[svn.haxx.se] · SVN Dev · SVN Users · SVN Org · TSVN Dev · TSVN Users · Subclipse Dev · Subclipse Users · this month's index

Re: Feature request - configurable intellisense prompts

From: Peter Mounce <pmounce_at_narrowstep.com>
Date: 2007-06-06 10:11:45 CEST

> Stefan wrote:
> Since this is for the commit dialog, I'd rather have that file stored
> directly in the property itself. I guess you already know that I don't
> like any repository (or other remote) access in the commit dialog :)

Yeah - I was hoping it would get downloaded just the once and cached. Actually,
though, it never even occurred to me to store it inside the property itself
- the issue I have there is that I manage the source control for my company
(read, "introduced to" ... :) ), and we follow the one repository-per-project
scheme. Storing it inside the property would mean managing it in ... 34
(and counting) places at once, which I _really_ wouldn't look forward to.
 If it's a file at a URL, once the property is set up, I can edit just one
file.

As an aside: are there any multiple-repository management tools that would
facilitate setting up a common set of properties at a common URL inside each
repo? I saw insurrection, and it looks like it might do that, but I struggled
and couldn't get it to work on my Windows/Apache setup :(

 
>> The XML file could look like:
>>
>> <promptLists escapeValue="$">
>> <prompts trigger="state:">
>> <prompt>Open</prompt>
>> <prompt>InProgress</prompt>
>> <prompt>InVerification</prompt>
>> <prompt>Closed</prompt>
>> </prompts>
> That's not too complicated (apart from being in XML - maybe some clean
> config-file like syntax would do the job too).
>
>> <prompts trigger="time:">
>> <prompt>$hh$h:mm</prompt>
>> <prompt>$mm$m</prompt>
>> <prompt>$h.m spent$:h.m remaining</prompt>
>> </prompts>
>> </prompts>
> This however might be a little bit more complicated.

{snipped my explanation, but not the XML example}
 
> Sorry, but I don't understand how this should work.
> I guess you want to have the autocompletion feature show
> "Open"
> "InProgress"
> "InVerification"
> "Closed"
> as soon as the user types 'state:'.
> But what should be done exactly when the user types 'time:' ? And what
> would pressing the escapeValue char do?

The idea is to present the user with a number of choices (as in, a list prompts)
in the auto-complete list, when they type a specific thing. So, in the tool
we use, it's possible to change the state of an entity by saying "state:{name
of new state}" (curly braces are mine). The point, here, is to publish the
prompt-list of valid states to the user, so they can pick one and not type
it.

Similarly, it can track time. Here, though, it accepts a number of different
patterns in the prompt-list, which do slightly different things:
* h.m - "I spent h hours and m minutes on this entity with this commit"
* h.m:h.m - "I spent h.m on this entity, and I think there's h.m remaining"
...

The $ (or, rather, escapeValue - maybe a bad name!) here, is not meant to
be typed, but indicate either
* cursor position ($h.m:h.m - "offer 'h.m:h.m' to the user, with the cursor
placed at the beginning), or
* selection ($h.m spent$:h.m remaining would be "offer ... to the user, and
if he picks that, select 'h.m spent' so he can immediately type over it).

(It's made slightly more complicated, in our use-case, in that different
entities have different state-names; I've thought how that might get handled,
but that's probably a feature too far...)

Cheers
Pete

---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe@tortoisesvn.tigris.org
For additional commands, e-mail: users-help@tortoisesvn.tigris.org
Received on Wed Jun 6 10:13:03 2007

This is an archived mail posted to the TortoiseSVN Users mailing list.