Ammo Type: Flechette about acf

[Bubbus]

Sounds like a ton of fun, had the idea a while back but didn't get around to it.
Reduced propellant sounds good - just waste some space and attribute it to "wadding". Balancing flechettes as an antiaircraft round could work as there aren't any rounds useful for that at the moment, so increased drag would counter that.
Explosive flechettes with a timed fuse would make for a great antiperson mortar round, and I'm betting it'd look cool too. Please thanks!

[Fervidusletum]

I've done a bit more modification and testing. Video demonstrating the round a bit: http://www.youtube.com/watch?v=rEUYmN-eJ7E

I bumped up the limit vel from 800 m/s to 900 m/s (flechettes have tiny cross sections relative to mass). minimum spread is 10, maximum is 60.

Currently 20% of propellant is either wasted or replaced with wadding. Although there is less propellant available, the flechettes still achieve considerable velocity (best pen at 1050 m/s) due to 30% decrease in overall projectile mass (wasted space from packing circles in circles).

With 3 flechettes per round, pen is up 15% over AP pen, but overall damage is 25% less. (50mmC, 110 pen ap, 3x 127.5 pen fl, fired at a non-angled 150mm rha plate with 180 hp, -3.87mm rha with ap hit, -2.87mm rha with 3x fl hit)

I'll do a bit more polish, then submit. Let me know if there's anything that needs tweaking

[Bubbus]

Damage isn't as important for AP as pen, if flechettes have more pen than AP then they'll become the new AP round, all I use AP for is ammo-racking and seat sniping anyway, it'll become a lot easier to do so with a tight cone of "super AP" rounds.

Ideally flechettes would have less pen than an AP round so there's still incentive to use AP rounds, and it'd relegate flechettes to more of an AA/antiperson role. Maybe it should be canister shot instead of flechettes?

[Fervidusletum]

I'm not sure flechette would be 'the new AP'. Pen is important for sure, but the loss of damage means that it will take that much longer for the flechettes to actually wear down the armor enough to start penetrating. The flechettes are only really effective as AP if the armor is close to being penned anyways -- the larger the difference in armor, the more effective AP rounds are, enforcing the idea that flechettes are for dealing with lighter armored, faster targets.

That being said, there are some options to discourage flechettes as a primary anti-armor round. Minimum spread is a quick way to reduce the effectiveness of penning weak spots reliably -- with 3 flechettes and say 15 or 20 degree spread, it's fairly difficult to pen that small weak spot at most ranges.

Alternatively, the minimum number of flechettes can be increased, which significantly drops the pen potential per flechette. This does pose a slight problem with the smaller SA calibers, as they fire too fast to allow much of an increase over the minimum of flechettes.

I'm thinking a combination would be best. Small SAs could be limited with wide min spread, and the rest would have a min flechette increase by 2 or 3, enough to drop flechette pen below AP pen.

[Edit]
I've bumped up the min spread to 20 degrees (covers roughly the size of a 96x96 sprops plate from 5 100mm cannon lengths away). Min flechettes is 6, penetration is 5% less than an equivalent AP round. Flechette rounds took 2 extra shots to destroy a high armor prop compared to AP rounds. (500mm rha plate @ 56 hp, 100mm cannon 203 pen ap, 6x 193 pen flechette)

[Bubbus]

That sounds good. The reason why I'm concerned about pen is ammo racks: they explode based on a random chance. That chance is based upon damage but if you get up close and smack a crate with a bunch of flechettes, it's quite likely to have a bad time regardless of damage dealt. Because you've increased the minimum spread by so much, this would make for an interesting new class of close-combat shotgun tanks though. I'm convinced to add these to an experiment branch and let people play with them - if the feedback is positive then it can go straight into master. Thanks for the good work.

[Fervidusletum]

I'm having problems with making a clean commit, could you please contact me via steam (http://steamcommunity.com/id/fervidusletum/) or email ([email protected])?

Problem stems from trying to keep my fork up to date with this repository -- when I fetch the upstream changes and merge, any future pull requests try to add the merged updates as part the pull request.

[Bubbus]

Alright, I got that patch uploaded to the flechettes branch. Took ages because I had to figure out the patch needed to be converted to UTF-8 and unix newlines.
If you'd like to announce the flechettes on FP and get some feedback on them, you can make pull requests to this branch and after we're sure it's balanced we can merge it into master.

