Hook's next cloud experiment-- Google redux. Questions.

[RickAgresta]

@Hook: On my Sharp Zaurus, the software that came with it included import/export options for Contacts and Activities, that allowed you to map where Field 1 of the Zaurus should go, w.r.t. the exported file structure -- and eve allowed renaming, iirc -- it has been an awfully long time.

@r0k: fwiw, thats a very similar approach I've seen taken, during data migrations; to say it worked like a charm would be a touch of an overstatement, but not because of the fundamental principles involved -- it was those pesky carbon-based life-forms that kept getting things wrong

[Hook]

@Hook: On my Sharp Zaurus, the software that came with it included import/export options for Contacts and Activities, that allowed you to map where Field 1 of the Zaurus should go, w.r.t. the exported file structure -- and eve allowed renaming, iirc -- it has been an awfully long time.

Which is actually similar to what Palm Desktop tried to do when importing Outlook CSV. I could export as CSV from Gmail, but that is way too much work. The Nokia and Outlook sync very smoothly via USB using Nokia PC Suite (probably the finest desktop I have ever encountered for a handheld). And, in fact any of the single point to point syncs in this are quite good. Outlook to Google is great, Google to Nokia is great. The other direction for each is also good. It is when thay all operate together the problems arise.

That's why I suspect flag setting being the issue.

[Hook]

I know I said I was done, but...

I guess, although these recent mentions DO present some options, they still give me more than a little heartburn. If MY data is supposed to be a "commodity," then why should I pay a fee just to get it pumped from one of my own devices to another? And why, in the name of all that is decent and good in the world, should I EVER have to start importing / rearranging / exporting blah blah blah to get the same data from one device to another. Then entire concept, in all honesty, seems downright stupid to me.

I don't mean to sound so negative, but with each "new option" presented, I see more and more holes. The time and energy required to manage all of this stuff would offset any benefit I might get by an order of magnitude. Perhaps if, as Rick suggested (simple use,) one's use of contacts was nothing more than a single name, a single phone number, and a single email, some of this might work. Then I look at dozens of contacts I have with dual full addresses, 2 or 3 email addys, half a dozen phone numbers, and some notes. It sounds as though one would be better off trying to sort and align a truckload of cooked spaghetti.

I've got to be missing something here, since the more I read the more horrific the entire situation seems to be.

The problem you are seeing is coming from Palm Desktop - Outlook- Datebk type apps --basically hard drive data management apps-- to the cell phone culture which sees data as out there to be searched rather than managed. I don't know that either way is "right," it's whatever works for you, but boy if you are used to managing your own data there is just nothing out there that is going to pull you into using the cloud as your hard drive unless you really have a purpose for it. You don't. I have some nice-to-have things I'd like from it, but it just can't pull off what I want. People like QF, r0k, Brick and Drillbit have real uses for it and it works pretty well for them to varying degrees. we always struggle with our technology to some degree. We put up with the struggles that get us what we want and have no tolerance for the struggles that have little benefit.

If you are waiting for someone to convince you, you might as well give up. You really have no use for it.

[JRakes]

It's not so much being "convinced," for me, as it is just seeing any possibilities. The more I look, the farther those possibilities fade off into the distance rather than becoming more realistic. Data management via the cloud isn't becoming more of a tool for my uses - It's morphing into a bigger and bigger burden. Given that, and with no real use for it, it's just not worth pursuing.

I suppose, in a way, it's mostly just disappointing (to me) that the so-called "march of technology" isn't adding anything of use and, in fact, seems to be moving away from anything of value to me. When 10+ year old systems and technology seem more effective and useful (albeit not as "pretty,") it makes me wonder why I should even bother trying to keep up.

OTOH, I'm a curious old S0B - I can't help watching, even if it does look somewhat like a train wreck at times.

[Antoine Wright]

Just looked over mail2web and realize I did look over it before. It is just too email-centric (I don't need/want email, have that covered) and push oriented (I want to sync, I don't need push). In general, despite the fact that they do contacts and calendar, that is my impression of most Exchange service sites.

That's because most people want to sync all of those. Again, you can choose to sync everything but email with Mail for Exchange on your mobile. What you sync is controlled at that level.

