I’d like to talk about a subject that has become quite important to me and gather your opinions on it.

Last year, it was announced that iThoughts was being terminated and abandoned by its developer. https://www.outlinersoftware.com/topics/viewt/10522/0/ithoughts-discontinued
There was no explanation apart from the fact that the company was closing.

This came as a bit of a shock to me then. It was an app that I relied a lot on, I spent years making dozens of elaborate maps about projects, plans, guides , etc.
I considered it to be by far the best mindmap software in the Apple environment, having almost perfectly identical versions on Mac/iPad/iPhone, and it seemed to be quite successful commercially as well. It also had a Windows version that was released more recently.

The idea that it could just disappear like that , one morning, without any warnings never even crossed my mind.
I found myself scrambling to find an alternative, testing pretty much every mindmap available on the market. The OMPL export from iThoughts never got properly imported into any other app. I settled on MindNode which did the best job at importing the iThoughts maps, but I still had to spend days reconstructing the maps properly, copying by hand some of the stuff that just didn’t get imported , relinking nodes that didn’t relink properly , etc. It was an infuriating process.

That experience really did make me think about what might happen to the massive amount of data I have in various applications , especially Curio, Tinderbox, Devonthink and Obsidian.
I hope all these developers have a long productive life, but we have to be prepared to the fact that one day we might just wake up and find that the developer of an app we deeply rely on has decided to become a reclusive monk living alone on top of a mountain with a goat and a plant of tomatoes. Or any other scenario that just leads to the company and the application being completely abandoned.

The best-case scenario would be that it would be sold to another developer, or the developer(s) have a plan that if they abandon the app, it would be released as open-source so that other volunteer developers would keep it running for at least enough time for people to figure out how to get their data into another software.
But I’m afraid it’s an unlikely scenario. It will probably just be abandoned , then at one point it wouldn’t even be downloadable.

I have no idea how I would get some of the massive documents I built over years ( decades even in the case of Tinderbox/Curio/Devonthink ) into other compatible apps. It’s just not going to happen. Maybe the Curio docs could be exported as limited PDFs then reconstructed ( over weeks ? months?) into something like OneNote . I also have no idea how I can rebuild Tinderbox maps anywhere else…

Only Obsidian seems enough bullet-proof in case of extinction. Any of the apps ( and a lot of others ) I mentioned could become extinct at any moment. This made me really reconsider the use of these types of apps. Their idiosyncrasies are exactly what make them unique and powerful, yet they are also their Achilles heel, because the way the data is presented and structured makes it nearly impossible to be reconstructed somewhere else, no matter what export format they use ( HTML, OPML, JSON, etc.).

So what to do ? Use only apps from big companies like Apple or Microsoft , that are unlikely to disappear in the short/medium term ? That would be a shame, as the most creative apps come from small indie devs. Or use only apps that are built as much as possible on standards like Markdown ( Obsidian ). Or … just live dangerously…

What are your thoughts on this ?

Yes, this is a constant and recurring theme for me when I consider my workflows and data security. I've experienced a lot of hassle over the years, with programs such as Ariadne, ConnectedText, WhizFolders and DocxManager being abandoned or just suddenly no longer working. And, like you, I have never been able to smoothly import the data into a new application even when the format is fairly standard, such as docx or html. There has always been a need to tweak. It took me a year to get my largest project into ConnectedText and then recently 2 months to get a slim-lined version of that project into RightNote. But again, as you say, often the best software for my needs is from small independent developers, so the risks are high.

I try to pick up on signals that a developer is moving towards abandonment - bugs not being investigated, forum posts being ignored and so on - and act accordingly. This is by no means fool proof! I avoid programs that work only online or depend on somebody else's servers to operate. For one project I duplicate all work in two programs as the chances of both going under at the same time is small (and also, as one is Microsoft Word, the chances of it outliving me are high). For other data I have a system of exporting to standard formats at regular intervals so that the amount of data I would lose if it stopped working would always be limited. That's a lot of extra work, though, as my data is scattered around a lot of different programs. I try to reduce that number to simplify my life, but that also increases the impact if any of the programs used get abandoned. Sometimes I choose software from a vendor which I know is unlikely to stop supporting it, such as OneNote.

I would also be interested to know how others tackle this perennial issue.

[Note: There are no answers in what follows; just musings on the topic of data portability.]

Not telling you anything you don't know, but information comes in a variety of formats. When I bought my first PC over 40 years ago, the first software I purchased was a flat file database. I transcribed the list of books I'd read, which I had recorded in a composition notebook. I've been adding to the list for 40 years now. It has followed me to at least a dozen computers and in as many different applications, always exporting to a CSV file to be imported into the new application. Knock on wood, I've never lost any data from this list. Today it resides in Airtable.

