Help choose my project for Ruby Mendicant

[Gregory Brown]

Hi folks,

So far, the Ruby Mendicant (rubymendicant.wikidot.com) fundraising
campaign has been going great.

== News About Donations ==

Over the last couple weeks, I've received contributions from 52
people, running the gambit from close friends to people I've never
even spoke with before.

With the help of RubyCentral's donation matching, I've raised enough
funds to work on open source Ruby as my primary responsibility for at
least 12 weeks, which is a decent chunk of time!

== The Projects ==

It's now time to figure out how to best spend this community funded
time. To simplify things, I've limited my project ideas list to the
three that have received the most community support:

* A clean, fast, light-weight rewrite of PDF::Writer.
* Ruby 1.9 compatibility work for a number of RubyGems
* 'Uncovering Hidden Gems' documentation project, writing tutorials
and quick references for lesser known Ruby libraries.

You can see more details of these projects here:
http://rubymendicant.wikidot.com/projects

I need your help selecting a project, but the way the projects will
be
selected has changed, so please read on.

== What About Voting? ==

When I started this project, I considered doing proportional voting
based on donation amounts. Unfortunately, I've been getting some
(mostly private) negative feedback about this.
The general suggestion is that if Ruby Mendicant is meant to serve
the
whole Ruby community, decision making shouldn't be tied directly to
donations.

For this reason, I am going to try to select a project based on
consensus, rather than doing an official 'vote'. Because of this, we
need to start the discussion much earlier,
rather than just opening up votes for a week or so after April 1st.

Starting now, I encourage folks to go over these three projects and
see if you have any strong opinions (good or bad) about them.
Then, please get involved by discussing your thoughts on the Ruby
Mendicant mailing list.

http://groups.google.com/group/rubymendicant

If there is a clear winner from the above ideas, we'll easily go with
that.

If the consensus becomes something to the effect of "any of these
projects are good", I'll select the project I like best.

If people have a hard time agreeing on a project, I'll need to do
some
sort of voting, but I hope it doesn't result in that.

Right now, we have some time to discuss this as much as necessary.
I
won't be selecting a project until after the 1st, and possibly not
until April 10th or so.
However, the quicker we can come to an agreement about this, the
faster I'll be able to start planning out my work for the coming
months.

I hope that people can appreciate the community-centric approach
here,
but please get in touch with me if you donated and this deeply
concerns you.
If you haven't donated yet and this concerns you, please contact me
*before* donating. :)

== Show Your Support ==

This pretty much sums up all the news on Ruby Mendicant. Though
planning will surely be underway soon, we're still in the fund
raising
stage for now.

There are still 7 more days where you can make a contribution, any
amount is welcome!

http://pledgie.com/campaigns/571

You can also link back to the campaign or tell your friends. So far,
this has been enormously effective in spreading the word, and I thank
anyone who's stuck a badge or did
a quick post on their blogs, I see a ton of people coming in from
your sites!

[znmeb]

On Mar 24, 9:54 am, Gregory Brown <gregory.t.br...@gmail.com> wrote:
> * A clean, fast, light-weight rewrite of PDF::Writer.

I like this one, and it fits well with Ruport. It's also of pretty
broad appeal independent of Ruport, and it's a focused effort.

> * Ruby 1.9 compatibility work for a number of RubyGems

Well ... I'm really on the fence about this one. It seems to me that
the project that owns the gem should/would have done the recruiting,
rather than expecting a talented volunteer to arise. And I think it's
scattered. Now if there's *one* gem that really needs this kind of
love, for example Ruport, Ruport-Util or PDF::Writer, by all means go
for it. :)

> * 'Uncovering Hidden Gems' documentation project, writing tutorials
> and quick references for lesser known Ruby libraries.

Again I suspect this is scattered and would dilute your focus. Let the
projects in question recruit tech writers.

So ... I guess I would say rewrite PDF::Writer, make sure it's
compatible with 1.8.x, 1.9.x and jRuby, and if you have cycles left
over see if you can get it working with Rubinius.

[rogerdpack]

My vote for rails ruby 1.9 compatibility *wink* (oh throw in mysql and
sqlite3 in there).
Thanks for doing that stuff.
-R

[Gregory Brown]

On Tue, Mar 25, 2008 at 2:53 AM, rogerdpack <rogerp...@gmail.com> wrote:
>
> My vote for rails ruby 1.9 compatibility *wink* (oh throw in mysql and
> sqlite3 in there).
> Thanks for doing that stuff.

Rails 1.9 compatibility is not going to happen, I don't have the
domain expertise, but ActiveRecord and sqlite3 would be possible in
1.9 work, especially because they're optional deps of Ruport....

-greg

[Nick B.]

Greg,

I think the ordering and wording of projects says a lot and I think
that PDF::Writer is a much better project to spend time on. Ruby lacks
a good PDF library and that's just one of the many reasons why some
people pick other frameworks/languages over it. Java folks have iText,
Python people have ReportLab etc. With some attention, PDF::Writer
could surpass all of the other PDF libraries.

The other proposal, Ruby 1.9 Field Medic, involves fixing other
people's code and I think you'd waste a lot of time on it. The reason
is that original authors and maintainers of those packages have the
necessary skills (and plans) to upgrade the packages themselves... and
only you and Mike really know how to fix PDF::Writer's issues. I also
think that focusing on one package would be more productive than
trying to do too much and work on all those other projects... quality
vs quantity issue.

Anyway, just some thoughts and best of luck no matter what you decide
to do... it's your time after all!

/nick

[Gregory Brown]

On Wed, Mar 26, 2008 at 10:54 AM, Nick B. <nik...@gmail.com> wrote:
>
> Greg,
>
> I think the ordering and wording of projects says a lot and I think
> that PDF::Writer is a much better project to spend time on. Ruby lacks
> a good PDF library and that's just one of the many reasons why some
> people pick other frameworks/languages over it. Java folks have iText,
> Python people have ReportLab etc. With some attention, PDF::Writer
> could surpass all of the other PDF libraries.
>
> The other proposal, Ruby 1.9 Field Medic, involves fixing other
> people's code and I think you'd waste a lot of time on it. The reason
> is that original authors and maintainers of those packages have the
> necessary skills (and plans) to upgrade the packages themselves... and
> only you and Mike really know how to fix PDF::Writer's issues. I also
> think that focusing on one package would be more productive than
> trying to do too much and work on all those other projects... quality
> vs quantity issue.

Well, the problem is we don't really know how to fix PDF::Writer's
issues, it's almost entirely Austin Ziegler's code, ported from a PHP
library.

However, given some dedicated time, I could build a very clean, very
Rubyish (non-ported) PDF library that works to support Ruby 1.9 and
m17n from the ground up.

Though my code name for this project is Prawn, as soon as it is fully
capable, I'd like to see it become PDF::Writer.

It seems like lots of people are behind this idea, where the other two
projects have had mixed opinions.

I'll drop a seperate post on this, suggesting that PDF is currently
the most popular project I've mentioned.

Thanks for sharing your thoughts!

-greg