Third-Party Plugins Integration in Evernote as a premium feature

[Rafael B 26]

I'd like to suggest an enhancement that could take Evernote to an even higher level in terms of functionality and customization: the integration of third-party plugins.

 

Observing the success of platforms like Anki and Trello, which allow the inclusion of plugins, I have noticed how this approach can significantly enrich the user experience. By opening the platform to external developers, Evernote could benefit from numerous extensions and improvements, including but not limited t

• Niche-Specific Features: Facilitating the development of specialized plugins to meet the specific needs of different professionals or user niches.
• Integration with External Tools: Enabling closer integration with other popular tools such as Anki, Trello, TickTick, Alfred and others for a more unified experience.
• Customized Interface: Allowing users to choose from a variety of plugins to customize the interface according to their individual preferences.
• Enhanced Audio Player: The possibility of integrating plugins that offer advanced audio playback features, facilitating the review of audio recordings and notes.
• Productivity Improvements: Developing plugins that optimize workflows and bring additional features to improve productivity.

By adopting this approach, Evernote could transform into a more dynamic and adaptable platform, catering to an even wider range of user needs.

 

 

I understand that security and stability are fundamental priorities. Therefore, I suggest that the approval and integration process for plugins be rigorous, ensuring quality and security for all users.

 

 

 

Rafael, from Brazil

[AlbertR 690]

13 hours ago, PinkElephant said:

The APIs have become sort of a battleground with many apps,

Yep - and if you (as a software vendor) do not have a suitable API, you will fail because your product has to live on an island.

13 hours ago, PinkElephant said:

competing services have used the API to drain data

OK, mass downloads of note content might be disallowed to avoid competitors to implement tool switches. But if anyone else is interrested in reading EN data, he is free to develop a 100% ENML importer - without an API.

13 hours ago, PinkElephant said:

And about paying or non paying accounts: Up to now this has never played a role.

Bad argument to think about the future...

14 hours ago, PinkElephant said:

Because the API usually doesn‘t communicate directly with the user, I think it will be sort of difficult to communicate the message „I am here, but I want your money before I give you the data“. There is always another app or service in between.

Yep - and already true these days on EN site: If you want to use the API, you have to create an authorization token. It's a trivial task to disallow creating such tokens for free (low low-paying) accounts.

If you want an other service or app to access your EN data, the service or app has to create such an authorization token with your help by allowing access for a specific time period. If this time period ends, the token gets invalid an the service or app cannot access your EN data any more.

It's also a simple task for EV developers to invalidate autorization tokens of a specific account immediately if such an account changes from payed to a level with no API access.

All-in-all:

• Technology to track API access along with account plans is available.
• Binding API access to payed plans generates income.
• Income can be used to improve the API
• An improved API helps to compete with competitors (more: is really necessary!)

[PinkElephant 8,744]

There are note apps on the market that rely on plugins. If they fare better, anyone can try.

3 frequent technical issues:

1. Plugins may use properties that are altered or removed from the app later, probably without a chance for the apps devs to see that a plugin uses a piece of code. Then the plugin fails - we have seen this happen with integrations or 3rd party services in the past years.
2. Active plugins may alter the content of the core apps data in unexpected ways. This can lead to difficult to troubleshoot support cases.
3. Plugins usually have a hard time to monetarize their services. This can lead to popular plugins being removed suddenly, or given up by their devs, leading to all sort of problems (in extreme to severe bugs and security breaches). Many plugins need to request full access to the notes - without transparency about how the extracted information will be processed, and if privacy is maintained.

I don't think that going down this road is really promising. In addition to these technical problems, several plugins installed can rapidly lead to a bloated, blundered and inconsistent UI. You just need to cross check your browser plugins for this to see what I mean.

It's open for discussion, but I don't see this approach as leading anywhere.

[Rafael B 26]

I recognize the validity of the arguments and share similar concerns. However, today there are already plugin-like features in Evernote. What is synchronization with Google Drive, Google Callendar and the like if not plugins? Everything in this life is about considering the balance between positives and negatives, and software development is no exception.
 

