This is from Alexandre... I think he has some good points and wanted to forward this on to the list.
Cheers, Josh
---------- Forwarded message ---------- From: Alexandre Prokoudine <alexandre.prokoudine@...23...> Date: Wed, Feb 2, 2011 at 12:23 PM Subject: possible paid project To: "Joshua A. Andler" <scislac@...23...>
Hi Josh,
This mail summarizes my current ideas regarding possible paid project to improve Inkscape. It's also somewhat useful if you decide to forward it to the Inkscape committee.
In the past the committee was strictly against paid work on Inkscape. However we've been participating in GSoC from the very beginning of the program in 2005, which is a lot about paid development, and last year we had an extremely successful project with Tav to improve the Text tool.
Right now we have a little over 13K on our account and we don't make *any* use of it for the good of the project. The $316 that were spent on travelling is really nothing in comparison to the whole sum, contrary to what we state at the Donate page and to what I stated in the interview to Packt Publishing last week. With all respect due, this is nonsense. We are not some kind of a dragon that sleeps on a hurd of gold.
So my proposal is to try doing a couple of paid projects like the one we did with Tav. The money on the account aloows us to do that. But we don't necessarily have to use the money on the account: SFC actually can help us with another donation campaign, it's what Bradley told me regarding the whole LinuxFund story (when the donations bypassed SFC).
If we do several projects, then we need both SVG compliance, performance and new features projects to cover all angles.
SVG compliance project is an "old" idea to do a second project with Tav. He was interested in that.
Performance is partly covered by Krzysztof's GSoC2010 project, but he never really did OpenCL part of it that he was going to do, to the best of my knowledge. At the same time some more bits of Inkscape could still be improved. E.g. there are still canvas related issues like infamous slowdown when doing rubberband selection over a bitmap.
"New features" project could be anything. We have geometry constructions unfinished, PowerStroke LPE started, diffusion curves started etc. I'd rather see one of our former GSoC students work on such a project. We also have over 100 blueprints.
Why former students: because they are very nearly our only serious force these days. The old team is mostly not around or busy working on 2geom. At least 1/2 of new features and refactoring changes come from GSoC projects. The former GSoC students already know their way round the codebase too.
I took liberty to talk to Maximilian Albert for a start. He still sort of owes us finishing geometry constructions tool, but you know that he had rather serious and understandable reasons to not finish the project. He also wrote an interesting proposal for perspective drawing that was later implemented in Adobe Illustrator CS5 in a rather verbatim way. Max is interested in a project, but he wants to know more about organizational part of work: amount of time to spend on the project a week, scope of a possible project etc.
The other former students would be:
Johan Engelen. He was interested in a paid project before, but I'm not sure about how much time he has now. Needs asking.
Krzysztof Kosiński. He recently wrote he was going to take an internship, so maybe he won't be available till autumn.
Jasper van de Gronde. Not sure about how much time he has either. Needs asking too.
Another possible project could be long overdue color separated PDF exporting, but this means a prerequsuite of Cairo being able to do that. It's a considerably expensive project (bulia estimated it it as $30K back in 2008), and, what's also important, is that Jon Cruz is currently on it, which is good and bad at the same time, because Jon has the right understanding of what needs doing, but, with all respect due again, I have my doubts that he'll be able to finish it any time soon.
In my opinion if the decision will be to try doing two or three paid projects, it'd be a good idea to wait till we know what GSoC projects we have this year (if any), because e.g. Tav expressed willingness to mentor a student re SVG compliance.
Proposed sequence of actions in case paid projects get green lights:
1. Talk to other former students whether they have time for a project, and what project. 2. Talk to Tav if he's still ready to do an SVG compliance project in case no GSoC students pops up. 3. Figure out how much money we can spend on each possible project, scope and other criteria considered, whether donation campaign is required. 4. Given p3, talk to Bradley@...28... about what exactly they can do regarding donation campaign. 5. Wait till we know all about GSoC this year. 6. Consider 1-5 and take action.
