Open Source need not be shiny

A very interesting comment on the tragedy of the lurkers (my concern that Blue Obelisk software is heavily used by people who do not show up in the community.)…

1. daen Says:
   October 4th, 2006 at 9:50 am eI can think of several reasons for this, Peter. I downloaded OpenBabel some time ago, wrote a interface layer (which treats C++ object instances as handles), built a DLL from the source using MinGW and wrote a wrapper for Delphi around the whole thing. It was very much a rush job and I have never gone back to clean it up. It was a quick hack but it kind of worked enough for what we were trying to do at the time (doing SMILESMOLFILE conversion driven from an Access database). In my opinion, there’s an ethical issue in contributing code which you know to be sub-standard and have neither the time nor inclination to redact.

Rather than post a reply I’ll expand here.
Firstly, Thanks very much. I understand and appreciate this attitude. However you have now already contributed, simply by announcing that you exist! That, in the first instance, is what we want. It has motivated me to make this post! Your contribution ipso facto enhances Openbabel. It gives us moral support to know that what we are doing is useful. It shows that people want to interface to Access databases. Maybe if 3 people also had done the same the OB project would seriously consider an interface to Access…
But by its nature OpenSource does not regard what you have written as substandard. It doesn’t have to be shiny. When I first saw Jmol it wa a million miles from where it is now. If Dan Gezelter (hope that’s right) had felt it wasn’t worth exposing we wouldn’t have it.
So we have mechanisms for taking code of all sorts. It might get worked on now, it might lie fallow for a year or more. Someone working on it later might even throw it all away. But that doesn’t mean it’s not useful
No one takes contributions to an Open Source project and regards them as “substandard”. They are simply contributions of varying quality and use. They may be useful and buggy or thoroughly tested and irrelevant (apparently) or even possibly both or neither. It’s worth checking beforehand: “I have written soime routines that do X… we are not likely to do any more work on them – would they be useful?” You should always get a courteous and considered reply from the guru.
Contributors are always honoured in an OS project, often in alphabetical order. This survives even if (or often when…) their code is refactored or removed so that not a word of the contribution remains lexically. But the contribution has still been made.
If contributors want to remain anonymous – and there is no shame in that – they can contact the guru privately and, if necessary, use an alias on SF. (My name on SF is petermr – hardly an encryption, admittedly – but I could have called myself zaphod237 or whatever).