Paging down or up does not page far enough

[grahamperrin]

At https://riot.im/develop/#/room/#riot:matrix.org/ for example, before paging down:

– the last visible line, which I read, was:

get rid of count at a point that everything is muted)

After keying Page Down:

– I have pixelised the four paragraphs, plus the name of the user, that certainly need not be seen twice.

I guess that it's also unnecessary to show, twice, the other lines that are immediately above

get rid of count at a point that everything is muted)

I reckon, the last line that is entirely visible before paging should be the first line that is visible after paging.

[lampholder]

I'd expect that there's a defined behaviour dictating the amount of "context" we see from the previous screenful after a page up/down.

If that context were defined as a number of messages rather than a number of lines, a large message could make the end result look funny.

[lampholder]

I called this a bug at first, but I had a change of heart - to my mind this is really a suggestion of an enhancement to the current behaviour, which is probably just suboptimal for large messages.

[lampholder]

I think that highlights the issue pretty clearly.

Gitter's page jump is the whole page minus ~2 lines; what was the bottom line becomes the second-from-top line.

Slack on the other hand does scroll the whole page - what was the bottom line becomes the first line off the top of the screen. It animates it pageup/down scroll, though, whereas gitter and riot are instantaneous.

Riot's scroll seems more like half a page (which, on reflection, I actually find makes it very hard to maintain good awareness of where I am as I page up/down because I'm expecting a larger jump).

Two thoughts occur to me:

• gitter's jump size seems more natural
• I think we could generally have more flexibility if we animated the jump - it makes it soooo much easier to follow along

So maybe:

• impl the animation
• increase the jump size to match gitter's

@ara4n - do you have strong opinions on this?

[grahamperrin]

Thanks!

I'm not sure about animation.

I sometimes use the keyboard to page down very rapidly. Might the animation feel contentious in that case?

---

Firefox 56 with the Scroll Position Highlight extension: https://youtu.be/QcPR35noYbs

Using a legacy extension to partially ease a keyboard-oriented issue with the Riot client for Matrix, …

[lampholder]

I just tried it in slack and found it comfortable to scroll quickly with the animation.

Then I loaded a long wikipedia page in Chrome (I picked https://en.wikipedia.org/wiki/List_of_Advanced_Dungeons_%26_Dragons_2nd_edition_monsters) and page up/downed that and discovered that Chrome animates its page up/downs anyway, so my guess is that it would be fine even when paging down at speed :)

[elfez]

It would really nice to see this addressed. I find it really jarring when reading to have to search where on the page I should continue reading from. It's fairly unintuitive that PgDn actually means Half-PgDn in this context.

[grahamperrin]

@claell thanks for looking at this.

---

From #8455

… almost impossible to scroll through them slowly …

This #3763 is debatably worse. Beyond almost impossible, it's consistently impossible to page normally :-(

Incentives

I'm begging for a fix and will gleefully throw … let's say £200 at a developer.

Is Bountysource still a popular option, or is something else likely to be more attractive? (The array at e.g. https://en.wikipedia.org/wiki/Category:Crowdfunding_platforms is dizzying.)

Patreon, maybe?

TIA

---

Postscript (for future reference): https://wiki.snowdrift.coop/market-research/other-crowdfunding