Discussion: Your documents, up in the clouds?

There’s a lot of discussion about ‘cloud computing’ and storing your data ‘in the clouds’. Some new applications are based on this paradigm and are either built ‘in the cloud’ or use a servers cluster for synchronizing multiple clients and for giving web access to your data. While we clearly see the necessity for synchronizing a database between two or more Macs and maybe iPhones, we’re still unsure about what implications full web access to your documents has. My top issues are:

  • data integrity
  • and data safety.

The first one, data integrity, is a matter of programming but bug could lead to documents disappearing and being lost forever. There is no bug-free software, and one example is MobileMe where server-side problems together with client-server synchronization had lead to complete email archives being lost forever if there was no backup. But, as said, it’s possible to get this under control.

Data safety is way more critical. Every server that is connected to the Internet can be broken. There is no system that is 100 percent safe. And if it can be broken, your data can be compromised. Even if one stores all data encrypted at least the web interface that you use to access it needs to know the decryption algorithms and keys, e.g. for convenient re-login.

I’d like to know if you’d use a web service that would sync with your databases and host them, give access to them, etc. including all the named security risks? Or would you rather go for a sync service that, e.g., uses the MobileMe iDisk as the hub? What is your personal attitude towards ‘cloud computing’? And how large are the databases that you would send across the public network?

Side note: Our license code server which holds all your details including your name, address, phone number, and email address works in ‘poll mode’ and is not directly accessible from the Internet.

23 Responses to “Discussion: Your documents, up in the clouds?”

  1. I’d use a web-based model if it were for simple, expendable stuff (like I put in Yojimbo) but for mission-critical, must-have-access=anytime data (the sort I put in DevonThink Pro Office) well, no, the cloud wouldn’t work for me.

    I would like syncing between multiple macs (which I now do by exporting my database and putting it on the 2nd machine via USB key) but my data zips up at around 400MB so I’m still able to do it that way.

  2. Kenb says:

    I wuld definitely like the option and would use the cloud model to provide access to my most important information anywhere. To me data confidentiality would be the issue, because I think there are adequate solutions to the integrity issue. To overcome this I would certainly be selective about what information I ut in the cloud. I would really love to be able to seemlessly access a DT database, easily and simply (with a native app please!) from an iPhone for example. I would guess that all kinds of professionals (journalists, law, medical etc) who need easy and quick access to information would find such a feature invaluable.

  3. Tim Murray says:

    I’d like to see both options. First I’d like to see MobileMe sync with access from the iPhone. Omni is doing this with OmniFocus.

  4. Like anything else, there’s trade-offs. For some data such as calendars that need to be synced with others, or files accessible from anywhere, the convenience is worth it.

    OTOH, I really am amazed at how many folks are thinking of using Google, etc., to host documents such as spreadsheets, etc., that I would assume include all sorts of private financial and business information.

    I can’t imagine putting all that stuff up in the cloud. Maybe I’m just more skeptical than most, but I think a responsible company would have a lot of concerns about taking the liability for privacy, integrity, etc. That’s more trust than reasonable to ask, IMHO.

    Of course, we’re currently being asked here in the U.S. to just write a $700 billion (or more) blank check to Wall Street (and trust them, no questions asked), so maybe I’m just too old-fashioned about this stuff!

  5. Justin says:

    I’d be interested in both options as well.

    I’m increasingly appreciating cloud computing. (I don’t deal with highly confidential information, however.) I just appreciate the convenient access and consistent backup aspect of it.

    My DT DB (with files) is around 500 MB–though I wouldn’t expect uploading that everyday–just syncing the changed parts as needed.

  6. Patrick says:

    Having your data available through a cloud service could sometimes and for some information be quite convenient but I don’t think I would use it for my DT database.

    My main database currently is around 2GB in file size and there are all kinds of information that are quite confidential to me.

    Another reason I wouldn’t use it is that I’m using a notebook as my main PC which I almost have with me all the time so there’s no need for accessing my data online.

  7. As far as DT goes, I’m very interested in having it sync to a mobile device. I guess you could either do that directly or sync to the cloud and access it that way.

    It would seem worth adding a parameter to every DT file, page, etc. (not sure what you call it) that could be marked private so it wouldn’t be put in the cloud.

    It might get hard to load most everything in a big DT DB directly on a mobile device, but maybe not, with memory continuing to increase.

  8. Cameron says:

    I currently sync my DTPO database to MobileMe. It is larger than 6 GB. I do that so that I always have an online backup of my office data in case of computer and local backup failure. I would like to see an improvement in the syncing, however, It seems as though that the entire DB has to be re-uploaded every time a change is made. This takes a very long time and often fails. If it could be changed so that only the changes to the DB are uploaded, that would be much better.