For me, if it's possible to have something from Google integrated into Evernote, it's possible to have other services. What's more, softwares like VS Code, Trello and Anki are packed with plugins/addons, and that's no problem for them.

[PinkElephant 8,744]

The mentioned existing methods to link content are no plugins. They are provided as an integral component by the app itself, maintained by EN devs.

About not having problems with plugins this could be discussed - there has been similar attempts in the past, and they failed (Example: Bubble Browser), or nearly failed (Example: Filterize). Not every ecosystem is plugin-friendly - EN users have not been keen to open their pockets for plugins in the past, which lead to commercial failure even when the additional function was welcomed.

The API is there, for devs to use it and create new integrations. If this doesn't happen, it tells a lot about the missing perspective of such endeavors. This is not a technical problem.

[AlbertR 690]

10 hours ago, PinkElephant said:

The API is there, for devs to use it and create new integrations. If this doesn't happen, it tells a lot about the missing perspective of such endeavors. This is not a technical problem.

The API works on server data - but it has some restrictions:

1. It's incomplete (cannot get|set all the note times),
2. it's not really RTE enhanced so far (cannot wait for a RTE room to be available)
3. has access restrictions (cannot get more than 128 note by one call and fires RATE_LIMIT_REACHED exceptions if call comes to often in a specific (uncodumented) time frame)

(2) and (3) are the reasons for Filterize's dead (beside some internal problems I guess 😉). There are some other products that might replace some Filterize functionality (i.e. Make (formerly known as Integromat)) - but they have to solve the same two problem areas which will not be possible without EN's willing help (that's not seen so far 😞)

About opening EN's UI for local Plugins: I'm not sure this would help in the current (unstable and unfinished) product lifecycle stage.

[Jon/t 1,640]

When the API gets updated to work with the new backend and any new features I think opening it up for 3rd party devs would be a good idea.

Having some sort plugin system like Obsidean/Joplin would  be a very bad idea! I do WordPress consulting and issues ALWAYS come from third party plugins and themes. Most are hobby projects and when devs lives change the project gets dumped and folks websites break.

I think using the API to build upon EN would be a good idea but not building within the app.

[Dave Green 260]

17 hours ago, AlbertR said:

Filterize's dead

Filterize is still operating, and things have gotten much better, although not perfect regarding RTE.  I am able to use it to update TOCs, provide a list of random notes to look at, and clear checklists.  I tend to leave those pages when I know updates will be applied plus I have a special tag "!filterize-fire" to have it look at a page that I have changed or will change due to the various rules.

[AlbertR 690]

6 hours ago, Dave Green said:

Filterize is still operating

... until December 31st 2023 😞 - I also use it for TOCs and some other tasks...
I've got an "End of Life" eMail last week. All subscriptions will be terminated in December.

Citing from this eMail: "Mounting technical challenges in the Evernote integration make keeping Filterize alive an impossibility at this point."...

Maybe "at this point" should give some hope - but this should come from EN 🤔

[Dave Green 260]

19 hours ago, AlbertR said:

... until December 31st 2023 😞 - I also use it for TOCs and some other tasks...
I've got an "End of Life" eMail last week. All subscriptions will be terminated in December.

Citing from this eMail: "Mounting technical challenges in the Evernote integration make keeping Filterize alive an impossibility at this point."...

Maybe "at this point" should give some hope - but this should come from EN 🤔

Wow.  I have not received such an email.  Sorry to hear this.

[Boot17 1,532]

It will be interesting to see where Bending Spoons goes with the API. My guess is they’ll either nix it (too much effort to support for the return on investment) or start charging for it — like X and Reddit. Bending Spoons is about getting compensation for their product and I’ll bet they have little motivation to continue hosting a free API. 

[Boot17 1,532]

On 11/20/2023 at 2:22 AM, Jon/t said:

When the API gets updated to work with the new backend and any new features I think opening it up for 3rd party devs would be a good idea.

Have you heard things through your channels that they are going to update it or maybe just taking an educated guess about it?

