Author Topic: Lix 0.7.22 experimental  (Read 13370 times)

0 Members and 1 Guest are viewing this topic.

Offline Simon

  • Administrator
  • Posts: 3860
    • View Profile
    • Lix
Re: Lix 0.7.6 experimental
« Reply #15 on: June 20, 2017, 05:23:01 PM »
Lix 0.7.6 experimental.

Download 0.7.6 for Windows -- keep separate from the stable 0.6 installation
Branch exp in lix-unstable

Physics bugfix: Duplicated brick are gratis, they don't decrease number of bricks left. This didn't impact replays; we still need 37 lemforum and 29 Rubix replays.

-- Simon
« Last Edit: July 01, 2017, 02:56:20 PM by Simon »

Offline Simon

  • Administrator
  • Posts: 3860
    • View Profile
    • Lix
Re: Lix 0.7.8 experimental
« Reply #16 on: July 01, 2017, 02:56:08 PM »
Lix 0.7.7 and 0.7.8 experimental.
  • Physics in 0.7.8: Walker always turns, even if it has cancelled a worker. I don't know whether I want to keep this. I've built both executables for Windows, 0.7.7 (with the 0.6 walker) and 0.7.8 (with the always-turning-walker).
  • The backroute fixes from 0.6.39, and geoo's lemforum level changes from the last few posts.
  • Some work on overtime, but that should be invisible. I want overtime in 0.7 because it's important to end drawish games, even if rarely necessary. Neutral lixes should wait at least until 0.8.
I'll be away for a week. You can play around with the walker physics. I'm still 50:50 on what it'll be, maybe slightly pro-always-turn after Rubix's most recent post that he likes them.

Would like feedback on the batch-verify button in the replay browser. It's crude because it merely prints text; it doesn't mark replay file entries in the browser. Still helpful?

-- Simon
« Last Edit: July 17, 2017, 08:36:54 PM by Simon »

Offline mobius

  • Posts: 2747
  • relax.
    • View Profile
Re: Lix 0.7.8 experimental
« Reply #17 on: July 02, 2017, 09:44:50 PM »
I downloaded the 0.7.8 exp build. (I don't currently have any other version on this PC)

some kind of glitch happened when I had tabbed out of Lix to another window my cursor got stuck in the top right corner of my monitor (not the Lix window, my whole screen). Only closing Lix solved the issue. It hasn't happened again-- can't say what caused it.
everything by me: https://www.lemmingsforums.net/index.php?topic=5982.msg96035#msg96035

"Not knowing how near the truth is, we seek it far away."
-Hakuin Ekaku

"I have seen a heap of trouble in my life, and most of it has never come to pass" - Mark Twain


Offline Simon

  • Administrator
  • Posts: 3860
    • View Profile
    • Lix
Re: Lix 0.7.8 experimental
« Reply #18 on: July 03, 2017, 03:50:13 AM »
Thanks for reporting. I reposition the mouse cursor to allow infinite movement, strange that this bug hasn't surfaced so far. I haven't seen that before and couldn't repro either. Will examine the source and guess.

In a pinch, Shift+Esc kills the game without questions, losing all unsaved data.

-- Simon

Offline Simon

  • Administrator
  • Posts: 3860
    • View Profile
    • Lix
Re: Lix 0.7.9 experimental
« Reply #19 on: July 17, 2017, 08:33:25 PM »
Lix 0.7.9 and 0.7.10 experimental.
  • Physics change: Reverted to 0.7.7 physics, i.e., the walker skill either cancels, or turns. This is the behavior from 0.6 and earlier. I've rejected my proposed walker change. It was a hard decision.
  • Shotgun-debugged mobius's mouse problem by applying a Linux workaround for all systems. I couldn't test, hope this helps.
  • Proxima's changes to lemforum levels from his posts above.
  • The lemforum pack is 233/240 covered. Thanks!
-- Simon
« Last Edit: July 30, 2017, 11:02:46 PM by Simon »

Offline Simon

  • Administrator
  • Posts: 3860
    • View Profile
    • Lix
Re: Lix 0.7.10 experimental
« Reply #20 on: July 17, 2017, 09:02:18 PM »
Updated to 0.7.10, Vicious/_order.X.txt mentions the new One-Lix Showdown instead of Lixes in Arms.

