Open letter to EN: The correct way to replace a product's code base

[Rob Freundlich 146]

Reading all of the posts here and in the EN reddit over the last few days, I've had a sinking feeling in my chest. I was so excited that the folks at Evernote were going to a single code base! As a developer, I understand this. It would make bug fixing and adding features so much easier! All of the platforms would finally have the best of all worlds!

Instead, what we've gotten is, well, what we've gotten. I'm very disappointed, both as a user and as a developer.

I've been working on a similar project for the last two or three years - porting a Flash-based application to Typescript and HTML. This was not optional; at the end of this year, browsers will stop supporting Flash, and my company's customers' business-critical applications need to continue working. Dozens, if not hundreds of enterprise customers, and tens or hundreds of millions of dollars of revenue, are on the line.

We started releasing the port a year or more ago, while still supporting the Flash version. With every release, we included a clear list of what worked and what didn't, with as complete a road map as we could of when the incomplete features would be complete. We made a few usability changes along the way, and dropped only one feature that I can think of. Before deciding to drop that feature, we polled account managers to see if it would be a problem, and were told it would not be. Even with that dropped feature, we are working to implement the spirit of it in another way.

Customers received the feature complete version a month or two ago, and even the ones who were slow to adopt it are now waking up and realizing they need to. Bug reports are rolling in, and we are prioritizing them over everything else, putting out patches every couple of weeks.

In addition to the bug fixing, we are communicating all of this to the customers. They know what we're doing, and they know that when Flash shuts down, they will have the best replacement possible for the product they have built into their business processes.

They are thrilled with us. Universally.

This is how you replace the code base beneath a product.

[DTLow 5,736]

3 minutes ago, Rob Freundlich said:

We started releasing the port a year or more ago, while still supporting the Flash version.

I continue to use legacy versions of Evernote; there's official Mac/Windows versions.    
Not clear if there will be any support

[Rob Freundlich 146]

2 hours ago, DTLow said:

I continue to use legacy versions of Evernote; there's official Mac/Windows versions.    
Not clear if there will be any support

That's my plan for now as well. But I'm pretty sure they said there won't be any feature development in the legacy version (kind of implicit in "legacy"), and I think I saw something that said there won't be any support. I could be wrong on the second point (and hope I am!)

[PinkElephant 8,115]

We will have to wait a little to see what is going to happen.

First what will be released by the EN team in the next weeks. This means missing functions as well as bug fixes.

Second if the legacy versions will continue to work. The fastest development has been with browsers, so probably the WebClipper will come under pressure first.

For me the next major question will be if the legacy client will still work under MacOS BigSur.

[mlidi 23]

The new version for windows is a complete disaster. Put aside the missing features, it's so slow! it takes literally 5-10 seconds for a new empty note to open in a new window. Continuing to use the legacy version and will start looking for alternatives

[Rob Freundlich 146]

20 hours ago, PinkElephant said:

We will have to wait a little to see what is going to happen.

Part of my point is that we shouldn't have to wait to see. If you (a company with a very mature product and a large user base) are going to release a complete rewrite that doesn't contain all of the functionality of the product it is replacing, the responsible, correct way to do it is with lots and lots of communication up front.

You don't leave your users wondering "where did my critical feature go?" and "when, if ever, will it come back?".

You don't leave them wondering if the legacy app is going to live on or suddenly disappear.

You don't leave them wondering if they are going to have to completely change the way they have done things for the last 5 years, 10 years, or even longer.

Even if your policy on the past had been to not talk about future plans, you put that aside, because this is a critical juncture in your product's history, and you need to recognize that business as usual isn't going to work. You have to give your customers the information they need to make intelligent, reasoned decisions.

Doing it any other way is irresponsible and shows a tremendous disregard for your customers.

[MarcSant 223]

At least in my point of view, they are unifying the product to have a single development team, reducing costs by allowing the Windows team to have the same development base for Apple. 

This forces them to remove features that are highly dependent on the operating system, such as sending a note via email, changing the font to a list larger than just 5, among other things. 

What they did was: "we will implement a multi-platform version, release it with what we could provide for both ecosystems, and over time we will implement back what we removed." 

From a business point of view, it makes sense because you save money and focus in one direction only, with shorter development cycles.

