Skip site navigation (1)Skip section navigation (2)
Date:      05 Apr 2002 10:57:13 -0800
From:      swear@blarg.net (Gary W. Swearingen)
To:        Antoine Beaupre <anarcat@anarcat.ath.cx>
Cc:        freebsd-libh@FreeBSD.org
Subject:   Re: libh learning curve, etc.
Message-ID:  <siadsi2cty.dsi@localhost.localdomain>
In-Reply-To: <370BA4FE-48A2-11D6-945B-0050E4A0BB3F@anarcat.ath.cx>
References:  <370BA4FE-48A2-11D6-945B-0050E4A0BB3F@anarcat.ath.cx>

next in thread | previous in thread | raw e-mail | index | archive | help
Antoine Beaupre <anarcat@anarcat.ath.cx> writes:

> If you ever worked on a already-started project, you would know that
> it's much harder to document and promote code that's already written.

Nobody expects detailed source code documentation.  Almost no free
software has that.  But some current (and prospective) high-level
documents would benefit the project.  Of course, I realize that many
people join projects for reasons other than the benefit of the project,
and I respect that, but it seemed clear that the -libh project was
devoting almost none of its time to Public Relations and that sort of
thing and very much needed to be "reminded about it" (as you put it).

As a former Linux web index maintainer, I've seen a lot of project web
sites, and think I have a feel for their importance to the project.
(No evidence, but a feel, nonetheless.)  For instance, the early GNOME
web site was a marvel of good, simple layout, covering all the bases
and looking good without being flashy.  And I suspect that it was
responsible for encouraging the entry of many developers to the project.

> I think it might be a good idea to put a tarball of a port snapshot on
> the website thought.

Good idea.  And links to web sites for TVision and TCL and Qt (and to
licenses to those and your own).  And links to articles like the one JKH
did last September for the 5.0 to-do list.  People can use Google, like
I did, but you should make it easy for them.  Google returned 10 pages
on "freebsd libh" but there were maybe 5 pages of good value.

> > -- The CVSWEB tool at freebsd.org has only a few, uninteresting libh
> > files.
> 
> That's because libh is not part of the main freebsd CVS repository.

There's a few files there.  Enough to waste 10 or 15 minutes figuring it
out.  You should say on your website what you just told me.

> > -- The mailing list archives are mostly mind-numbing commit messages.
> 
> *That's* not constructive.

Why not?  I can envision a project with a separate list for such notices.

> 1- get sponsorship
> 2- get people to work on libh
> 
> The problem with [1] is that noone seems to be interested in sponsoring
> a tool so unimportant (!). And the problem with [2] is that we need to
> work on documentation and design documents, and to get there, we need
> [2].

From the looks of the list archives, you have some coders.  I just
wanted to remind them that you can generate more code by attracting more
coders than you can by coding.  Of course, personal purposes are a
factor and you have to let yourselves spend most of your time doing what
benefits yourself, not the project.  But hopefully not 99.94% of it.

> There was some good criticism, but I'm getting tired of seeing people
> just pass by the page and say: "Hey, this looks like it needs some help,
> let's complain to -libh@!".  Yes the code is under-documented, yes, we
> need some solid design documents. Thank you for reminding us.

Forget about the code being under-documented.  Document at a high level.
Turn the JKH memo into a document that at least looks like it reflects
the current design.  Give people a few diagrams to look at.  Etc. 
 
> I hope you'll take this as constructive criticism; it wasn't meant to
> hurt feelings though it probably will.

No hard feelings here; I expected a worse reaction and hesitated to
cause it.  But thought it worthwhile, even if it does damage my name
before I've totally abandoned any plans to join the project.
 
> Maybe I'm just ranting and I should shut up as I had the same reaction
> as you the first time I got interested in libh.

I'd say the project was very lucky to get someone as willing as you were
to work past the reaction and work hard enough to learn whether you
wanted to be a part of the project.

To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-libh" in the body of the message




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?siadsi2cty.dsi>