Hacker News Reader

Comment by Arathorn | Hacker News Reader

Speaking as the project lead for Matrix (and Element) - we are doing everything we can to improve Element's UX. If you're interested in the details, go watch the first 5 minutes of this week's Matrix Live (our weekly podcast), where Amandine and I spell out how important this is; it's literally an existential threat to the project; and what we're doing to fix it: https://matrix.org/blog/2020/11/13/this-week-in-matrix-2020-...

Our goal is of course to use it as a replacement to WhatsApp / Telegram / Discord for social stuff, even though it's dominated by developers today. And we will get there (despite all the negativity on this thread :|)

View on HN

I really like Element. And I tried to get my friends away from WhatsApp. Over half of the people asked me how to send voice messages. I came across this ticket while searching for the feature: https://github.com/vector-im/element-web/issues/1358 From then on, it was clear to me that I could stop my attempts to bring people to Matrix. To be honest, the discussion in the ticket frustrated me deeply. The ticket is open since 2016 and the topics discussed in it are not oriented towards the users of this world, but towards some almost completely uninteresting technical details. Over the years, several people have intensively pointed out how important this feature is. Those responsible must pull themselves together and find a solution for it.

> Over half of the people asked me how to send voice messages.

This is a very regionally desirable feature, in my experience.

I have never (deliberately) sent a voice message on any platform, nor do I know anybody (well/any more) who would. But I went to a university with a high proportion of Chinese students, and it was really striking how much that feature was used (in apps that supported it obviously).

I'm just speculating, but I wondered if it was at least partly because of the tonal nature (roughly speaking: four intonations for each vowel, rendering completely different words, not just subtle changes in tone/emphasis) of mandarin, and relatively large number of different characters, making vocal communication more efficient than written; to an even greater extent than that may be true of English, for example.

I'm not really sure what I'm saying. Maybe it's not good to burn one market, maybe it is good to focus on excelling in others if (if!) the alternative is mediocrity everywhere, I don't know. Your comment just reminded me and I thought it was vaguely interesting.

Do you message older people? I interact with multiple and they find audio messages to be indispensable. (Who can blame them when recording a voice message is vastly easier than getting your glasses and typing out a message on a small screen with even beady on-screen buttons). In addition to old people, I also message a disabled person to whom speech and recorded messages are her primary mode of communication.

In short, I don't think it's regional. It depends on the person with whom you need to communicate.

Yes, and no older people use it here in Sweden. I do not know anyone in Swden who uses voice messages.

My grandmother has an iPhone and my parents would be baffled if she sent them a voice message.

I'm not denying it's helpful/required for accessibility, I'm just talking about 'in general' large scale use of it (or not).

Maybe it's less about glasses (if you have to find your glasses to type a message I expect you also need to in order to unlock your phone, open the app, and tap the voice message button) for older people than it is about accustom to IM - the brief chatty format is more like a telephone call than a letter.

(But those I've known to use smartphones haven't used voice messages - as distinct from telephone voicemail - anyway, as I said.)

I'm not doubting your (or your sibling's) claims, but I find it extremely difficult to imagine any older person preferring a virtual keyboard to voice messages (provided they are properly introduced to the feature, of course). Maybe this preference stems from how simple telegram (the app we use) makes taking and hearing voice messages? Perhaps it really is a regional thing after all. We need more data.

From my anecdotal/regional experience, people over 60 prefer to use voice messages almost exclusively while younger people prefer to text.

Maybe someone at telegram or whatsapp can provide statistical data to shed more light on voice message usage.

I have a family member with bad internet reception where she lives. Realtime voice communication usually cuts out, so voice messages are a really good alternative because they are uploaded at once and can be listened to on demand.

Voice clips are extremely used in South America. I'd say some people even send more voice clips than text messages.

Any app lacking them has no chance of gaining traction there.

I think it's because we dumped SMS (which has no voice clips) early and moved to WhatsApp (I think it's been the norm for like 8 years now in Argentina).

Yeah, you're right, it's different. One half does not use voice messages, the other half cannot live without this feature. My peer group is mostly from Germany.

I think it depends on your style of using devices. I am on my PC 99% of times, so I have access to a full-sized keyboard, so I can type longer messages with no issue. But on a phone, typing anything longer than few sentences is a miserable experience, so people prefer to just use voice messages and say what they want to say.

