On Tue, Jan 26, 2010 at 6:57 PM, Yuen Hoe Lim <yuenho...@gmail.com> wrote:
> What about giving the user a choice when a duplicate is found? As a user, I
>> wouldn't like the software to create two copies of the same project (maybe
>> different versions of it) and then me having to delete one of them. Lets
>> just give a "Overwrite, Rename, Skip" option.
>> What do you guys think ?
>>
>
> My vote is against, for the following reasons:
>
> - True duplicates (the kind we don't want to have) will only occur in
> one of two scenarios.
> a) I export all projects, then import all projects on the same PlasMate
> instance, which is a senseless thing to do. Or I import twice onto the same
> instance, again senseless.
> b) I export my projects and then go to another computer where someone
> else has imported one of my projects (from online, say). This is gonna be
> rare, and isn't really a true duplicate. You may actually want to keep
> both.
> The vast majority of conflicts are likely going to be the kind where
> you unintentionally name two different things with the same name.
>
> - Even in the rare case that true duplicates occur, the remedy is
> trivial. One button-click away.
>
> - Prompting with "Overwrite, Rename, Skip" is bad because, as I have
> repeatedly reiterated, folder names are invisible to the user. ie, when
> there is a folder name conflict - **the user cannot see it**! If the
> user doesn't see a conflict but PlasMate prompts and insists that there is
> a
> conflict, we are definitely going to confuse the user. Furthermore if he
> decides to rename but doesn't see any visible effect from the renaming
> (because folder names are not visible!), he is going to become even more
> confused. And worse - he will probably report it as a bug.
>
> Er, there will be a folder name conflict only if there is a project name
conflict, right? Only If I already have a project "abcd" and I import which
already contains "abcd", there will be a conflict. Correct me if i'm wrong.
>
>
>
> If you guys insist I guess we could use a hash check to ignore "pure" true
> duplicates, so the algorithm reads something like this:
>
>
> for each folder f in import archive {
> for each folder g in the target location {
> if hash(f) == hash(g) { // a "pure" duplicate
> delete f from archive;
> goto end;
>
> }
> }
> while f also exists in the target location {
> generate a new name for f and rename f in the archive;
> }
> end:
>
> }
> extract archive to target location;
>
> But particularly for the third reason above, I really don't like the idea
> of prompting.
>
>
> ----
> Jason "moofang" Lim Yuen Hoe
> http://yuenhoe.co.cc/
>
>
>
>
> _______________________________________________
> Plasma-devel mailing list
> Plasma-devel@kde.org
> https://mail.kde.org/mailman/listinfo/plasma-devel
>
>
--
Shantanu Tushar (UTC +0530)
http://www.shantanutushar.com
_______________________________________________
Plasma-devel mailing list
Plasma-devel@kde.org
https://mail.kde.org/mailman/listinfo/plasma-devel
