Clint Heyer pro­jects


Mobile so­cial soft­ware, an ex­plo­ration in light­weight group com­mu­ni­ca­tion, co­or­di­na­tion and shar­ing.
PhD pro­to­type, 2004-2007.

Rhub was de­signed for light­weight, group-ori­ented so­cial co­or­di­na­tion and shar­ing, mostly by way of SMS mes­sages. I de­signed, im­ple­mented and eval­u­ated it as part of my PhD stud­ies (2004-2007). There was also a soph­sti­catd web­site for rhub, and you could also in­ter­act with it via in­stant mes­sag­ing pro­to­cols and email.

A few core prin­ci­ples evolved in rhub:

Some his­tor­i­cal con­tex­tu­al­i­sa­tion is nec­es­sary. When rhub was be­ing de­signed, iPhone and Twitter did­n’t ex­ist and Facebook was just be­com­ing avail­able. It was the days be­fore apps and the ex­plo­sion of so­cial and mo­bile sys­tems.


While it was pos­si­ble to send mes­sages di­rectly to a per­son, rhub’s core value was in group mes­sag­ing. For ex­am­ple, if you’re in a group of ten­nis bud­dies called tennis’ you can send a mes­sage to them all by SM­S­ing rhub:

tennis: Anyone want to play pairs this afternoon?

Rhub would re­ceive this mes­sage and for­ward it to the group mem­bers, via SMS, in­stant mes­sag­ing or a no­ti­fi­ca­tion if they were brows­ing the web­site. Replying to a group mes­sage would dis­trib­ute the re­ply to every­one - al­low­ing peo­ple to stay in touch with­out hav­ing to ne­go­ti­ate tech­nol­ogy use.

Interestingly, a form of mi­cro-co­or­di­na­tion quickly evolved. People were able to or­gan­ise events at the last-minute, and ad-hoc changes could be made with the min­i­mum of has­sel. For ex­am­ple, friends used rhub to con­verge on a des­ti­na­tion for go­ing out on a Friday night, and then when one of them ar­rives early and sees a large queue to get in, no­ti­fies every­one who are then able to con­verge at an­other lo­ca­tion.

This kind of com­mu­ni­ca­tion was sim­ply not pos­si­ble at the time rhub was made.

Another as­pect of the push’ mes­sag­ing was that the group had much bet­ter aware­ness of shared ac­tiv­i­ties. If you weren’t go­ing out, you could check up on the mes­sages in the morn­ing to get a sense of what peo­ple were up to. The con­stant buzzing of phones was the source of some an­noy­ance as well. One user coined the term rhub rage” for when you get an­noyed by oth­ers send­ing too many mes­sages.

Lightweight, ad-hoc

Throughout the de­sign of rhub, thought was put into how to make it as sim­ple as pos­si­ble to par­tic­i­pate, and start us­ing the sys­tem. Because the sys­tem was used by a rel­a­tively small num­ber of peo­ple (180ish), it was pos­si­ble to de­sign it with­out wor­ry­ing too much about abuse. For example, in the in­ter­ests of re­duc­ing use fric­tion, you were au­to­mat­i­cally added to a group if in­vited.


Rhub was also novel in that it spanned a va­ri­ety of tech­nolo­gies. A single group con­ver­sa­tion could span web, SMS, MSN Messenger, email and MMS.

Rather than be the site where so­cial ac­tiv­ity is sup­pose to take place, rhub is de­signed to sup­port so­cial ac­tiv­ity where it hap­pens best: the every­day, phys­i­cal world. That is, rhub aims to grease the wheels’ of so­cial­is­ing, mak­ing it eas­ier for friends to arrange ac­tiv­i­ties and main­tain aware­ness, and once they are to­gether, for the tech­nol­ogy to re­cede into the back­ground. Post-activity, rhub lets peo­ple col­lect fol­low-up com­ments, pho­tos and so on.

Rapid Agile Iterative Design (RAID)

The de­sign of the sys­tem was led and in­formed by the use of the sys­tem in a it­er­a­tive feed­back loop. Especially for so­cial sys­tems, a lon­gi­tu­di­nal ap­proach is nec­es­sary in or­der for users to build try and ap­pro­pri­ate it.

This ap­proach de­mands care­ful con­sid­er­a­tion that a broad enough de­sign space was ex­plored and that the de­sign pro­ceeded in a prop­erly-in­formed man­ner. These days, this ap­proach is ac­tu­ally rather com­mon in in­dus­try, with the idea of launch­ing a minimum vi­able prod­uct”, but in 2003, it was a novel idea.

Other in­ter­est­ing things…

I ex­per­i­mented with other fea­tures and ideas in the live sys­tem us­ing the RAID ap­proach to see whether there was any value in the idea and how it might be de­signed.

(Again, much of these fea­tures are taken-for-granted now)