Alexandre Prokoudine http://libregraphicsworld.org
On Feb 2, 2011, at 12:39 PM, Josh Andler wrote:
Another possible project could be long overdue color separated PDF exporting, but this means a prerequsuite of Cairo being able to do that. It's a considerably expensive project (bulia estimated it it as $30K back in 2008), and, what's also important, is that Jon Cruz is currently on it, which is good and bad at the same time, because Jon has the right understanding of what needs doing, but, with all respect due again, I have my doubts that he'll be able to finish it any time soon.
FYI, the cairo changes needed for print can perhaps start to show up a bit sooner rather than later. Had some good talks with Carl, Eric and others and things are looking up for a decent API to begin.
However, the one thing we need to be careful about is "separations" and "color separated". Those have special connotations that include converting from RGB to CMYK or CMYKOG to the degree of controlling lpi, screen angle, dot shape, etc. All are very complex things that we could avoid if possible.
On Wed, 2011-02-02 at 12:39 -0800, Josh Andler wrote:
In the past the committee was strictly against paid work on Inkscape. However we've been participating in GSoC from the very beginning of the program in 2005, which is a lot about paid development, and last year we had an extremely successful project with Tav to improve the Text tool.
Right now we have a little over 13K on our account and we don't make *any* use of it for the good of the project. The $316 that were spent on travelling is really nothing in comparison to the whole sum, contrary to what we state at the Donate page and to what I stated in the interview to Packt Publishing last week. With all respect due, this is nonsense. We are not some kind of a dragon that sleeps on a hurd of gold.
I'm still against doing paid projects with the foundation's money. I think that introducing having some paid developers and others not is difficult. GSoC is different in that it's Google handling the money and it's also limited to a relatively restricted group of developers (students). I think that the amount of animosity caused by the money is not offset by the amount of work completed.
That being said, yes, we should spend the money. I'd rather spend it on hackfests where the foundation would cover travel and expenses for developers to work on specific projects. So, for instance, do a color separation sprint the week before LGM. Or, perhaps a webpage boot strapping hackfest the few days before SCALE in LA.
--Ted
On Thu, Feb 03, 2011 at 09:19:39AM -0600, Ted Gould wrote:
I'm still against doing paid projects with the foundation's money. I think that introducing having some paid developers and others not is difficult. GSoC is different in that it's Google handling the money and it's also limited to a relatively restricted group of developers (students). I think that the amount of animosity caused by the money is not offset by the amount of work completed.
That being said, yes, we should spend the money. I'd rather spend it on hackfests where the foundation would cover travel and expenses for developers to work on specific projects. So, for instance, do a color separation sprint the week before LGM. Or, perhaps a webpage boot strapping hackfest the few days before SCALE in LA.
I'm with Ted, $13k doesn't go very far, even for student projects (equivalent to perhaps two SoC students). Unless there was a very unsexy project and the right person we would be better off using the money for tee shirts and travel.
njh
On Thu, Feb 03, 2011 at 03:41:35PM +0000, Nathan Hurst wrote:
On Thu, Feb 03, 2011 at 09:19:39AM -0600, Ted Gould wrote:
I'm still against doing paid projects with the foundation's money. I think that introducing having some paid developers and others not is difficult. GSoC is different in that it's Google handling the money and it's also limited to a relatively restricted group of developers (students). I think that the amount of animosity caused by the money is not offset by the amount of work completed.
That being said, yes, we should spend the money. I'd rather spend it on hackfests where the foundation would cover travel and expenses for developers to work on specific projects. So, for instance, do a color separation sprint the week before LGM. Or, perhaps a webpage boot strapping hackfest the few days before SCALE in LA.
I'm with Ted, $13k doesn't go very far, even for student projects (equivalent to perhaps two SoC students). Unless there was a very unsexy project and the right person we would be better off using the money for tee shirts and travel.
My vote's with Ted and Nathan. I think Ted's idea of a hackfest or development sprint is brilliant. Or even just offsetting LGM travel costs (like, up to $500) for as many Inkscape contributors as possible.
Bryce
participants (5)
-
Bryce Harrington -
Jon Cruz -
Josh Andler -
Nathan Hurst -
Ted Gould