Re: Matrix IRC bridge considered harmful

[Matthew Hodgson via desktop-devel-list]

On 13/02/2020 14:27, Neil McGovern wrote:


On Wed, 2020-02-12 at 20:21 +0100, Alexandre Franke wrote:

My concern would be the "federal" nature of matrix where people
don't need a
gnome.org specific chat account to join a room. Whilst there are a
lot of
arguments for this I'm increasingly convinced it's an anti-feature
especially
if we want to enforce CoC (which, of course, we do)
  
That was a concern for Mozilla too. I don’t know the details, but

they have a solution for that it seems. See e.g. the Community safety
section in the [annoucement](
https://discourse.mozilla.org/t/synchronous-messaging-at-mozilla-the-decision/50620
).


The solution they're using is federation turned off.


Just a heads up that there is now a lot more information about why 
Mozilla's federation is turned *on*, and why they picked Matrix, up at 
http://exple.tive.org/blarg/2020/03/06/brace-for-impact/


M

--

Matthew Hodgson
Matrix.org

___
desktop-devel-list mailing list
desktop-devel-list@gnome.org
https://mail.gnome.org/mailman/listinfo/desktop-devel-list

Matrix IRC bridge fixed? (was: Re: Matrix IRC bridge considered harmful)

[Matthew Hodgson via desktop-devel-list]

On 15/02/2020 16:08, Matthew Hodgson via desktop-devel-list wrote:

Meanwhile, we've gone and published the IRC room list as the public 
room directory on https://gnome.riot.im.


Other than GNOME branding and a sensible home somewhere in 
https://*.gnome.org, the dedicated server & bridge should now be good 
to go.


In terms of performance, it's unrecognisably snappy.  From my user as 
@matthew:gnome.org, i see ~50ms to IRC and back again.  (I had graphs 
to prove it, but the mailing list moderation kicked in due to the 
images.)


Please let us know if you have further problems with the bridge. We 
want Matrix to work well for everyone using it, and the IRC bridge 
(when run on a non-overloaded server) should work really well.


Since this, we've done a bunch more work on the gnome.org Matrix server 
- untangling various rooms which had been incorrectly double-bridged, 
getting rid of the remaining vestiges of the old bridge, fixing up room 
aliases to have sensible names, upgrading rooms to the latest version of 
the Matrix protocol, etc.  There was a bit of disruption a few weeks ago 
when we finished this off, but since then the bridge and server look to 
be working (at last) as intended, and continue to be very performant and 
stable.


So please do give https://gnome.riot.im a go, and let me know any 
problems in #general:gnome.org.


thanks,

Matthew

--
Matthew Hodgson
Matrix.org

___
desktop-devel-list mailing list
desktop-devel-list@gnome.org
https://mail.gnome.org/mailman/listinfo/desktop-devel-list

Re: Matrix IRC bridge considered harmful

[Matthew Hodgson via desktop-devel-list]

On 15/02/2020 01:28, Zander Brown wrote:


This has now been fixed as of 00:50 GMT or so.

Around the same time people seem to have suddenly left rooms and/or had
permissions stripped


Yes, that was the bridge restarting to flush out the problems in #gnome. 
things should now be back in sync.  Meanwhile we're working on a next 
gen of the bridge which lets it restart without dropping the IRC 
connections: https://github.com/Half-Shot/irc-conntrack


Meanwhile, we've gone and published the IRC room list as the public room 
directory on https://gnome.riot.im.


Other than GNOME branding and a sensible home somewhere in 
https://*.gnome.org, the dedicated server & bridge should now be good to go.


In terms of performance, it's unrecognisably snappy.  From my user as 
@matthew:gnome.org, i see ~50ms to IRC and back again.  (I had graphs to 
prove it, but the mailing list moderation kicked in due to the images.)


Please let us know if you have further problems with the bridge. We want 
Matrix to work well for everyone using it, and the IRC bridge (when run 
on a non-overloaded server) should work really well.



--
Matthew Hodgson
Matrix.org

___
desktop-devel-list mailing list
desktop-devel-list@gnome.org
https://mail.gnome.org/mailman/listinfo/desktop-devel-list

Re: Matrix IRC bridge considered harmful

[Zander Brown]

> This has now been fixed as of 00:50 GMT or so.

Around the same time people seem to have suddenly left rooms and/or had
permissions stripped


signature.asc
Description: This is a digitally signed message part
___
desktop-devel-list mailing list
desktop-devel-list@gnome.org
https://mail.gnome.org/mailman/listinfo/desktop-devel-list

Re: Matrix IRC bridge considered harmful

[Matthew Hodgson via desktop-devel-list]

On 14/02/2020 21:08, Matthew Hodgson via desktop-devel-list wrote:


and a bridging problem that's specific to the #gnome room.

This is because (for historical reasons) there are two instances of 
#gnome on Matrix - one is a 'portal' into the IRC channel 
(#_gimpnet_#gnome:gnome.org) and the other is a plain Matrix room 
(#gnome:matrix.org) which has been 'plumbed' into #gnome on gimpnet.  
Irritatingly there are a bunch of legitimate Matrix users in both 
rooms, and each see the activity of the other Matrix users bridged via 
IRC, which indeed looks like 'garbage'.


We'll fix this shortly by unplumbing #gnome:matrix.org; migrating the 
users in the #gnome:matrix.org room into the portal room (and pointing 
#gnome:matrix.org at the portal room), and then upgrading the portal 
room to flush out any idle users.



This has now been fixed as of 00:50 GMT or so.

--
Matthew Hodgson
Matrix.org

___
desktop-devel-list mailing list
desktop-devel-list@gnome.org
https://mail.gnome.org/mailman/listinfo/desktop-devel-list

Re: Matrix IRC bridge considered harmful

[Matthew Hodgson via desktop-devel-list]

On 14/02/2020 18:46, Michael Catanzaro wrote:

We're currently working with Matthew to resolve several additional 
quality issues with the bridge, including an issue where Matrix users 
are disconnected from IRC rooms but don't notice because they continue 
to receive updates in Matrix, 


This was because the bridge had hit a soft-limit on the IRC network, 
which we've worked around as of 21:00 GMT - the behaviour should now be 
resolved (and should solve many of the other problems folks have hit).



and a bridging problem that's specific to the #gnome room.


This is because (for historical reasons) there are two instances of 
#gnome on Matrix - one is a 'portal' into the IRC channel 
(#_gimpnet_#gnome:gnome.org) and the other is a plain Matrix room 
(#gnome:matrix.org) which has been 'plumbed' into #gnome on gimpnet.  
Irritatingly there are a bunch of legitimate Matrix users in both rooms, 
and each see the activity of the other Matrix users bridged via IRC, 
which indeed looks like 'garbage'.


We'll fix this shortly by unplumbing #gnome:matrix.org; migrating the 
users in the #gnome:matrix.org room into the portal room (and pointing 
#gnome:matrix.org at the portal room), and then upgrading the portal 
room to flush out any idle users.


I vaguely remember us spotting this back in April but hoping that folks 
would congregate in one room or the other, and then it fell off the 
radar until it was reported just now.


Hopefully we can solve these and keep the bridge running, so that 
Matrix and IRC users can continue to participate in one unified 
community without feeling pressured to switch from one to the other. 
(Of course, even nicer if you join the Matrix side of things, so you 
can talk directly to our federated friends at 
Mozilla/KDE/Igalia/Purism as well as to IRC users, but with a bridge 
that's working properly, that would of course remain optional. :)


:)

-

Matthew Hodgson
Matrix.org

___
desktop-devel-list mailing list
desktop-devel-list@gnome.org
https://mail.gnome.org/mailman/listinfo/desktop-devel-list

Re: Matrix IRC bridge considered harmful

[Michael Catanzaro]

On Wed, Feb 12, 2020 at 4:35 pm, Michael Catanzaro 
 wrote:
Here is my suggestion: fellow Matrix proponents, let's turn off the 
IRC bridge ASAP. All we've accomplished by running the IRC bridge is 
convincing GNOME devs that Matrix is awful. I'm pretty sure that all 
of this negative feedback is about the IRC bridge.


OK, my main complaint is already resolved: I've tested and verified 
that proper errors are returned when I try to message an IRC user who 
is offline.


My suggestion to shut down the bridge was based on an incorrect 
assumption that it was unmaintained and not a priority to fix. We're 
currently working with Matthew to resolve several additional quality 
issues with the bridge, including an issue where Matrix users are 
disconnected from IRC rooms but don't notice because they continue to 
receive updates in Matrix, and a bridging problem that's specific to 
the #gnome room. Hopefully we can solve these and keep the bridge 
running, so that Matrix and IRC users can continue to participate in 
one unified community without feeling pressured to switch from one to 
the other. (Of course, even nicer if you join the Matrix side of 
things, so you can talk directly to our federated friends at 
Mozilla/KDE/Igalia/Purism as well as to IRC users, but with a bridge 
that's working properly, that would of course remain optional. :)


Michael


___
desktop-devel-list mailing list
desktop-devel-list@gnome.org
https://mail.gnome.org/mailman/listinfo/desktop-devel-list

Re: Matrix IRC bridge considered harmful

[Matthew Hodgson via desktop-devel-list]

On 14/02/2020 17:00, Michael Catanzaro wrote:

On Thu, Feb 13, 2020 at 11:52 pm, Matthew Hodgson via 
desktop-devel-list  wrote:
1. The original issue that Michael Catanzaro reported (Matrix->IRC PM 
going missing) was a legitimate bug in the bridge.  The bridge is 
meant to display an error if you try to talk to an absent IRC user; 
this was fixed today and will be deployed tomorrow: 
https://github.com/matrix-org/matrix-appservice-irc/pull/978. Sorry 
that you got bitten by this :(


Thanks a lot Matthew! I appreciate the fast response here.


np :) the fix was deployed on all the bridges at 09:00Z this morning.

M


--
Matthew Hodgson
Matrix.org

___
desktop-devel-list mailing list
desktop-devel-list@gnome.org
https://mail.gnome.org/mailman/listinfo/desktop-devel-list

Re: Matrix IRC bridge considered harmful

[Michael Catanzaro]

On Thu, Feb 13, 2020 at 11:52 pm, Matthew Hodgson via 
desktop-devel-list  wrote:
1. The original issue that Michael Catanzaro reported (Matrix->IRC PM 
going missing) was a legitimate bug in the bridge.  The bridge is 
meant to display an error if you try to talk to an absent IRC user; 
this was fixed today and will be deployed tomorrow: 
https://github.com/matrix-org/matrix-appservice-irc/pull/978.  Sorry 
that you got bitten by this :(


Thanks a lot Matthew! I appreciate the fast response here.

