If before a checkin you just recursively add everything fossil very nicely 
doesn't add things in twice.  I.e. if it's already in the repository it doesn't 
put it in again.  The new stuff will be added and your problem is handled.  The 
only problem is adding in files you don't want in the repo (i.e. obj files or 
the like).  But if all your (say source) files were in a specific tree then 
recursively adding that wouldn't hurt.

        --jim

On 10 Feb, 2010, at 8:33, Benjohn Barnes wrote:

>       Problem
> 
> I forget to add in new files that I create. This leads to me creating  
> one or more check-ins that aren't a complete representation of the  
> working set of files. Worse, the new files are often the ones that I'm  
> working on most actively because they're new.
> 
> 
>       Suggestion
> 
> Could fossil flag these untracked files to me, and in fact, not allow  
> me to check-in until I've either added the files to the project, or  
> told fossil that I don't want them added?
> 
> 
>       Discussion
> 
> 1) Yes – I realise this is just me being stupid :-) But I'm a human,  
> and I'm fallible, and I like tools to help me improve on that.
> 
> 2) Maybe fossil already does this and I just don't know about that  
> setting.
> 
> 3) I'm not necessarily asking for fossil features here – I'd be happy  
> with another approach. However, I like the fossil philosophy of doing  
> all the stuff it should do with minimum fuss or configuration, and  
> without external tool requirements.
> 
> 4) In my working set, I have a small number of files that I do not  
> want to place under revision control. They're essentially settings  
> files in use by my IDE. I've made some effort to keep this set of  
> files small (I have two, in fact), keeping build artefacts in another  
> location. I think it is a normal situation for the working set to be  
> polluted by such files (it's been the case with all IDE and build  
> environments I've used). I'm assuming that it is reasonable for these  
> to be there. This means (I think) that I require a way of marking  
> files (and perhaps patterns of files, although in my personal current  
> case I don't need patterns) as explicitly not being under revision  
> control.
> 
> 5) Some or many people might not want this feature in some or many  
> situations, so it should probably be possible to disable it with a  
> setting and on an ad-hoc basis with a force.
> 
> 6) When I perform "fossil extra", I'd want it to only list the files  
> that are not tracked and not to be ignored, although it would be  
> useful to be able to list all extra files too.
> 
> 
> Thanks,
>       Benjohn
> 
> 
> 
> 
> _______________________________________________
> fossil-users mailing list
> [email protected]
> http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

_______________________________________________
fossil-users mailing list
[email protected]
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users

Reply via email to