I know there is nothing instructive about this example, as a CSV file is about as basic as it gets. I think the reality is that if you want to take advantage of some of the great features in the more artesian applications, you're going to be vulnerable to losing your data someday. But then, having an established company involved is no guarantee. I remember when Symantec bought out my favorite app of all time, Grandview, then promptly killed it. The same is true for when Netmanage took over EccoPro. The truth is that anytime an app is handed from the primary developer to another, there's a good chance it is already doomed. Not always, but usually. Even Obsidian has its vulnerabilities, especially if you rely on any third-party plug-ins, which are as vulnerable as any of these specialty apps -- or so it seems to me.

The good news is that some of these artesian applications do hang on for a long time. Tinderbox has been going strong for over 25 years. Curio for almost 20. Same for Scrivener. Zoot has been around 30 years.

It's rare that even a discontinued app will just stop working (unless it is strictly web-based), which gives you time to do what you can to bring as much of your work as possible into a new app. I suspect the best that you can do in a lot of cases is bring static PDFs into the new app, especially if the data is visual rather than textual.

In my experience, manipulating the data in these visual apps is where the value comes in. That is, thinking things through while building the maps, is where I get most of the benefit of using the app. So maybe there is some solace in the fact that you've already gotten good value from the now-abandoned application. And if you need to rebuild the map in a new app, it is a new opportunity to re-examine your efforts and perhaps develop new insights.

Steve Z.

Fear of abandonment is the chief reason I refrain from apps that don’t work properly without store my data in the company’s cloud servers. That includes PWA apps like AmpleNote, which misleadingly markets itself as “local-first” but actually operates on the premise that any disconnection from the server would only be temporary.
Abandoned local software will, at the very least, continue to work on an older system. Abandoned cloud-based software is not predictable in any way. They could give me only two months’ notice to export data, and permanently delete everything after that.[^src1]

[^src1]: For instance, the read-it-later app Omnivore did exactly that in November 2024. https://web.archive.org/web/20241029154734/https://blog.omnivore.app/p/omnivore-is-joining-elevenlabs

I also learned that an app being open-source is by no means a guarantee of longevity. The real guarantee is:

- You have the data stored locally, in an accessible format (that is, not just in your browser cache).
- You have tested the export capabilities, and is OK with the results. Not needed for markdown/plain text. (I test monthly all my apps where export is potentially useful, right after testing the integrity of my 3-2-1 backups. Always test your backups and exports.)
- You have a general understanding of the data store format, as well as some idea of what other tools can be used to access/manipulate the data store.

My recollection of what happened to the developer of ConnectedText is that he had a bad bicycle accident. He made some initial assurances that he would get back to ConnectedText, but never did. One can only sympathize with his plight. I mention this as an example of how random events can hit. There are no guarantees that you can rely on a small shop continuing or winding down in a way that protects your data, no matter how ethical the developers are.

So what to do ? Use only apps from big companies like Apple or Microsoft , that are unlikely to disappear in the short/medium term ?

In my experience big companies kill off products as often as small companies, if not more so. Google is famous for killing products.

For longevity:

* Downloadable software beats web software. Web software dies the second the vendor turns off their server. Downloadable software might keep going for decades.

* Windows applications beat Mac applications. Microsoft puts a lot of effort into backward compatibility. Apple nukes their entire ecosystem from orbit every few years.

* Choose software that uses (or can export to) an text based standard format, such as CSV.

As a vendor I hope to keep my applications available to customers for some years after I stop selling them, if I possibly can.

And, if you are having to reshape/reformat data to move it from application to another, maybe consider my Easy Data Transform software. ;0)
https://www.easydatatransform.com

Ps/ Nothing lasts forever. Even the sun is going to burn out, eventually.

I think I'd add Devonthink to the list of future proof, or at least future tolerant, apps. At first glance, it looks like everything disappears into a proprietary database. In reality, the database file is a Mac package, which is to say it's a normal directory perceived as a single unit by the Finder. Inside the package (directory) all your documents are there in original form, somewhat similar to an Obsidian vault.

You lose the groups, but not the documents, if you don't have Devonthink.

Devonthink makes its own folder hierarchy, which is how it handles duplicate names in a group.

If you want to keep the groups, export your database as files and folders or as a web site as part of your backup rituals.

I like Curio. It would be tougher, I think, to extract information from it without the app. A long time ago, I took an export from Curio and converted it to an Aeon Timeline-friendly import file. It didn't take much. I put Aeon fields like Arc, Observer, and Participant in Curio note attachments. That's a nice Curio feature. Anything can have a note attachment.