> I have never (deliberately) sent a voice message on any platform

same here.

really dislike how people essentially just ramble on and on for what could have been one sentece or a few words.

so it costs the recipient a lot of time and concentration to get the gist of the message.

maybe a cultural thing because my wife and here friends use voice msg a lot.

I tried to use Matrix with a few friends, and while it's impressive how far it's come, it's not there yet. The features are all great, now "all" that's needed is to make them work. We had jankiness like keying problems, messages that wouldn't send or receive, and general problems that made it not "just work".

My personal pet peeve is that I have no idea what the difference is between "people" and "rooms", and why when I try to talk to a friend our chat is sometimes in one and sometimes in the other. Maybe you'd like to take a page out of the Telegram/Whatsapp book and not distinguish between the two, since they're all chats, they just sometimes have two people and sometimes more.

Best of luck, I'm really hoping to switch to Element for all my chats one day!

I appreciate all the work that you and your team have put into Element and Matrix over the last few years. While I still think there's plenty to go, I think there's been clear communication about what the project has worked on, what you're trying to do with it, and acceptance that certain features, while important, aren't as important as more core features. (My personal pet one is custom emoji, MSC1951).

I still think that this is one of the more important open-source future-facing projects around and will continue supporting it, and urge others to, as well. If you dislike the hegemony of centralized services, then throw some money towards projects that support ideals you believe in.

Thanks :) In terms of custom emoji, it may be worth checking out our new public roadmap for Element (which is new this week and hasn't yet been announced): https://github.com/vector-im/roadmap/projects/1. On the plus side, custom emoji are on the roadmap. On the minus side, they're on the 'Later' column - so you can see where they stack up relative to other work. Agreed that they would make a massive difference. We will get there eventually though (or perhaps folks will contribute them sooner :)

I'd like to agree with the comment above yours that how communicative you are being and the openness about issues combined with the transparency of the project really raises my and probably many others' opinion of the project and that if you continue this way, it's really looking bright!

I think my anecdotal experience with getting casual users to switch to Element might be helpful to you:

I am a member of a social group that varies widly in age and technical experience. When I tried switching two of our members over to Element, I noticed them quickly growing upset with the app because the found it to be missing key features they expected like easily recording audio messages or the visual seperation of messages that are more than a few minutes apart.

From my personal experience with switching people from one chat app to another, people expect a replacement app to:

1. have 1:1 feature parity with the app it's replacing, 2. display all the information they've grown to rely upon, 3. and both feel and work like the app it's replacing. (even cosmetic things like text bubbles and larger font sizes)

Failing the first two points lead to the immediate dismissal of the app. Failing the last point is seen as a deficiency in the new app. Inversely, succeeding in all three leads to smooth adoption. It is why all my group is using telegram right now instead of whatsapp.

I'll have another go at switching my group to Element when/if the app gains the ability to send audio messages, but without Element looking and feeling similar its compition, I'm skeptical of my chances to succeed.

By the way, I really like Matrix and appreciate the hard work and care you're putting into this project. Thank you and keep up the good work.

The biggest grab for Discord is anonymous accounts. You can put in a username (even a conflicting one due to their #number system) and get going straight away.

Sign-up is a metric that your marketing team loves to track, but is a crazy user bottleneck.

Even if you only do it for, e.g. 6 months-1 year (spam concerns), let people chat with no account/sign-up. Let me give people a link that will let them chat with me, with no sign-up.

I agree. Users want to jump into a discussion first and foremost. Discord facilitates this without adding any unnecessary friction.

Matrix has always supported "Guest" accounts, where a user can immediately join a room and start talking in there with a generated user_id. They have limitations, but for the "get going straight away" factor it was quite nice.

Element (back when we called it Riot) removed support for this a while ago BUT we are trying to design it back into the app, which you can follow https://github.com/vector-im/element-web/issues/9264

And yeah as a developer on the project, getting users into rooms quickly and chatting is a desperate goal of mine.

I'm glad to hear that. You guys really seem to be willing to be wrong, and seriously listen to ideas. If you fail you will color me the strongest shade of surprised: everything you are doing is optimal.