[Jon/t 1,640]

2 hours ago, Boot17 said:

Have you heard things through your channels that they are going to update it or maybe just taking an educated guess about it?

Not heard anything. I can't imagine they wouldn't update the API so I'm hoping once the new backend is finished they'll work on it, at least to include tasks.

[AlbertR 690]

10 hours ago, Boot17 said:

I’ll bet they have little motivation to continue hosting a free API. 

They might open it only for paying accounts. Another argument to pay for the service at all...

[PinkElephant 8,744]

The APIs have become sort of a battleground with many apps, because partially competing services have used the API to drain data and income away (for example 3rd party apps for Xitter, formerly known as Twitter). Because EN doesn‘t rely on advertising as income, I don‘t think this is a motivation here.

And about paying or non paying accounts: Up to now this has never played a role. Because the API usually doesn‘t communicate directly with the user, I think it will be sort of difficult to communicate the message „I am here, but I want your money before I give you the data“. There is always another app or service in between.

[Jon/t 1,640]

I think the biggest benefit for Evernote of an open API would be getting more stuff into the app and increasing usage. If you can link other software to automatically send stuff to EN then it instantly becomes more valuable. Better links with IFTTT, Gmail, Outlook etc and Evernote devs won't need to build these connections, others can.

[AlbertR 690]

3 hours ago, Jon/t said:

the biggest benefit for Evernote of an open API would be getting more stuff into the app and increasing usage.

Totally agree - but we have to distinguish between an API to the client app (on Windows, Web, Mac, Android) and an API to data on server site.

The API to server data is available and should be enhanced soon to match new functionality (tasks, some note properties and usability)

The API to the client apps is MUCH more difficult to implement and may cause many problems because masses of not-so-well-trained users try to use it in unknown environments. Support will be overloaded. I would not recommend to do it for now...

[Jon/t 1,640]

@AlbertR Yeah, just the server side plus good documentation as well! I'm building a training hub/portal thing for a client on top of Zoho CRM and the API is very powerful, I can do almost anything but the documentation is hard work 😂

[AlbertR 690]

Just to pop-up this discussion...

• EN uses The scalable creator and customer community platform - Invision Community as its Forum software.
  • Invision Community offers a REST API to handle Forum postings (and all other stuff around managing the forums itself): REST API Documentation
  • EN itself does not provide API keys to customers to access EN forum - which is OK. It's not our (user's) main task to manage discussions by an API😉
     
• But: It's our main task to manage our note within EN - and therefore we really need a complete and reliable API interface to our accounts.
  • The current API interface (see Documentation - Evernote Developers) is somewhat outdated (call it "Legacy" if you want...). There are 12 (in words: TWELVE) different documents available to describe access from "supported" operating systems and programming languages😞
     
• My suggestion: Implement and offer a REST API to out accounts
  • REST is state of the art around nearly all Software products that use any server components
  • By using REST there's no need to provide different Developer APIs
  • Use Invision's implementation and documentation as a well known example😉

BTW - here a very interessting discussion from Invision forum

• https://invisioncommunity.com/forums/topic/467056-hump-day-new-live-community-features/?do=findComment&comment=2892554
• they talk about
  • bigger invision customers (whereas EN is mentioned),
  • backgound of business strategies (... it's better to focus on quality than quantity...)
  • and got IV's managment to get involved (exemplary!)

[AlbertR 690]

Another add to our ideas:

• A new server API has to include EN Tasks for sure!
  • After working with tasks since app. 2 weeks (and beeing satisfied so far since the task pane contains a notebook tab now 👍), the only missing feature is the ability to automate Task creation, modification and readability.
  • Here are some suggestions following the current (old) methodology:
    • ENTask myPlainTask = new ENTask();
      myPlainNote.<property> = <value>;
    • ENTaskRef myPlainTaskRef = ENSession.SharedSession.UploadTask(myPlainTask, ...);
    • List<ENSessionFindTasksResult> myTasksList = ENSession.SharedSession.FindTasks(...);
    • ...
  • But better implement it in a REST API like available for the forums 😉