Michael


___
desktop-devel-list mailing list
desktop-devel-list@gnome.org
https://mail.gnome.org/mailman/listinfo/desktop-devel-list

Re: Matrix IRC bridge considered harmful

[Matthew Hodgson via desktop-devel-list]

Hi Carlos, all,

Sorry if it came across as overdramatic; it's just a bit frustrating 
that the service isn't working as intended, and we've been stuck for 
almost a year on resolving it (and that it's been impacting both Matrix 
& GNOME as a result).  The fault of the delay lies originally on the 
Matrix side (we were dealing with operational issues in Apr 2019) and 
more recently on the GNOME side (the lack of response to my various 
emails pleading to finish standing up the service).


To be clear, if people were using the server we stood up at 
https://gnome.modular.im (which was intended to be 
https://chat.gnome.org, had the process been finished), then the vast 
majority of the Matrix<->IRC issues would be nonexistent and everyone 
would be much happier.


Perhaps the solution to this is for folks to just go use 
https://gnome.riot.im (or connect their Matrix client to the 
https://gnome.modular.im homeserver) for now if they want use the 
dedicated GNOME matrix server, and perhaps in future it can be given 
proper *.gnome.org branding.


thanks,

Matthew

On 14/02/2020 13:40, Carlos Soriano wrote:

Matthew,

On April 19th 2020 we completed the server set up configured as 
agreed. After that, we though everything was done and ready, and as 
you probably remember we did actually informed the community about the 
improved services [0]. That the previous answer to this thread make it 
sound like it has been on hold because of us since then is not correct 
and I believe it has more drama on it than it needs to be.


We do truly appreciate the services, because you are right that it's 
beneficial for both organizations and we want GNOME contributors to 
have the best experience regardless of the service they are using. 
However, I hope you understand our careful consideration on what steps 
we follow here, as we care about not putting the community in a 
position that would be difficult to go back from. This is specially 
true around branding, external services and official recommendations.


I don't see a reason why we couldn't make the IRC bridge performance 
ok with this requirements in place, so let's continue working on 
making that happen as we have been doing.


Thanks,
Carlos

[0] 
https://mail.gnome.org/archives/desktop-devel-list/2019-March/msg00015.html


On Fri, 14 Feb 2020 at 01:00, Matthew Hodgson via desktop-devel-list 
mailto:desktop-devel-list@gnome.org>> 
wrote:


Hi folks,

Sorry for the delay in response here - the last 24 hours have not
been fun.

Trying to address the main bits of feedback here:

1. The original issue that Michael Catanzaro reported (Matrix->IRC
PM going missing) was a legitimate bug in the bridge.  The bridge
is meant to display an error if you try to talk to an absent IRC
user; this was fixed today and will be deployed tomorrow:
https://github.com/matrix-org/matrix-appservice-irc/pull/978.
Sorry that you got bitten by this :(

2. In terms of: "We currently have loads of garbage IRC users in
the channels after the bridge hosted at matrix.org
 was replaced with one hosted at the gnome.org
 homeserver." - I thought we'd cleared this up
in the days following the migration, and this is the first I've
heard of it still being a problem.  It sounds like the old bridge
created IRC users on the Matrix side, and then the new bridge
bridged them back over to IRC.  It should be trivial to kick out
the old bridge's IRC users - please can you give me an example
channel/room where this is happening to look at?

3. In terms of bridge performance: we set up a dedicated bridge
and server for gnome.org  powered by modular.im
 back in April 2019.  The bridge got put live,
but the server was never publicised because we never got a
greenlight to announce its existence (plus the go-live was
eclipsed by some unrelated security dramas on the matrix.org
 homeserver).  As a result, the majority of
users have been using the bridge via the public matrix.org
 server, which is (unfortunately) often
overloaded thanks to the exponential growth we've been dealing
with. However, if folks were actually using the dedicated GNOME
server, then it would be an *excellent* experience - much like the
one that Mozilla is enjoying currently.  It's worth noting that we
provided the GNOME server for free because we want to support the
project, but the running costs are significant - it's been very
frustrating that the server has not been used.  (It seems that
some people have found it anyway over the course of today, to
quote someone in #general:gnome.org : "OMG the
IRC bridge is SO much faster on this instance.").  I'm hoping that
we can get a greenlight to point people at the Gnome.org server,
as right now the 

Re: Matrix IRC bridge considered harmful

[Carlos Soriano via desktop-devel-list]

Matthew,

On April 19th 2020 we completed the server set up configured as agreed.
After that, we though everything was done and ready, and as you probably
remember we did actually informed the community about the improved services
[0]. That the previous answer to this thread make it sound like it has been
on hold because of us since then is not correct and I believe it has more
drama on it than it needs to be.

We do truly appreciate the services, because you are right that it's
beneficial for both organizations and we want GNOME contributors to have
the best experience regardless of the service they are using. However, I
hope you understand our careful consideration on what steps we follow here,
as we care about not putting the community in a position that would be
difficult to go back from. This is specially true around branding, external
services and official recommendations.

I don't see a reason why we couldn't make the IRC bridge performance ok
with this requirements in place, so let's continue working on making that
happen as we have been doing.

Thanks,
Carlos

[0]
https://mail.gnome.org/archives/desktop-devel-list/2019-March/msg00015.html

On Fri, 14 Feb 2020 at 01:00, Matthew Hodgson via desktop-devel-list <
desktop-devel-list@gnome.org> wrote:

> Hi folks,
>
> Sorry for the delay in response here - the last 24 hours have not been fun.
>
> Trying to address the main bits of feedback here:
>
> 1. The original issue that Michael Catanzaro reported (Matrix->IRC PM
> going missing) was a legitimate bug in the bridge.  The bridge is meant to
> display an error if you try to talk to an absent IRC user; this was fixed
> today and will be deployed tomorrow:
> https://github.com/matrix-org/matrix-appservice-irc/pull/978.  Sorry that
> you got bitten by this :(
>
> 2. In terms of: "We currently have loads of garbage IRC users in the
> channels after the bridge hosted at matrix.org was replaced with one
> hosted at the gnome.org homeserver." - I thought we'd cleared this up in
> the days following the migration, and this is the first I've heard of it
> still being a problem.  It sounds like the old bridge created IRC users on
> the Matrix side, and then the new bridge bridged them back over to IRC.  It
> should be trivial to kick out the old bridge's IRC users - please can you
> give me an example channel/room where this is happening to look at?
>
> 3. In terms of bridge performance: we set up a dedicated bridge and server
> for gnome.org powered by modular.im back in April 2019.  The bridge got
> put live, but the server was never publicised because we never got a
> greenlight to announce its existence (plus the go-live was eclipsed by some
> unrelated security dramas on the matrix.org homeserver).  As a result,
> the majority of users have been using the bridge via the public matrix.org
> server, which is (unfortunately) often overloaded thanks to the exponential
> growth we've been dealing with.  However, if folks were actually using the
> dedicated GNOME server, then it would be an *excellent* experience - much
> like the one that Mozilla is enjoying currently.  It's worth noting that we
> provided the GNOME server for free because we want to support the project,
> but the running costs are significant - it's been very frustrating that the
> server has not been used.  (It seems that some people have found it anyway
> over the course of today, to quote someone in #general:gnome.org: "OMG
> the IRC bridge is SO much faster on this instance.").  I'm hoping that we
> can get a greenlight to point people at the Gnome.org server, as right now
> the situation is indeed terrible and it's hurting Matrix's reputation as
> well as hurting GNOME :((
>
> 4. Mozilla *are* running their homeserver federated (as of this week) -
> c.f. https://twitter.com/littledan/status/1227603567722319873. They're
> countering abuse by using the arsenal of anti-abuse tooling we've been
> working on over the last year as per
> https://matrix.org/docs/guides/moderation/ and
> https://github.com/matrix-org/matrix-doc/blob/msc2313/proposals/2313-moderation-policy-rooms.md
> etc.
>
> You may also be interested that the core Matrix team is starting to look
> seriously at the work going on around Fractal, particularly around E2E
> encryption, to accelerate E2E encryption for Rust clients in general.
> Specifically, we're porting pantalaimon (the Matrix daemon which offloads
> E2E encryption) from Python to Rust, and we hope that the resulting
> official E2EE-capable Matrix Rust SDK will be directly usable by Fractal
> and help the project along massively as a first class native Matrix GTK
> client (assuming they want to use it! :)
>
> So, TL;DR: we've had a solution to much of the Matrix<->IRC problems since
> April 2019, we just need to actually use it.
>
> I'm sorry this has taken so long to sort out - I genuinely hadn't realised
> that things were so bad.
>
> thanks,
>
> Matthew
>
>
> On 13/02/2020 20:28, Carlos Soriano via 

Re: Matrix IRC bridge considered harmful

[Matthew Hodgson via desktop-devel-list]

Hi folks,

Sorry for the delay in response here - the last 24 hours have not been fun.

Trying to address the main bits of feedback here:

1. The original issue that Michael Catanzaro reported (Matrix->IRC PM 
going missing) was a legitimate bug in the bridge.  The bridge is meant 
to display an error if you try to talk to an absent IRC user; this was 
fixed today and will be deployed tomorrow: 
https://github.com/matrix-org/matrix-appservice-irc/pull/978. Sorry that 
you got bitten by this :(


2. In terms of: "We currently have loads of garbage IRC users in the 
channels after the bridge hosted at matrix.org was replaced with one 
hosted at the gnome.org homeserver." - I thought we'd cleared this up in 
the days following the migration, and this is the first I've heard of it 
still being a problem.  It sounds like the old bridge created IRC users 
on the Matrix side, and then the new bridge bridged them back over to 
IRC.  It should be trivial to kick out the old bridge's IRC users - 
please can you give me an example channel/room where this is happening 
to look at?


3. In terms of bridge performance: we set up a dedicated bridge and 
server for gnome.org powered by modular.im back in April 2019.  The 
bridge got put live, but the server was never publicised because we 
never got a greenlight to announce its existence (plus the go-live was 
eclipsed by some unrelated security dramas on the matrix.org 
homeserver).  As a result, the majority of users have been using the 
bridge via the public matrix.org server, which is (unfortunately) often 
overloaded thanks to the exponential growth we've been dealing with. 
However, if folks were actually using the dedicated GNOME server, then 
it would be an *excellent* experience - much like the one that Mozilla 
is enjoying currently.  It's worth noting that we provided the GNOME 
server for free because we want to support the project, but the running 
costs are significant - it's been very frustrating that the server has 
not been used.  (It seems that some people have found it anyway over the 
course of today, to quote someone in #general:gnome.org: "OMG the IRC 
bridge is SO much faster on this instance.").  I'm hoping that we can 
get a greenlight to point people at the Gnome.org server, as right now 
the situation is indeed terrible and it's hurting Matrix's reputation as 
well as hurting GNOME :((


4. Mozilla *are* running their homeserver federated (as of this week) - 
c.f.https://twitter.com/littledan/status/1227603567722319873 
. They're 
countering abuse by using the arsenal of anti-abuse tooling we've been 
working on over the last year as per 
https://matrix.org/docs/guides/moderation/ and 
https://github.com/matrix-org/matrix-doc/blob/msc2313/proposals/2313-moderation-policy-rooms.md 
etc.


