Any interest in all sequences/open sequences?

[Greebley]

I just recently updated all open ended sequences (March 20 - 9233 of them after 940470 terminated)
I also have an updated list of all sequences (see below) below one million.
Would anyone like me to make either list available via sendspace?

----
The list of all sequences is in a special format. There is no program to read this but it never duplicates values. The format is:
1) There is a SeqDesc.out file in special format with all numbers 1 to 999999 (. means the number, # is include another sequence).
a)It uses a format like 18: . 21 #11:0 which means .=18, 21, zeroth term of 11=11, then follows 11 (which is 11: . 1 for 11 1)
b) 21 will have the format 21: #18:1. It has no '.' Since 21 itself is part of a previous sequence.
c) Cycles can refer back to themselves like 6: . #6:0
d) This covers all factors for sequences that never reach 2^64 before merging.
e) For sequences like 138 and 276, the format is 138: #. which has special meaning to look for an alq_138.enh file.
2) There is an alq_<seq num>.enh file for each #. in SeqDesc.out. It is the same format as an *.elf file except that it can end with #include line to tell how the sequence continues: For example, 237552.enh ends with
133 . 3975989072 = 2^4 * 11 * 31 * 113 * 6449
#include 70740:28

which means the sequence continues with the 28th index of seq 70740. It can refer back to the SeqDesc.out file - for example 138 ends with #include 59:0 for 59.

For open ended sequences the enh file is the same as the elf file. There are also a few terminating sequences that don't have a #include because they end in a larger prime, like alq_2136.enh ends with 73 . 15559 = 15559. For this special case the 1 term is not included.

This information comes from the database. There are a few broken sequences that don't output correctly, but the data is there.

[kar_bon]

I can upload such file on my Aliquot-pages or in the Download-page on www.rieselprime.de, so they will reachable not only for a few weeks like on sendspace.

[Greebley]

What format would be best for you kar_bon? The 7z format is less bytes than the zip format. Not sure if I should use it though as it might be less common. It is maybe 70 MB less in the 7Z format.

Is it better as one zip file, or 3 or 4 files of 300 MB or less?

The way I know to get it to you is to break it up into pieces and use sendspace - unless there is a way I could ftp it directly?

for the zip format the full set of all sequences using the enh format is just about 900 MB, and the open sequences are about 800 MB.

I could make it self extracting but that would likely work only for windows.

[Greebley]

There are 4 files:
seq.zip, seqz01, seq02 - these 3 are needed for the open sequences
all.zip - this is the rest of the sequences.

http://www.sendspace.com/filegroup/K...W7x8%2FElRYSow

[kar_bon]

I've downloaded all files and unpacked w/o any problems.

The description-file (SeqDesc.out) I can upload on my page (~7MB zipped) but the others (*elf and *.enh) are too big (download time too long).

Perhaps there's another option to make them available for longer terms than at sendspace et al.

Some remarks to SeqDesc.out:
Looking at the graph of seqs ending in (for example) 41, I suggest some changes there.

yours:
41: . 1
48: . 76 64 63 #41:0
63: #48:3
64: #48:2
76: #48:1

I suggest like this:
41: . 1
48: #76:0
63: #41:0
64: #63:0
76: #64:0

So the line 48 -> 76 -> 64 -> 63 -> 41 is straight foward instead of giving 63 as #48:3.
This will also shorten the file, if more of such changes could be done
(just another example the line "938: . #138:171", better "938: . #694:0").

Other changes:
- The ':' can be omitted (~ 1MB less!): first number of line is seq-number (~5,9MB less if seq-number is not listed at all (line-number = seq-number) but not much user-friendly and unreadable)

- The '#' can be omitted, too: if a value is given as "xx:y" it means seq xx at index y, otherwise (without ":y") the value only (~920k less characters)

- The same for '.'.

Such changes would shorten SeqDesc.out from 19,8MB to about 12MB or even lesser.

[schickel]

An update on my sequence summary file has been uploaded here.

I didn't get around to fixing the listings for sequences that have problems, so check the main list against the sequences posted here that are known to have problems (basically there may be more, so if you find some, please post them in the FactorDB thread.....)

[henryzz]

I could seed a torrent theoretically.