Log In  

so, I modify my sprites and export the cart as html. works ok in firefox, edge etc. but not in chrome: the exported cart is stuck with a four-days-old version of the spritesheet!

long story short, this can actually happen on ANY html-export of any cart:
the js file starts with

var _cartname=['whatever.p8'];

now if I change this for

var _cartname=['breakout.p8'];

then the cart pulls old data from thin air! the data that was in "breakout.p8" FOUR DAYS AGO! I don't even have any file featuring the outdated data.

so:

  • four days ago the data from "breakout.p8" got cached "somewhere" for some reason.
  • the cart finds some cached data labeled "breakout.p8", and loads it instead of the data in the js file.
  • gfx & sfx as far as I can tell, but code is ok.
  • tried other carts names, no problem. something special happened with "breakout.p8".
  • only in chrome. cleared the cache etc. no luck.
  • seems related to the multicart system?

solved the problem by renaming my project file, but it's gotta come and bite me again...

semi-related question: where do the html exports save their cartdata?

P#42167 2017-07-03 19:29 ( Edited 2017-07-04 19:51)

ok. I can reproduce that with cstore.
cstore() from an html export will save all the cart data somewhere else (where?) and that data will prevail.

I guess the cstored data should be temporary? it seems to just stay there in chrome.

still the weird thing is I didn't even use cstore at all in my breakout project!

(edit) btw, it was about chrome on windows10 but the same thing happens with chromium on linux mint. firefox is ok everywhere.

P#42168 2017-07-03 21:19 ( Edited 2017-07-04 09:42)

Look up in the manual where PICO-8 stores the config.txt file. (It's different for every OS.) That folder has many things in it you maybe weren't aware of. For example, backups of carts you've been working on are stored there, splore data, and so forth. Data stored there is likely where your problem is stemming from.

P#42174 2017-07-04 02:41 ( Edited 2017-07-04 06:41)

@MBoffin I had checked there, nothing special. the html exports do their own stuff though. also they are (should be) self-contained.

P#42176 2017-07-04 07:05 ( Edited 2017-07-04 11:05)

If its just happening on Chrome do you think it's a caching issue? Have you tried clearing cookies from Chrome? Is the copy you're running at the same directory / does it have the same filename as the previous version?

I've run into issues with Chrome storing old states for other things that cause an updated version of a site to run oddly. Clearing the offending cookies could work. You can see a list of items if you click to the left of the url.

P#42182 2017-07-04 14:29 ( Edited 2017-07-04 18:29)

thanks enargy! I never clicked there it seems :)
it says "0 cookie in use", but if I block them and refresh, the cstored data is not fetched back. that's where it's stored then, a permanent "hidden" cookie...?

oh well, for all I know it differs between chrome and firefox, so I guess it's up to zep to sort this out now...

P#42183 2017-07-04 15:46 ( Edited 2017-07-04 19:52)

[Please log in to post a comment]

Follow Lexaloffle:          
Generated 2023-01-30 22:10:53 | 0.008s | Q:14