  9. Tricia says:

    I am very much liking Evernote and its ability to sync across platform, but since I don’t keep confidential information in Evernote, that part isn’t an issue.

    The integration of iPhone, Mac, and PC is amazingly wonderful. It almost makes the framework invisible, and I can see the danger of someone not even considering where the information is stored, since it’s always right *there*.

    My research in DTPO is also more important to me than anyone else, so I would be happy with a solid sync solution, with a handy iPhone app on the side. I use iDisk to sync quite a few things, so one more (if reliable) would be fine with me.

    That said, I would also be happier with a secure solution than with an insecure one, because I can see me using DTPO for more confidential info once the multiple databases feature comes out.

  10. psj says:

    There are very successful examples of “cloud” services, those can provide interesting insights.

    For example, think about IMAP. Would you hand all your personal correspondence to a complete stranger? Yes, you do!

    Is there a possibility of data loss? Sure, it is, but usually mail clients are smart enough to keep your mail buffered locally, and mail server admins are usually doing a fabulous job at keeping the backups intact.

    So, back to Eric’s other concern: privacy (or call it integrity or security or whatever). That, too, can be addressed in IMAP. Simply add PGP or S/Mime encryption on top! Many mails that I send through Gmail are encrypted, and thus useless to google. Of course this means that I cannot read them on the go, but isn’t that the goal, after all: only being able to access them in a safe and sound environment?

    To sum this up:

    a) The “cloud” is a term for transparent server application scaling, not for data management.
    b) “cloud” solutions have been around for a long time.
    c) Collaboration needs compatibility. Compatibility needs interoperability. What we need are standards (open or not) for syncing and sharing data.

    I personally consider a solution to be seriously usable once there is something that I can get both as a hosting service and as software to install on my own server. (WebDAV is in such a situation). Then I will add transparent encryption to make sure that my data is only decrypted on my local machine, and voila.

    PS: Something like this might be viable with rsync/unison onto an encrypted DMG image on a SSHFS on MacFuse, but yes, that would probably be unbearably slow.

  11. Panxatony says:

    I am using Dropbox to synchronize one of my Devonthink databases between my macs. but only for my “knowledge base” database.

    But maybe someday I will store my important data on an encrypted sparseimage and sychronize the image with dropbox. Dropbox is using Amazon S3, for this reason it should be safe, but who knows.


  12. CO2 says:

    Though MobileMe is an easy convenience, it has had a lot of growing pains. I’m somewhat hesitant to completely trust one sync solution, but would prefer perhaps a choice of MobileMe and some other platform.

    Personally I am currently using Amazon’s S3, it’s about a close to “the cloud” as one could imagine since it is suppose to be distributed and have multiple redundancies.

    I’ve brought this up before in the forums, meaning the idea of using Amazon S3 as a means of backup for our databases. So I’m going to be favorable to having this as an option.

  13. darwin says:

    I think I don´t need that, simply because most of the time I´m working at home. There is where I need my data for an article. I don´t need my DB outside.

  14. Petra says:

    I wouldn’t take my money and stuff it into a mattress tossed in some woods in another state, thinking that was a safe way to protect it. What I know about that mattress is the same as what I know of places in the network clouds. SLAs, privacy policies, and the like, are ephemeral. On the other hand, if there was utility in it for me, I can imagine taking a picture of my money and putting that in the old mattress in the woods. (This metaphor isn’t holding up, but you get the point.) Same with my data. I’ll put a copy of some of the less critical stuff into the cloud if I need to access the data from different places or in different ways. In any case, local storage is cheap, and I’ll keep several backups in my office, at home, and in a vault.

  15. talazem says:

    I have recently started using SugarSync between my Mac and iPhone. And I must say — I am hooked. I can see any document, and listen to any MP3, from my iPhone that has been uploaded from my laptop or home computer. Recently, while at a conference, someone called me asking for a document. I was nowhere near my computers. Using SugarSync on the iPhone, I sent him the document right then and there. I bought a license for SugarSync the next day. Now, I have universal access to my stuff, as well as everything backed up, with 5 levels of versions for each file if I choose.

    That all being said, I personally don’t use DTPO the same way — at least not in its current version of 1.x. I primarily use DTPO for its “see also”, and OCR features. I have tried to use it as a server; I set it up on my home computer, but the UI was too limiting; on my iPhone (or other computer’s web browser) I want to be able not only to SEARCH my files, but also to browse them.

    It would be interesting to see what 2.0 brings to the table, in this regards.

    In sum, I am convinced by the utility (though I am wary of the security, even as I use it) of cloud computing, especially in the way SugarSync has done it. It would be interesting to see what Devon-tech brings to the table, especially in light of the AI features.

