On Mon, Aug 13, 2012 at 8:39 AM, Peter Bittner <
peter....@gmx.net> wrote:
> Anthony,
>
> I've opened an issue on that very task:
>
https://github.com/pyjs/pyjs.org/issues/9
>
> Basically, I compiled what we have already discussed about the
> "examples index" and "broken examples" roughly in May this year.
>
> 2012/8/13 C Anthony Risinger <
ant...@xtfx.me>:
>> yes quite a few examples are either crappy, outdated, or completely botched.
>>
>> ... this is just one of the many maintenance issues that has arisen
>> from a former "add add add everything anywhere" approach.
>>
>> i will probably be axing many examples in good time, or publishing
>> only a select few on the website.
>
> We shouldn't be repetitively condemning the past. I believe "no
> examples" are still worse than some that "usually work", or work
> partly. One task of examples still is giving an idea of what it is
> like using the technology in question. The several code pieces give an
> idea, and if they are from different people, we may see different
> approaches or coding styles, and learn from it. We should be careful
> on which efforts to throw away.
weeell, i don't think i'm repetitively doing anything (i'm not sure
i've even alluded any problem from the past), but i don't agree much
with your sentiments :-( ... the repository is absolute chaos with the
obvious rotting bits all over the place, and i'm more inclined to
slash first and resurrect later if useful.
IMO anyway, 2 busted examples in a row would likely cause me to move
on, or at the very least, reconsider/search-elsewhere-harder ... no
published examples means lack of time and/or resources; knowingly
publishing 50% busted examples is essentially incompetence and/or lack
of pride.
the examples are written by many people for sure, but IMO many are
poorly written and lacking any consistent style whatsoever -- in the
HTML or the Python code -- giving an absolute "scatterbrain" feel to
the whole thing without providing good context for WHY each example
exists or is useful at all.
inconsistency is the enemy of sustained forward development -- any
example surviving the gauntlet must eventually conform to the
established style (roughly PEP8), and a more rigorous entrance exam.
the first task however is converting all examples to packages,
importable from the `examples/__init__.py`.
> Yes, I do agree on consolidating! What needs to be done is find out
> which examples roughly cover one and the same topic, and turn them
> into ultimately "great examples" that look great in both, presentation
> and source code. To start with we could cluster the existing examples
> to get a better overview of what is out there already.
sounds fantastic, and a good project for anyone looking to take the plunge.
--
C Anthony