For "M100 Elan" the problems for me are anticipating imcdb-user needs. Finding current model sets easily from lists is one issue - in general terms things should be where people expect to find them. This has to be a subjective judgement, and is open to discussion of different ways of expressing the model titles. However the bigger test is what happens when future entries occur - maybe next week or maybe in a couple of years time. If the protocol is not what folk expect, then things often get entered with the "wrong" title/code and the collection fragments into different groups, and it can be very difficult to realise when this starts happening and herd them back into the one pot. People come and go from imcdb - I/you/whoever can watch for specific fragmentation while we're active, but if that watcher moves on, then unfamiliar or particularly contrived designations are vulnerable to losing their guardians. Whatever we decide now needs to be robust enough for 5 or 10 years time and whoever the imcdb contributors and admins are by then and whatever level (or lack of) knowledge they might bring. This anticipation is definitely not scientific, but we have to keep trying.

We can and do change naming systems as we go, but for me, if you have a protocol which has several years of successful use without problems, you need a very clear advantage to justify changing it and educating everyone and subsequent monitoring/ensuring the new system takes root completely.

I ain't no Lotus expert but sort of watch over it to mop up spillages etc - there may well be others who do the same in our happy band so I'm not claiming any particular status - but what I can say is the current system on Elans chugs along quite happily - things get entered correctly without needing much additional intervention or correction. Ditto the [Mxxx] system. So I'm wary of changing to a new system unless it's really needed and the old system is definitely wrong by several units of measurement. Refinements are however much easier so 121 for US Elise could be added and probably taken up without much problem (although it seems we already have 121 for the 2006 Europa - if it's a double code, that's fine - we have loads of overlaps already).

I hope that all makes sense - sometimes ease and familiarity of use becomes dominant over strict accuracy because our systems have be simple to pick up by newbies and to work consistently. For instance, we recently discovered a huge swathe of Renault codes, but had to back off them as just not workable - http://imcdb.opencommunity.be/forum_topic-7755-58044-R20__R30_website.html#p58044 was one strand of a bigger thread (which I can't find at the moment).

For the Kia as M100 question, we can not link [codes] across different manufacturers on the current database, but it is one of the features anticipated for the next version, so useful to plug gaps now as we come across them to be ready for the new era.