From owner-metabolic-reg@net.bio.net Fri Sep 10 22:29:00 1999 Path: biosci!NCGR.ORG!pjm From: pjm@NCGR.ORG ("Pedro Mendes") Newsgroups: bionet.metabolic-reg Subject: Re: Reasons...... Date: 10 Sep 1999 16:29:28 -0700 Organization: BIOSCI International Newsgroups for Molecular Biology Lines: 56 Sender: daemon@net.bio.net Distribution: world Message-ID: <199909102322.RAA08454@madison.ncgr.org> References: <7rbt3g$dch$1@dyfi.aber.ac.uk> NNTP-Posting-Host: net.bio.net Well, well, We (MMFF group) already discussed most of this binary vs. text story. And to assure everyone in this list, I repeat what Herbert said in his original email: we are also going to define a text format too. So if you can wait a bit longer, you will see it happen! However, I'd like to express my personal opinion that this discussion is rather futile: No one ever edits a text file directly (i.e. the strings of 7 bits inside each byte). One uses a *program* to do so. This program displays things on screen that we then percieve as letters of the alphabet and other characters. It turns out that a text file is as binary as anything else. Now, rather than text, let's think of a pathway. There is no way of representing a pathway in a computer other than by strings of zeros and ones (like anything else). So, to make the parallel with text, what we need is a pathway editor program which (like the text editor) allows one to view and edit pathways in a convenient way. And that is what everyone (me included) ever uses to manipulate pathways with a computer. The file format is totally irrelevant because one always uses a program to read it. The portable metabolic text file format will only be needed by those who prefer using a text editor to manipulate a pathway (using some language). That's all. So you emacs, vi, notepad, pfe, etc. fans should wait a little longer because we will cater for you too. The main objective of PMB files is for programs to communicate with the same file format. So that if I use Gepasi to define a pathway model I can then send it to Herbert and he can read it with Scamp straight away, just like that. We hope that other programs will adopt the format such that users can pass data between them without any trouble. In fact our (NCGR's) future metabolic database, named PathDB, will also produce these files as output. You will be able to query it over the net and get on your computer a PMB file such that you can immediatly load it into Scamp, Gepasi, etc. and play away. Again, what the file format looks like is of no relevance really, since one can use any of these programs to edit it. (By the way, nobody seems to care much if word files are ASCII or binary since one ends up using word to edit them anyway). But to finish on a positive note: those that really like text files should wait a while because we will also produce a text-based format. So be patient... Douglas made a different point in his email that I will reply independently (since it is very important). Pedro Dr. Pedro Mendes http://gepasi.dbs.aber.ac.uk/pedro/prmhome.htm National Center for Genome Resources, 1800-A, Old Pecos Trail, Santa Fe, NM 87505, USA http://www.ncgr.org .