#1
|
|||
|
|||
Large Database Strategies
Like so many on this forum, I have tried so many different PIMs. I really like the structure, capacity and power of Ultra Recall, but I do have a concern for the long term. It is not a limitation of UR, but more of my own limitation. I would like to hear how the various users are planning for the long term with thousands, tens of thousands of pieces of information. How many levels of parent-children? How many items under a single branch before it should be broken down? How to keep the interface uncluttered as much as possible? There are many, many more question than this of course. I am looking for practical strategies.
|
#2
|
|||
|
|||
One of the things I use UR for is to keep a daily log. I create a text note, and give it a name like this: YYMMDD. This differs from the typical MM/DD/YY format for dates, and has the advantage of allowing you to keep things sorted sensibly. I then use this kind of hierarchy for the journal:
Journal -> Year -> Month -> Week -> File At the end of every week, I review the week's journals, make note of any open issues, ideas or other items that I want to continue to keep on my radar screen, and make sure the information is reflected in any other appropriate places (via links or whatever). Not earth shattering or anything, but certainly conducive to large collections of these things. |
#3
|
|||
|
|||
This looks like a good thread to revive I'm very interested in how other people organise their stuff, especially how you combine 'what to do' with 'how to do it'.
I got myself into a lovely mess with the tree over the past few years: lots of information tucked away I'd entirely forgotten was there. After spending a couple of hours moving things to the right places, and then moving them again when I thought of another level of organisation... I realised I need a different system. Here's what I'm planning to do - 1) Use user-defined keywords (ctrl+k) to 'tag' everything related to (eg) product development. (You can select multiple items and allocate the same keyword to them all at once - useful if you already have them more-or-less organised in folders.) 2) Create a search on user-defined keywords only (that's a checkbox at the bottom of the search page) for 'product development' 3) Lock this search and copy it to the overall 'projects' folder. Now every time I add a new note, journal entry, idea, task, web page etc related to product development I remember to 'keyword' it as such. And then no matter where I dump it in the tree, it'll show up in the saved search folder in the right place. I'll do the same, of course, for product development > membership site and product development > membership site > Amember, and build a new 'tree' of searches. So I can still have a visual view of the hierarchy, which I like, but I don't have to worry about losing things because I forgot to store them in the right folder. Instead I just have to worry about remembering to tag things... Of course saved searches can do more than just replace or supplement a static folder structure. I can just as well save one for 'all membership site tasks due in the next two weeks' or 'all I Ching readings about membership site development'. Oh, Kinook...? This leads to a feature suggestion/ question. Would it work for search results to appear not just in the search results pane, but also in the tree as virtual child items of the search? It'd make it easier to browse their contents in a full-size window without losing site of all the results. And it'd turn saved searches into 'smart folders'. |
#4
|
|||
|
|||
Don't let your database become unnessarily large.
Instead of Storing external files inside the UR database, create a sub-folder structure below the UR database location, then Link (not Store, Copy or Move) to the external file. The file path (UR URL Attribute) will be relative when importing from child folders of the UR database location, Links outside the UR data path will be absolute (full drive path). If you need to transport, distribute, or backup your database this folder stratgey makes things a whole lot easier. Granted, it does nothing to shrink the amount (filesize) data you must deal with (UR database + Linked files); I'm just a firm believer of not putting something (phyiscally) into a database unless you really have to. |
#5
|
||||
|
||||
Quote:
|
#6
|
||||
|
||||
Quote:
File storage systems associated with modern operating systems tend to be very robust. I'm basically restricted to MS for most of my work so i use NTFS which at an age of over 10 years and version 3.1 does IMHO a v.good job of file storage. Please follow ashwken's suggestion regards having a data storage sub-folder relative to your UR database. Quote:
Enhance searching ... really really well done in UR already and a little tweaking like Open in New Tab or Open in New Window - yum yum. (Would you like to think of a few searching enhancements then start a new thread in the "Suggestions" forum?) |
#7
|
|||
|
|||
I can't think of any search enhancements! Apart from the option to have results appear as subfolders of the search. Does that merit a whole new thread?
If/when you do post your organisational notes, could you add a link here to the new thread, so I don't miss it? Thanks. |
#8
|
||||
|
||||
Quote:
More details on your suggestion please, (although I just remembered we're in the wrong forum)! What if it were only the Search Pane that had this special behaviour? So, in the Search Pane you run a search and the results appear in a tree structure below the search item. The Child Items Pane at this point is empty. When you click on a "result" item, the Child Items Pane then becomes populated appropriately. Quote:
|
#9
|
|||
|
|||
Actually I did follow your suggestion and start a new thread. Let's be extravagant with those electrons. Here it is:
http://www.kinook.com/Forum/showthre...&threadid=2731 |
|
|