You may also be interested that the core Matrix team is starting to look 
seriously at the work going on around Fractal, particularly around E2E 
encryption, to accelerate E2E encryption for Rust clients in general.  
Specifically, we're porting pantalaimon (the Matrix daemon which 
offloads E2E encryption) from Python to Rust, and we hope that the 
resulting official E2EE-capable Matrix Rust SDK will be directly usable 
by Fractal and help the project along massively as a first class native 
Matrix GTK client (assuming they want to use it! :)


So, TL;DR: we've had a solution to much of the Matrix<->IRC problems 
since April 2019, we just need to actually use it.


I'm sorry this has taken so long to sort out - I genuinely hadn't 
realised that things were so bad.


thanks,

Matthew


On 13/02/2020 20:28, Carlos Soriano via desktop-devel-list wrote:

Hi folks,

We been in contact with Matthew from Matrix for some time already. I 
lately didn't have much time to invest on this, so we had have some 
delays on answering. However, it's our expectation that with the set 
up that we have right now the IRC bridge should perform as its best, 
as we are using servers from modular.im , the 
company that maintains paid severs with matrix services on them. We 
believe our set up is correct, but there might be some miss 
configuration somewhere, or the current situation it's already the 
best that can be offered.


We'll update you as soon as we have more information.

Cheers,
Carlos

On Thu, 13 Feb 2020 at 17:16, Michael Catanzaro > wrote:


On Wed, Feb 12, 2020 at 4:15 pm, Britt Yazel mailto:bwya...@gnome.org>> wrote:
> Attached is an image of the compact mode + dark theme. Just for the
> record.

The thing is, it really comes down to personal preference. I
suspect we
have a lot of people who like web clients, and a lot of people who
just
don't. With open protocols like IRC, XMPP, or Matrix, where lots of
client choice exists, you can use whatever you prefer. It's no
problem
if you don't like any particular client because 

Re: Matrix IRC bridge considered harmful

[Carlos Soriano via desktop-devel-list]

Hi folks,

We been in contact with Matthew from Matrix for some time already. I lately
didn't have much time to invest on this, so we had have some delays on
answering. However, it's our expectation that with the set up that we have
right now the IRC bridge should perform as its best, as we are using
servers from modular.im, the company that maintains paid severs with matrix
services on them. We believe our set up is correct, but there might be some
miss configuration somewhere, or the current situation it's already the
best that can be offered.

We'll update you as soon as we have more information.

Cheers,
Carlos

On Thu, 13 Feb 2020 at 17:16, Michael Catanzaro 
wrote:

> On Wed, Feb 12, 2020 at 4:15 pm, Britt Yazel  wrote:
> > Attached is an image of the compact mode + dark theme. Just for the
> > record.
>
> The thing is, it really comes down to personal preference. I suspect we
> have a lot of people who like web clients, and a lot of people who just
> don't. With open protocols like IRC, XMPP, or Matrix, where lots of
> client choice exists, you can use whatever you prefer. It's no problem
> if you don't like any particular client because you can just use a
> different one.
>
> Myself, I like to see GNOME clients: polari, dino, fractal, chatty.
> They look good next to our other apps, and vindicate the potential of
> our desktop platform. But if we're going to have a web client, at the
> very least do it using WebKitGTK, like Revolt does, to stick with GNOME
> technologies and avoid bundling Electron.
>
> I'll also suggest: whatever we use, it'd be nice to select something
> with the potential to become a widely-accepted standard, like IRC used
> to be. Chat has become a failed disaster area of fragmented walled
> gardens, and when we have a choice between an emerging standard and yet
> another silo, I think there's tremendous value in choosing the standard.
>
> Michael
>
>
> ___
> desktop-devel-list mailing list
> desktop-devel-list@gnome.org
> https://mail.gnome.org/mailman/listinfo/desktop-devel-list
>
___
desktop-devel-list mailing list
desktop-devel-list@gnome.org
https://mail.gnome.org/mailman/listinfo/desktop-devel-list

Re: Matrix IRC bridge considered harmful

[Michael Catanzaro]

On Wed, Feb 12, 2020 at 4:15 pm, Britt Yazel  wrote:
Attached is an image of the compact mode + dark theme. Just for the 
record.


The thing is, it really comes down to personal preference. I suspect we 
have a lot of people who like web clients, and a lot of people who just 
don't. With open protocols like IRC, XMPP, or Matrix, where lots of 
client choice exists, you can use whatever you prefer. It's no problem 
if you don't like any particular client because you can just use a 
different one.


Myself, I like to see GNOME clients: polari, dino, fractal, chatty. 
They look good next to our other apps, and vindicate the potential of 
our desktop platform. But if we're going to have a web client, at the 
very least do it using WebKitGTK, like Revolt does, to stick with GNOME 
technologies and avoid bundling Electron.


I'll also suggest: whatever we use, it'd be nice to select something 
with the potential to become a widely-accepted standard, like IRC used 
to be. Chat has become a failed disaster area of fragmented walled 
gardens, and when we have a choice between an emerging standard and yet 
another silo, I think there's tremendous value in choosing the standard.


Michael


___
desktop-devel-list mailing list
desktop-devel-list@gnome.org
https://mail.gnome.org/mailman/listinfo/desktop-devel-list

Re: Matrix IRC bridge considered harmful

[Michael Catanzaro]

On Thu, Feb 13, 2020 at 11:14 am, Benjamin Berg 
 wrote:

One could do this comparison properly. But it would need setting up a
private Matrix server for GNOME (possibly without Federation) and then
checking how well it holds up when compared to Rocket.Chat.


gnome.modular.im is already our "private" Matrix server for GNOME. 
Presumably there are no performance issues there?



___
desktop-devel-list mailing list
desktop-devel-list@gnome.org
https://mail.gnome.org/mailman/listinfo/desktop-devel-list

Re: Matrix IRC bridge considered harmful

[Adrian Perez de Castro]

On Thu, 13 Feb 2020 14:27:21 +, Neil McGovern  wrote:
> On Wed, 2020-02-12 at 20:21 +0100, Alexandre Franke wrote:
> > > My concern would be the "federal" nature of matrix where people
> > > don't need a
> > > gnome.org specific chat account to join a room. Whilst there are a
> > > lot of
> > > arguments for this I'm increasingly convinced it's an anti-feature
> > > especially
> > > if we want to enforce CoC (which, of course, we do)
> > 
> >  
> > That was a concern for Mozilla too. I don’t know the details, but
> > they have a solution for that it seems. See e.g. the Community safety
> > section in the [annoucement](
> > https://discourse.mozilla.org/t/synchronous-messaging-at-mozilla-the-decision/50620
> > ).
> > 
> 
> The solution they're using is federation turned off.

Mozilla has turned on federation on their Matrix instance just this week.

Cheers,
—Adrián


signature.asc
Description: PGP signature
___
desktop-devel-list mailing list
desktop-devel-list@gnome.org
https://mail.gnome.org/mailman/listinfo/desktop-devel-list

Re: Matrix IRC bridge considered harmful

[Neil McGovern]

On Wed, 2020-02-12 at 20:21 +0100, Alexandre Franke wrote:
> > My concern would be the "federal" nature of matrix where people
> > don't need a
> > gnome.org specific chat account to join a room. Whilst there are a
> > lot of
> > arguments for this I'm increasingly convinced it's an anti-feature
> > especially
> > if we want to enforce CoC (which, of course, we do)
> 
>  
> That was a concern for Mozilla too. I don’t know the details, but
> they have a solution for that it seems. See e.g. the Community safety
> section in the [annoucement](
> https://discourse.mozilla.org/t/synchronous-messaging-at-mozilla-the-decision/50620
> ).
> 

The solution they're using is federation turned off.

Neil
-- 
Neil McGovern
Executive Director, The GNOME Foundation

___
desktop-devel-list mailing list
desktop-devel-list@gnome.org
https://mail.gnome.org/mailman/listinfo/desktop-devel-list

Re: Matrix IRC bridge considered harmful

[Alberto Fanjul Alonso via desktop-devel-list]

The only reason I use matrix is that allows me to not loose any comment
while I'm not connected. If we can recover the IRC log not, there's no need
for it.

If we can log off to avoid people talk to ghost users that's another
improvement.

Hands up everybody that claim IRC is just perfect but have a personal
bouncer running 24x7!

Cheers,
Alberto

El jue., 13 feb. 2020 3:15, Christian Hergert 
escribió:

> On 2/12/20 4:15 PM, Britt Yazel wrote:
> >
> > If I remember correctly our conversation last month, you said you didn't
> > want a web browser open as it provided tabs to distractions. At which
> > point I mentioned the electron Flatpak (which contains no such tabs),
> > but you weren't having it.
>
> I share Nirbheek's viewpoint on this but also find the features that get
> added from having a browser engine to be disorienting. (Animated gifs,
> video, inline images, etc).
>
> The ability to disable all that would allow me to contribute without
> disorientation.
>
> I'm looking for minimal distraction, immediacy when present, high
> signal-to-noise ratio and the ability to walk away and come back later
> without loss of information.
>
> It doesn't have to be IRC.
>
> I would be extremely happy if someone had time to make Polari but on
> rocket.chat. The API at first glance doesn't look terrible, but someone
> needs to have time to take on that battle.
>
> I have enough battles already.
>
> -- Christian
> ___
> desktop-devel-list mailing list
> desktop-devel-list@gnome.org
> https://mail.gnome.org/mailman/listinfo/desktop-devel-list
>
___
desktop-devel-list mailing list
desktop-devel-list@gnome.org
https://mail.gnome.org/mailman/listinfo/desktop-devel-list

Re: Matrix IRC bridge considered harmful

[Benjamin Berg]

On Wed, 2020-02-12 at 16:35 -0600, Michael Catanzaro wrote:
> On Wed, Feb 12, 2020 at 2:09 pm, Britt Yazel  wrote:
> > I have had horrible experiences with Matrix/Riot.im. I'm not sure 
> > which of those is due to the IRC bridge or which is due to Matrix 
> > itself, or which is due to the clients, but I really shouldn't 'have' 
> > to know the chat system at that level. My experience has been awful.
> 
> Here is my suggestion: fellow Matrix proponents, let's turn off the IRC 
> bridge ASAP. All we've accomplished by running the IRC bridge is 
> convincing GNOME devs that Matrix is awful. I'm pretty sure that all of 
> this negative feedback is about the IRC bridge.

