@Hubzilla Support Forum
Just wanting to give a short feedback about the results of me pitching hubzilla to the switter admins.
First of all it failed .
They didn't liked the UI and found it in all to complicated.
That they could change it based on their needs they've been a bit reserved about.
One of their main argument also was:
That the channels can't have their individual branding within the url.
They wanted myadress.blabla instead of hubzilladress.org/channel/@yourname
But it's actually possible to have this, or?
Just wanting to give a short feedback about the results of me pitching hubzilla to the switter admins.
First of all it failed .
They didn't liked the UI and found it in all to complicated.
That they could change it based on their needs they've been a bit reserved about.
One of their main argument also was:
That the channels can't have their individual branding within the url.
They wanted myadress.blabla instead of hubzilladress.org/channel/@yourname
But it's actually possible to have this, or?
show all 32 comments
I have to agree with their UI position, what we have is functional, but not easy to use or inspiring.
I've been thinking that a new UI would be the way forward, ignoring redbasic and starting from scratch with more modern UI paradigms.
Here's an example of what I mean
There are a number of prototyping tools available to build something cool.
Are there any workflow diagrams concering Hubzilla?
I've been thinking that a new UI would be the way forward, ignoring redbasic and starting from scratch with more modern UI paradigms.
Here's an example of what I mean
There are a number of prototyping tools available to build something cool.
Are there any workflow diagrams concering Hubzilla?
hmm, about 'modern' I'm always a bit reserved. It's for me a metaoher for,a what's new and trending.
With what I agree is that we don't need to limit the creation of a different UI to the current status quo.
The example, even so it might be modern, I don't like very much, at least when viewing it with mobile.
Ah, also that might be an interesting sidenode.
One of the most famous german blogs looks like this:
https://blog.fefe.de/
With what I agree is that we don't need to limit the creation of a different UI to the current status quo.
The example, even so it might be modern, I don't like very much, at least when viewing it with mobile.
Ah, also that might be an interesting sidenode.
One of the most famous german blogs looks like this:
https://blog.fefe.de/
Personally I love minimal styles and fixed fonts, but the question is: do we want to address a wider public?
And that is always the question. It seems to me that right now is a great time to encourage people to migrate, and an easier UI for newcomers could be a good gateway method.
I guess this HomeSocial theme looks reasonably modern and could be updated to be fully compatible with Hubzilla 3.2+
https://gitlab.com/zot/hubzilla-themes
https://gitlab.com/zot/hubzilla-themes
Thanks for reaching out @paulfree14 I agree that the UI needs some work to be less confusing to newbies. For example, it worries me when I type "@paulfree14", and "@paulfree14+[number]" is substituted. Is that a bug? What does that number mean?
I've promoted my sexworker forums a fair bit on sex worker forums on reddit and switter. But I don't think many actual sex workers have posted: it's mostly been Hubzilla advocates posting about how to reach out to sex workers:
https://gerzilla.de/channel/sexworkers
(provider-side)
https://gerzilla.de/channel/gentlemensclub
(client-side)
I think part of the issue is that it's hard to get people to pay attention to anything new. They're busy with their own lives, and they don't want to spend a lot of time learning the UI of some new forum.
I think that it's also difficult to communicate "nomadic identity"--it's easy to get bogged down in confusing details, at least for me.
Here's some ides for making HZ more appealing to sex workers:
* Make a sex worker theme for Hubzilla, optimized for their needs.
* Recruit some sex workers to test the new theme, adjusting the UI based on their feedback.
* Once you're happy with the theme, then reach out to some sex worker thought leaders, like Liara Roux. If well-known, well-respected sex workers start using it, then others may follow.
* Reach out to sex worker orgs, such as SWOP, BaySWAN, and try to recruit them to use it. If respected sex worker non-profits promote it, then sex workers may be more likely to use it.
* Hire a UI/UX expert to do mockup.
* Do member polls, asking what UI feature was most confusing. Fix the ones that are common problems.
I've promoted my sexworker forums a fair bit on sex worker forums on reddit and switter. But I don't think many actual sex workers have posted: it's mostly been Hubzilla advocates posting about how to reach out to sex workers:
https://gerzilla.de/channel/sexworkers
(provider-side)
https://gerzilla.de/channel/gentlemensclub
(client-side)
I think part of the issue is that it's hard to get people to pay attention to anything new. They're busy with their own lives, and they don't want to spend a lot of time learning the UI of some new forum.
I think that it's also difficult to communicate "nomadic identity"--it's easy to get bogged down in confusing details, at least for me.
Here's some ides for making HZ more appealing to sex workers:
* Make a sex worker theme for Hubzilla, optimized for their needs.
* Recruit some sex workers to test the new theme, adjusting the UI based on their feedback.
* Once you're happy with the theme, then reach out to some sex worker thought leaders, like Liara Roux. If well-known, well-respected sex workers start using it, then others may follow.