Really sorry to see all the negativity. I use Matrix every day and don't get most of what people are complaining about. For some reason lots of HN threads on it seem to hate it irrationally, despite Matrix by far and away being the most advanced protocol and implementation in its space.

"The most advanced protocol" counts for fuck-all if the actual client is missing features compared to the competition.

Whatsapp: 1) Install 2) it works. Even my 60+ relatives can handle it.

Discord: 1) install 2) it works. It also has voice/video chat rooms and a huge ecosystem of bots.

Telegram: Works, has bots, channels can be moderated, sticker packs, gifs etc fun stuff.

Whatsapp: Sends all phone numbers on your phone to a third party.

Discord: Works as long as Discord does not run out of VC money.

Telegram: Sends all phone numbers on your phone to a third party.

What is your point? That Matrix can never be popular as a phone messenging app because it doesn't collect phone numbers, and doesn't raise VC capital?

All three of those services are centralized though. I'm talking about things I can host.

Whatsapp is the worst chat system I use. It doesn't support reactions (which are essential if you don't want a chat with many people to devolve into incessant pointless notifications), nor does it allow easy editting of the message you just sent (another essential feature, given how often small mistakes can totally change the meaning of text). I can't collapse pictures/gifs that annoy me, I can't chat with people on other services. It won't let me use whatsapp web on multiple other devices simultaneously, it won't let me use whatsapp web when my phone is out of battery (the exact time I most need it!). I can't invite people to chat with me on whatsapp without them having to create an account and give up their phone number (and usually their contact address book too).

It's very strange to me to hear people describe Whatsapp as having good UX.

WhatsApp has good onboarding and brand recognition. (YMMV, this is highly dependent on which country you are in). You don't need to create accounts or register. You just need to know your phone number.

Having to give up your phone number and contacts is a feature for most people. They can instantly see that Marge from the office also uses WhatsApp so they can call/message them using that instead of paying for calls and SMS.

The limitations of WhatsApp web are mostly due to it being end to end encrypted. The "web client" is basically just using the phone as a proxy. It _is_ possible to have E2E encryption and allow multiple separate clients, but it's not a trivial problem to solve, thus most won't do it. (Telegram's E2E chats are device to device and not available on other clients than the one that initiated the connection.)

It's still crap, don't get me wrong. I'd rather use anything else, but at least where I live, it has critical mass and is the de facto standard.

I don't disagree with what you've written, but a chunk of it is exactly what people are getting criticised for elsewhere in this thread - giving a technical reason that I as a user probably don't care about for bad UX.

> it has critical mass and is the de facto standard.

Yeah, it's the critical mass thing that is the most significant reason here, although it originally caught on back when simple onboarding was very unusual and that was its killer feature.

I would love to have a easy and fast deployable Element server for my companies internal network. IT tried it twice already and failed somewhere along the way which took them too long to fix the problems they had with it. Mattermosts Omnibus installation is something we'd be looking for.

If your IT group is alright with docker, the matrix-docker-ansible-deploy[1] is excellent quality and very quick and easy to setup/manage.

[1]: https://github.com/spantaleev/matrix-docker-ansible-deploy

I’m don’t mean this in a mean way, but if I were you that would not give me a great deal of confidence in my IT department’s level of talent. I’m assuming you are talking about setting up a Synapse server, not the Element web client.

I’ll admit, the documentation isn’t perfect. But it’s pretty good, and they should be able to set it up. There’s a few things I think could be more explicit in the docs, but even a mediocre set of sys-admin skills should be able to fill in the gaps.

I’m not an IT guy, or a sys-admin. I’m just a lowly, mid-level web developer :) But I was recently able to successfully setup my own Synapse server as per the docs, the very first attempt.

However, to your point. A more streamlined, or ‘automatic’ installation option would be fantastic!

In my case, it’s because I heard (last year) that e2e was a planned feature, and not implemented yet. Is that information out of date?

Also wary of it wanting to upload my contacts list - could you comment on how it’s used? Or is that entirely up to the specific app?

E2EE has been implemented since 2016, and turned on by default as of May 2020.

Uploading contacts is strictly optional (and is missing on some platforms like Android currently, as per complaints elsewhere on this thread). They're used purely to discover which of your contacts are already on the platform - the privacy policy can be found at https://github.com/vector-im/policies/blob/master/docs/ident...

Reply on news.ycombinator.com