The problem you are seeing is coming from Palm Desktop - Outlook- Datebk type apps --basically hard drive data management apps-- to the cell phone culture which sees data as out there to be searched rather than managed. I don't know that either way is "right," it's whatever works for you, but boy if you are used to managing your own data there is just nothing out there that is going to pull you into using the cloud as your hard drive unless you really have a purpose for it.

This was funny. I deal with this all day. And mainly because I'm trying to get people out of their silos so that they can do with information what they've been trying to do for years.

If your data (input) is structured (clean), then managing data is easy - that is, you don't need to. All you need to do is query it.

On the other hand, if the data (input) is structured, but not clean for use (meaning that it sits in a system or set of structures that's not malleable for future use) - while being clean for collection, you end up managing the data - or like the the other example, the data is actually managing you because you can't shift towards seeing just data since silo and process are attached to how you manipulate it.

Anytime that you cannot separate the data from the system that it sits in, you have a problem. And this makes it harder to move forward/sideways when that data needs to do more than what it is doing. Palm and Google are trying to make this change, but they are running against categroized, but structurally dirty data (data structures weren't designed for the fluid use that users ascribe themselves to now, see the change that the Social Mobile Framework that the Symbian Foundation is employing for reference).

So then you have two methods of holding data - one big blob that's queried against (webOS), or several smaller blobs that need to be indexed for a query to be useful (Outlook).

Both methods are correct, but both cannot be supported as we go forward with mobile/web/data. As a user, you have to make the chose as to how you will (a) collect, (b) organize, and (c) use that information. If its more important for you that collections are spatially structured, then an Outlook/Palm Desktop approach works. If its more important that you can just find it - and then the query's output is already structured spatially (think Google and the way it treats email and the content within your email), then you go that route.

Systems are moving towards the latter not because there's less control, but because there are more data points and the former method is less efficient for I/O systems and processes. And while our brains adapt, they usually have a hard time making the jump from one to the other.

When you are looking at an older system of organization (categories and groups of Outlook for example), you see this play out when a person moves from Outlook to something *seemingly* less structured. The key is for the person to ask themselves if they really need the level of spatial organization they had with Outlook, or if they need to re-normalize the way they use and interpret that data.

Personally, I made this decision when I moved to Symbian from Palm. My info was structured, but not manageable because of where it was structured. That's no longer an issue, and ironically, I manage more data than I had previously.

[JRakes]

So, a serious question: I have my data in Outlook. It was moved there from Palm Desktop with a simple, clean, and utterly painless process (one time action) when I simply installed Pocket Mirror, then synced my Palm OS device with a "handheld overwrites..." So, if I want to put that same data, intact and equally as useful, into another system, say, a Google-something, what simple, clean, and utterly painless process (one time action) will do it? And to be clear, this is to include everything. All of my PIM data. Totally and completely. No "restructuring" of data or format or display or anything. No exporting / screwing around with / importing. Just do it: Click > Done.

If the technology is advancing, then it should be "smart enough" to know how. If I want to find a piece of contact info for somebody, I simply won't do any more than I do now to get it - A few taps and boom: There it is. In the format I expect it. What am I supposed to restructure in order to do that? I simply want to have / use my existing data as cleanly and simply as I've always done.

[Hook]

AW's point is more or less that you have a choice. You can give up the precision of the structure --this information goes in this slot in this item that belongs to that category-- and go to a system where information is tagged and analyzed by keyword and retrievable more flexibly but less precisely in a manner that can more easily allow surprising relationships in data to emerge.

Think of it this way. Email. Is it easier to look for a particular email in a set of categorized folders or by searching with a couple of keywords. Depends on how much email you have and how good your categories are. I keep very little email but I instantly appreciated the value of search and tags in the gmail model. If I really kept most of my email, this would be invaluable to me.

However, as nice as the search model is, for me, for better or for worse, Contacts are a structured data type. I do need this information in that field in that item as part of this category or group. Maybe if I had hundreds or thousands of contacts, I would think differently. I don't. I don't need hundreds or thousands of contacts. But I need the hundred or so I do have to be kept properly updated as precise information in precise fields change.

