Performance with very large documents

I read in this thread that a 2.8MB OPML document makes the mac app slow and unresponsive.

Also on a recent blog I read that more than 1000 items would do the same thing on mobile apps (though as I understand this is a bug).

My WorkFlowy OPML is 8MB, so desktop and mobile apps would be unusable if I was to import it currently unfortunately.

May have to migrate only part of it into Dynalist. Would breaking it up into separate documents help performance?

Also can anybody comment on their experience with the web app performance with very large documents?

Yes, it definitely will, but you need to load the WorkFlowy import in order to break it down, so that would still be slow.

The slowness mainly comes from having too many things visible. That means, if you move a chunk of things into other documents, or simply collapse or zoom in, it would improve performance by a lot.

I’ve poked around and didn’t find a direct answer to a question I’ve had for a while, so I’ll ask it by tagging onto this old question.

Are there practical limits to how large one should limit your Dynalist file(s)?

The answer to the second question may be obvious from the answer to the first question, but if one should keep an eye on document size, what’s an easy way to do that?


1 Like

It depends on the the device you’re using. Generally, 1 MB is already a lot.

To give an idea how much is 1 MB, this huge list on our demo is only 731 KB (around 73% of 1 MB):

So 1 MB should be a pretty practical size limit for most documents.

If your document is indeed huge, you can break it down to multiple documents, and use internal links to link between them to hold the structure together.

Constantly delete things you don’t need is a way to keep the size down. When you check things and they get hidden, they’re still part of the document and gets loaded every time you open the document. The “Delete checked items” menu option makes this simple.

Export and download the OPML version and check the file size.

As I mention above, periodically pruning deleting items also helps a lot.

1 Like

Thanks for the helpful response!

I’ve just been using Dynalist a short time but finding it very useful. So my question was more looking towards the long term rather than an immediate need.

My main Dynalist file is 115 kb at the moment, but I can see it easily growing to 10X that in less than a year by extrapolating the growth over the last couple months.

I tried the sample file, but perhaps I’m confused, as it seems to be separated into several documents each of fairly modest size. Does the performance of this group of documents demonstrate the performance of a single large document?

Clearly I’m confused and hopefully you can straighten me out.

[Edited a few minutes later]…I may have figured out the point. The document under Scholars > Propædia is 936 KB when I export it as an OPML, so that’s apparently the document you wanted me to see to demonstrate performance of a ~1MB file.

I see that it does respond with a bit of lag when all branches are open and the document has just been loaded. As I page down thru the document, it takes a moment for each new page to load. If I expand/collapse all nodes it takes about 5 seconds to respond to that command.

But if I collapse to the top level nodes and just open them one at a time, response is very snappy and there’s virtually no lag. Also, after the document has been loaded and used, it’s a bit more responsive than when first loaded. Finally, and most critically for me, searches are very responsive and quickly narrow down the text to a topic of interest.

Now that I see what you were trying to show me, the demo does answer my question very effectively. I can see directly how a 1 MB OPML file responds.

As I would never have cause to have all nodes open simultaneously, I can see from the demo the kind of response I would actually experience in a scenario that’s realistic for me.

Thank you so much for answering my question so well. My concerns on document size have been put in perspective and I doubt I’ll worry about it much again.


Glad that you figured it out! :slight_smile:

FYI…I imported the 1 MB Propædia file into my Dynalist account to check performance and it was so responsive — even with all the nodes expanded, which is a more challenging use case than I would ever experience in real world use — that I’ll have to say I no longer have any concerns about the practicality of Dynalist for me.

Thx again!

1 Like


To be honest, the performance of lists always depend on the user’s device (how much memory and CPU is available). So I’m glad you tried things out yourself and got an idea how it would work out for you :slight_smile:

Hi friends,

I am looking to transfer my archive on manuscripts from the 600 sermons I’ve preached over the last 12 years. The reason is I’d like to tag sections by theme to aggregate concepts across manuscripts.

I’ve imported the sermons into both Dynalist and WorkFlowy. It’s approximately 6mb of plain text. (80k lines).

In workflowy, it takes a few seconds to load the outline, but then it’s super responsive and quick.

In Dynalist, it loads quickly, but performance is extremely sluggish and even times-out and becomes unresponsive. Even when I split it up into multiple documents (which is less than ideal because I’d like to select search results across all lines), which does improve performance, but still not to the level of workflowy. This happens in the browser, the Mac app, and the iOS app.

Perhaps the two services have a different mechanism for loading the outline into memory or something.

I MUCH prefer the community here and dislike the idea of using workflowy over dynalist, but the performance difference is massive.

Any advice to make this work in Dynalist would be so great!

@Shida: could you take a look at @David_Eiffert’s case? Thanks!

We’ve released a recent update that addressed a bug which caused significant performance issues. It’s currently live on web, and the desktop app version 1.2.3.

Let me know if it helps, but if not I’d be happy to take a look at your case. Most likely we’ll ask for a performance log that you might have to take on Chrome, and we’ll be able to analyse the log on our machine.

1 Like

I don’t have a lot of data on my list (about 3K items), but I definity feel much faster. Thanks.

1 Like

Yeah seems much faster for me as well! Thanks so much.

1 Like