[Fervidusletum]

Since nobody seems interested in testing the balance of flechettes, I suggest merging flechettes into the master, and putting a healthy EXPERIMENTAL warning and directing anyone concerned about balance to post on the FP thread or here on github, in the ammo description.

[Bubbus]

If we're gonna push changes like this, we could use an ACF news box that shows up to admins only once every time the addon is updated. It'd explain the changes and contain a link to the FP thread and github issue tracker, along with instructions to reset to a previous version. This way we can post warnings to the entire ACF userbase and make sure we don't miss anyone. If we implement this or something similar, I'd feel pretty comfortable with pushing experimental changes to master. I'll set up a branch called acfnews on here so we can work on it.

[nrlulz]

We could bring back that changelog on the menu tool
On Jun 22, 2013 12:14 PM, "Bubbus" [email protected] wrote:

If we're gonna push changes like this, we could use an ACF news box that
shows up to admins only once every time the addon is updated. It'd explain
the changes and contain a link to the FP thread and github issue tracker,
along with instructions to reset to a previous version. This way we can
post warnings to the entire ACF userbase and make sure we don't miss
anyone. If we implement this or something similar, I'd feel pretty
comfortable with pushing experimental changes to master.

—
Reply to this email directly or view it on GitHubhttps://github.com//issues/26#issuecomment-19863112
.

[Bubbus]

Yeah, if we got that thing to read git commits from this site it'd be great. We could use the github api if gmod can use https.

[nrlulz]

I'm not familiar with the api, but even if we just have a text file and grab it from raw.github.com and it would work just like before, all we'd have to change is the URL.

Just tested with https://raw.github.com/nrlulz/ACF/master/readme.txt and it worked so I guess https works.

[Bubbus]

API would be great because it'd require zero effort to keep up-to-date once we've laid it down in code. We could even replace the version check with it, I swear half of our commits are caused by that number. Good to hear that it works, I could drop my missile stuff for a day or 2 and mess with the api - sounds like fun.

[nrlulz]

If you can figure out how to get a "version number" (number of commits) or even the id of the latest commit via the api that would be great for the version check. Right now it's fetching the entire webpage and doing a string match for "### commits" which I think is probably a little bulkier than necessary...

[Bubbus]

Combining http://stackoverflow.com/a/4546755 with http://developer.github.com/v3/repos/commits/ looks like a good solution, just need to be sure that commiters have the precommit hook installed. Shame we can't drop that precommit hook onto github itself.

[nrlulz]

I was playing around with the pre-commit hook that Nebual dropped in wiremod. It seems like it would work, but everyone who commits would have to have it installed, pull requests would have to be merged manually to update the version, etc. So I kinda just gave up on that idea...

[Bubbus]

Still a step up on what we have. Alternatively if someone has a server to
spare we can get github to send it a postupload hook and it can bounce back
a modified commit - this would be too much effort though. I'd rather go
with the local hook, seems like a good balance.

A better option would be to make ACF record the first date it's been run
on, and then compare that timestamp against github's latest commit. The
only problem with this is when a person leaves a long time between updating
and first-running, during which a commit sneaks in. We'd also need some
kind of modifiable marker in the addon folder which we can toggle during
that first run. This will be awkward because garry won't let us use the
data folder in the addon folder.

[Fervidusletum]

This is similar to what you suggested, but what about using a text file or something in the garrysmod/data/acf folder (not addons/acf/data/acf ) that keeps track of the last commit / version number the user was notified about.

When acf is started up, it checks to see if there's a data file; if no file then create one and pop up a description of the last few updates. If file exists, it compares its current version with the last notified version from the file, and pops up a description for all commits between last notified and current version, then marks current version as notified. Perhaps limit the number of commits described to 5 or something, so users don't get spammed with commit descriptions.

[Bubbus]

The problem is in knowing the current version, this requires manual updating or a precommit hook. If there's a way of programmaticaly determining current version at runtime, then all problems are solved. Otherwise the timestamp idea is a close approximation of that, and the precommit hook is a version-based solution except in the case of pull requests. I'll try all of these today.

[Redreaper]

Propose we rename to canister. Should have a sharp limit on how many flechettes we can carry, too, since it'll lag, and pen should be very sharply limited. Flechettes have high point impact KE transfer but poor terminal ballistics, why SPIW and the like never caught on.

Make it like you're shooting 14.5mm MGs from a 203 maybe lol

[Fervidusletum]

Added in commit 5fce1b7.