On Wed, Jul 04, 2001 at 02:21:42PM -0400, Chris Armstrong wrote:

On Wed, Jul 04, 2001 at 10:43:42AM -0700, Kevin Turner wrote:

...

Please look at Evolution too. I'm sure you can find something to not like about it, but there's been a lot of sweat put into it, so you should at least be able to learn something from their efforts.

The biggest problem with Evolution (or anything) is that it's not Twisted enough.

Oh, as I said, I have no doubt that there are problems, from biggest to littlest. I was merely suggesting the possibility that they might have got something *right* too -- and if so, that's good information to have when you're setting out to achieve something with many of the same goals. I don't think it hurts to know what the wheel looks like before you reinvent it.

...

Now, forget I said that phase, because it doesn't mean much. ;) I just raise a few more questions: What if the resource you are working on is not an emacs buffer, but an Excel spreadsheet? Can you incorporate resources into this system that you, the Twisted.pim designer, did not anticipate the existance of?

If you want to make a new type of information, make a new subclass of the base InfoData class. [...] The "integration" he's talking of isn't at the UI level, it's at the base. In other words, our integration isn't the same integration that Desktop Environments are going after -- Look&Feel integration. We're going for data that will be integrated.

This is why I told you to forget I said the "d. e." phrase =) I think you underestimate miguel's ambition -- Twisted folks aren't the only ones trying to take over the world. ;) Having a cohesive environment does not just mean that the dialog boxes all look similar and the context-sensitive pop-ups feel the same, you also need to be able to incorporate the product of one application into another -- e.g. your plotter into your spreadsheet, your spreadsheet into your marketing document, your documents into your conferences and presentations. You have not convinced me that this is different from your "data integration". [ http://www.ximian.com/tech/bonobo.php3 ] [ The "why corba sucks" flamewar may now commence. ] [ Feel free to understand a "How little Acap understands things" flame too ;) ]

...

Now, we all know that since PB is being published under the Twisted label, it's got to be the greatest thing since spreadable cheese. But even so, but I'd sleep better at night if I read a memo comparing its suitability for "Integration." to [...] corba [etc...].

(The following is me paraphrasing radix, not quite his original words:) } PB has nothing to do with integration between the PIM components. } That will be done by our /imaginary/ [emphasis mine ;) ] 'associator' } module. PB will just serve as the portocol for [remote] access to the } information. See issue my #2 in my other post from today, <20010705201609.A3233@troglodyte.menefee> Cheers, - Acapnotic (Kevin Turner)

...

...

...

...

"Acapnotic" == Kevin Turner writes:

> This is why I told you to forget I said the "d. e." phrase =) I > think you underestimate miguel's ambition -- Twisted folks > aren't the only ones trying to take over the world. ;) Having a > cohesive environment does not just mean that the dialog boxes > all look similar and the context-sensitive pop-ups feel the > same, you also need to be able to incorporate the product of one > application into another -- e.g. your plotter into your > spreadsheet, your spreadsheet into your marketing document, your > documents into your conferences and presentations. You have not > convinced me that this is different from your "data > integration". > [ http://www.ximian.com/tech/bonobo.php3 ] [ The "why corba > sucks" flamewar may now commence. ] [ Feel free to understand a > "How little Acap understands things" flame too ;) ] Given the constraints of GNOME's environment (i.e. "using C"), I think ORBit/Bonobo dont suck _too_ badly. At some point in the future (after at least one rev each of Twisted and GNOME), I think it'd be worth looking at getting Twisted to handle Bonobo components intelligently. That'd make a _lot_ of client-side integration stuff easier -- and incidentally help along twisted.pim. ;) At this point though, there's not a whole lot we can do with integration with 'desktop tools' (which _is_ what something of this scope should do) -- and I dont think writing our own is all that great of a solution either. So, for now, I think it's worth concentrating on stuff we can write and use _now_ (and an integrate easily into existing twisted modules -- the biggest example i'm thinking of now is a webpage-annotation system a la crit.org, but personal and not written in horrid perl) rather than do-everything frameworks that depend on stuff we dont have yet. -- Allen Short Programmer-Archaeologist shortwa@auburn.edu "optimization is hacker crack. " -- Johan Kullstam, comp.lang.lisp