The fact is, my data is very easy to manage if various apps would leave it alone. Maybe if there were a desktop app that handled data equally fluidly I might change my mind, but where contacts are concerned, I don't think so. In this thread, and my experiments, that is where the problem keeps coming up. Otherwise, my needs are very minimal. For email I have my own private cloud and calendars everyone seems to handle well.

The social mobile networks has no interest to me. I'm not particularly social and my job doesn't encourage me to be. It doesn't serve any real need I have and, as I've hinted before, if Nokia and Symbian become to steeped in it, I'll move on.

[RickAgresta]

so I could achieve (potentially) similar results by having all my contacts in one big text file, with the presumably minor structure of a unique entry on its own line (word-wrap turned off ), possibly with a header line -- nah, wouldn't really be needed, if the unique lines each contained appropriate tags (aka, m.e.t.a. data) -- the header line could be applied to the *presented* search results, after they were found. in this example, we have Notepad as a db; who'd've thunk it?!

*m.e.t.a. gave me **** on first pass, when I left out the periods...

[Hook]

so I could achieve (potentially) similar results by having all my contacts in one big text file, with the presumably minor structure of a unique entry on its own line (word-wrap turned off ), possibly with a header line -- nah, wouldn't really be needed, if the unique lines each contained appropriate tags (aka, m.e.t.a. data) -- the header line could be applied to the *presented* search results, after they were found. in this example, we have Notepad as a db; who'd've thunk it?!

*m.e.t.a. gave me **** on first pass, when I left out the periods...

Of course, silly. "m-e-t-a" is a dirty word in the search world.

[Antoine Wright]

so I could achieve (potentially) similar results by having all my contacts in one big text file, with the presumably minor structure of a unique entry on its own line (word-wrap turned off ), possibly with a header line -- nah, wouldn't really be needed, if the unique lines each contained appropriate tags (aka, m.e.t.a. data) -- the header line could be applied to the *presented* search results, after they were found. in this example, we have Notepad as a db; who'd've thunk it?!

*m.e.t.a. gave me **** on first pass, when I left out the periods...

Yes, and this is exactly the structure that Palm/Garnet OS was developed after. Its one of the reasons that it was so flexible and fast. Unfortunately, it was way ahead of its time.

Metadata can be written as one word and get around that filter right?

So, a serious question: I have my data in Outlook. It was moved there from Palm Desktop with a simple, clean, and utterly painless process (one time action) when I simply installed Pocket Mirror, then synced my Palm OS device with a "handheld overwrites..." So, if I want to put that same data, intact and equally as useful, into another system, say, a Google-something, what simple, clean, and utterly painless process (one time action) will do it? And to be clear, this is to include everything. All of my PIM data. Totally and completely. No "restructuring" of data or format or display or anything. No exporting / screwing around with / importing. Just do it: Click > Done.

If the technology is advancing, then it should be "smart enough" to know how. If I want to find a piece of contact info for somebody, I simply won't do any more than I do now to get it - A few taps and boom: There it is. In the format I expect it. What am I supposed to restructure in order to do that? I simply want to have / use my existing data as cleanly and simply as I've always done.

I feel you, but the tech is only as smart as the person programming it. A lot of times - not all, just a lot - the idea of thinking towards what is smart is beyond a simple engineering approach. Its more or less taking what's possible from the engineering end and applying some kind of usage to it.

This is what Apple does well as its the culture of how they develop products. Palm is the same. Google is quickly catching onto this as well.

Developing intelligence in either kind of structure is a mixed bag. Because not only do you need to make sure that the initial structures support what it is that you are trying to do now, but you want to make sure that it also can extend far enough out that when you have to redo it later that the pain is a bit less. Unfortunately, not enough people saw data intelligence and analytics moving this fast on the mobile side (its about as fast as things have been moving in the network and server areas); and so we are left with the solution/problem that Hook has started this thread with.

Ideally, every data structure is going to use similar formats, and from there life is easier. Humans are involved in that decision, therefore its much easier said than done.

For us, supporting open (standardized) standards is a key towards helping this along. We are ahead of the curve in respect to this piece; but we would quickly fall behind the mainstream adoption if we don't figure out quickly the best practices here. Yes, you just want to use your data, but you (speaking of this thread's audience) are way ahead of others.