Yeah, I also think most of the Matrix issues that people see are either
related to IRC bridging or that the public servers we rely on were
overloaded.

So, said differently, I would expect that anyone using a walled garden
Rocket.Chat instance (i.e. chat.gnome.org) without all that baggage
will have a great user experience in comparison. But, unfortunately,
that tells us nothing about which chat system is superior.

One could do this comparison properly. But it would need setting up a
private Matrix server for GNOME (possibly without Federation) and then
checking how well it holds up when compared to Rocket.Chat.


I have no idea which option is superior. And that can be a hard
question as it might even differ depending on whether your focus is on
e.g. short term experience vs. long term technical viability. That
said, I would love to see arguments here (for oragainst each chat
system) that I can compare in a useful manner.

Benjamin

PS: I had a nice chat with the Rocket.Chat person at FOSDEM who was
keen on getting IRC bridging up and running on their side. He said that
when I mentioned that Red Hat had experimented with it. If someone is
serious about this, I can pass on the contact.

> (We also need to fix the fractal bug that causes it to create private 
> rooms set to allow participants to view only messages sent after they 
> have joined the room. I guess fractal is sending the wrong permissions 
> enum value when creating rooms, or something similar to that.)
> 
> On Wed, Feb 12, 2020 at 2:09 pm, Britt Yazel  wrote:
> > So, the last thing I'll say is this. As a project that is trying to 
> > attract more users, many of whom are young, new to FOSS, and or are 
> > non-technology skilled professionals such as artists, designers, 
> > writers, etc, is Matrix really the best option? Or do you just want 
> > it to be the best option?
> 
> It's really the best option.
> 
> The problem with Rocket.Chat is that with only a web client, I doubt 
> very many developers would actually be willing to use it. (At least, I 
> don't think I'm the only one who would be hard no to a web client.) And 
> honestly I have no reason to believe Rocket.Chat will exist in five 
> years. Alexandre says it's another silo, rather than an 
> extensively-documented backwards-compatible protocol like Matrix 
> (although since Rocket.Chat is open source, I suppose it might be the 
> best walled garden among walled gardens). Rocket.Chat doesn't seem 
> designed to unify online communication in the same way that Matrix is, 
> and honestly without a desktop client I'd say that alone leaves it far 
> behind IRC. We need to select something that we can really unify our 
> community behind, something that everyone will like, not something 
> that's only going to be used by people who like web clients. In 
> particular, we don't want to wind up with one chat community on 
> Rocket.Chat and another on IRC, which is where we're heading currently 
> if we keep chat.gnome.org online.
> 
> Michael
> 
> 
> ___
> desktop-devel-list mailing list
> desktop-devel-list@gnome.org
> https://mail.gnome.org/mailman/listinfo/desktop-devel-list
> 


signature.asc
Description: This is a digitally signed message part
___
desktop-devel-list mailing list
desktop-devel-list@gnome.org
https://mail.gnome.org/mailman/listinfo/desktop-devel-list

Re: Matrix IRC bridge considered harmful

[Tobias Mueller]

Hi,

On Wed, 2020-02-12 at 15:32 -0800, Britt Yazel wrote:
> Can you explain to me what the big issue with web clients are? I keep
> hearing over and over again that developers don't want to use web
> clients, either in browser or with Electron, but I don't recall ever
> hearing a "why" in there.
One of my reasons for disliking anything Browser-based is economic. It
takes noticeably longer to scratch all the bits of a Web browser off my
disk than anything else that has been tailored for the purpose.

A corollary is the gigantic trusted computing base. Not only do the Web
projects that I have to deal with pull in ridiculous amounts of
dependencies (most of which of questionable quality or reputation), they
ultimately also require a fully fledged Web browsing engine with all
bells and whistles. The number of lines of code used for some mundane
functionality as IRC is staggering.

And as big code bases tend to produce big binaries, we also need to have
a lot of memory to run those apps. I'm happy to be able to run my
Firefox and Evolution in parallel. I'm dreading to open any other
Browser-based thing because I know that my machine will start swapping
(yes, I do have swap space, because I sometimes I need to have Firefox,
Evolution, and something else running..).

So I'd rather have a bad Gtk (or "native") client than a good Web
client. Also because fixing Gtk-based apps is well within our skill set.

Cheers,
  Tobi


___
desktop-devel-list mailing list
desktop-devel-list@gnome.org
https://mail.gnome.org/mailman/listinfo/desktop-devel-list

Re: Matrix IRC bridge considered harmful

[Tobias Mueller]

Hi,

On Wed, 2020-02-12 at 14:09 -0800, Britt Yazel wrote:
> this is not going to be an academically backed response, just my
> personal take.
> 
> I have had horrible experiences with Matrix/Riot.im.
Too bad you missed out on actually mentioning what your experience was.
So it's very hard to relate to anything you're mentioning in this
thread.

Cheers,
  Tobi

___
desktop-devel-list mailing list
desktop-devel-list@gnome.org
https://mail.gnome.org/mailman/listinfo/desktop-devel-list

Re: Matrix IRC bridge considered harmful

[Christian Hergert]

On 2/12/20 4:15 PM, Britt Yazel wrote:
> 
> If I remember correctly our conversation last month, you said you didn't
> want a web browser open as it provided tabs to distractions. At which
> point I mentioned the electron Flatpak (which contains no such tabs),
> but you weren't having it.

I share Nirbheek's viewpoint on this but also find the features that get
added from having a browser engine to be disorienting. (Animated gifs,
video, inline images, etc).

The ability to disable all that would allow me to contribute without
disorientation.

I'm looking for minimal distraction, immediacy when present, high
signal-to-noise ratio and the ability to walk away and come back later
without loss of information.

It doesn't have to be IRC.

I would be extremely happy if someone had time to make Polari but on
rocket.chat. The API at first glance doesn't look terrible, but someone
needs to have time to take on that battle.

I have enough battles already.

-- Christian
___
desktop-devel-list mailing list
desktop-devel-list@gnome.org
https://mail.gnome.org/mailman/listinfo/desktop-devel-list

Re: Matrix IRC bridge considered harmful

[Michael Gratton]

On Wed, 12 Feb, 2020 at 12:30, Michael Catanzaro  
wrote:
(b) Replace our IRC with an actual Matrix server, so we get native 
Matrix. Matrix is very nice as long as you're not using the abysmal 
IRC bridge, which is unfit for purpose.


+1

The usability of IRC is terrible for normal people and when there's a 
TZ difference. In fact, having a web interface is actually a great boon 
for non-technical people, since at most people using Linux are able to 
create an account on a web site.


Using something that isn't libre and that doesn't have a CoC is going 
to inevitably end up being the same kind of disaster as GitHub and 
GitLab when the company ends up being trash and/or they pull the 
freemium carpet out from under us.


Further, anything that isn't libre and that doesn't have a native GNOME 
client is a massive show of no confidence in our own platform. Fractal 
isn't perfect, but it's better than no client.


Personally I'd like to see XMPP being used, but we don't have any 
decent, modern, native clients (although Dino is shaping up pretty 
nicely - https://dino.im/) and it seems to have been put in the 
too-hard basket by too many people.


So despite it's failings, I'd rather see us use Matrix rather than IRC 
or something like RocketChat.


//Mike

--
⊨ Michael Gratton, Percept Wrangler.
⚙ 


___
desktop-devel-list mailing list
desktop-devel-list@gnome.org
https://mail.gnome.org/mailman/listinfo/desktop-devel-list

Re: Matrix IRC bridge considered harmful

[Britt Yazel]

Huh? RocketChat Experimental + the dark theme it comes with is pretty
fabulous IMO. I genuinely like the React Native app they have for IOs and
Android.

On Wed, Feb 12, 2020 at 4:20 PM Zander Brown  wrote:

> > I do not use it mobile that much, but enough to notice Riot is not
> > mature yet. I have not tested RocketChat mobile app. YMMV.
>
> I suggest you continue in your innocence :-)
> ___
> desktop-devel-list mailing list
> desktop-devel-list@gnome.org
> https://mail.gnome.org/mailman/listinfo/desktop-devel-list
>
___
desktop-devel-list mailing list
desktop-devel-list@gnome.org
https://mail.gnome.org/mailman/listinfo/desktop-devel-list

Re: Matrix IRC bridge considered harmful

[Britt Yazel]

I appreciate your candor in sharing your reasons. I'd also like to make it
clear that I wasn't digging for any uncomfortable sharing of information.
That said, what makes an electron application (since we're avoiding web
browsers here, which I understand) significantly more distracting than an
IRC window? If you use the RocketChat compact theme + dark mode, is that
significantly more distracting that an IRC window? That said, would a
Matrix window be significantly more distracting than an IRC window as well?
So is your stance here that you aren't going to move away from IRC, or that
you need whatever the option is to be able to be stripped down to something
as bare bones as IRC? Or is that you only can tolerate one chat app being
open, and therefor whatever it is has to be universal?

If I remember correctly our conversation last month, you said you didn't
want a web browser open as it provided tabs to distractions. At which point
I mentioned the electron Flatpak (which contains no such tabs), but you
weren't having it.

Attached is an image of the compact mode + dark theme. Just for the record.

https://imgur.com/a/79rnzKc

(appologies to Christian who got this email twice)

On Wed, Feb 12, 2020 at 3:59 PM Christian Hergert 
wrote:

> On 2/12/20 3:32 PM, Britt Yazel wrote:
> > Can you explain to me what the big issue with web clients are? I keep
> > hearing over and over again that developers don't want to use web
> > clients, either in browser or with Electron, but I don't recall ever
> > hearing a "why" in there.
>
> I'll repeat what I told you last month.
>
> I don't want a browser open when I work, yet I need to collaborate with
> other GNOME developers on code and platform issues. I keep Builder open,
> a terminal, and IRC. That's it.
>
> For those of our community with ADHD and similar neuro-divergence, this
> can be a necessity to get things done. The only way I get through my
> todo list is to reduce all possibilities for distraction.
>
> I suspect you wont get many people telling you that though, because to
> do so requires a level of emotional labor to share with the world their
> medical history.
>
> -- Christian
> ___
> desktop-devel-list mailing list
> desktop-devel-list@gnome.org
> https://mail.gnome.org/mailman/listinfo/desktop-devel-list
>
___
desktop-devel-list mailing list
desktop-devel-list@gnome.org
https://mail.gnome.org/mailman/listinfo/desktop-devel-list

Re: Matrix IRC bridge considered harmful

[Zander Brown]

> I do not use it mobile that much, but enough to notice Riot is not
> mature yet. I have not tested RocketChat mobile app. YMMV.