Let's cover all of lemforum, and as many Rubix levels as possible, until end of July August 10th. I'd love to declare Lix stable with Raymanni's tilesets then!

-- Simon
« Last Edit: July 31, 2017, 07:38:44 AM by Simon »

Offline Simon

  • Administrator
  • Posts: 3860
    • View Profile
    • Lix
Re: Lix 0.7.10 experimental
« Reply #21 on: July 20, 2017, 05:36:44 PM »
Lix 0.7.10 replacement executable with mouse code from April 2017. Proxima had trouble with the mouse, let's see if exe fixes that on Windows. If still bad, I'll get even older mouse code.

-- Simon

Offline Simon

  • Administrator
  • Posts: 3860
    • View Profile
    • Lix
Re: Lix 0.7.10 experimental
« Reply #22 on: July 24, 2017, 04:14:24 PM »
Strategic plan for 0.7.
  • Debug overtime. I had a nasty build problem where the server wouldn't initialize on my home machine, but looks fixed now but I was an idiot, I had an old instance running that already listened on the port. I can finish the physics myself, we don't have to schedule test sessions yet. Physics must be good, everything nonphysics can come after the stable 0.7 release.
  • Look at möbius's levels. Wait for Proxima to decide what goes in and cover lemforum 100 %. Should be done around August 10th. (Not August 1st because Proxima is on visite surprise en France.)
  • Rubix: Have you covered some of your 29 levels that need extra replays? In two weeks, I'd like to release Lix 0.7 as stable. I can hold back the uncovered levels and include all covered.
  • Have a multiplayer level design contest after the stable 0.7 release. I'd like to encourage new levels with Raymanni's tilesets or the new Oriental set.
-- Simon
« Last Edit: July 24, 2017, 04:35:16 PM by Simon »

Offline Proxima

  • Posts: 4562
    • View Profile
Re: Lix 0.7.10 experimental
« Reply #23 on: July 24, 2017, 07:21:15 PM »
More coverage replays: Survival of the Craftiest, From the Other Side, Toccata.

The following levels currently have problems (intended solution does not work):
Dream the Impossible Dream: problem with floater/tumbler behaviour. May be fixable with a terrain change.
Metal City Mayhem: problem with builder/blocker interaction. Should certainly be fixable by changing the SI.
« Last Edit: July 24, 2017, 09:39:45 PM by Proxima »

Offline Simon

  • Administrator
  • Posts: 3860
    • View Profile
    • Lix
Re: Lix 0.7.12 experimental
« Reply #24 on: July 30, 2017, 11:55:27 PM »
Lix 0.7.12 experimental.

(Ask me if you want this before the 0.7.13 physics changes, I'll build you this.)
  • lemforum pack: Simple renamed to Lovely. Some backwards-compatibility code that converts progress checkmarks during loading the user file.
  • lemforum/wip-0.7 has all levels in consideration for lemforum pack's stable release with 0.7.x in 2 weeks. I hope all levels are up-to-date.
  • Raymanni's 3 water tiles have higher trigger areas. (Lixes drown earlier and higher up than before when they drop into the water.)
  • images/geoo/construction/chain1.png added temporarily. This is 57 pixels high. In case we keep a lemforum level that requires chain1: Please revise the tile and give it a multiple of 8 for its height!
  • Zoom 1x is always allowed, even if the level doesn't fill the available area.
  • Fix #206: Editor browsers show the innermost directory of hatches, goals, traps, and other gadgets. (The bug was: All hatch buttons are called "hatch", that's zero information.)
  • Goals are locked during the nuke. The nuke is purely for decoration and to finish replays in style.
  • Mouse code reverted to the 0.6 mouse code. If anybody has trouble, tell me.
  • Keyboard key duration before first repeat for moving selected editor pieces: 0.33 seconds -> 0.2 seconds.
  • Lots of multiplayer nuke code that handles conceding and calling for the nuke. When you press the nuke button, you lose control over your lix, overtime starts, but your lix don't explode yet. We'll see how it feels in 2-3 weeks. Let's get the singleplayer into shape this week!
  • Editor allows to set overtime in multiplayer levels. Existing multiplayer levels are tagged with the correct player number.
-- Simon
« Last Edit: August 02, 2017, 02:21:17 AM by Simon »

