Interesting post here http://al3x.net/2009/01/31/against-everything-buckets.html that I suspect many of the people here won't agree with (I certainly use Opal for as much as I can, but it's not intended to be an everything bucket and I don't try to use it as one).

Interesting and provocative, but ultimately I think the author is wrong. Not entirely of course. I think most of us have come to the conclusion that there isn't a holy grail application, one that handles all our information management needs. But the best of these applications do things that make information gathering and organizing much more efficient than the system described by the author of the article. First, they help us clip and collect information quickly and efficiently. How many steps would it take to capture this paragraph by the method the author suggests? Several -- copy the block of text, open your text editor, create a new document, and save the file after you've navigated the file system to find the proper folder. With Zoot, for instance, I can clip this paragraph and save it in an appropriate Zoot database in less than 10 seconds -- and even more quickly on my MacBook using Yojimbo.

The best of these programs allow you to view your data in different ways. Again, talking about Zoot, I can see my information as a table sorted on various attributes, and I can view individual entries.

There are efficiencies with multipurpose applications as well. With Zoot I can clip an e-mail message, add it to a project database and assign a tickler alarm to it so I do not forget to attend to it. You can't do that using the file system.

I also think his basic premise is wrong. That is, he implies that applications like DevonThink, Yojimbo, etc... promise to handle all your information, and I don't believe that's true. Just today I was reading the forum for Curio, and the developer encouraged people to use OmniOutliner for long lists and outlines. In fact, I rarely see any developer claim that you can put everything in their application without need of other programs.

Steve Z.

Many of his arguments are off the mark.

I do agree that proprietary database "everything buckets" are probably not a good idea -- we've seen enough PIMs die that this argument makes sense. However, the trend among PIMs is clearly to use the filesystem rather than a proprietary database. Together, EagleFiler, DevonThink 2, etc. all use the underlying filesystem so you're not locked in.

The other big problem with his argument is that you could do everything just using the filesystem, which is true (e.g. you could do DevonThink-style "tagging" using Unix symlinks), but so what? The user interface is what matters, and symlinking as a form of tagging just takes too many clicks to bother dealing with. These programs add value through their interfaces. Those with artificial-intelligence algorithms for classification, etc. also add value in other ways.

That said, I do agree that using the filesystem together with filesystem metadata is a valid alternative for information management for many people and is worth considering.

-- Chris

Chris Thompson wrote:

...I do agree that using
the filesystem together with filesystem metadata is a valid alternative for
information management for many people and is worth considering.

-- Chris

On the Mac there is a number of applications that specialise in tagging within the filesystem, based on the Open Meta standard, with few other bells and whistles. One of the most recent but simplest is called - appropriately - Tags: http://gravityapps.com/tags/overview/

This site, for those who don't know it, is a very good way of keeping up with Mac tagging developments, by the way: http://tagamac.com/

Here's a freeware file tagging program for Windows, currently in beta:
http://lunarfrog.com/

I haven't tried it or tagging in general yet, but am warming up to the concept.

Alexander Deliyannis wrote:

Here's a freeware file tagging program for Windows, currently in
beta:
http://lunarfrog.com/

I haven't tried it or tagging in general yet, but am
warming up to the concept.

Very Mac-looking!

I can see the value of tagging, but I can also see its major pitfalls, especially for someone like me. It seems that to be truly effective using tags, you need to be extremely disciplined in their use -- not only making sure you tag EVERYTHING, but also making sure you are consistent in your tagging. But, see, the problem is, if I were that disciplined, I would have settled on one PIM long ago, learned everything about it and used it relentlessly. Or, to put this another way, if I were disciplined, I wouldn't have such a bad case of CRIMP.

Don't get me wrong, I use tags where possible in Mac applications. They are handy. But there is no way that I can see myself tagging every file I download or save.

Steve Z.

but the tags does not seem to be stored in the files ... are they?

Alexander Deliyannis wrote:

Here's a freeware file tagging program for Windows, currently in
beta:
http://lunarfrog.com/

I haven't tried it or tagging in general yet, but am
warming up to the concept.

The tags are stored in a "Tags.wft" file in the program's directory. Not really a clean solution.

-- Chris

quant wrote:

but the tags does not seem to be stored in the files ... are they?

Alexander
Deliyannis wrote:
>Here's a freeware file tagging program for Windows, currently
in
>beta:
>http://lunarfrog.com/
>
>I haven't tried it or tagging in general
yet, but am
>warming up to the concept.
>

The use of the file structure in Windows to simulate a PIM has been discussed before. See, for example, http://www.outlinersoftware.com/topics/viewt/880 . In that topic, I suggested that tags be entered into the "properties/custom" or "properties/summary" section of the file and then these tags could be searched for using any of a number of standard search engines. However, if I recall correctly, I later found some file types where this might not be possible. However, it's worth another look.