I suggest you continue in your innocence :-)


signature.asc
Description: This is a digitally signed message part
___
desktop-devel-list mailing list
desktop-devel-list@gnome.org
https://mail.gnome.org/mailman/listinfo/desktop-devel-list

Re: Matrix IRC bridge considered harmful

[Nirbheek Chauhan via desktop-devel-list]

On Thu, Feb 13, 2020 at 5:29 AM Christian Hergert  wrote:
>
> On 2/12/20 3:32 PM, Britt Yazel wrote:
> > Can you explain to me what the big issue with web clients are? I keep
> > hearing over and over again that developers don't want to use web
> > clients, either in browser or with Electron, but I don't recall ever
> > hearing a "why" in there.
>
> I'll repeat what I told you last month.
>
> I don't want a browser open when I work, yet I need to collaborate with
> other GNOME developers on code and platform issues. I keep Builder open,
> a terminal, and IRC. That's it.
>
> For those of our community with ADHD and similar neuro-divergence, this
> can be a necessity to get things done. The only way I get through my
> todo list is to reduce all possibilities for distraction.
>
> I suspect you wont get many people telling you that though, because to
> do so requires a level of emotional labor to share with the world their
> medical history.
>

Thanks for sharing this important perspective with us, Christian. :)

RocketChat also has a standalone Electron client, do you foresee any
issues with using that?

Personally, I don't want more Electron things on my desktop since it
kills performance on Wayland for me. In theory, we can "just" write a
native client for it; does Rocket Chat have a documented and versioned
API?

I am also hesitant for us to start using another open-core/enterprise
product where the company will be interested and enthusiastic in the
beginning because getting us to use it will be good for publicity.
Then once we're dependent on their product, they will start ignoring
us and our issues completely — leaving us with missing steps in our
workflow (hey, first-time contributor, please tick the "allow
maintainer commits checkbox") or forcing us to write bots.

I'm sure eventually we'll end up having to fork. Would rather not be
in that soup with two critical bits of infrastructure.

Cheers,
Nirbheek
___
desktop-devel-list mailing list
desktop-devel-list@gnome.org
https://mail.gnome.org/mailman/listinfo/desktop-devel-list

Re: Matrix IRC bridge considered harmful

[Christian Hergert]

On 2/12/20 3:32 PM, Britt Yazel wrote:
> Can you explain to me what the big issue with web clients are? I keep
> hearing over and over again that developers don't want to use web
> clients, either in browser or with Electron, but I don't recall ever
> hearing a "why" in there.

I'll repeat what I told you last month.

I don't want a browser open when I work, yet I need to collaborate with
other GNOME developers on code and platform issues. I keep Builder open,
a terminal, and IRC. That's it.

For those of our community with ADHD and similar neuro-divergence, this
can be a necessity to get things done. The only way I get through my
todo list is to reduce all possibilities for distraction.

I suspect you wont get many people telling you that though, because to
do so requires a level of emotional labor to share with the world their
medical history.

-- Christian
___
desktop-devel-list mailing list
desktop-devel-list@gnome.org
https://mail.gnome.org/mailman/listinfo/desktop-devel-list

Re: Matrix IRC bridge considered harmful

[Britt Yazel]

Can you explain to me what the big issue with web clients are? I keep
hearing over and over again that developers don't want to use web clients,
either in browser or with Electron, but I don't recall ever hearing a "why"
in there.

On Wed, Feb 12, 2020 at 3:24 PM Michael Catanzaro 
wrote:

> On Wed, Feb 12, 2020 at 2:41 pm, Britt Yazel  wrote:
> > No offense Michael, but the argument that you are making seems a
> > whole lot like "we're not going to use RocketChat, or even consider
> > it a legitimate option, so the rest of you can either fracture the
> > community (which will be on you) or you can get on board Matrix"
> >
> > Unfortunately, you've done nothing to quell my anxiety of using
> > Matrix, the complexity therein, or explain how it's a better
> > experience than RocketChat, but rather placed an ultimatum upon me
> > (and the foundation). Do you see an issue here?
>
> Perhaps my language was too strong.
>
> Honestly I think you should just poll the community to see whether it
> wants to use Rocket.Chat or not. I'm assuming support will be pretty
> low, because I don't think many devs will want to keep a web browser
> window dedicated to chat where a GNOME client would otherwise work. But
> if I'm wrong and people like it, then I will shut up, and you can
> proceed. :)
>
> Basically anything would be better than status quo, messages going to
> nowhere
>
> Michael
>
>
>
___
desktop-devel-list mailing list
desktop-devel-list@gnome.org
https://mail.gnome.org/mailman/listinfo/desktop-devel-list

Re: Matrix IRC bridge considered harmful

[Germán Poo-Caamaño]

On Wed, 2020-02-12 at 14:09 -0800, Britt Yazel wrote:
> [... ] I could see an argument being that Matrix has Fractal and
> therefor is a nice GTK client, but, unfortunately as it is to say, my
> experience with Fractal was a bit iffy at best. I cannot even count
> the number of messages just dropped into the ether using Fractal.

I must say that this is the same reason I stopped using Fractal.
However, there are alternatives like Revolt:
https://github.com/aperezdc/revolt

UI-wise, RocketChat feel more polished than Riot, but I have used
RocketChat on and off.

Regarding to the mobile client: Riot's client is not mature yet.
The stable version for Android does not support the latest Riot
protocol, which allows edits, and reactions. As a consequence, if
someone edits a message say 5 times, you see 5 times the same message
(with a star at the beginning to know it is an edited message), and no
reaction. The development version supports edit, but it lacks several
features already available in the stable one.

I do not use it mobile that much, but enough to notice Riot is not
mature yet. I have not tested RocketChat mobile app. YMMV.

-- 
Germán Poo-Caamaño
https://calcifer.org




signature.asc
Description: This is a digitally signed message part
___
desktop-devel-list mailing list
desktop-devel-list@gnome.org
https://mail.gnome.org/mailman/listinfo/desktop-devel-list

Re: Matrix IRC bridge considered harmful

[Michael Catanzaro]

On Wed, Feb 12, 2020 at 2:09 pm, Britt Yazel  wrote:
I have had horrible experiences with Matrix/Riot.im. I'm not sure 
which of those is due to the IRC bridge or which is due to Matrix 
itself, or which is due to the clients, but I really shouldn't 'have' 
to know the chat system at that level. My experience has been awful.


Here is my suggestion: fellow Matrix proponents, let's turn off the IRC 
bridge ASAP. All we've accomplished by running the IRC bridge is 
convincing GNOME devs that Matrix is awful. I'm pretty sure that all of 
this negative feedback is about the IRC bridge.


(We also need to fix the fractal bug that causes it to create private 
rooms set to allow participants to view only messages sent after they 
have joined the room. I guess fractal is sending the wrong permissions 
enum value when creating rooms, or something similar to that.)


On Wed, Feb 12, 2020 at 2:09 pm, Britt Yazel  wrote:
So, the last thing I'll say is this. As a project that is trying to 
attract more users, many of whom are young, new to FOSS, and or are 
non-technology skilled professionals such as artists, designers, 
writers, etc, is Matrix really the best option? Or do you just want 
it to be the best option?


It's really the best option.

The problem with Rocket.Chat is that with only a web client, I doubt 
very many developers would actually be willing to use it. (At least, I 
don't think I'm the only one who would be hard no to a web client.) And 
honestly I have no reason to believe Rocket.Chat will exist in five 
years. Alexandre says it's another silo, rather than an 
extensively-documented backwards-compatible protocol like Matrix 
(although since Rocket.Chat is open source, I suppose it might be the 
best walled garden among walled gardens). Rocket.Chat doesn't seem 
designed to unify online communication in the same way that Matrix is, 
and honestly without a desktop client I'd say that alone leaves it far 
behind IRC. We need to select something that we can really unify our 
community behind, something that everyone will like, not something 
that's only going to be used by people who like web clients. In 
particular, we don't want to wind up with one chat community on 
Rocket.Chat and another on IRC, which is where we're heading currently 
if we keep chat.gnome.org online.


Michael


___
desktop-devel-list mailing list
desktop-devel-list@gnome.org
https://mail.gnome.org/mailman/listinfo/desktop-devel-list

Re: Matrix IRC bridge considered harmful

[Alexandre Franke]

On Wed, Feb 12, 2020 at 10:53 PM Michael Catanzaro  wrote:
> So it seems we have two GNOME clients for Matrix

Since I know some of the people who keep using IRC are actually using
something like irssi (maybe with tmux/screen) or weechat and care more
about a terminal option than a GNOME one, let’s mention there are
several terminal clients available too, including a weechat script.
Granted, if the motivation for that method is to “always be connected”
then a GNOME client would do the job as well for Matrix.

-- 
Alexandre Franke
GNOME Hacker
___
desktop-devel-list mailing list
desktop-devel-list@gnome.org
https://mail.gnome.org/mailman/listinfo/desktop-devel-list

Re: Matrix IRC bridge considered harmful

[Britt Yazel]

Here's my two cents, granted this is not going to be an academically backed
response, just my personal take.

I have had horrible experiences with Matrix/Riot.im. I'm not sure which of
those is due to the IRC bridge or which is due to Matrix itself, or which
is due to the clients, but I really shouldn't 'have' to know the chat
system at that level. My experience has been awful.

Rocket.Chat hosted at chat.gnome.org on the other hand has been pretty much
consistently awesome since it first spun up, and has gotten better
consistently over the last few months with weekly or bi-weekly updates. On
top of that, I don't want to put words in the mouths of the sysadmins, but
it seems like it's been really easy on their end to keep running and
maintained. There was a small issue with NTP not working right on the
server, but that was sorted out by Bart in not much time.

So, y'all are talking about bridge this and federation that, this client,
that client, yada yada. This isn't making me yearn to use Matrix. I really
don't want to have to have a MS degree to use a basic chat tool
(exaggeration to make a point). RocketChat works, is simple, has nice
mobile apps, has tons of features, and Matrix has to offer me a LOT of
benefit over RocketChat to pull me in that direction.

So my question to you is this. Why is Matrix a technically and
usability-wise superior option over RocketChat? And in that question, I'm
not asking for theoretical benefits, I really am interested in practical
use-case benefits. I could see an argument being that Matrix has Fractal
and therefor is a nice GTK client, but, unfortunately as it is to say, my
experience with Fractal was a bit iffy at best. I cannot even count the
number of messages just dropped into the ether using Fractal.


So, the last thing I'll say is this. As a project that is trying to attract
more users, many of whom are young, new to FOSS, and or are non-technology
skilled professionals such as artists, designers, writers, etc, is Matrix
really the best option? Or do you just want it to be the best option?

