On Sun, 17 Jan 2010 21:42:59 -0500, El Smeggo <el@smeggo.org> wrote:
:>A good part size is 128 MB (1/8 GB). This is what JimB0ss is using (among
:>others).
:
:Okay, good. I did an initial RAR of 100 but I'll do another at 128.
:
:Thanks for telling me about the yEnc line size, too. I did a test at
:the default setting of 3,000 but I'll do another at 10,000 or so.
:
:It's no problem using the precise settings, but I'm assuming nobody
:will really quibble much if I use the "rounded off" ones.
The precise settings would mostly benefit you. If your part size is really 128
MB, then 8 of them would equal 1024 MB which is exactly 1.0 GB. (32 of them
would be 4.0 GB... it just makes it easier to tell the size of the post, and
whether it's a DVD5 or DVD9.)
:And regarding a PAR set: What is the recommended Redundancy setting?
:My test was done with 10%, which generated 241 blocks. This would
:lead to a maximum of five recoverable files. Should it be higher?
:Lower?
I'm sure everyone is holding their breath on this one, but I'm not going to
recommend doing what I do. I give new meaning to the word 'frugal' when it comes
to .PAR2's. In fact, I'm downright stingy! (I post about 1%.)
Generally, 4% is considered 'Generous', 3% 'Standard'. This assumes that you are
posting 'Universal' .PAR2 blocks. That means that one .PAR2 block is equal to
the size of one yEnc article (plus overhead). To do that, the 'Source Block
Count' slider must be set to maximum. Also, it is a good idea to set the format
to 'Variable (Powers of Two Sizing Scheme)'. Lastly, things work out more
elegantly if you choose a precise number of blocks to create from the 'Powers of
Two Minus One' series, to wit: 1, 3, 7, 15, 31, 63, 127, 255, 511, 1023, 2047...
etc. (One of those numbers will always be close to the % of .PAR2's you wish to
create.)
Isis
:Thanks!
:Smeggy
:
:
:>His part size is set to 128000000 Bytes and his yEnc line setting is set to
:>10000. These are the 'decimal' settings, and although they aren't precise, they
:>have the distinct advantage of being easy to remember and you can do the
:>arithmetic in your head.
:>
:>The more precise 'binary' settings would be a part size of 134217728 Bytes, and
:>a yEnc line setting of 10592. This will give you a part size of 128 MB as
:>measured on your hard drive (magnetic media).
|
| Follow-ups: | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 |
|