Quite likely the API of the libs could have changed, and it just takes
one such thing to force you to call in Mycroft Holmes to solve it.
Remembering the highest-paid consultant editing an IBM mainframe OS
binary in hex at NASA once when it was the only way to get the system back.
On 12/16/20 12:49 PM, Kenneth Huang wrote:
> Hi Ryan,
>
> Would the current libraries contribute that much of a difference? I don't
> expect them to be identical to say 3rd or 4th decimal place, but even
> considering rounding and precision, shouldn't it be similar? Disclosure-
> I've never tried testing Amber compilations across different libraries.
>
> My very rough grasp of its timeline makes me think 1990-1996? The Amber7
> manual is dated to 2002, so I'd guess it'd have been built on Debaian 1.0
> or maybe even earlier?
>
> Best,
>
> Kenneth
>
> On Wed, Dec 16, 2020 at 12:35 PM Ryan Novosielski <novosirj.rutgers.edu>
> wrote:
>
>> -----BEGIN PGP SIGNED MESSAGE-----
>> Hash: SHA1
>>
>> I think that's probably unlikely. Even in cases where cases where a
>> piece of software compiles fine on either, if it builds against a
>> specific version of a library (whatever.so.5), it's probably not going
>> to work out on the newer one. Not sure when Amber 4 was from,
>> though/what OS would have been current. I don't think I've come across
>> older than 11 myself. Would that then be from 2004? So I guess RHEL
>> 2-3/Debian 3.0 timeframe?
>>
>> On 12/16/20 12:54 AM, Kenneth Huang wrote:
>>> Hi Bill,
>>>
>>> Yeah, that was my next try, since I naively thought sander
>>> shouldn't have changed too much that present day libraries may
>>> work? Though I've had little luck in finding any one with the
>>> source code or much documentation from that far back- I knew it was
>>> a very, very long stretch, given how far back I'm digging.
>>>
>>> Best,
>>>
>>> Kenneth
>>>
>>> On Tue, Dec 15, 2020 at 11:20 PM Bill Ross <ross.cgl.ucsf.edu>
>>> wrote:
>>>
>>>> I'd find a copy of the source and compile it. Probably
>>>> sander/cpu wouldn't be a lot of work to do. A binary from then
>>>> might expect old libs to be in place.
>>>>
>>>> Bill
>>>>
>>>> On 12/15/20 2:37 PM, Kenneth Huang wrote:
>>>>> Hi all,
>>>>>
>>>>> A very strange question- does anyone happen to have, or know
>>>>> anyone who
>>>> has
>>>>> the old binaries for Amber 4 gathering dust on a hard disk?
>>>>>
>>>>> I'm having something of a debate about whether the minimization
>>>>> procedure in what was shipped in Amber 4 (assuming done in
>>>>> sander?) would be comperable to minimization in contemporary
>>>>> Amber (specifically sander in Amber 16,18, 20) builds given the
>>>>> same setup, parameters, etc. Given
>>>> sander
>>>>> is one of the older parts of Amber, my thought is that it
>>>>> shouldn't have changed much?
>>>>>
>>>>> Barring the unlikely chance anyone has still code that old
>>>>> around still, would there be any good way to test to see if the
>>>>> minization changed in a signficant way? My first thought was to
>>>>> find a minimized test structure
>>>> in
>>>>> the source code that might date back from that time, but I
>>>>> haven't had
>>>> much
>>>>> luck in sifting through the test folder so far.
>> - --
>> #BlackLivesMatter
>> ____
>> || \\UTGERS, |----------------------*O*------------------------
>> ||_// the State | Ryan Novosielski - novosirj.rutgers.edu
>> || \\ University | Sr. Technologist - 973/972.0922 ~*~ RBHS Campus
>> || \\ of NJ | Office of Advanced Res. Comp. - MSB C630, Newark
>> `'
>> -----BEGIN PGP SIGNATURE-----
>>
>> iF0EARECAB0WIQST3OUUqPn4dxGCSm6Zv6Bp0RyxvgUCX9pFSAAKCRCZv6Bp0Ryx
>> vjZYAKCV6iHVkW4/N5LcKYCraChPzL1VlwCeP7wYHTlkj2bLcRfgKfY8dMD1Nxk=
>> =xEl6
>> -----END PGP SIGNATURE-----
>>
>> _______________________________________________
>> AMBER mailing list
>> AMBER.ambermd.org
>> http://lists.ambermd.org/mailman/listinfo/amber
>>
>
--
Phobrain.com
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Dec 16 2020 - 15:30:02 PST