On Wed, Feb 12, 2020 at 1:53 PM Michael Catanzaro 
wrote:

> On Wed, Feb 12, 2020 at 4:23 pm, Georges Basile Stavracas Neto via
> desktop-devel-list  wrote:
> > The Riot application is hard to use. It took me days to figure out
> > how to connect
> > to a GNOME room. It doesn't allow me to log out of the servers.
>
> These are all problems with the IRC bridge, not with normal Matrix. I
> agree the quality of the IRC bridge is catastrophic. Joining rooms is
> extremely difficult, and it is a Hotel California bridge (you can check
> out, but you can never leave! that's why we have all these trap ghost
> accounts from Matrix who never see the PMs you send them, basically the
> inverse of the problem I started this thread to complain about).
>
> Anyway, normal Matrix has none of those problems. Please don't judge it
> based on the quality of the IRC bridge.
>
> It is a shame that the Matrix developers continue to operate IRC
> bridges that are clearly serving to harm the reputation of Matrix. I
> know there's value in bridging to IRC, but it should be done well if
> it's going to be done at all. No doubt the Matrix developers have
> limited time and competing priorities, just like we do ourselves
>
> > Fractal is nice,
> > as I really like native clients, but Polari feels more polished.
> > Matrix apparently
> > doesn't allow turning off federation, and to me that's a no-go aspect
> > of it. At
> > last, I have a strong impression that Matrix suffers from feature
> > bloat.
>
> Fractal needs some love, for sure. I wouldn't want Matrix to be judged
> by the quality of Fractal today. (In particular, all PMs you send to
> users are silently discarded until the other user joins the room, and
> there is no UI to indicate the user has joined the room. You really
> have to create PMs from Riot, which is pretty awful.) But Fractal's
> problems are all well within the skills of our community to solve.
>
> Jan-Michael also likes the Chatty app from Purism. I didn't realize
> that was a Matrix client until today, so I haven't tried it and can't
> vouch for it myself, but it looks similar to Polari or Fractal. So it
> seems we have two GNOME clients for Matrix, zero for Rocket.Chat
>
> > Rocket.Chat has been apparently more responsive to out contact, and
> > even
> > accepted a few pull requests from us. I believe it has a brighter
> > future, specially
> > if a native GTK client shows up.
>
> I've never seen any core Rocket.Chat developers flying to us to give
> talks at GUADEC, like Matthew has done. I think Matthew has perhaps
> stopped focusing on GNOME due to perceived lack of interest on our side
> (and competing time pressures; I guess keeping Mozilla and the France
> government connected is not easy. ;)
>
> Basically my opinion would be: there's a pretty clear industry leader
> here, other open organizations are selecting Matrix after investigating
> available options, why 

Re: Matrix IRC bridge considered harmful

[Alexandre Franke]

On Wed, Feb 12, 2020 at 10:43 PM Michael Catanzaro  wrote:
> The problem there is simply that you can never log off once you join the IRC 
> bridge

Ha! So that must be what Georges was talking about! I didn’t get that
it was about *IRC* servers.

> (or, if such a way exists, it's so hard to discover that the GNOME
> community is not using it).

There actually is a way. Once more it involves the GIMPNet IRC Bridge
status conversation. People can start a new conversation with
@gimpnet-irc:gnome.org by the way if they don’t have it anymore. The
bot has a !quit command and its help says:
`!quit : Leave all bridged channels, on all networks, and remove your
connections to all networks.`

> So if you try the IRC bridge once just to
> see what it's like, you're forever left  with a ghost user account that
> people will send private messages to without realizing you can't read
> them (or even public messages, if your real account is offline).

Removing inactive account is kind of a hard problem (would you be
happy to find out you lost your nick and history if for some reason
you came back a long while after your last use?) but I *think*
matrix.org actually has a policy to clean them up after several
months, which would reduce that issue.

-- 
Alexandre Franke
GNOME Hacker
___
desktop-devel-list mailing list
desktop-devel-list@gnome.org
https://mail.gnome.org/mailman/listinfo/desktop-devel-list

Re: Matrix IRC bridge considered harmful

[Michael Catanzaro]

On Wed, Feb 12, 2020 at 4:23 pm, Georges Basile Stavracas Neto via 
desktop-devel-list  wrote:
The Riot application is hard to use. It took me days to figure out 
how to connect

to a GNOME room. It doesn't allow me to log out of the servers.


These are all problems with the IRC bridge, not with normal Matrix. I 
agree the quality of the IRC bridge is catastrophic. Joining rooms is 
extremely difficult, and it is a Hotel California bridge (you can check 
out, but you can never leave! that's why we have all these trap ghost 
accounts from Matrix who never see the PMs you send them, basically the 
inverse of the problem I started this thread to complain about).


Anyway, normal Matrix has none of those problems. Please don't judge it 
based on the quality of the IRC bridge.


It is a shame that the Matrix developers continue to operate IRC 
bridges that are clearly serving to harm the reputation of Matrix. I 
know there's value in bridging to IRC, but it should be done well if 
it's going to be done at all. No doubt the Matrix developers have 
limited time and competing priorities, just like we do ourselves



Fractal is nice,
as I really like native clients, but Polari feels more polished. 
Matrix apparently
doesn't allow turning off federation, and to me that's a no-go aspect 
of it. At
last, I have a strong impression that Matrix suffers from feature 
bloat.


Fractal needs some love, for sure. I wouldn't want Matrix to be judged 
by the quality of Fractal today. (In particular, all PMs you send to 
users are silently discarded until the other user joins the room, and 
there is no UI to indicate the user has joined the room. You really 
have to create PMs from Riot, which is pretty awful.) But Fractal's 
problems are all well within the skills of our community to solve.


Jan-Michael also likes the Chatty app from Purism. I didn't realize 
that was a Matrix client until today, so I haven't tried it and can't 
vouch for it myself, but it looks similar to Polari or Fractal. So it 
seems we have two GNOME clients for Matrix, zero for Rocket.Chat


Rocket.Chat has been apparently more responsive to out contact, and 
even
accepted a few pull requests from us. I believe it has a brighter 
future, specially

if a native GTK client shows up.


I've never seen any core Rocket.Chat developers flying to us to give 
talks at GUADEC, like Matthew has done. I think Matthew has perhaps 
stopped focusing on GNOME due to perceived lack of interest on our side 
(and competing time pressures; I guess keeping Mozilla and the France 
government connected is not easy. ;)


Basically my opinion would be: there's a pretty clear industry leader 
here, other open organizations are selecting Matrix after investigating 
available options, why not go with what everybody else is doing? Goal 
should be for GNOME developers to only need *one* chat app to do their 
jobs. Let's go with whatever has the best chance of obsoleting IRC, 
which looks like Matrix.


That said, I don't know much about Rocket.Chat. Honestly, I don't think 
I'd heard of it before I rediscovered chat.gnome.org a couple weeks 
ago. So if Rocket.Chat really is open protocol like Matrix -- a 
backwards-compatible protocol that encourages the creation of diverse 
clients, rather than something we just have to hope doesn't change and 
break a future GNOME client -- and if we really seriously plan to write 
a GNOME-style client, then I guess that might turn out just fine, and 
allow us to finally move on from IRC. (I really want to see a GNOME 
client though; a web client is just not good enough.) But Matrix is 
good today. I guess all we need to do is redirect chat.gnome.org to 
gnome.modular.im, turn off GIMPNet, and call it a day. Bonus points if 
we can keep the string "poop" out of any of the domain names.


Michael


___
desktop-devel-list mailing list
desktop-devel-list@gnome.org
https://mail.gnome.org/mailman/listinfo/desktop-devel-list

Re: Matrix IRC bridge considered harmful

[Michael Catanzaro]

On Wed, Feb 12, 2020 at 7:07 pm, Zander Brown  wrote:
My concern would be the "federal" nature of matrix where people don't 
need a
gnome.org specific chat account to join a room. Whilst there are a 
lot of
arguments for this I'm increasingly convinced it's an anti-feature 
especially

if we want to enforce CoC (which, of course, we do)


What's the problem with federation? Can't we just ban individual users 
when there are CoC problems? Surely federation doesn't prohibit 
moderation? I would be far more worried about spammers (a rare problem 
on GIMPNet) than abusive users, and in any case the solution is to give 
moderator rights to more trusted users. It seems that whenever we have 
spammers in GIMPNet (fortunately, a rare occurrence) the IRC ops are 
asleep; simple solution is to just have more ops in more timezones.


I used to use my workplace Matrix account to join the fractal Matrix 
room. Federation worked quite nicely. If not for federation, I would 
not have been able to join the room at all, because fractal does not 
support use of multiple Matrix accounts; obviously, if I have to choose 
between my work account and a GNOME account, I have to pick my work 
account. So I'd say we definitely want federation.


Federation also makes it easier for people from unrelated communities 
to join. E.g. libvirt developers might join the Boxes room, or WebKit 
developers might join the Epiphany room, something like that. It's a 
tough sell to create a new user account to join one particular room for 
a quick conversation, but with federation you can just pop in and 
there's basically no cost.


Michael


___
desktop-devel-list mailing list
desktop-devel-list@gnome.org
https://mail.gnome.org/mailman/listinfo/desktop-devel-list

Re: Matrix IRC bridge considered harmful

[Michael Catanzaro]

On Wed, Feb 12, 2020 at 8:45 pm, Jan Alexander Steffens via 
desktop-devel-list  wrote:
We currently have loads of garbage IRC users in the channels after 
the bridge hosted at matrix.org was replaced with one hosted at the 
gnome.org homeserver. The old bridge left its Matrix users in the 
rooms and the new bridge dutifully bridges them over to IRC. Maybe 
they tried to contact one of these users.


This is another longstanding issue, but it's unrelated. The problem I'm 
complaining about occurs when people try to contact my actual IRC 
account, not any ghost account.


Yes, we have lots of fake users with [m] after their name, and that's 
another big problem, since PMs to them also go to nowhere. The problem 
there is simply that you can never log off once you join the IRC bridge 
(or, if such a way exists, it's so hard to discover that the GNOME 
community is not using it). So if you try the IRC bridge once just to 
see what it's like, you're forever left  with a ghost user account that 
people will send private messages to without realizing you can't read 
them (or even public messages, if your real account is offline).


The solution to both problems is to shut down the IRC bridge.

Michael


___
desktop-devel-list mailing list
desktop-devel-list@gnome.org
https://mail.gnome.org/mailman/listinfo/desktop-devel-list

Re: Matrix IRC bridge considered harmful

[Michael Catanzaro]

On Wed, Feb 12, 2020 at 8:16 pm, Alexandre Franke  
wrote:

Not immediately relevant to the issue but would help me as I’m a bit
confused: weren’t *you* on Matrix rather than IRC?


No, I was using IRC.


