever on the side of the prudent.

fix for @Stan 's comments. Also replace all "buildings" with "structures" and replace "factions" with "civilizations".

Successful build - Chance fights ever on the side of the prudent.

How does it behave for Delenda Est which has a probably much more complex tree ?

Successful build - Chance fights ever on the side of the prudent.

More fixes by @Langbart and @Imarok comments.

@Nescio would you like to make some comments before adding it to the game?

Successful build - Chance fights ever on the side of the prudent.

Successful build - Chance fights ever on the side of the prudent.

Successful build - Chance fights ever on the side of the prudent.

Successful build - Chance fights ever on the side of the prudent.

Successful build - Chance fights ever on the side of the prudent.

Add macro

Successful build - Chance fights ever on the side of the prudent.

The added new pics to zip file in D4107 look great (war dogs, fishing, formations, ...).
Optional:

• war_elephants: I would include an Indian and an African version, since it is described in the text - war_elephants.txt.
• For city walls one could add some soldiers on to the walls since it is described in the text.

Restore the first loop, I still don't see how to do the second one in a similar fashion

Successful build - Chance fights ever on the side of the prudent.

Successful build - Chance fights ever on the side of the prudent.

Successful build - Chance fights ever on the side of the prudent.

Successful build - Chance fights ever on the side of the prudent.

Successful build - Chance fights ever on the side of the prudent.

Successful build - Chance fights ever on the side of the prudent.

Successful build - Chance fights ever on the side of the prudent.

Fix mistakes found by @ValihrAnt.

In D4114#176000, @s0600204 wrote:

I agree that the structree needs to handle files existing but being unreadable - that's my mistake for not considering that edge case. (Probably should be its own revision though.)

Fair.

I'm having a little problem understanding why the engine is no longer able to display all of the structree's gui elements at once, however. It used to be able to - heck, as mentioned in D2734 the structree used to have ~3964 more gui elements, and they could all be visible at the same time without a crash. So what changed?

Vladislav changed the rendering in rP25229 from a "iterate all and draw" to a "collect and draw" model, similar to the Scene rendering. The benefit is that this will allow better batching & faster rendering in the future.
However, for performance, this uses an std::vector with a max memory limit of 128Kb (I think this bit wasn't expected, but it's a consequence of the design), which the StructTree happens to go over when it fails right now.

Successful build - Chance fights ever on the side of the prudent.

Thanks for the fix.

Successful build - Chance fights ever on the side of the prudent.

Thx, for the patch.
Wording is good.

Successful build - Chance fights ever on the side of the prudent.

Fixes after rP25504, @wraitii is right there is no SDL there, so we'd have to rely on platform implementations. So just fix the TODO instead

Fix notes

Successful build - Chance fights ever on the side of the prudent.

With @s0600204's proposed changes and running apt install libfreetype6-dev it compiles on Ubuntu 20.04 and reports:

"freetype": "2.10.1",

That is the correct version.

Remove whoopsie, add more context

In D4083#175117, @wraitii wrote:

Wouldn't it be better as a property of the option as a whole?

Successful build - Chance fights ever on the side of the prudent.

The code should load dds when png are not available. this breaks the release build. Raising in case we forget.

Update as Langbart suggested. Tried to delete the unneeded files again, this time using the command line and not the tortoiseSVN interface.

Works for me it brings the popup back. Thanks for the patch

Successful build - Chance fights ever on the side of the prudent.

Actually the same bug can be achieved and before the patch, but it requires additional step.

It does not create a segmentation fault for me anymore (rP25733+D4113).
PS: Besides testing it for Capture the Relic also tested it with 'Wonder`, Regicide, Conquest Units, Conquest Structure, Conquest Civic Center and Conquest.

With the suggested changes below applied on an ArchLinux system, userreport_hwdetect.txt reports:

The .txt file for Iphicrates, Pericles, Savanna Biome, Scout Towers, Themistocles and Viriato still exits but are empty --> leading to be still displayed.

Forgot one. Also even better version of the parchment.

No, handmade from the old one.
I've seen that I have forgotten one .txt so I will update it again.

Where is that parchment from? If it's DE it's not compatible.

Successful build - Chance fights ever on the side of the prudent.

Note that so far I can tell, the actual crash is caused by Trying to recover the original scriptInterface from the private data of the JSContext.