General feedback: your prioritization seems wrong to me

⁉️ Ask us anything!
#1

I’m sorry if this post will sound harsh. I’m just giving you my opinion straight. You have a great product idea and great potential, but the execution is lacking, and I hope you can fix that.

It seems to me that your efforts prioritize adding features rather than fixing your bugs. There are so many bugs, so many instabilities. Just look at the feedback you got on Hacker News. I was hoping these issues would have gone away. But I can see that even today you’re not that interested in correctness. This problem is critical: you’re dealing with our lifelong data. We cannot rely on this app long-term unless we can trust your reliability.
As a long-time developer myself, I recommend that you do a feature freeze and you clean up all your technical debt and fix all your critical reliability issues. It’s not normal for people to have their database get wiped out randomly. It’s not normal for imports to freeze all the time (or go on infinitely). It’s not normal that I can’t get local backups working on macOS. I know that these are tough technical problems, but that doesn’t mean you should ignore them. If your current developers can’t solve these problems, then hire some better ones. These are showstopper bugs, as they do not inspire confidence.
“Move fast and break things” doesn’t mean you should leave things broken.

Good luck. I want you to succeed and I want to love this product

6 Likes
#2

Wow, thanks for the time and bravery to publicly post such an honest and direct feedback.

Appreciate this very much. :pray:

We are indeed now focusing more on fixing those bugs one-by-one, as we recognise that the many bugs we have is really pissing people off (us too).
The thing with our setup is that Memex works in a very hostile environment (offline-first in the browser) where there are so many different setups of people that are hard to foresee. E.g. we have to roll out all new updates to every user at once, and can’t, unlike in traditional cloud setups only migrate a few users for testing. Also the release cycle of Chrome is so long that it takes usually 3-7 days for each update to release.

The problem with the bugs you mention is that they are actually much out of our control :confused:

  1. The lost data is because browser databases are not very reliable. They may eject data out of the blue, even though Memex requests permanent and unlimited storage. We are working on moving the application and storage logic of Memex outside of the browser, but it’s a gigantic task that we can’t do all at once. In the upcoming 2.2 version we have introduced a potential fix but we are not sure because they browser docs are so vague and we should have already permanent storage by setting the right value in the manifest.json.
  2. The backup app we have investigated a couple of times but we can’t find the bug. The assumption is that some people have additional software somehow blocking that port, either from the browser or local.
    Do you have any special configuration that may cause that the local app can’t be found?
  3. The freezing issue is explained here and is also outside of our control, there is currently no workaround for this, other than the one described.

Anyway these should just explain the situation we are in, but no excuses to act on your feedback.
Indeed there are a couple of things in our control and that we should focus on:

  1. Making the importer more reliable (counter is off and sometimes some urls cause issues), and also show the failed items, and more reliable error reporting so you know what happens if things get stuck.
  2. Improving the Backup app (its an old legacy piece that need some refactoring)

Thanks again for the straightforward feedback. That sunk in.

The database is too fragile for (supposedly) long-term data
#3

So, we’ve made some updates to the backup helper app.
You can get a new version here: https://github.com/WorldBrain/local-backup-server/releases/tag/0.5.0

Does it now work for you?

#5

@BlackForestBoi Thanks so much for addressing my points.

Using your numbering of my bugs:

  1. I hope this doesn’t happen to me often and I can rely on the backups to be restored properly (while we wait for your new outside-browser solution)

  2. As I mentioned in the backup thread , 0.5.2 solved my problem. Thanks for that! That was probably the most showstopper.

  3. As for the freezing issue, I followed the instructions you guided me to above, but that didn’t help. Chrome settings have changed a bit so maybe I’m not doing exactly what the instructions say. Here’s what I switched:
    image
    Does that look right?

#6
  1. We have some new leads to fix 1 and will attempt that next week.
  2. great happy - and thanks for the pressure
  3. can we have a call going through this together? calendly.com/worldbrain
    I have some ideas on how to investigate this, but I can also guide you here.
    A few questions:
  • How many items do you have to download?
  • how does the freezing manifest itelf? Does the whole tab freeze or just the download stops?
  • What’s the last item in the “failed items” list when the freezing happens? (you find that list by clicking on “Show error details” in the download process)
#7

@BlackForestBoi Sure! I submitted through calendly

To answer your questions: the download just stops at something like:

image
image.png740×557 35.8 KB

#8

@BlackForestBoi Hmm, I tried again and import of (browsing history, bookmarks, or Pocket) no longer freezes on me anymore. I guess one of your recent updates must have fixed the problem.
Thanks!

#9

We recently updated some of the importer logic that should have fixed a couple of issues. The freezing is still weird. Let me know if you run into this at some point again.

Are we still having the call to go through the restore problem together?

#10

Yes, sorry for the late reply