I’m pretty sure one gets at least a notification in the GIMPNet IRC
Bridge status conversation that goes along the lines of:

Received an error on irc.gimp.org: err_nosuchnick
["AlexandreFranke","someircnick","No such nick/channel"]

At least I did a couple weeks ago when I tried to send a message to an
IRC user. I would then agree that it is far from ideal, but it would
also clearly not as dramatic as you paint it.


Hm, if so, then that doesn't seem so bad. But this is the first I've 
heard of that. I've had multiple users tell me their messages are being 
dropped without any warning. I've just asked Jan-Michael to look at 
your comment and his response is: "no, nothing," meaning he's not 
receiving any such error message. That said, he uses fractal and 
chatty, not Riot, so *maybe* it could be a deficiency with those 
clients, but I suspect the newer contributor who had trouble messaging 
me was using Riot.



The Matrix folks offered to host our instance on
[Modular](https://modular.im/) just like they already do for KDE and
now Mozilla too[1], so sysadmin time is not a problem, is it?


modular.im would be amazing! I didn't know about gnome.modular.im. If 
it's good enough for Mozilla, surely it must be good enough for us.


Michael


___
desktop-devel-list mailing list
desktop-devel-list@gnome.org
https://mail.gnome.org/mailman/listinfo/desktop-devel-list

Re: Matrix IRC bridge considered harmful

[Bartłomiej Piotrowski]

On 12/02/2020 19.30, Michael Catanzaro wrote:
> Personally, I think native Matrix would be a *lot* nicer than IRC, if we
> have sysadmin time to get it set up, but I'm not going to be picky here.
> I'd just like us to be able to trust that we're not missing important
> messages.

I have reservations if deploying and maintaining fourth chat solution is
the best use of sysadmin time just because there are people who dislike
browser-based service which is functionally IRC, sans federation which
we don't really want.

I see mentions of Modular.im, but we strongly insist on self-hosting
services, even if vendor offers a hosted service for free. There's
always a concern that one day such sponsorship will end, leaving us in
an unpleasant situation and prospect of rushed migration elsewhere. I'm
pretty sure the more experienced half of sysadmin team can bring more
reasons why.

Bart
___
desktop-devel-list mailing list
desktop-devel-list@gnome.org
https://mail.gnome.org/mailman/listinfo/desktop-devel-list

Re: Matrix IRC bridge considered harmful

[Zander Brown]

I've seen this happen more than once

In one room we even had a feedback loop for a while before we mass-kicked the
ghost users (python matrix module all the things!)

On Wed, 2020-02-12 at 21:07 +0100, Alexandre Franke wrote:
> On Wed, Feb 12, 2020 at 8:46 PM Jan Alexander Steffens
>  wrote:
> > We currently have loads of garbage IRC users in the channels after the
> > bridge hosted at matrix.org was replaced with one hosted at the gnome.org
> > homeserver. The old bridge left its Matrix users in the rooms and the new
> > bridge dutifully bridges them over to IRC. Maybe they tried to contact one
> > of these users.
> 
> Ok, I’ll investigate that issue. Thanks for letting me know about it.
> 


signature.asc
Description: This is a digitally signed message part
___
desktop-devel-list mailing list
desktop-devel-list@gnome.org
https://mail.gnome.org/mailman/listinfo/desktop-devel-list

Re: Matrix IRC bridge considered harmful

[Alexandre Franke]

On Wed, Feb 12, 2020 at 8:46 PM Jan Alexander Steffens
 wrote:
> We currently have loads of garbage IRC users in the channels after the bridge 
> hosted at matrix.org was replaced with one hosted at the gnome.org 
> homeserver. The old bridge left its Matrix users in the rooms and the new 
> bridge dutifully bridges them over to IRC. Maybe they tried to contact one of 
> these users.

Ok, I’ll investigate that issue. Thanks for letting me know about it.

-- 
Alexandre Franke
GNOME Hacker
___
desktop-devel-list mailing list
desktop-devel-list@gnome.org
https://mail.gnome.org/mailman/listinfo/desktop-devel-list

Re: Matrix IRC bridge considered harmful

[Jan Alexander Steffens via desktop-devel-list]

On Wed, Feb 12, 2020, 20:16 Alexandre Franke  wrote:

> Not immediately relevant to the issue but would help me as I’m a bit
> confused: weren’t *you* on Matrix rather than IRC?
>

We currently have loads of garbage IRC users in the channels after the
bridge hosted at matrix.org was replaced with one hosted at the gnome.org
homeserver. The old bridge left its Matrix users in the rooms and the new
bridge dutifully bridges them over to IRC. Maybe they tried to contact one
of these users.

>
___
desktop-devel-list mailing list
desktop-devel-list@gnome.org
https://mail.gnome.org/mailman/listinfo/desktop-devel-list

Re: Matrix IRC bridge considered harmful

[Alexandre Franke]

On Wed, Feb 12, 2020 at 8:23 PM Georges Basile Stavracas Neto via
desktop-devel-list  wrote:

> It doesn't allow me to log out of the servers.
>

I’m not sure what you mean by that. What would that action achieve as a
result?


> Matrix apparently
> doesn't allow turning off federation, and to me that's a no-go aspect of
> it.
>

It does. Plenty of deployments are internal and not federated. I disagree
that we should not federate, just as I don’t think we should block any non
gnome.org email address to sent email to gnome.org email addresses, but it
can be done and it’s trivial.


> At last, I have a strong impression that Matrix suffers from feature bloat.
>

I’m sure you would get angry at a user for saying the same thing about
GNOME. You can probably do better than that.


> Rocket.Chat has been apparently more responsive to out contact,
>

Quite the unbacked statement when Matthew has been actively participating
in Matrix related discussions on our lists (and I don’t recall ever seeing
any such discussions for Rocket.chat here).


> and even accepted a few pull requests from us.
>

I have no idea what this is about, but the Fractal team is in constant
contact with the rest of the Matrix ecosystem. Some of our contributors
have contributed to Matrix, some Matrix contributors have contributed to
Fractal…

-- 
Alexandre Franke
GNOME Hacker
___
desktop-devel-list mailing list
desktop-devel-list@gnome.org
https://mail.gnome.org/mailman/listinfo/desktop-devel-list

Re: Matrix IRC bridge considered harmful

[Alexandre Franke]

On Wed, Feb 12, 2020 at 8:20 PM Link Dupont  wrote:
> On Wed, 2020-02-12 at 20:16 +0100, Alexandre Franke wrote:
> > The Matrix folks offered to host our instance on
> > [Modular](https://modular.im/) just like they already do for KDE and
> > now Mozilla too[1], so sysadmin time is not a problem, is it?
> >
> > [1] they just switched, details at
> > https://discourse.mozilla.org/t/synchronous-messaging-at-mozilla-the-decision/50620
>
> https://gnome.modular.im already exists. Do we know who created it or
> who at Modular.im can clue us in?

Yes, Matthew. I’m sure he’ll jump in here at some point. He’s probably
not reading d-d-l as closely as we are, so give him a bit of time. ☺

I’ll poke him if we have a specific question that needs his immediate
attention. In the meantime, I guess the ball is the GNOME court.

-- 
Alexandre Franke
GNOME Hacker
___
desktop-devel-list mailing list
desktop-devel-list@gnome.org
https://mail.gnome.org/mailman/listinfo/desktop-devel-list

Re: Matrix IRC bridge considered harmful

[Georges Basile Stavracas Neto via desktop-devel-list]

Personally, my experiences with Matrix have been catastrophic so far. So
much
so that I'm convinced that Rocket.Chat, with all it's flaws and
misbehaviors, is
a better option on the long run.

The Riot application is hard to use. It took me days to figure out how to
connect
to a GNOME room. It doesn't allow me to log out of the servers. Fractal is
nice,
as I really like native clients, but Polari feels more polished. Matrix
apparently
doesn't allow turning off federation, and to me that's a no-go aspect of
it. At
last, I have a strong impression that Matrix suffers from feature bloat.

Rocket.Chat has been apparently more responsive to out contact, and even
accepted a few pull requests from us. I believe it has a brighter future,
specially
if a native GTK client shows up.

Em qua., 12 de fev. de 2020 às 16:07, Zander Brown 
escreveu:

> 
>
> I've used the matrix bridge for years now (I'm generally only on irc "for
> real"
> to fix things after the bridge does crazy things like de-op me or change my
> nick without warning...)
>
> Matrix isn't perfect. matrix.org, the main "homeserver", regularly has
> high
> latency further exacerbated by the bridge. Hopefully hosting our own would
> avoid that
>
> I know there are some (possibly even the majority of people on this list)
> that
> will never move away from IRC for one reason or another so it does seem
> reasonable to allow IRC access to matrix (rather than the current
> matrix-to-
> IRC). I guess this would still have disappearing PMs but at least it has a
> chance of getting status right giving you a fighting chance
>
> RocketChat is a really nice idea but so far only the web/mobile clients are
> available which leave a lot to be desired whereas Fractal does the job for
> matrix (personally I'm a riot-in-firefox person though)
>
> My concern would be the "federal" nature of matrix where people don't need
> a
> gnome.org specific chat account to join a room. Whilst there are a lot of
> arguments for this I'm increasingly convinced it's an anti-feature
> especially
> if we want to enforce CoC (which, of course, we do)
>
> Zander
>
>
> On Wed, 2020-02-12 at 12:30 -0600, Michael Catanzaro wrote:
> > Hi,
> >
> > I just got an email from a new-ish contributor: "I sent you some PMs
> > about a week ago but I think you weren't online when I sent them so I'm
> > assuming you didn't receive anything." Problem is the Matrix IRC bridge
> > presents all IRC users as online, even when they're not. If an IRC user
> > is offline, it lets you send private messages, but they get *silently
> > dropped*. From Matrix, it appears as if the message was successfully
> > delivered, but it was never actually sent to IRC.
> >
> > Basically our chat has broken down into a dystopian scenario where
> > users message other users, thinking they've successfully sent messages
> > that were never actually sent. We've been living with this for a couple
> > years now and it's just not OK that we tolerate it. I have no way of
> > knowing how many messages I've missed due to this issue, but I'm sure
> > it's causing problems for newcomers who don't realize their messages
> > aren't being delivered.
> >
> > WORKAROUND: Matrix users should ask "you there?" whenever starting a
> > conversation, and assume your message was dropped unless you receive a
> > response. If you get a response back, then a human is reading, at least
> > initially. This applies to all stages of a conversation: if I sign off
> > IRC partway through a conversation, the Matrix user has no way of
> > knowing, so Matrix users must assume all messages sent to IRC after the
> > last message received from IRC may be unread.
> >
> > Anyway, a workaround is not a solution. Can we please either:
> >
> >  (a) Shut down the bridge to Matrix and force everyone to use IRC,
> > which actually works properly; or
> >  (b) Replace our IRC with an actual Matrix server, so we get native
> > Matrix. Matrix is very nice as long as you're not using the abysmal IRC
> > bridge, which is unfit for purpose.
> >
> > Personally, I think native Matrix would be a *lot* nicer than IRC, if
> > we have sysadmin time to get it set up, but I'm not going to be picky
> > here. I'd just like us to be able to trust that we're not missing
> > important messages.
> >
> > (Note I don't include rocketchat in the list of options because I don't
> > consider it a serious option when compared to Matrix, which has become
> > extremely popular and has a variety of client options: desktop Riot,
> > mobile Riot, fractal, or whatever UI you prefer, sure to keep almost
> > everyone happy. Or use the reverse IRC bridge, which I can only hope is
> > not as awful as the bridge we're using now.)
> >
> > Michael
> >
> >
> > ___
> > desktop-devel-list mailing list
> > desktop-devel-list@gnome.org
> > https://mail.gnome.org/mailman/listinfo/desktop-devel-list
> ___
> desktop-devel-list mailing 

