On Mon, Jan 23, 2012 at 3:00 AM, Eric S Fraga <e.fr...@ucl.ac.uk> wrote: > Max Mikhanosha <m...@openchat.com> writes: > >> At Fri, 20 Jan 2012 12:04:51 -0600, >> John Hendy wrote: > > [...] > >> Generally I think the way to tackle this is to take advantage that you >> are working with plain text and not with Word document, and use >> standard Emacs/Unix tools for working with text. > > Agreed! >
Thanks to both of you for input! >> Some ideas: >> >> Before updating each project, cut-n-paste it into the new >> revision.. Org mode makes it easy to cut-n-paste trees, for myself >> duplicating a headline is simply pressing Y then P over a folded >> headline (viper/vimpulse user) > > Is it not easier to simply make use of any of the revision control > systems (git, mercurial, svn, even RCS) that are out there? You can > easily tag a particular revision based on milestones and then see diffs > between the current content and any previous version. > > In terms of the original questions, I use a combination of hierarchical > structure that is filled in as a project develops, with > revision control to allow me to see progress, together with a log based > recording of activities (e.g. meetings, deliverables delivered, issues > raised). That is, I mix both of the approaches mentioned by John in his > initial email. > This is intriguing. I don't suppose you have a sample file of sorts? Specifically, I'm interested in how you mix 'n match hierarchical/topical vs. time-based organization. I really struggle with this and my purely time based stuff is *definitely* not the way to go in my opinion, as I have tons of related things in separate trees with time stamps, which makes finding some little tidbit silly since I'm looking through different dates for it. Also, I'm a super git newb. The furthest I've gotten to is setting up a repo for sharing stuff between work and home, for example, and simply doing `git pull` from each location when I want to work on something and then `git push` when I'm done. I'm assuming I could set up some sort of cron job to `git commit; git push` each day/week or something? And then learn more git commands to show progress? Is there a way to spit git timeline based output into separate file revisions? Sorry if this is getting too off-track from org. I should probably look into the power of git on my own... > The key, as John has already stated, is to record everything! With > emacs, I can usually pull out what I want *if* the information was > recorded in the first place. > > Finally, tags can be very useful for quick searching as well. I still need to figure out a better tag system as well. Currently, I just have a project acronym under each main header: ----- * Tracking Odds and ends tasks * Project 1 :proj1: Stuff for project 1 * Project 2 :proj2: Stuff for project 2 * References :ref: Misc things I need to refer back to now and then (project reporting IDs, etc.) ----- That's quasi helpful, but not really... Do you tag by type of data stored? Project/task type/name? Thanks again, John > > -- > : Eric S Fraga (GnuPG: 0xC89193D8FFFCF67D) in Emacs 24.0.90.1 > : using Org-mode version 7.8.03 (release_7.8.03.192.g32af) >