* Reach out to sex worker orgs, such as SWOP, BaySWAN, and try to recruit them to use it. If respected sex worker non-profits promote it, then sex workers may be more likely to use it.
* Hire a UI/UX expert to do mockup.
* Do member polls, asking what UI feature was most confusing. Fix the ones that are common problems.
They wanted myadress.blabla instead of hubzilladress.org/channel/@yourname
But it's actually possible to have this, or?
Nope.
Well, Hubzilla is a decentralized pro tool for securely sharing digital content throughout the internet. It might be not what they are looking for.
It's true that the UI/UX still isn't perfect, but we will get there eventually
@Cat Gun I believe what you describe is a strategy that has a high chance of being succesfull.
Just speaking for me, it's a way over my skills.
On the same I'm even kind of anti-comercial. Which means that my personal interest isn't very high at all in that subject (beside countering the discrimination)
I believe(d) that hubzilla could be of great use for sexworkers, regarding to counter the opression comming from SESTA/FOSTA. And that was what motivated me tryn to introduce it. It seems they are prefering using a different plattform. So they are still countering SESTA/FOSTA but with using something else then hubzilla. What plattform they use actually doesn't matter to me (as long it's free software ).
that's a very good idea and in general usefull.
I've got some connection to fuckforforrest, an 'ecoporn' collective. I'll see if I get a chat somewhen the next days with one of them. Maybe they've got more interest, also with hubzilla's current state.
But in general if I'm getting time/energy to do a larger campaign I'd be focusing on getting 'all' the 'major' anarchist/left activists/journalists from german speaking countries etc. to join and let it create by that a network effect that will spread wide behond such circles.
Just speaking for me, it's a way over my skills.
On the same I'm even kind of anti-comercial. Which means that my personal interest isn't very high at all in that subject (beside countering the discrimination)
I believe(d) that hubzilla could be of great use for sexworkers, regarding to counter the opression comming from SESTA/FOSTA. And that was what motivated me tryn to introduce it. It seems they are prefering using a different plattform. So they are still countering SESTA/FOSTA but with using something else then hubzilla. What plattform they use actually doesn't matter to me (as long it's free software ).
Do member polls, asking what UI feature was most confusing. Fix the ones that are common problems.
that's a very good idea and in general usefull.
I've got some connection to fuckforforrest, an 'ecoporn' collective. I'll see if I get a chat somewhen the next days with one of them. Maybe they've got more interest, also with hubzilla's current state.
But in general if I'm getting time/energy to do a larger campaign I'd be focusing on getting 'all' the 'major' anarchist/left activists/journalists from german speaking countries etc. to join and let it create by that a network effect that will spread wide behond such circles.
Yeah seems there is a lot of interest in hubzilla in recent weeks. I myself decided to give it a try and guess what. I fell inlove I dont know why every time throughout the years i tried it I always run away too quick.
Well one reason is... UI
Nice to see more people are looking into it. We're also trying to come up with a nice looking theme before our hub goes live. Thanks for the links I'll check whats out there already.
I think our initial idea is to base it on redbasic with possibility for people to adjust some settings (just like they can on original redbasic) but try to make it look more slick and easy for new comers. Later we would probably like to re-do the entire UI based on some framework like @Galette Saucisse suggested.
BTW
how do I properly mention people?
Well one reason is... UI
Nice to see more people are looking into it. We're also trying to come up with a nice looking theme before our hub goes live. Thanks for the links I'll check whats out there already.
I think our initial idea is to base it on redbasic with possibility for people to adjust some settings (just like they can on original redbasic) but try to make it look more slick and easy for new comers. Later we would probably like to re-do the entire UI based on some framework like @Galette Saucisse suggested.
BTW
how do I properly mention people?
@muppeth, if you're connected to them, you can use the "@[name]" notation. Type the @ symbol, and a popup will appear with names that you can select from as you type.
If you're not connected to them, the "@" notation doesn't work. I'm not sure the best way to mention someone in that case.
If you're not connected to them, the "@" notation doesn't work. I'm not sure the best way to mention someone in that case.
@Christopher M. Rasch I see... I hope it can be fixed. It took diaspora quite few years to be able to mention poeple properly in comments.
This is not a bug. You can only mention the people that you are connected to, because it requires a permission. When you mention someone, you want them to get a special notification, and that kind of communication is something that the mentioned individual might not want to allow.
Yes understand its not a bug, though it would be nice to be able to mention people in the same thread. makes it easier to notify people someone he was discussing something in the same tread mentioned him. Without the need to connect ot him.
I like that hubzilla as oppose to other federated platforms generates real and long discussions under posts. In threads that grow up to 50 and more posts, being able to mention someone would make it easier to answer directly, to someone.
I like that hubzilla as oppose to other federated platforms generates real and long discussions under posts. In threads that grow up to 50 and more posts, being able to mention someone would make it easier to answer directly, to someone.