  16. spk says:

    I have my DT Pro Office database living in my DropBox account. I access it from home, the office and on the road; I can add and delete – no problems.

    Although this may all change when I see DT 2.0!

  17. TomasF says:

    i think web-accessible versions of DT databases are the way to go. long-term a better online system is a MUST. the elephant in the room here is evernote which is clearly taking off and becoming very popular. and definitely being hyped up in the blogoshpere. for me, i’m not about to replace DT with evernote, but the iphone integration and accessible anywhere approach are really really tempting. saves me lugging my laptop everywhere. i’ve flirted with it but it just doesn’t have enough features to tempt me yet. in a year or two though, i think it could well be a different story…

  18. David Bosman says:

    I store my DT database (~900Megs) in my Dropbox folder and it works fine (it also works with SugarSync, btw). Only changes seems to be synchronised… But I only use DT for 2 or 3 weeks.

    I like the simplicity of Dropbox. It just works & it is fast. I own MobileMe but find it way too slow (at least, here in France) to do heavy syncing.

    I do not worry too much about safety or privacy of the files I do synchronise online:
    1/ I have backups — years of digital paranoia 😉
    2/ Those files can be read by anybody and, to some extend, they will be: it is mainly articles that will be published.

    For more confidential files, there is no online syncing at all.
    I’ld gladly pay for _real_ online security/privacy, but is this possible through the Internet (without even considering local legislations that may question the privacy of data stored in the servers of such or such company) ?

  19. Felix Martinez says:

    I certainly believe that we are all well aware of the dangers involved with placing information in the cloud. I believe that those with extremely sensitive information will and should avoid the cloud at all cost. So to each his own.

    Having a cloud options is of outmost importances, since most information is not of that sensitive nature. It is not like there wont be safe guards anyways.

    Being able to synchronize general databases, that have general information is very important. I cannot speak for all, but I have a work iMac, a Macbook Pro, and an iMac at home. It is terribly difficult to keep my information synchronized! And because of the reliability that I have with Devon-Think, it is one of those missing links!

    As far as iDisk or another other I say, as long as one of them works properly…

  20. macpug says:

    Count me amongst those who keep DT databases sync’d between computers with DropBox. It has worked nicely keeping my iMac and MacBook in harmony. I stopped using iDisk a long time ago…it was just painfully slow.

    I’d forgotten hearing about SugarSync…looks very similar to DropBox with the extremely appealing addition of being able to also keep iPhone files in sync. I’ll be checking into it after I post this message. I’m still using the freebie DropBox account at the moment, but I stay at about 99% capacity and have been needing to pay to upgrade for the past couple of months. Looks like SugarSync might be a little better since it costs about the same and throws in syncing the iPhone as a bonus 🙂

    I must say I’ve been experimenting with using Evernote –which I LOVE– to keep files on all 3 devices in sync. I would much prefer one integrated solution of keeping files on multiple computers along with the iPhone and/or iPod Touch securely in sync, as with Evernote. I do like the option of being able to publish a notebook in Evernote so colleagues can have access to it as well. Very nice feature when there’s collaboration on a project. I’ve already been able to take a business trip without my laptop for the first time ever because of Evernote. I had everything I needed right there on my iPhone. I agree with TomasF that I won’t be abandoning DTPO for Evernote anytime soon, but there are times where I’ve traded off and used Evernote instead lately and it has been really nice. I’m really looking forward to seeing what v2.0 brings to the table. But you better hurry, because Evernote is really sucking me in.

  21. cgrunenberg says:

    I’m with Richard & Larry:


    Sounds like the next hype…

  22. Karen says:

    Since I’m not collaborating with anyone, I have no real need for cloud computing. I understand that an off-site server that is professionally maintained is less likely to lose data than my extra harddrive stored offsite, but I just can’t get excited about paying monthly fees to store my data. One of my databases is over 2GB and I don’t think I could find a service at my price point, i.e. free. There’s also the issue about what would happen to my data if I missed a payment. I do use Evernote, but I am looking to mostly dump it when 2.0 rolls around. I guess I might keep it for a few things I might need to access away from my desktop. Export from Evernote is too difficult and I don’t want to tie up anything important in it if I’m not sure I can get it out.

  23. Jeffrey Long says:

    My data management is in two stages. The first is evernote. Evernote is where all my current projects and todos go. The second stage is devonthink. What is beautiful about Evernote is that I can access my data anywhere. In my work environment, I do not keep my personal macbook. So my Devonthink database is not available to me. I would _love_ for devonthink to follow this model. For my database to live in the cloud and be syncronized with a desktop client. The paid version of evernote includes ssl encryption which I’ve considered ponying up the $5 a month for. Devonthink could do the same thing to keep my data secure.