DevonThink Consulting

I believe this happened once in the entire history of Devonthink - and they implemented a fix very rapidly which restored almost everyone’s data.

It’s hard for me to think of any software or hardware with a better track record than that. Even Amazon AWS - at a much higher price point - has suffered unrecoverable data loss. To me, if I put my critical data in a DT3 database and then zip that database and archive it on a NAS drive and somewhere in the cloud, that’s as close to rock-solid data storage as I can possibly imagine and fit for anyone other than maybe a national intelligence service.

8 Likes

and following up on that, on their forum it is uncountable how many times the response to a suggestion or bug report was “will be in the next version” or something like that.

2 Likes

Advantages and Disadvantages of moving files into the database vs just indexing.

OBTW, Thanks for your input.

According to them, it has not been fixed. They have a workaround that relies on your having a backup. They do not know what caused the problem.

From this thread:

We still have not located the root cause of this effect but firmly believe that it is caused by an issue in earlier versions of DEVONthink To Go 2.

  1. Identify the corrupted documents and try to replace them from your backup, best on a Mac or on another iOS device not affected.

And here’s a very long thread discussing the problem.

Due to this reminder of data loss, today I moved all my data out of DT, deleted all databases, and set up a DT_safe_store folder. Into that folder, I have ChronoSync do an hourly one-way copy of data from my working folders. I then created a new database that only indexes the data within those folders. This allows me to continue to use DT’s “See Also” function, without it having access to my working data.

3 Likes

I’m having a hard time understanding why you don’t stop using it given the risk your perceive and report.

1 Like

See Also is helpful.
As I’ve said, it’s fully quarantined now, and I can continue to get some value for the $300 or so that I’ve spent on it.

1 Like

Your quotes are from February 16th (when the problem reared it’s head). Since then there have been ~ eight DTTG updates and I don’t believe the problem is still an issue

3 Likes

Let’s hope.

The second thread I linked above shows it’s still a problem as late as June. There may be others who haven’t discovered their data is corrupt.

If I were them and had a locked post about my software causing data loss, I would certainly update it when the cause was found. But I don’t think the cause was ever found (I’m happy to be updated). To me, that is disconcerting. If they reused code or logic, that same problem could be in DTTG 3 or elsewhere, but at this point I don’t think they know.

2 Likes

I would trust DEVONThink with my life…and, indeed do. :grinning: I have 60+ years of daily diary (journal) entries in one database—and many other databases with other critical information. I have never epxerienced any data loss.

Stephen

7 Likes

Treating the underlying source of files as the determinant of duplicates is not the expected behavior from the majority of our user base, hence the default behavior (noting the behavior is from us listening to our user base, not merely an individual opinion).

However, there is a hidden preference coming in the next release to index the raw source of Markdown files.

PS:

1 Like

Have you read this in DEVONthink’s Help or manual: Help > Documentation > In & Out > Importing & Indexing, especially the Indexing and the filesystem section?

This is a must read, especially if you’re indexing items in cloud-sync locations.

There was one the day before you made this post. DTTG 3 is corrupting files - #270 by OogieM - DEVONthink To Go - DEVONtechnologies Community

More properly, workarounds, as the cause is still unknown.

Finally, the problem I reported with duplicates involved files that had different links with the same display text.

So files that differed like this:

[bookends](bookends://sonnysoftware.com/77131)

and

[bookends](https://www.amazon.com/dp/B08PTWF1DF)

would be treated as duplicates.
If I, for instance, saw those two files listed as duplicates, I might delete one or the other thinking they are actually duplicates, when in fact one was correct and the other not.
I’m surprised the majority of users would want this behavior.

3 Likes

These type of issues are why I stick to a straight-forward “anything bucket” program like EagleFiler. The documents and other files that I entrust to it remain normal files in the Mac operating system, so they can be be edited by the best app of that file type on my Mac, keep their version history, and do not lock me into a proprietary system. The program may not be as fancy as DT3 , or DT2 which I used to use, but it offers good import options, flexible organization, search, and tagging. There is no syncing which some see as a downside, but its simple design makes it is hard to see how I could ever lose anything.

2 Likes

I have been finding myself in similar spots and working to isolate a better solution.

Having been through some of what you are experiencing here are a few discoveries to keep in mind.

Anything from the iOS camera roll with DevinThinkToGo (DTTG) will be slow…it only lets you import one blasted photo at a time…might be as many as 7 taps to net a single photo…deal breaker on my side.

I lost information years ago when a similar program went under and I never had an export. DevinThink will not have that as bad of a problem and I mention because it forced me to come up with my own system that has no dependencies…and it took time and effort.

I think you might be able to find opportunities to accomplish the same needs with less effort that you will need to put in for DevonThink, especially if you are already looking to hire someone to get you set up.

DevonThink isn’t a magical answer to manage data, it is one tools to help.

1 Like

see https://www.devontechnologies.com/blog/20210302-photos-devonthink-ios

1 Like

Update - there is an iOS shortcut that sort of addressed the need - but through links via markup - messy at best.

Company specializes in data and document management and UX and workflow so cumbersome.

Nice idea but do you ever run into problems reindexing after you update the DT working files?

So far, so good.
DT doesn’t know if ChronoSync copied over a modified file, or if it was “me”.
Yesterday I had DT re-OCR a file in that folder, then after verifying its contents, copied back to my working folder.

Not true. To me it happened to more than 500 files. The “fix” does NOT fix it it only now alerts you to when it happens again thus giving you a chance to recover from a recent backup. The real problem probably happened years ago so in my case it was off the end of my backups which went back over 2 years.

The files appear fine until you actually access them. Then even though the title is ok the contents are erased.

Almost everyone getting their data restored is not good enough for a program whose primary function is the storage and retrieval of data. Especially if you are the one who has lost hundreds of items.

As to bug reports. I’ve given up on that. I’m tired of the cavalier attitude towards users on the forums and have just decided to bail. Like @JohnAtl I will try to recoup some of my rather large investment in DEVONThink software by continuing to use the one feature that I find really useful, its web clipper. But then I immediately move those files out of DT into markdown with image links and get them out of the clutches of the DT ecosystem.

One of my biggest complaints about the entire mess is that the tools available to verify and repair a database and to rebuild a database did not give any indications of a potential problem. So I blissfully went along dutifully verifying my databsae and a periodic rebuild as suggested in the documentation only to get blindsided when important items were lost.

I was one of the biggest fans of DT for years, I had been using it extensively for over 11 years. For me to give up on a system I have invested so much time and effort in took a major disaster like this bug. I’m not going back.

1 Like

Not sure how it’s “messy at best”. And bear in mind, the shortcut was made due to discussions with users on our forums and support system. If it doesn’t suit you, that’s okay too.

1 Like