IRC logs for #openttd.dev on OFTC at 2014-05-24
⏴ go to previous day
01:14:41 *** tycoondemon has joined #openttd.dev
04:37:47 *** tycoondemon2 has joined #openttd.dev
07:22:00 *** tycoondemon has joined #openttd.dev
08:34:11 *** Alberth has joined #openttd.dev
08:34:11 *** ChanServ sets mode: +v Alberth
10:45:50 *** frosch123 has joined #openttd.dev
10:45:50 *** ChanServ sets mode: +v frosch123
10:48:13 *** tycoondemon has joined #openttd.dev
11:32:38 *** Supercheese has joined #openttd.dev
11:32:38 *** ChanServ sets mode: +v Supercheese
13:35:13 *** tycoondemon has joined #openttd.dev
16:11:34 <frosch123> looks like showsavepresetinwdow and sumgrfid can be declared static
16:12:13 <Alberth> good point, didn't think of that
16:13:46 <frosch123> i would call sumgrfid something like getgrfidhash, and add a constant GRFID_HASH_MAX instead of 8*15+1
16:15:05 <Alberth> indeed, it looks a bit magical now
16:16:37 <frosch123> +STR_NEWGRF_SETTINGS_UPGRADE_TOOLTIP :{BLACK}Upgrade NewGRF files that have a newer version available at the system <- how about "Upgrade NewGRF files for which you have a newer version installed"
16:16:54 <frosch123> i fear some confustion about "upgrade" not downloading anything :p
16:18:51 <Alberth> I tried to cover that with "available at the system", but your idea is even more precise
16:22:19 <frosch123> the save preset window has no resize box
16:22:50 <frosch123> and i would expect it to highlight a line when clicked
16:23:51 <frosch123> possibly preselecting one when the load dropdown does not display "custom"
16:23:56 <frosch123> i.e. when you edit just after loading
16:25:18 <Alberth> perhaps call it "update" instead of upgrade?
16:25:55 <Alberth> tell that to debian :p
16:26:09 <Alberth> but the newgrf window uses "upgrade"
16:26:20 <frosch123> well, upgrade is more correct here, than update
16:28:02 <frosch123> actually, how about two buttons?
16:28:46 <frosch123> "check for newer version" -> "open content download restricted to used grfids" (not sure whether that works already, currently we only do that with md5sum)
16:28:50 <frosch123> "use newer version"
16:29:02 <frosch123> but, well separate patch :p
16:29:17 <frosch123> "use newer version" is kind of long for a button text :)
16:29:46 <frosch123> well, just call it update or upgrade, people will figure it out :p
16:30:00 <frosch123> they do not read anyway, but will just click all buttons
16:34:31 <frosch123> hmm, let's try with my 1.4k grfs
16:35:07 <frosch123> the hash is likely to use the numbers in the middle most of the time
16:37:43 <frosch123> looks fast enough :)
16:40:25 <Alberth> otherwise you probably have to sort an array on grfid, and do a binary search
16:43:40 <Alberth> perhaps add a "disable upload newer version" flag for newgrfs :)
16:44:17 <frosch123> you mean an auto-upload feature? :p
16:44:41 <frosch123> auto upload any grf that bananas does not have... i cannot imagine the rant :p
17:22:51 <Rubidium> I don't care about the rant, I care about the shitload of takedown notices
17:24:32 <Rubidium> now people rarely uploading stuff that they shouldn't upload is probably due to having to create and account and all stuff. If they now have a magic button that just does that... someone will just crawl their 1k+ NewGRFs to find ones that aren't there yet
17:25:32 <Rubidium> 10: +void ShowSavePresetWindow();
17:25:55 <Rubidium> that should be static (unless something else needs it later on)
17:28:34 <Rubidium> + size->height = max(min(this->presets.Length(), 20), 5) * resize->height;
17:34:40 <Rubidium> + char key[512], params[512];
17:34:58 <Rubidium> (in 20) key should account for MAX_PATH
17:39:06 <Rubidium> in 30 I wonder whether the trickery with creating a "hash" of the GRF ID is really useful
17:41:38 <Rubidium> *if* you want to do something, use std::map<uint32, const GRFConfig *c>. Easier and probably faster
17:42:27 <Rubidium> at least less collisions, and no need for further looping
17:49:32 <Alberth> added static already, aded ClampU
17:52:33 <Alberth> /* Hex grfid (4 bytes), "|", hex md5sum (16 bytes), "|", file system path. */
17:52:33 <Alberth> char key[4 * 2 + 1 + 16 * 2 + 1 + MAX_PATH];
18:00:50 <Rubidium> that seems okay; maybe replace "x bytes" by "x bytes in nibbles"?
18:29:26 <Alberth> changed the comment text, and added the std::map, makes finding a newgrf much more compact
18:41:00 <frosch123> either you have to make it a mulitmap
18:41:08 <frosch123> or already select the newest upon inserting
18:41:26 <frosch123> i guess the latter is more useful
18:41:40 <frosch123> insert of insert, find and compare version
18:43:20 <Alberth> euhm, I throw the this->actives in, which have a unique grfid
18:44:20 <frosch123> it works the other way around :)
18:44:29 <Alberth> and copy grfs from the _all_grfs list, updating the value of the map
18:47:19 <Alberth> doing it the other way around would enable reducing copying versions, in case you encounter a newer but not newest version
18:47:52 <Alberth> but the multi-map would make things more complicated again, so it won't save much, I think
20:04:55 *** Alberth has left #openttd.dev
21:14:12 *** Klanticus has joined #openttd.dev
23:07:51 *** tycoondemon has joined #openttd.dev
23:39:21 *** tycoondemon has joined #openttd.dev
continue to next day ⏵