[darcs-users] Re: Frustrations diffing against the last change to a file

[Thomas Zander]

On Saturday 02 April 2005 01:47, you wrote:
> On Fri, Apr 01, 2005 at 07:25:55PM +0200, Thomas Zander wrote:
> > when you do a darcs pull you can view all changes per patch in the pull
> > interface; pressing the 'x'.
> > Easy enough to see what happened when you were gone since pull is
> > probably the first thing you do.
>
> Pulling is something which happens in the past, I am not going to
> remember everything that I pulled. 

You missed that my answer was to the situation of Gerhard wanting to view 
all changes after his vacation.

> > I'm not sure why you want to look at diffs per file;  the whole concept
> > of changesets is chosen because a patch rarely stands alone..
>
> Because if one is working on fileX then one wants to see the changesets
> which include fileX.  Those change sets might include other files but I
> only want to see those which include fileX.  If fileX is not included in
> the change, I don't care about it.

Please try the commands I pasted before hitting reply;  I wrote
   darcs changes fileX
in my last mail (well it was Makfile then :)  which does exactly this.

> And the assertion that patches rarely stand alone is not true.  In a
> large

If that is the case; then you just made the point for me;  that you still 
want to look at changesets instead of change per file because in your 
projects that the same thing.  So use darcs very good ways to look at a 
particular changeset and don't worry to much about files.


> If I have to go down to the XML interface for a trivial operation like
> looking at the last change to a file something is Wrong.

You naming each and every fix you make 'fixed bug' is also wrong.

> Ahh, do not confuse "how do I do this *better*?" with "how do I do this?"
> I think if you read the discussions as "yes, I can do this way but I feel
> it can be improved" they won't seem quite so pointless.

Adding revision numbers to a system that is designed in such a way that 
revision numbers are inherently unneeded an counter productive is not an 
improvement, I'd call it a step back.
This has been said and argued; and all you keep doing is saying you don't 
want to copy paste and you want to keep naming your patches "fix". 
On top of that you say that its ugly and wrong to use the low-level stuff; 
while that may be; I suggest you write a little GUI using perl to show this 
patch for you.

To lift this stalemate what about this;
darcs changes/annotate/diff/etc learn to handle a unique code that is a 5 
character (alfaneric) hash of the comment/date of each patch which is 
unique enough to keep collisions low, and have the advantage of being 
consistent after pulls/unpulls.

darcs changes --hashcode build.xml:
kJ6xM: Sun Jan 30 18:51:43 CET 2005  Thomas Zander <zander at kde.org>
  * Fix unit-test framework again

so you can type:
darcs diff --hashcode kJ6xM

Any darcs-developers want to comment on this?  Michael; will this solve your 
problems?
-- 
Thomas Zander
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: not available
Url : http://lists.osuosl.org/pipermail/darcs-users/attachments/20050402/abe0def6/attachment.pgp