-c

Cassius wrote:

In that
topic, I suggested that tags be entered into the "properties/custom" or
"properties/summary" section of the file and then these tags could be searched for
using any of a number of standard search engines.

I tried this once but couldn't find a way to search for a specific tag using the MS search engine. For instance, "keyword: outliner" didn't work and I couldn't find anything on the net that indicated that there was a way to do it. Its been a while since I tried so perhaps Vista provided a solution for this problem?

Tom S.

Tom S. wrote:

Cassius wrote: In that topic, I suggested that tags be entered into the "properties/custom" or
>"properties/summary" section of the file and then these tags could be searched for
>using any of a number of standard search engines.

I tried this once but couldn't find a way to search for a specific tag using the MS search
engine. For instance, "keyword: outliner" didn't work and I couldn't find anything
on the net that indicated that there was a way to do it. Its been a while since I tried so
perhaps Vista provided a solution for this problem?

Tom S.

-------------
TOM,

I apparently was wrong about "properties/custom." However, try entering a tag in "Properties/Summary/keywords." This seems to work in Win XP SP2. One problem, .htm files do not have a "Properties/Summary" tab, but curiously, .mht files do!

One can save plain .htm files as .mht files, but it seems that the .mht version may be 10 times larger in kb.

-c

The MHT file embeds the images, javascript and css in the file. That is why it can be larger. If you save your web page in HTM with all referenced files, the two will be comparable in size

Tom S. wrote:

Its been a while since I tried so perhaps Vista provided a solution for this problem?

Reference:
Tag files and save searches in Windows Vista
http://lifehacker.com/232891/tag-files-and-save-searches-in-windows-vista

http://lifehacker.com/232891/tag-files-and-save-searches-in-windows-vista

Beautiful. Thanks.

Tom S.

Pierre Paul Landry wrote:

The MHT file embeds the images, javascript and css in the file. That is why it can be
larger. If you save your web page in HTM with all referenced files, the two will be
comparable in size.

I don't think so. I saved a file in .htm format. It's properties are:
Size: 26 kb
Size on disk: 28 kb.

I then opened this saved .htm file and re-saved it as an .mht file. The .mht file properties are:
Size: 286 kb
Size on disk: 288 kb.

-c

Cassius, were you online when you opened and re-saved the file? If so, it's quite possible that IE downloaded extra info from the web.

I just tried converting a .html file to .mht while offline. There was a difference, but not significant (8.44 Kb to 10.6 Kb) and this could be explained by the presence in the .mht file of empty 'containers' for the various outside files.

Alexander

More views on this topic, related to Tinderbox and DevonThink:

http://www.markbernstein.org/Feb09/EverythingBuckets.html

http://www.devon-technologies.com/scripts/userforum/viewtopic.php?f=7&t=7452

I'm sympathetic to the idea that the development of structure can sometimes only be incremental, or even emergent. (But of course that's not in itself an argument for tagging as opposed to a pre-defined hierarchy; it is an argument for an inbox...)

Alexander Deliyannis wrote: Cassius, were you online when you opened and re-saved the file? If so, it's quite possible that IE downloaded extra info from the web.

I just tried converting a .html file to .mht while offline. There was a difference, but not significant (8.44 Kb to 10.6 Kb) and this could be explained by the presence in the .mht file of empty
'containers' for the various outside files.

Alexander

============================================
Alexander, you appear to be correct. When I turned off my wireless connection, opened a saved .htm file, and then tried to save it in .mht format, one of two things happened:
1. Like you, my .mht file was not much larger than the .htm file, or
2) The .htm file would not save. I got a message that the file (or something) could not be found. Perhaps some .htm files retain some links/hooks to what Pierre referred to and cannot be saved without accessing them from the net.

My apologies to Pierre and thanks for proving me wrong.

-c

Cassius wrote:

My apologies to Pierre and thanks for proving me wrong.

No harm done. An MHT file is essentially the same as an EML file (i.e. an email message) without the email header (to, from, etc). It contains the HTML, and a base64 encoded section for each dependant content (image, js, css, embedded object). If you look at the HTML page source, you'll see near the top or embedded in the text, references to these. I know MHT quite well, as InfoQube can grab content in HTML and in MHT formats.

MHT and RTF can both be used as standalone documents. RTF is widely used of course, mostly for word processors, MHT/EML has traditionally been limited to emails and web clipping. They are very different internally and conversion between the two has always been problematic. IMO, the big big plus of MHT, is that it is embeddable. So you can easily combine multiple documents into a single one, something impossible with RTF (AFAIK). When editing MHT documents in MS Word, just about every feature is supported, excepts versions.