On Fri, Jul 06, 2001 at 12:36:06PM -0500, carmstro@twistedmatrix.com wrote:

I also think you don't quite understand what we're trying to do here - We're not trying to integrate "everything under the sun",

I'm glad I got that across :) no, I don't quite understand, so I'm trying to get you to define scope (something Nafai has confessed to struggling with). What is "PIM data"? Which things are "personal information" that you're managing, as opposed to "apersonal information"? You have ToDo entries, grouped in lists, some of which have times associated with them and thus get called "appointments". Does that cover it? Then why didn't you say so? ;) -- The moon is waning gibbous, %99.15 full. (15.6 days old) phase: 0.529400 Angular diameter as seen from the earth: 0.4949 degrees.

Greetings, Palm's latest desktop app has a new look: http://www.palm.com/software/desktop/ A Twisted app which could match Palm's level of personal information organization would be good, one that could surpass it would be great. I've used paper-based Franklin Planners, the original Apple Newton, the Palm III, and currently loving the Kyocera Palm/cellphone combo unit (with Internet access, oh yeah). I spend 80% of my time with the Palm in the Memos section, 15% in Addresses, 5% in Calendar, and 0% in Todos. Not that I don't keep todos, it's just that the one built into the Palm suite is too limited. (I keep them in various memo categories.) When I'm sitting in front of a computer, I have the Palm desktop app open. When I get up, I sync and have the Palm with me. The Palm bridges data between computers that I use. Activity logs, ideas, project meta-information, notes, web links are all there, part of the seamless information space formed between the desktop apps and the portable device. From this perspective, the device is far less important than the information space, it can be seen as merely one view into that space. In addition to the basic Palm apps, I use a piece of software called 'Secret!', which has a Palm-hosted app, a desktop-front end, and a conduit to connect the two. It acts as a (simple) memo app, but it encrypts that data for true privacy: http://linkesoft.com/english/secret/ Strong encryption is a requirement for any personal information application. Also, there's some really interesting functionality being developed by the oddly named folks, Llama Graphics: http://www.llamagraphics.com/ Their app is called "Life Balance", which claims to be "the first personal productivity application designed to help Palm users go beyond conventional planning systems to not only manage their time but to achieve balance between the often conflicting demands of career and personal life." They don't have the desktop front-end ready, but will RSN. When they do, it will probably be a tool worth evaluating. There are still good ideas to extract from the Franklin Planner people, especially their diverse set of forms for various specialized purposes. Hmm... Checking the site now, they have a new app called 'One Place' which does some interesting indexing and cross-referencing of your hard drive and documents, behold the intensely cheesy demo: http://www.franklincovey.com/oneplace/demo.html Jason

On Fri, 6 Jul 2001 carmstro@twistedmatrix.com wrote:

On a side note, we really have to stop using the term "associator module" ;) The association code will be in the InfoData class itself, and there will probably be another module that just keeps track of lists of all the different kinds of InfoData. Not sure what this should be called -- anyway, go read TwistedPim, I'm babbling =)

I really have completely stopped understanding what you're talking about here. When you're talking about associations, is it a relational database? That's a fairly heavy degree of complexity, but it's necessary for some of what you're describing. The cool thing about Twisted is that it just uses the "the world is a directed graph" feature of Python to maintain associations. I don't see why you need to be more complicated than that. ______ __ __ _____ _ _ | ____ | \_/ |_____] |_____| |_____| |_____ | | | | @ t w i s t e d m a t r i x . c o m http://twistedmatrix.com/users/glyph