Remove stale frontend elements

[parkan]

A lot of the peerpad.net site components have become stale, were never implemented, or aren't well-aligned with the immediate goal of using the tool for internal notetaking. Let's cull those so we can focus on the core product:

• Landing Page: the landing page clearly had a lot of work put into it, but it's out of date (still talks about Y.js), doesn't bring much value to internal users, and creates an extra step before you get your pad. Other *pad products tend to only deploy a landing page of any kind of if they want you to pay or log in, otherwise booting to a new document is common (this can be prefilled with some explanatory copy, e.g. https://dillinger.io/). Let's bypass this page during internal testing and come back to it fresh afterwards.
• "PeerPad is Alpha" banner: important warning but well-understood by internal users (and kind of demoralizing), let's make it dismissable
• Files/Settings/Shortcuts: never implemented, can hide until and unless we build them out
• "Math Pad" mode: I guess this was meant to be a different editor UI (maybe like LaTeX/overleaf?) but never seems to have been implemented. Removing the mode choice also makes booting directly to a document unambiguous.

Most of these are just deletions in react templates or otherwise simple work, probably a couple of hours at the most. I know a lot of work has gone into the original prototype version, but it is no longer well-aligned with our immediate goals for the product.

Completing them will clear the path to #25, which will provide us with invaluable user feedback to inform further UX work, and a cleaner base for new designs on (#31).

[parkan]

instead of the landing page, it might be useful to have a simple "about" page linked from the footer that explains that this is no ordinary pad (could just be a link to a blog post, too)

[jimpick]

I agree that it would be good to simplify it a bit... especially if the primary target for now is just internal users.

I know we do want to keep it open for public users to also serve as a demo for peer-star-app. That's probably the biggest challenge for the messaging on the site. Should we be trying to funnel developers to IPFS / peer-star-app? Or should the messaging be more consumer-focused?

[parkan]

my view is that we should dial in performance and core UX before trying to convert external users

it's fine for the app to be publicly accessible for the curious developer, so any messaging should focus on that (though, again, I'm in favor of keeping it relatively minimal)

[victorb]

One option could be to leave peerpad.net as-is right now (production branch in the GitHub repository) and exclusively work from master branch. Master branch gets deployed to dev.peerpad.net, and there is where we can streamline it (open directly to new pad, remove verbose warnings and similar)

[victorb]

PR here: peer-base/peer-pad#238

[victorb]

PR is currently waiting review. In the meantime, you can review the new versions UI here: https://ipfs.io/ipfs/QmWMDrh8yHuyqhFYUvH4BJ27pKBbCUuKYzpP3KgRAH1QQV/ < checks all the boxes in the opening issue here

[victorb]

The PR has been merged now and the changes mentioned in this issue have been deployed to js.peerpad.net! Unfortunately, something is wrong with the certificates on that site, working on fixing that.

I'd like to have confirmation on the following:

One option could be to leave peerpad.net as-is right now (production branch in the GitHub repository) and exclusively work from master branch. Master branch gets deployed to dev.peerpad.net, and there is where we can streamline it (open directly to new pad, remove verbose warnings and similar)

^ would mean that for internal testing, we tell people to use dev.peerpad.net instead, and we leave peerpad.net in it's current state, for visitors to understand peerpad before using it.

[parkan]

do we have any analytics on the live site? I'm curious if it gets a meaningful amount of non-PL usage

[parkan]

The PR has been merged now and the changes mentioned in this issue have been deployed to js.peerpad.net! Unfortunately, something is wrong with the certificates on that site, working on fixing that.

js.peerpad.net doesn't seem to be running a web server, do you mean dev.peerpad.net?

would be nice to have a shorter redirect url for this, do we have an internal shortlinker?

I'd like to have confirmation on the following:
...
^ would mean that for internal testing, we tell people to use dev.peerpad.net instead, and we leave peerpad.net in it's current state, for visitors to understand peerpad before using it.

how about this: we can tag f29acedbb76ddca1f208df5fad7257a9c1585311 as "production" and leave it deployed on the main site with the explainer landing page, etc

I don't think it's worth maintaining it as an active branch/backporting work from master on an ongoing basis, unless we see significant usage (or a really bad vulnerability)

[victorb]

@parkan Ah, yes, I meant dev.peerpad.net. Don't think it's long enough to warrant a shorter link...

My suggestion is to leave the production as-is today, which means being locked to commit peer-base/peer-pad@2727041

Agree with not worth backporting fixes unless we see high impact security issues.

[parkan]

oops, you're right, I had the wrong parent, 2727041 is right

reasoning for shortlink (or more unique prefix) is to minimize keystrokes to load the site, there's probably a lot of dev... prefixes in most people's histories

not a huge deal tho! marking this as done

[victorb]

Yeah, my fear about link shorteners is that eventually they die, and all the redirects die with them. Usually just better to use the proper link directly. We can add aliases to test.peerpad.net or whatever in the future.

Missed this question before:

do we have any analytics on the live site? I'm curious if it gets a meaningful amount of non-PL usage

AFAIK, we don't. We do have Rollbar but that only gives us the errors.

Once the pinner has metrics, we can get some basic information from that, first total peers and total collaborations, in the future we can expand it to contain more information.

[parkan]

sgtm