不明绑定

华夫饼全是黑种吗？

不懂路过

帮顶。我的核弹头都用不了

不明

AR所用的系统是不太一样，我印象中AR、UHDbits一类的属于不同的设置，最后我有试出可行的方法。我找找记录，待会儿贴上来。

贴吧一直吃掉我写的详细设置…
白打了两次，每次贴完，重载后文章就不见了...

IRC帐号的密码及电邮栏位是否设置。西方的不同PT的设置不太一样，简言之，有三种排列组合：
1. 帐号、密码、电邮皆填入。
2. 填入帐号、密码，不填电邮。
3. 填入帐号，不填密码、电邮。
三种之中会有一种成功，不成功的话就三种设置都试一次，总有一种成功。最好是先用IRC手动注册IRC autodl帐号专用的密码及电邮。
以下是我先在IRC设置好IRC autodl帐号后，试后AR可行的autodl.cfg设置。

同場加贈AR適用的高效率灌上傳的filters:

上图中的热门美剧字串如下：
The Walking Dead, The Flash, Arrow, The Expanse, Stranger Things, The Big Bang Theory, The Expanse, Legion, Suits, The Blacklist, Marvels Agents, Supernatural, Legends of Tomorrow, The Magicians, Its Always Sunny in Philadelphia, The 100, Criminal Minds, Modern Family, Talking Dead, The Daily Show, Colony, Family Guy, Blindspot, Supergirl, Elementary, Hawaii Five
（适用于现在，过个半年一年，热门的美剧就会变了。）

有需要RED或apollo的刷ratio filter 设置再跟我说，我在RED上也是刷完收工了。

再補充一下：
AR 每二、三個月會有一次全站免費，那時如果有盒子，隨便無腦刷都隨便爆上傳，一律全下載即可。因此不急著在平日刷AR。

以下是apollo充ratio的設置：

戲法人人會變，巧妙各有不同，可自行修改。大原則就是用IRC autodl下載較多人下載的種。千萬不要用RSS，那太慢了，連灰都吃不到。
音樂站通常需要國外的盒子才刷得動，而且要用Deluge。要不然就是永遠作種，自然也會慢慢累積。
但要注意，apollo bot 常常掛掉，那麼自然沒辦法自動下載。
另一個重要技巧：apollo 及 RED的下載目錄最好設同個目錄，因為兩邊的檔案很多是重覆的，都放在同個目錄，先前已下載過的話，Deluge 會直接輔種。因此我的apollo ratio在３以上。

其实最好是先不用autodl，先用irc客户端，把认证进频道走一遍，看到通告信息，再用autodl

在以上各filters设置中，关键都是在于挑较多人下载的种子。
为什么要挑较多人下载的种子？因为越少人下载的种子，平均ratio越低于1；越多人下载的种子，平均ratio越接近1。
另一方面，上传及下载的总和几乎是个「零和游戏」（zero sum game），盒子的速度较快，就容易在获利的一边。
这背后的数学我思考过，解释如下（由我在RED写的文章贴过来）：
In terms of the mathematics behind buffer/ratio for the autosnatchers, AFAIK unless there's freeleech, or except for the people uploading new content, the upload or ratio of each file in nearly a "zero sum game".
The average ratio for people downloading/uploading the same file should be close to 1. This means that for one guy who gets > 1 ratio from a torrent, there's another guy getting < 1 ratio. How high ratio we can get from each torrent is affected by three factors:
1) How our server competes in the swarm. (This includes factors such as connectivity between the downloaders - therefore the European seedboxes have an advantage because they are inter-connected better.)
2) How many people downloaded the torrent, both during the initial swarm and afterwards.
3) Freeleech, including Freeleech token, which stimulates the economy. To simplify the matter, we can exclude this factor in the following analysis.
The mathematics can be easily proven. Here is a simple example:
Assuming there is only the uploader and 1 downloader in total, how the upload and download would look like? The uploader would have exactly the same size of the torrent as the upload buffer s/he gains from this upload. The downloader would have 0 KB upload and the size of the file as his/her download from this file. Mathematically, total upload amount = total download amount. Ratio of (sum of all upload) / (sum of all download) = 1.
If there is an additional downloader, the situation gets better. The uploader at least gets his guaranteed amount of upload, plus some additional upload. Each downloader on average would get less than or equal to half of the file size as the upload amount.
If there are 3 downloaders, each downloader on avarge would get <= 2/3 of the file size as the upload amount.
If there are 4 downloaders, each downloader on average would get <= 3/4 of the file size as the upload amount.
The more downloaders there are, the closer the total average ratio of the downloaders would be to 1. Therefore, autosnatching the files with more autosnatchers (2017 FLAC lossless CD with log with cue) in general results in higher ratio than autosnatching the files with few or no downloader. However, unless there is some kind of freeleech, it can never be 1 for the downloaders because the uploader is guaranteed for at least the file size as his upload.
Therefore, it's a zero sum game. Just that those who get in earlier would be on the profitable side, and those who get in later would have ratio < 1.
During the initial swarm, those with fast servers and great connectivity will get in earlier than others, thereby having high ratio. Those with slow servers or poor connectivity will get in later than others, thereby having low ratio. The people who download manually hours or days later will not have a good ratio, until people who download even later download from them.
I used to try to "optimize" the autosnatch setting to achieve the highest ratio possible. After thinking it through, I no longer want to maximize my ratio otherwise people with slower servers would suffer. After all, all the additional buffer is useless if we don't use it, and promotion of user class here is only based on total upload, not requiring extremely high ratio.
Nevertheless, autosnatchers do help the economy grow. Because of the autosnatchers who keep their files after the initial swarms, every downloader afterward can download the file much faster from multiple sources. Also, the more downloaders there are, the closer to 1 the average ratio becomes.
In summary, autosnatchers can unite to help the uploaders by gradually expanding the downloading criteria, if we can afford. The autosnatchers who barely get the ratio 1 is safer to use the most popular setting (2017 FLAC lossess CD with log and cue), while the autosnatchers who have got enough buffer can expand their setting to include unpopular downloads to help the uploaders and help the community grow. Except the upload acquired by providing new content, too big a buffer from autosnatch won't help anyone including ourselves and could just make other people suffer low ratio.