Re: Maintaining / Helping out on Dia

[Zander Brown <zbrown gnome org>]

On Wed, 2018-12-05 at 12:59 -0500, Alejandro Imass wrote:

On Wed, Dec 5, 2018 at 12:25 PM Zander Brown <zbrown gnome org> wrote:

On Wed, 2018-12-05 at 11:34 -0500, Alejandro Imass wrote:

[...]

 

You seem to know exactly what needs to be done to get through the GTK 2 bump, and it awesome that at least someone does and understands the urgency to AT LEAST bump it up to GTK 3. We use DIA on daily basis and don't want to see it die. So we could cough up some cash and finance some of this effort.

I'll investigate what the policy for that is (don't want to get anyone is trouble esp myself)

What I was thinking is that we could channel some funds to a University and have them work on it. My native country (Venezuela) is in pretty bad shape but top-notch Universities are still operating despite the crisis. This would be a win-win IMHO.

Worth investigating

 

In your opinion is is too crazy to suggest separating the UI/UX from the logic and salvaging the C code, templates and all the other non-UI stuff  as a "backend". 

It seems reasonable to have a more formal split between Dia & libdia but I'm not sure it's worth reimplementing the UI in JS.

For sure not right now, and it's quite obvious to me that a straight GTK 3 upgrade is the logical next step for DIA. But if we work towards the goal of at least foreseeing that additional UIs to be able to bolt-on, then I think it would be great for DIA. Especially a web-based version that competes with draw.io.

Not sure directly competing with draw.io would work especially well

To implement the UI in anything other than C/Vala would make the build quite complex and we would lose the advantage of compile time checks however we should be able to somewhat simplify the UI implementation by taking advantage of GtkBuilder (once we make it to Gtk3) to define our UI in XML but until the Cairo port is done we are largely talking hypotheticals here

Was there any particular reason you wanted to see Dia's interface reworked in JS/JSX?

No reason at all. Just thought that since the effort is relatively big to modernize DIA to GTK 3 it would be worthwhile to at least discuss it. Moreover, I think a future Web-based interface to DIA (even if not the primary/native one) will be important for the long-term survival of DIA from a product standpoint.

Personally I'd like to see Dia remain a 'native' application (which seems to be the general view of others on the list as well) and plugable interfaces generally get very messy very fast (esp when we do a lot of custom rendering) so I'd rather Dia remains true to it's roots as a Gtk application

-- 

Zander Brown <zbrown gnome org>
GNOME

Attachment: signature.asc
Description: This is a digitally signed message part