Nisus surprised me. It's been catatonic for months without any communication from the company. You can still buy it through automated systems. Customer support is unreachable. Fortunately, Nisus documents, while not standard RTF, are standard enough any word processor can read them.

Scrivener should be considered pretty much future proof. There's an XML file describing how everything relates, and then all the documents, index cards, etc., are either plain text or RTF. Creating a legit Scrivener project without access to Scrivener isn't all that hard. The XML file and inner structure makes perfect sense, and Literature and Latte has a developer's document they give out to anyone who's interested.

Mellel (Mac word processor) has a proprietary native document format, sort of. It's a zip file. BBEdit will directly open Mellel documents.

Images are in the zip file in original form. The document itself is in an XML file, easily interpreted by guesswork, with all content in plain text. Although some minimal effort is required, it doesn't take much to get the important parts out of a Mellel document without using Mellel. I'd call it future proof.

If an app uses standard things, like XML, or if data is in any easily discernible format, I'd call it future proof. Everything doesn't have to be pure plain text.

Andy Brice wrote:

Ps/ Nothing lasts forever. Even the sun is going to burn out,
eventually.

This was a scary thought. Then I checked.

Finding a running computer might be tough after Sol wheezes out, but Easy Data Transform itself is software. It's not solar powered.

We're good.

I still miss Ariadne which was a compact, elegant (for its time) progam. I sometimes wonder what it would be today if someone breathed new life into it.

By the way, what about Pocket's planned demise? I have thousands of items which will need to be transferred.

I am wondering what program to use as a Pocket replacement. Any suggestions? FWIW I am looking at Readwise and Readise's Reader.

Are there better options,

Thanks.

Daly

Instapaper (for online storage), or Obsidian w/ the Obsidian web clipper, which is quite good.

Daly de Gagne wrote:

By the way, what about Pocket's planned demise? I have thousands of
items which will need to be transferred.

I am wondering what program to use as a Pocket replacement. Any
suggestions? FWIW I am looking at Readwise and Readise's Reader.

Are there better options,

Thanks.

Daly

Speaking of Apple nuking everything from orbit, I give you macOS 26:

https://www.apple.com/os/

Its not clear to me yet what effect this is going to have on existing macOS apps.

--
Andy Brice
https://www.hyperplan.com
https://www.perfecttableplan.com
https://www.easydatatransform.com
https://www.successfulsoftware.net

MacOS 26? I guess my version 12 is pretty far in Apple's rear view mirror.

It's the newest version I can run. I'm starting to think about moving my life out of certain key Mac apps into things I can run under Linux. My machines are 10 years old, so clearly Apple hardware lasts. Upgrading takes a bite out of the budget, though.

Amontillado wrote:

My machines are 10 years old, so clearly Apple hardware lasts.

The trackpad on my MacBook Air only lasted 2.5 years. Warranty was only 1 year. Not impressed, given the price.

Wait what, so mac apps don't work on older versions of macOS or iOS? This is total news to me, so what do devs have to do to keep the business running for a longer period then? Make significant changes for each new mac version? Approximately how much time can an average purchased software be expected to work in the iOS and macOS enviornment then? Apparently there has been a differently named OS every year in the recent years.
I was actually considering buying a macOS device just to use some of the most fascinating apps that get mentioned here but I pirate (unpopular opinion here, and understandably so) and a lot of those that can be found can be quite old. So they probably wouldn't even work on the new mac minis. That's really interesting, and something people seldomly mention for some reason.

A question for heavy mac users, to what extent has this been a problem and what do you usually do about this? Do you just buy the classic one-time purchase software, and the optional upgrades purchases cover the OS changes and make sure the application keeps working? What if you don't want to or can't purchase an upgrade and you need new hardware but then all your data is in the old OS that you can't use in the hypothetical situation. Do you then just export content and find another app that can import satisfactory?
And maybe this can partially explain why the subscription model took over?



Andy Brice wrote:

So what to do ? Use only apps from big companies like Apple or
Microsoft , that are unlikely to disappear in the short/medium term ?

In my experience big companies kill off products as often as small
companies, if not more so. Google is famous for killing products.

For longevity:

* Downloadable software beats web software. Web software dies the second
the vendor turns off their server. Downloadable software might keep
going for decades.

* Windows applications beat Mac applications. Microsoft puts a lot of
effort into backward compatibility. Apple nukes their entire ecosystem
from orbit every few years.

* Choose software that uses (or can export to) an text based standard
format, such as CSV.

