[reportlab-users] Python 3000 and 3 possible 'branches'

[The Helmbolds]

IMHO half measures utlimately are more trouble and effort than going full-bore to the goal. So I think your option (b) is significantly less desirable than either of the others. In fact, I favor your option (c) -- even if I have to stick with Python 2.x for a while longer.

Message: 6
Date: Fri, 5 Dec 2008 11:19:21 +0000
From: "Andy Robinson" <andy at reportlab.com>
Subject: Re: [reportlab-users] Python 3000
To: "Support list for users of Reportlab software"
    <reportlab-users at reportlab.com>
Message-ID:
    <956003ae0812050319rb4a7bf1hc4e6dbe5900d76b0 at mail.gmail.com>
Content-Type: text/plain; charset=UTF-8

I've been pondering this more.  I see 3 possible "branches"...


(a) the present codebase running on 2.3 to 2.6.  But we aim to remove
dead code and examples, document and package a bit better, and make
APIs a bit more explicit, so it's clearer what needs porting.

(b) a Python 3.0 port which aims to present the same behaviour (and
thus the same undocumented mishmash of APIs as we have now), but
needing some significant internal rewrites to achieve it

(c) a Python 3.0 port which aims to deprecate any stuff we don't need,
properly take advantage of the clear distinction between
natural-language text and byte-arrays, restructure anything we want
to, and maybe catch up a bit with recent PDF specs.    This would
probably start off by porting just the lowest layers (pdfgen/pdfdoc)
and we could try to build it up cleanly with good docco and test
coverage as we go, in no rush.

Does anyone here think (b) is worth pursuing to production quality,
given that we have limited resources, and time spent on it would cut
into time spent on (c)?


-- 
Andy

 The Helmbolds
2645 E Southern Ave A241
Tempe AZ 85282
Email: helmrp at yahoo.com
VOX: 480-831-3611 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://two.pairlist.net/pipermail/reportlab-users/attachments/20081205/7c654e4f/attachment.html>