[guardian-dev] Moving beyond XMPP

[Hans-Christoph Steiner]

On 12/06/2013 04:32 PM, Nathan of Guardian wrote:
> On 12/06/2013 02:57 PM, Hans-Christoph Steiner wrote:
>> I agree that I think a big focus on usability bugs would be helpful.  We get
>> quite a few downloads of it, but we have only about 25-30% ongoing users.  I
>> don't know how that compares to other Android apps but I think we should
>> improve that ratio.  Then we'll have more users, and more people recommending
>> it to others.
> 
> I think that it is so easy to install apps, that you have many, many
> drive-by users who try something once, and then are distracted and move
> on. That is not to say that we shouldn't do a better job of keeping
> people around, using our app, but I also think that while 25-30% seems
> low, that it isn't as bad as it may seem.
> 
> Alternatively, the reasons that people stay in a social network have
> more to do with social pressure, and not necessarily usability,
> specifically.
> 
>> It sounds like it would be a lot of work to integrate support for the
>> proprietary protocols, and it would be probably quite trivial for those
>> companies to specifically block ChatSecure on their networks.
> 
> There are open implementations of WhatsApp's protocol in Python already
> that are used without being blocked.
> 
> https://github.com/tgalal/yowsup
> https://github.com/tgalal/yowsup/wiki/yowsup-cli
> https://github.com/tgalal/wazapp
> 
>> In the long run, it probably makes sense to add support for as many protocols
>> as possible, like Pidgin/Purple/Adium do.  That will be a large undertaking
>> beyond what we are currently capable of.
> 
> Yes, we are back to that cat and mouse game of the AIM days, I agree.
> Still, if it is only a few tweaks to XMPP, then I am for it. If it is
> 100% custom, then not so much.
> 
> +n

Right now, I think its a distraction to work on adding protocols.  I think the
best plan would be first to make it a competitive client to WhatsApp.  We're
close to that.  Then when we have inline sound and images down and the general
usability down, then its a good time to add WhatsApp support.  If it just a
worse version of the WhatsApp client in the eyes of WhatsApp users, then we've
screwed ourselves.  If its a good replacement with real privacy, then we won't
lose people on their first impression.

Its good to see that people are developing libs for those new proprietary
protocols.  We could port this Blackberry C++ to JNI or Java:

https://github.com/tgalal/OpenWhatsappBB

There is also some activity related to Pidgin/Purple.  Reading up on that, my
guess is managing their crazy password could be quite a pain.
https://developer.pidgin.im/ticket/14850

.hc

-- 
PGP fingerprint: 5E61 C878 0F86 295C E17D  8677 9F0F E587 374B BE81