As a vendor I hope to keep my applications available to customers for
some years after I stop selling them, if I possibly can.

And, if you are having to reshape/reformat data to move it from
application to another, maybe consider my Easy Data Transform software.
;0)
https://www.easydatatransform.com

but I pirate (unpopular opinion here, and understandably so)

Sufficiently unpopular that I won't be taking the time to make a substantive reply.

Andy Brice wrote:

Speaking of Apple nuking everything from orbit, I give you macOS 26:

https://www.apple.com/os/

Its not clear to me yet what effect this is going to have on existing
macOS apps.

I’m not sure what you mean by “Apple nuking everything from orbit”. If you’re referring to the 26 denomination, they simply changed the way they named the OS ,now by their year of release instead of the usual incremental numbers from their first release.

The rest is just cosmetic change to the look ( more transparency ) .

Lucine wrote:

Wait what, so mac apps don't work on older versions of macOS or iOS?
This is total news to me, so
>

No, they just changed the way they named their releases ( now by Year of release )

I'm not sure what you mean by Apple nuking everything from
orbit

I am referring to the fact that Mac apps have to be continually updated to work with the latest macOS OS/processor.

I develop apps for Windows and Mac.

If you take a Windows release of my software from 20 years ago, there is a good chance it will run on the latest Windows OS.

If you take a Mac release of my software from 20 years ago, there is no chance it will run on the latest Mac OS. In fact a release from 5 years ago might not run on the latest Mac OS.

I understand macOS 26 is not going to support Intel chips. So you won't be able to update that expensive Intel Mac you bought in 2020. It looks like Intel apps will continue to run under Rosetta translation, for now:
https://forums.macrumors.com/threads/apple-to-phase-out-rosetta-2-starting-with-macos-28-as-intel-era-ends.2458631/

As a vendor:

On the plus side:

-Mac users are pretty much forced to upgrade their apps regularly if they want to stay on the latest version of macOS.

-apps that aren't being actively maintained disappear as competitors quite quickly.

On the minus side:

-It is a grind to keep on jumping through the hoops created by Apple's changes.

Andy Brice wrote:

I understand macOS 26 is not going to support Intel chips. So you won't
be able to update that expensive Intel Mac you bought in 2020. It looks
like Intel apps will continue to run under Rosetta translation, for now:
https://forums.macrumors.com/threads/apple-to-phase-out-rosetta-2-starting-with-macos-28-as-intel-era-ends.2458631/

Starting macOS 28 , not OS26. So, in the year 2028. It’s written in the title of the link you posted.
Also, support for Intel is not going to be dropped entirely in os28 ( year 2028 ) , as written in the link you posted :
“ Beyond this timeframe, we will keep a subset of Rosetta functionality aimed at supporting older unmaintained gaming titles, that rely on Intel-based frameworks.“

But yes, support for Intel processors will likely to be entirely dropped in 4 or 5 years from now. For anyone old enough to remember, the same thing happened when Apple transitioned from Motorola/IBM PowerPc processors to Intel, decades ago. In a similar way Rosetta 1 provided emulation for software written for PowerPc processors, and was dropped a few years later to only support Intel processors ( it lasted less than the actual Rosetta 2 , which will be providing full Intel support for at least 6 years, then probably a much limited support for a few more years beyond that ) .

Actually , as new releases of MacOS and iOS are traditionally announced ( and beta released) at WWDC in June, but publicly released around October, that means the limited support for Intel will begin closer to the year 2029…

I think 6 years ( beginning in 2020 ) is a reasonable time frame for developpers to transition their code from Intel to the Silicon Mx processors ?

Skywatcher wrote:

Actually , as new releases of MacOS and iOS are traditionally announced
( and beta released) at WWDC in June, but publicly released around
October, that means the limited support for Intel will begin closer to
the year 2029…

I think 6 years ( beginning in 2020 ) is a reasonable time frame for
developpers to transition their code from Intel to the Silicon Mx
processors ?

EDIT : sorry, MacOS 2028 will be released in late 2027, not 2028 as I wrote. So , full support of Intel will diminish starting October of next year.

( I wish there was a way to edit our existing posts instead of having to post new ones to rectify something in it )

sorry, MacOS 2028 will be released in late 2027, not 2028 as I wrote.

It is confusing, isn't it?

But it looks like macOS 26 (released 2025!) will continue to support Intel chip:

https://www.theregister.com/2025/06/10/apple_macos_26_last_intel_support/

My mistake.

I tried to address your questions in:

Abandoned Software and Lock-in
https://bicycleforyourmind.com/abandoned_software_and_lock_in
#macOS #software