But the problem is that they haven't thought about users at all: they've launched the new product and damn what the user thinks. They could have minimized all this stress by sending an email to all users talking about the new version, and in this email putting the link to the Legacy version in case the user didn't feel comfortable with the changes.

I live here in Brazil, I have always praised and spoken well of American Technology companies, how serious they are, competent and how they produce excellent systems and products that help our lives, but this time, sincerely gave me shame about what they did. 

It was very, very amateurish. I would expect this kind of posture from a backyard company, a beginner, which has a bunch of teenage nerds starting now, but, honestly, the role is inverted: The teenage nerds are producing each time more great products, creating applications that are innovating and changing the life of many people, and contrary to what I said, they don't take these decisions that irritate so many people in such a short time.

I think the company's board needs to look to this, it needs to look to what is happening, it is not just a bunch of angry people with a new function or new resource but essential functions that makes the product viable for the users.

Evernote as a company can't just believe that because we depend on them we won't migrate to another platform, we can and will do it! 15 years ago there wasn't alternative, today we have, and to have alternatives means that your product can't give margin for error. 

And this last change was a grotesque mistake, which is leading to a huge base of paying users to rethink Evernote as a loyal note companion. 

[CalS 5,280]

On 10/11/2020 at 10:38 AM, PinkElephant said:

First what will be released by the EN team in the next weeks.

I wonder what's the over/under on bug vs feature composition of V 10.1.  And the when.  Should be indicative of speed to market with the new infrastructure,  Hoping for the best.

[CalS 5,280]

2 hours ago, MarcSant said:

This forces them to remove features that are highly dependent on the operating system, such as sending a note via email, changing the font to a list larger than just 5, among other things. 

This may have been forced by the draconian approach of making all platforms the same.  Seems to me a "desktop" version and a "mobile" version with a defined overlap of function and stuff specific to the strengths of the two types of platforms would have made more sense. 

Create a table with three columns for Both, Desktop, Mobile, and fill in the blanks.  If something is harder on Windows or Mac than the other, tough, work harder and fix it. Pretty easy to communicate as well.  None of this something we are still considering babble, after 2 years, WTF?  Somebody architected this thing.  Sounds like oops to me.

[Dave-in-Decatur 3,959]

58 minutes ago, CalS said:

This may have been forced by the draconian approach of making all platforms the same.  Seems to me a "desktop" version and a "mobile" version with a defined overlap of function and stuff specific to the strengths of the two types of platforms would have made more sense. 

This.

[BrodocK 0]

There was an open promise that they would not use Electron, and would keep native interfaces for every platform. The new UI is just laggy and bugous. Every click now takes an extra "loading". On iPad there is a visible white flash when switching from one note to another. On Mac OS you see that flash on the "content area" where fixed UI text disapear when loading a new note. We lost the entire desktop topbar for this crappy weblike version of the application.

On iPad this is even worst, as Pencil support is horrible. It lags so much that every time you take the pencil out of the screen to, let's say, write the next letter/word, half of the time it doesn't register in time to keep up with what you were drawing... native notes app works as expected.

Evernote is not a cheap subscription, and note-taking is kind of commodity. The only reason I payed all this time for evernote was to support native development, because I value that. You now lost the entire differentiation you had.

[chronistin 187]

On 10/12/2020 at 2:35 PM, Rob Freundlich said:

You don't leave your users wondering "where did my critical feature go?" and "when, if ever, will it come back?".

You don't leave them wondering if the legacy app is going to live on or suddenly disappear.

You don't leave them wondering if they are going to have to completely change the way they have done things for the last 5 years, 10 years, or even longer.

Exactly! Lack of communication is EN's biggest problem. The resolution to give more information with the behind the scenes videos was a big disappointment. Going into the depths of GUI designing or search setup in a frontal teaching setup does not make up for lack of communication. 

As I see it, with every new version they are continuing to dumb down EN to make it more appealing to new users, abandoning or hiding features that made it unique for power users. I don't like that direction at all.

At this point, I'm only still using EN out of laziness. Migrating > 20000 notes is a task I would like to avoid. I've been using a double strategy for the past 2 years, continuing to file in EN, but parallel setup of notion / gdrive / keep. Still missing the elegant universal information vault EN used to be for me, but at this point, I'm ready to jump ship.