Re: Matrix IRC bridge considered harmful

[Alexandre Franke]

On Wed, Feb 12, 2020 at 8:07 PM Zander Brown  wrote:

> I've used the matrix bridge for years now (I'm generally only on irc "for
> real"
> to fix things after the bridge does crazy things like de-op me or change my
> nick without warning...)
>
> Matrix isn't perfect. matrix.org, the main "homeserver", regularly has
> high
> latency further exacerbated by the bridge. Hopefully hosting our own would
> avoid that
>

Yes, it would help a lot both ways:
* we wouldn’t be hurt by performance issues on the matrix.org HS (apart
from federation issue for people “from there”)
* we would reduce the load on the matrix.org HS

My concern would be the "federal" nature of matrix where people don't need a
> gnome.org specific chat account to join a room. Whilst there are a lot of
> arguments for this I'm increasingly convinced it's an anti-feature
> especially
> if we want to enforce CoC (which, of course, we do)
>

That was a concern for Mozilla too. I don’t know the details, but they have
a solution for that it seems. See e.g. the Community safety section in the
[annoucement](
https://discourse.mozilla.org/t/synchronous-messaging-at-mozilla-the-decision/50620
).

-- 
Alexandre Franke
GNOME Hacker
___
desktop-devel-list mailing list
desktop-devel-list@gnome.org
https://mail.gnome.org/mailman/listinfo/desktop-devel-list

Re: Matrix IRC bridge considered harmful

[Link Dupont]

On Wed, 2020-02-12 at 20:16 +0100, Alexandre Franke wrote:
> The Matrix folks offered to host our instance on
> [Modular](https://modular.im/) just like they already do for KDE and
> now Mozilla too[1], so sysadmin time is not a problem, is it?
> 
> [1] they just switched, details at
> https://discourse.mozilla.org/t/synchronous-messaging-at-mozilla-the-decision/50620

https://gnome.modular.im already exists. Do we know who created it or
who at Modular.im can clue us in?

___
desktop-devel-list mailing list
desktop-devel-list@gnome.org
https://mail.gnome.org/mailman/listinfo/desktop-devel-list

Re: Matrix IRC bridge considered harmful

[Alexandre Franke]

On Wed, Feb 12, 2020 at 7:31 PM Michael Catanzaro  wrote:
> Hi,

Hi,

> I just got an email from a new-ish contributor: "I sent you some PMs
> about a week ago but I think you weren't online when I sent them so I'm
> assuming you didn't receive anything." Problem is the Matrix IRC bridge
> presents all IRC users as online, even when they're not.

Not immediately relevant to the issue but would help me as I’m a bit
confused: weren’t *you* on Matrix rather than IRC?

> If an IRC user
> is offline, it lets you send private messages, but they get *silently
> dropped*. From Matrix, it appears as if the message was successfully
> delivered, but it was never actually sent to IRC.

I’m pretty sure one gets at least a notification in the GIMPNet IRC
Bridge status conversation that goes along the lines of:

Received an error on irc.gimp.org: err_nosuchnick
["AlexandreFranke","someircnick","No such nick/channel"]

At least I did a couple weeks ago when I tried to send a message to an
IRC user. I would then agree that it is far from ideal, but it would
also clearly not as dramatic as you paint it.

> Personally, I think native Matrix would be a *lot* nicer than IRC, if
> we have sysadmin time to get it set up, but I'm not going to be picky
> here. I'd just like us to be able to trust that we're not missing
> important messages.

The Matrix folks offered to host our instance on
[Modular](https://modular.im/) just like they already do for KDE and
now Mozilla too[1], so sysadmin time is not a problem, is it?

[1] they just switched, details at
https://discourse.mozilla.org/t/synchronous-messaging-at-mozilla-the-decision/50620

-- 
Alexandre Franke
GNOME Hacker
___
desktop-devel-list mailing list
desktop-devel-list@gnome.org
https://mail.gnome.org/mailman/listinfo/desktop-devel-list

Re: Matrix IRC bridge considered harmful

[Zander Brown]



I've used the matrix bridge for years now (I'm generally only on irc "for real"
to fix things after the bridge does crazy things like de-op me or change my
nick without warning...)

Matrix isn't perfect. matrix.org, the main "homeserver", regularly has high
latency further exacerbated by the bridge. Hopefully hosting our own would
avoid that

I know there are some (possibly even the majority of people on this list) that
will never move away from IRC for one reason or another so it does seem
reasonable to allow IRC access to matrix (rather than the current matrix-to-
IRC). I guess this would still have disappearing PMs but at least it has a
chance of getting status right giving you a fighting chance

RocketChat is a really nice idea but so far only the web/mobile clients are
available which leave a lot to be desired whereas Fractal does the job for
matrix (personally I'm a riot-in-firefox person though)

My concern would be the "federal" nature of matrix where people don't need a
gnome.org specific chat account to join a room. Whilst there are a lot of
arguments for this I'm increasingly convinced it's an anti-feature especially
if we want to enforce CoC (which, of course, we do)

Zander


On Wed, 2020-02-12 at 12:30 -0600, Michael Catanzaro wrote:
> Hi,
> 
> I just got an email from a new-ish contributor: "I sent you some PMs 
> about a week ago but I think you weren't online when I sent them so I'm 
> assuming you didn't receive anything." Problem is the Matrix IRC bridge 
> presents all IRC users as online, even when they're not. If an IRC user 
> is offline, it lets you send private messages, but they get *silently 
> dropped*. From Matrix, it appears as if the message was successfully 
> delivered, but it was never actually sent to IRC.
> 
> Basically our chat has broken down into a dystopian scenario where 
> users message other users, thinking they've successfully sent messages 
> that were never actually sent. We've been living with this for a couple 
> years now and it's just not OK that we tolerate it. I have no way of 
> knowing how many messages I've missed due to this issue, but I'm sure 
> it's causing problems for newcomers who don't realize their messages 
> aren't being delivered.
> 
> WORKAROUND: Matrix users should ask "you there?" whenever starting a 
> conversation, and assume your message was dropped unless you receive a 
> response. If you get a response back, then a human is reading, at least 
> initially. This applies to all stages of a conversation: if I sign off 
> IRC partway through a conversation, the Matrix user has no way of 
> knowing, so Matrix users must assume all messages sent to IRC after the 
> last message received from IRC may be unread.
> 
> Anyway, a workaround is not a solution. Can we please either:
> 
>  (a) Shut down the bridge to Matrix and force everyone to use IRC, 
> which actually works properly; or
>  (b) Replace our IRC with an actual Matrix server, so we get native 
> Matrix. Matrix is very nice as long as you're not using the abysmal IRC 
> bridge, which is unfit for purpose.
> 
> Personally, I think native Matrix would be a *lot* nicer than IRC, if 
> we have sysadmin time to get it set up, but I'm not going to be picky 
> here. I'd just like us to be able to trust that we're not missing 
> important messages.
> 
> (Note I don't include rocketchat in the list of options because I don't 
> consider it a serious option when compared to Matrix, which has become 
> extremely popular and has a variety of client options: desktop Riot, 
> mobile Riot, fractal, or whatever UI you prefer, sure to keep almost 
> everyone happy. Or use the reverse IRC bridge, which I can only hope is 
> not as awful as the bridge we're using now.)
> 
> Michael
> 
> 
> ___
> desktop-devel-list mailing list
> desktop-devel-list@gnome.org
> https://mail.gnome.org/mailman/listinfo/desktop-devel-list


signature.asc
Description: This is a digitally signed message part
___
desktop-devel-list mailing list
desktop-devel-list@gnome.org
https://mail.gnome.org/mailman/listinfo/desktop-devel-list

Matrix IRC bridge considered harmful

[Michael Catanzaro]

Hi,

I just got an email from a new-ish contributor: "I sent you some PMs 
about a week ago but I think you weren't online when I sent them so I'm 
assuming you didn't receive anything." Problem is the Matrix IRC bridge 
presents all IRC users as online, even when they're not. If an IRC user 
is offline, it lets you send private messages, but they get *silently 
dropped*. From Matrix, it appears as if the message was successfully 
delivered, but it was never actually sent to IRC.


Basically our chat has broken down into a dystopian scenario where 
users message other users, thinking they've successfully sent messages 
that were never actually sent. We've been living with this for a couple 
years now and it's just not OK that we tolerate it. I have no way of 
knowing how many messages I've missed due to this issue, but I'm sure 
it's causing problems for newcomers who don't realize their messages 
aren't being delivered.


WORKAROUND: Matrix users should ask "you there?" whenever starting a 
conversation, and assume your message was dropped unless you receive a 
response. If you get a response back, then a human is reading, at least 
initially. This applies to all stages of a conversation: if I sign off 
IRC partway through a conversation, the Matrix user has no way of 
knowing, so Matrix users must assume all messages sent to IRC after the 
last message received from IRC may be unread.


Anyway, a workaround is not a solution. Can we please either:

(a) Shut down the bridge to Matrix and force everyone to use IRC, 
which actually works properly; or
(b) Replace our IRC with an actual Matrix server, so we get native 
Matrix. Matrix is very nice as long as you're not using the abysmal IRC 
bridge, which is unfit for purpose.


Personally, I think native Matrix would be a *lot* nicer than IRC, if 
we have sysadmin time to get it set up, but I'm not going to be picky 
here. I'd just like us to be able to trust that we're not missing 
important messages.


(Note I don't include rocketchat in the list of options because I don't 
consider it a serious option when compared to Matrix, which has become 
extremely popular and has a variety of client options: desktop Riot, 
mobile Riot, fractal, or whatever UI you prefer, sure to keep almost 
everyone happy. Or use the reverse IRC bridge, which I can only hope is 
not as awful as the bridge we're using now.)


Michael


___
desktop-devel-list mailing list
desktop-devel-list@gnome.org
https://mail.gnome.org/mailman/listinfo/desktop-devel-list