Offline Simon

  • Administrator
  • Posts: 3860
    • View Profile
    • Lix
Re: Lix 0.7.13 experimental
« Reply #25 on: August 01, 2017, 09:22:36 PM »
Lix 0.7.13 experimental.
  • Physics: The basher ignores steel in its top 5/18 rows, not merely 2/18.
  • Physics: The digger ignores steel in its 4 outermost hi-res pixels on both sides, not merely in its outermost 2 on each side.
  • Physics: The digger digs every (16*n) frames following its assignment, not already every (16*n - 4).
  • Fixed bug #202: Small levels image-export with all stats again, not only with their skills.
I don't know whether I'll keep all these physics exactly. I'll probably keep them. The changed digger frame breaks many replays. :lix-cry: The changed steel sensitivity breaks very little. Reasons for these changes.

-- Simon
« Last Edit: August 02, 2017, 03:04:44 AM by Simon »

Offline Simon

  • Administrator
  • Posts: 3860
    • View Profile
    • Lix
Re: Lix 0.7.14 experimental
« Reply #26 on: August 02, 2017, 02:19:48 AM »
Lix 0.7.14 experimental.
Again, please try hitting steel at various heights with the basher. Startup animation for digger is still placeholder.

-- Simon
« Last Edit: August 09, 2017, 03:54:09 AM by Simon »

Offline Nepster

  • Posts: 1829
    • View Profile
Re: Lix 0.7.14 experimental
« Reply #27 on: August 02, 2017, 03:41:35 PM »
Here are some minor issues with the current experimental version 0.7.14:
1) Please use "\r\n" as a newline in both level files and replay files, regardless of the current OS. It is pretty annoying when I want to modify a replay file created by Simon on Linux and everything is displayed in a single line on my computer...
2) Some of the digger sprites still have numbers "0" to "4". Please do not forget to remove them for the new stable version.
3) Lix does not close the "log.txt" file properly. This means while Lix is running, I cannot delete the file or parts of it, but rather have to scroll to the very bottom every time I want to read the replay verified results.
4) Anyway, the replay verified results should probably not be written to "log.txt", but rather to a separate text file.
5) You mentioned at some point that the SI can still be displayed in some way. I have not yet found out, how to do this. This needs to be a lot more discoverable.

Offline Simon

  • Administrator
  • Posts: 3860
    • View Profile
    • Lix
Re: Lix 0.7.14 experimental
« Reply #28 on: August 02, 2017, 04:05:32 PM »
1) Please use "\r\n" as a newline in both level files and replay files

Unsure if I want to print \r\n, but will keep in mind and research what other programs would do.

Use a normal text editor, not Notepad with its 30-year-old bugs. >_>

Quote
2) Some of the digger sprites still have numbers "0" to "4". Please do not forget to remove them for the new stable version.

Yeah. I'd like to draw 4 unique frames of startup animation. These 4 new frames aren't looped, let's see how creative I get before the release.

Quote
3) Lix does not close the "log.txt" file properly. This means while Lix is running, I cannot delete the file or parts of it, but rather have to scroll to the very bottom every time I want to read the replay verified results.

Yeah, I should close it after any larger chunk of logs.

If you're desperate and want this before tomorrow, find a Windows version of tail and then tail -f log.txt in a shell. Or verify from the command line instead:
lix --coverage path/to/your/replay/dir

Quote
4) Anyway, the replay verified results should probably not be written to "log.txt", but rather to a separate text file.

Ideas for where? Next to where the image exporter writes images, export/nameOfDirectory.txt?

Quote
5) You mentioned at some point that the SI can still be displayed in some way. I have not yet found out, how to do this. This needs to be a lot more discoverable.

Hover with the mouse over a hatch. Good to know that this is undiscoverable. Probably mouse cursor isn't expected to interact with anything but the lixes on the map. Thanks!

-- Simon

Offline Nepster

  • Posts: 1829
    • View Profile
Re: Lix 0.7.14 experimental
« Reply #29 on: August 02, 2017, 04:17:14 PM »
Quote
4) Anyway, the replay verified results should probably not be written to "log.txt", but rather to a separate text file.
Ideas for where? Next to where the image exporter writes images, export/nameOfDirectory.txt?
A text file in data is fine, e.g. "replaycheck.txt". Just not the "log.txt" that already contains huge amounts of other stuff that is totally uninteresting for the average player.