I've been working on refactoring mentions the last couple of days. We can definitely do better. Our autocomplete library may need fixing or replacement before everybody is happy with it. The problem is scalability. If you try and autocomplete 'david' and there are two million david's known to the system, you need to truncate the results. Say you truncate it to 100K entries (currently it's set a lot lower than this). As you continue to type, the current autocomplete library never queries again to get more entries with the more complete search text. It just truncates its "in memory" list. So you'll never see the two millionth David even if you match his full name exactly.
This is why we normally restrict mentions to connections. The returned list of matches is then finite and we don't have to be worried about browser memory limits. No matter what we do, we want the results to be ordered with your friends on top of the list. There is also a hidden option to tag anyone, but it has some usability issues such as I've described so remains a hidden option until somebody works through the issues.
This is why we normally restrict mentions to connections. The returned list of matches is then finite and we don't have to be worried about browser memory limits. No matter what we do, we want the results to be ordered with your friends on top of the list. There is also a hidden option to tag anyone, but it has some usability issues such as I've described so remains a hidden option until somebody works through the issues.
I would also support allowing mentions of someone in the current thread, unless they've explicitly blocked you. (Of course, I understand that may be easier said than done.)
I see what you mean about being able to mention people already in a thread, @muppeth (case in point!). And I can imagine plenty of other cases where mentioning people you are not connected to would be useful and acceptable.
Mentioning is now highly improved on upstream code and will migrate to Hubzilla in due course. I did have to remove the autocomplete cache, which involves a tradeoff. The results will now be more accurate and can include anybody your system knows about, but it may not be as responsive on performance challenged hardware because it will hit the server for fresh results every time you type a character.
Any way it can use some sort of buffered cache (for lack of a better term)? Basically, type a letter, it sends off the request but will search based on the last result-set received until a new result set is received. If there is a result-set in transit, block requesting a new one, but build the request string for a new one in the interim. Once a new result set is received, rerun the search (if the user hasn't selected an answer) and fire-off the next request.
On fast systems, it should be nice and zippy. On slower systems, at least it won't get bound up with a new request if it hasn't finished returning the last one.
For me, my biggest problem isn't usually the speed of the hub, the speed of the link between me and the hub and high latency (I live in the middle of nowhere and use fixed wireless which is shared with who knows how many neighbors --- I have in all honesty seen ping times of 75000+ms.) If the UI is going to essentially freeze with every new character when I'm trying to do a mention, it's going to be painful.
On fast systems, it should be nice and zippy. On slower systems, at least it won't get bound up with a new request if it hasn't finished returning the last one.
For me, my biggest problem isn't usually the speed of the hub, the speed of the link between me and the hub and high latency (I live in the middle of nowhere and use fixed wireless which is shared with who knows how many neighbors --- I have in all honesty seen ping times of 75000+ms.) If the UI is going to essentially freeze with every new character when I'm trying to do a mention, it's going to be painful.
I commented out the existing query cache stuff, but it's still there if you want to figure out how to make it work. The developer who did the work originally is long gone and I have some difficulty following exactly what he was trying to do here. All I know is that it only makes one query for 100 items max based on the first character typed and never asks again. view/js/autocomplete.js starting at line 10. The autocomplete library also has its own cache mechanism and a way to lock requests. I haven't tried to see how that cache behaves and whether it is any better or worse.
Thanks for the pointers... I'll add it to my list - if it gets annoying enough, I may even do something about it. That is, unless someone else gets to it first!
Be careful what you wish for. At work, we have 211 000 members and in some dialogs you get to autocomplete search among all of them. It takes a loooong time, eeeeevry time, and that is.... annoying. As Mike wrote, there are hidden browser limits, they are at around 50000 results before you need to search again. For us, these hidden limits gave inconsistent results, like the person you were looking for were in the result some of the time, so we dropped them in some places. While Mike probably irons out all such stuff because he is really smart, I am not sure searching for non-contacts and at-tagging them is smart.
I am not sure searching for non-contacts and at-tagging them is smart.
I've had "tag_anyone" enabled for a few years now. I'm well versed in its problems.
When Umberto Eco wrote "The Name of the Rose", he chose to make the first 100 pages very difficult. He thought that they should be the monastery-wall which should protect his novel.
James Joyce did something similar in Ulysses. Perhaps new themes should be made with expanded literary allusions? "Easy Peasy Lemon Squeezy", "The Name of the Rose", "Ulysses", "The Mahābhārata"... The UX would be terrible, but nobody would dare to leave the party from fear of missing out.
#I am joking
#I am joking
Just to get this thread offtopic again, i have just pushed the latest version of textcomplete (jquery.textcomplete is discontinued) to dev. It has built in caching and it seems to work fairly well as far as i can see atm. There might be regressions though.
no. there's a new law in the US that is discriminating sexworker.
switter is a plattform that tries to counter this. They are using mastodon as a software.
switter is a plattform that tries to counter this. They are using mastodon as a software.