Talk:Microgrants/Microgrants 2020/Proposal/Potlatch 2 for desktop

From OpenStreetMap Wiki
Jump to navigation Jump to search

Potlatch 1 sunset?

Just wondering if this implies a sunset of Potlatch 1 by the end of the year, meaning we can remove amf controllers on the api side, and the (few) mappers will no longer be able to do an undelete? Mmd (talk) 20:11, 10 May 2020 (UTC)

-- Yes, absolutely. It might focus minds a bit on recreating the undelete functionality elsewhere! Richard (talk) 22:27, 10 May 2020 (UTC)

What a pity, PL1’s undelete ways was a key feature for so many years (although you can now find all kind of deleted objects with the overpass attic data function and josm lets you undelete objects, it still was dead simple and convenient for ways with PL).—Dieterdreist (talk) 20:32, 30 July 2020 (UTC)
Undelete is one of the topics of the Potlatch sunset issue: https://github.com/openstreetmap/openstreetmap-website/issues/2622 - although it's really unlikely that someone will implement it. Mmd (talk) 21:28, 30 July 2020 (UTC)

Clarification

"It continues to have features not available in other editors, and a user interface that for many tasks is more efficient.". I am curious as someone who never used Potlatch - what kind of tasks are much easier to do in Potlatch than any other tool?

For reference, I used JOSM, Vespucci, StreetComplete, custom-modified StreetComplete version, level0, custom scripts, MAPS.ME, iD, JOSM plugins (and maybe something else) and every tool had some unique strengths. But sadly I am unaware why I would want to use Potlatch and I wonder what I am missing.

Sadly Potlatch 2 page is not (yet) describing this unique strengths.

Mateusz Konieczny (talk) 09:50, 15 May 2020 (UTC)

There's a bunch of distinct functionality, but if I had to choose one thing, I think P2's distinguishing characteristic is that it's extraordinarily ergonomic for moderately familiar users, without ever having too complex an interface. With a handful of keyboard shortcuts you can zip about the map and perform common editing actions efficiently: for example, you can skip 10 nodes back/forward along a way with one keypress, jump to the other end of the way with another, then assign a set of multiple tags to the way (which you've previously recorded) with another single keypress. The TIGER fixup work I do, for example, would likely be significantly slower in another editor. It's particularly well suited to rural editing, probably less so to urban areas. --Richard (talk) 23:09, 16 May 2020 (UTC)
GPX waypoint support in P2 is significantly better than in other editors. Much of the edits I do are survey-based, and in each case all the details of what I want to update will be in GPX waypoints. I've asked a few "how to do X in JOSM" questions in the past such as https://help.openstreetmap.org/questions/6368/in-josm-is-it-possible-to-see-gpx-track-waypoint-details , but sometimes the answer is "you can't". SomeoneElse (talk) 14:34, 31 July 2020 (UTC)
Late to this, but Potlatch2's approach to tasks works rather better than Josm when reviewing muany items only a few of which may need editing. This is because one can return to previous items (something needed when reviewing a large number of items as one's mouse clicking finger can be slightly ahead of brain in processing the images). Miss something in a 2-300 item list in JOSM and either one has to reload the tasks or redo a pass once the first one is finished. SK53 (talk) 16:52, 12 December 2020 (UTC)

Can they really forbid using flash?

I mean, there will be some hack to enable it if you really need it, right? Maybe installing an older version of browser or something?

--SviMik (talk) 11:29, 15 May 2020 (UTC)

Why would you want to run old unsupported software with security issues left wide open? Doesn't really make much sense if you think about it for a minute. Mmd (talk) 11:42, 15 May 2020 (UTC)
Like I said, there are situations when you really need it. By your logic you wouldn't use a toilet if you see that the door lock doesn't have 2-factor authentication. Your comment doesn't really make much sense either. Not to mention we are getting off topic. --SviMik (talk) 12:37, 15 May 2020 (UTC)
Well, you could of course run Flash on your local machine, but it won't make any difference. Both web based Potlatch editors will be removed from osm.org by the end of the year. Then you're left to run those swf files somewhere on your own. All of which will be unsupported. Mmd (talk) 12:46, 15 May 2020 (UTC)

€500/day?

I'm sorry to be blunt, but... Are you from Silicon Valley or something? Also, if it takes just 5 days - it should be doable on weekends if you actually like this editor and not trying to make a profit out of the situation. TwilightOwl (talk) 01:26, 21 May 2020 (UTC)

This is a) an entirely reasonable day rate for a skilled contractor in the UK market (approximates to GBP 450 or a bit less: I was billing at slightly more than this rate in 1992!); b) Richard has invested vast amounts of his free time to building the editor in the first place & this is necessarily a sustained piece of work, not a series of small incremental patches, so is not appropriate for a weekend project; c) the work is time critical, it needs to be completed before the end of the year, and thus needs to be insulated from competing demands of Richard's other commercial work which pays the bills. SK53 (talk) 12:54, 2 August 2020 (UTC)

WebAssembly

I know this proposal is based on a desktop app. However, did you by chance evaluate if Potlatch could get ported to run on WebAssembly instead? That might be a better option for people who just want to start their editor and don’t bother with local installs. Time investment would for sure be significantly higher, though. Mmd (talk) 15:07, 21 May 2020 (UTC)