A connected user editing can send another user’s cursor to the top of the document

[codingisacopingstrategy]

I will have to do more research if this occurs always and with all browsers—but we’ve been running into this problem a lot.

It is kind of hard to test though, because you need to be with more people!

[enjalot]

interesting. what if we setup a client-side "monkey typewriter" mode that
would type random characters at a relatively steady rate.

does this happen with the pad example in racer or share? i've worked with
codemirror cursor control a bit, if we are replacing ranges of text we
might want to explicitly update the cursor's position accordingly
I did this with my Inlet project
https://github.com/enjalot/Inlet/blob/master/src/inlet.js#L47
example usage (click on a number or a hexcode to change it rapidly)
http://enjalot.github.io/Inlet/

On Sun, Feb 22, 2015 at 2:26 PM, Eric Schrijver [email protected]
wrote:

I will have to do more research if this occurs always and with all
browsers—but we’ve been running into this problem a lot.

It is kind of hard to test though, because you need to be with more people!

—
Reply to this email directly or view it on GitHub
#3.

Ian Johnson - 周彦
http://enja.org

[codingisacopingstrategy]

I’m all for the monkey!
Though I will have to look at this later given my timezone.

Cheers!E

[codingisacopingstrategy]

I fooled a bit around trying to set up a typewriter monkey, but had trouble synthesizing the keyboard events in JavaScript. Do you have an idea of what would the best approach for that?

[enjalot]

we could skip key events and directly set model data. this would
effectively simulate another browser typing.

btw i changed the logic to use setDiffDeep which doesnt do string diffs.
this means the updates are less granular but it solved some race conditions
i was seeing with codemirror auto formatting.
On Apr 29, 2015 8:33 AM, "Eric Schrijver" [email protected] wrote:

I fooled a bit around trying to set up a typewriter monkey, but had
trouble synthesizing the keyboard events. Do you have an idea of what would
the best approach for that?

—
Reply to this email directly or view it on GitHub
#3 (comment).