yafu bugs

[jwaltos]

Originally Posted by bsquared

..I didn't design it while considering either OS expectations or user convenience...

I have personally found YAFU to be one of the best pieces of software I've ever come across relative to its size and power. YAFU is also very easy to use, not perfect, but I think it's one of the best in its class. In case anyone has difficulty using/understanding this software perhaps they should whisk themselves over to my little blog area and administer a self test (or two) for which I have provided a link (as a service to this forum which could probably use a professional help section).

I just wanted to put a positive spin on this post-riposte and provide some helpful options.
(I won't be at all offended if this post is removed ;))

[jwaltos]

Originally Posted by jwaltos

..my little blog area and administer a self test (or two) for which I have provided a link (as a service to this forum which could probably use a professional help section)..

The link has been deleted but I still think a "self-test" area should exist within this forum. If someone has the capability to set up an account and write some posts then the capacity to take a simple Q & A "self-awareness" test also exists. This proposal should be mutually beneficial.

[BudgieJane]

Originally Posted by Aleksey

1. Your reasons are not related to the bug at all as they are related to the specification provided in yafu README rather than its completely wrong implementation in yafu.
"Except your case" -- it is not my case as this is the case created by the developer of yafu. The basics of operating systems should be fulfilled by a software developer to respect users and their convenience provided by OS features like the concept of current directory.

Well, I won't get those several minutes back again. What a load of rubbish.

Before I retired, I worked for a bank developing and supporting the programs used by their traders in their money-market operations. The way the programs were located was almost identical to how YAFU is located here, except, of course, we only had a single copy of the production code, and if more copies were required they were made by symbolic links.

If you've got two identical copies of the same program, ask yourself one question: "Why?". When you compare them next time and find that they are no longer identical, how can you tell which one is the correct one? How do you know that the one with the later date is the correct one? How long will it take you to determine which is the right one and how much money will you lose in the process?

[Happy5214]

Originally Posted by BudgieJane

Well, I won't get those several minutes back again. What a load of rubbish.

Before I retired, I worked for a bank developing and supporting the programs used by their traders in their money-market operations. The way the programs were located was almost identical to how YAFU is located here, except, of course, we only had a single copy of the production code, and if more copies were required they were made by symbolic links.

If you've got two identical copies of the same program, ask yourself one question: "Why?". When you compare them next time and find that they are no longer identical, how can you tell which one is the correct one? How do you know that the one with the later date is the correct one? How long will it take you to determine which is the right one and how much money will you lose in the process?

On this forum, I've used Prime95/mprime, LLR, OpenPFGW, srsieve and co., mtsieve, GMP-ECM, yafu, GGNFS, aliqueit, msieve, and PRPNet. Of those, Prime95 and LLR can be configured as Windows services and GMP-ECM has an install Makefile target (probably from autoconf). None of the others are set up to be installed to a central system-wide folder like /usr/local/bin. yafu's setup is not uncommon in this community.

[bur]

I was factoring a (more-or-less) random C110 from factordb when during the NFS stage this happened:

33480 100265208893 1203362500562179225106
hashtable: 1024 entries,  0.02 MB
elapsed time of 1881.5290 seconds exceeds 1900 second deadline; poly select done
nfs: commencing algebraic side lattice sieving over range: 1620000 - 1640000
nfs: commencing algebraic side lattice sieving over range: 1600000 - 1620000
Der Befehl ".." ist entweder falsch geschrieben oder
konnte nicht gefunden werden.
Der Befehl ".." ist entweder falsch geschrieben oder
konnte nicht gefunden werden.
nfs: could not open output file, possibly bad path to siever
nfs: could not open output file, possibly bad path to siever
fopen error: No such file or directory
could not open rels0.dat for reading

The same happened a few days back on a different machine for a different composite. Restarting yafu with -R succesfully recommenced the NFS stage, but after a short while it stopped with the same error.

The message sound like there's a problem with finding a file due to ".." being misinterpreted maybe? The german roughly says "Command ".." is either misspelled or couldn't be found". Which is a standard command.com error message since DOS days.

There is no rels0.dat in the yafu directory or in "yafu\..".

If any other log files than the one I attached are helpful, please let me know.

[bur]

After using more suitable search terms I found these posts.

Then with the help of this I hope it should work. Since I deleted the intermediate files I have to restart from ECM/PM1 stages, so it'll take a while until I know.

Sorry for the many edits...

[EdH]

Originally Posted by bur

. . . Since I deleted the intermediate files I have to restart from ECM/PM1 stages, so it'll take a while until I know...

You can use the -noecm flag to skip ECM.

[bur]

Thanks, I suspected that was possible, but hasn't checked yet.