Re: [Rosegarden-devel] BUGS (connection too unstable for sourceforge)

[Richard B. <bo...@bo...>]

On Wednesday 09 April 2003 17:51, Guillaume Laurent wrote:

> The problem is that the bank editor is not a dialog anymore, but I
> don't see how you can make it working as a dialog if you want both
> undo/redo to work sensibly and "applying" changes whenever you switch
> to another device.

I am confident that what you're doing will be a better solution 
eventually.  I was unfairly giving you a hard time there and I 
apologise.

> Rich, I'm trying my best to remain calm here. I've been doing nothing
> but fixing bugs for several months now. It's been exceptionnal that I
> get more than 6 hours of sleep a night for quite a while, because of
> this. I don't "refactor" for the pleasure of it, I claim to have
> enough professional experience to know better. And believe me, there
> are dozens of places where our code is ugly and needs some amount of
> polish. I don't do it. Given that you've also did just that,
> engineered work for yourself with the tooltip-like widget, I frankly
> feel insulted by your comment.

Well good.  It's about time I got you to react to something.  God knows 
I've been trying for long enough.

I'm not trying to insult your professional integrity but I wonder where 
all your effort is getting us?   What I was doing with the TextFloat 
was basically exactly the same thing as you're doing with the bank 
editor - doing something that's a nice piece of work to do rather than 
getting down and dirty with the important stuff.

Are we really tackling key problems and making sure that they don't come 
back again?  We're seeing multiple reversions at the moment and while I 
know that if we all had the time to hammer them together on a "big 
push" we could knock 'em all out in a couple of weeks - but I wonder if 
we'll have that time to prove a concerted effort again.

Yeah, you're working really hard - we all are - I just don't want any of 
us to have to work harder than we need to and retread old ground 
continuously to get to 1.0.

> What I'm doing on the bank editor is to correct bug 707757, namely
> "undo doesn't work". Now if you think you can get "undo" to work
> properly on the dialog as it was, just tell me. I'll revert all my
> changes, attribute the bug to you, and here you go.

I think that it's not as important as say the control rulers or a 
flexible method of defining controllers for the instrument parameter 
box (or both).  We should pay more attention to prioritisation as 
regards general usability.  A Bank Editor is something that the average 
beginner isn't going to want to play with - they might work out how to 
load a new set of banks from an existing rgd file (reminds me we need 
an SBLive! .rgd pretty badly) but they probably won't want to play 
around making their own for a while.

So I'm concerned we might be tackling the wrong bugs - I'm also 
concerned that we should minimise churn later (in testing towards 1.0) 
by finishing off our features as much as possible as soon as possible - 
even if the code doesn't work yet just get it in the build but make 
sure no-one can get to it.  So yeah, I am being a preachy, insulting, 
teach-your-grandmonther-to-suck-eggs pain in the arse right now but I'm 
doing it because I'm aware that we should be concentrating on the 
important stuff and that we shouldn't be getting bogged down 
frustrating, long-winded pieces of work.

The debate is in the important thing.

Personally I like it when other people do bugs that have been really 
annoying me for ages and I can't work up the energy to fix.   I suggest 
that we create a list of big bugs (based on the sf list) that we could 
raise priority on (in a text file or something away from SF) so we can 
concentrate on them.   Hold a quick poll of the most annoying bugs or 
lack of features.  We should also really try to have a release soonish 
as we've slipped basly on this one - I'm hoping to get audio complete 
enough for a release by the midde of next week - so maybe next 
Friday..?  We _really_ need a release.

In fact I needed this release.

B