ConwayLife.com

In recent years, oscillators have mainly been stamp-collected on the b3s23osc repository (within a single pattern). But as there are more and more discoveries, this collection is exceeding its limits and can no longer put everything in (due to limitations within hotdogPi's stamp collection scripts). Therefore, I and hotdogPi have been discussing a new period-divided jslife-style stamp collection on Discord since the beginning of this month; note that I haven't run the stamp-collection scripts yet, so this remains a bunch of apgcodes for now.

Repository:
https://github.com/carson-cheng/newjslife

Note: Even though the stator variants have been deduplicated (this reduced the size of the collection by a little bit, especially on the lower periods), rotor variants (like sparker reductions or alternate catalyses) are still not filtered out yet, as no scripts that do this are available. Therefore, the initial editions of the stamp collection will contain some duplicates.

Thanks for doing this! We definitely need some more complete collections so that it's a bit easier to check if something is already known.

How are you collecting the patterns? I notice that there are no metadata. Personally, I really want to see who found each oscillator and when, preferably with a link to a source for the information. If you're scraping the forums, I think you should at least record a link to the first post where the pattern occurred, maybe even with the post's associated username (although the username on the post might not be the discoverer).

Although not extremely important, I think it might be good to use a name that doesn't reference jslife. If anything, it's really an update of the all-osc collection, which was only merged into jslife in Dec 2012, around the time Jason Summers stopped updating his Life pages. On that note, I should probably change "jslife-moving" to something like "moving-objects", but since it's been like that for years I'm not sure what problems a name change will cause.

Also note that there is recent parallel work by Scorbie on building an oscillator database. Each project may benefit from lessons learned by the other project.

Sokwe wrote: August 17th, 2023, 9:01 am How are you collecting the patterns?

It's a combination of b3s23osc and a web scrape of the oscillator discussion thread, possibly plus existing jslife (don't know if that was included or not), then run through apgsearch (so that multiple side-by-side oscillators are separated), then reduced/deduplicated using rstatoropt (viewtopic.php?f=9&t=5468&p=137738&hilit ... or#p137738).

Side note: rstatoropt explicitly doesn't increase bounding box (as it was made for b3s23osc). It finds the minimum population that has the same or lower bounding box in each dimension as the input.

Since b3s23osc is part of the database, anything that exists there can have its discovery information and comments included in this database.

The RLEs and stamp collections have been added to the repository (link in the beginning of this thread) now; note that the metadata (discoverer information and links) is still unavailable yet.

carsoncheng wrote: August 17th, 2023, 6:44 am Note: Even though the stator variants have been deduplicated (this reduced the size of the collection by a little bit, especially on the lower periods), rotor variants (like sparker reductions or alternate catalyses) are still not filtered out yet, as no scripts that do this are available. Therefore, the initial editions of the stamp collection will contain some duplicates.

Would apgcodes (links) for variants be preserved in comments along with discovery information?

Sokwe wrote: August 17th, 2023, 9:01 am Although not extremely important, I think it might be good to use a name that doesn't reference jslife. If anything, it's really an update of the all-osc collection, which was only merged into jslife in Dec 2012, around the time Jason Summers stopped updating his Life pages. On that note, I should probably change "jslife-moving" to something like "moving-objects", but since it's been like that for years I'm not sure what problems a name change will cause.

Also note that there is recent parallel work by Scorbie on building an oscillator database. Each project may benefit from lessons learned by the other project.

Suggested possible names for the new collection/database: "all-osc-new" (an update of all-osc) or "all-osc-YYYYMMDD" (to clarify the current version in the name), or something like "b3s23-omniperiodic".

Reminder for myself to make discussions about the project in this thread:
https://github.com/conwaylife-dev/conwaylifedb

hotdogPi wrote: February 28th, 2024, 5:41 pm It appears to be known (it's in newjslife), but I have no discovery information.

I think it shouldn't be named "newjslife". As far as I understand it is a completely new collection/database.

confocaloid wrote: February 28th, 2024, 5:44 pm

hotdogPi wrote: February 28th, 2024, 5:41 pm It appears to be known (it's in newjslife), but I have no discovery information.

I think it shouldn't be named "newjslife". As far as I understand it is a completely new collection/database.

Do you have a better name? I'm open to renaming. It just happens to be the case that names intended to be temporary become permanent. (Despite being under Carson Cheng's folder, I did some of it.)

https://github.com/carson-cheng/newjslife/tree/main

hotdogPi wrote: February 28th, 2024, 5:50 pmDo you have a better name?

Maybe "all-osc-new" or "all-osc-updated"?
If it is intended to be versioned, version or date could be appended to the name. (Something like "all-osc-20240229")

confocaloid wrote: February 28th, 2024, 5:44 pm

hotdogPi wrote: February 28th, 2024, 5:50 pm

Discussion moved from the oscillator discussion thread to the thread discussing this particular oscillator collection.

Bump

Sokwe wrote: August 17th, 2023, 9:01 amAlthough not extremely important, I think it might be good to use a name that doesn't reference jslife. If anything, it's really an update of the all-osc collection, which was only merged into jslife in Dec 2012, around the time Jason Summers stopped updating his Life pages. [...]
Also note that there is recent parallel work by Scorbie on building an oscillator database. Each project may benefit from lessons learned by the other project.

hotdogPi wrote: February 28th, 2024, 5:50 pm

confocaloid wrote: February 28th, 2024, 5:44 pmI think it shouldn't be named "newjslife". As far as I understand it is a completely new collection/database.

Do you have a better name? I'm open to renaming. It just happens to be the case that names intended to be temporary become permanent. [...]

confocaloid wrote: February 28th, 2024, 5:53 pm

hotdogPi wrote: February 28th, 2024, 5:50 pmDo you have a better name?

Maybe "all-osc-new" or "all-osc-updated"?
If it is intended to be versioned, version or date could be appended to the name. (Something like "all-osc-20240229")

dvgrn wrote: March 7th, 2024, 11:26 am

hotdogPi wrote: March 7th, 2024, 10:21 am 2. The mid-2023 https://github.com/carson-cheng/newjslife/tree/main collection should be added, which is more comprehensive than b3s23osc.

What exactly do you think should be added from "newjslife" (or whatever it might get called instead, like "newosc-{date}")?
[...]

confocaloid wrote: March 24th, 2024, 3:19 am Bump

I don't think there have been recent plans to update the oscillator collection with the late-2023 and 2024 oscillators (so it's still open to naming and arrangement suggestions for now) -- but I think it could be versioned (e.g., all-osc-20240324), or maybe "cclife" (naming suggestion by C_R_116) could also be a decent choice. I think a versioned one could be better though, if someone in the future wants to dig up a collection of oscillators that are discovered before a certain date.