OSError [Errno 26] ?!?!

[Tim Chase]

CGI didn't stop working just because more powerful, or better,
alternatives now exist.

And for most exceptionally cheap hosting services, your choices are
usually limited to PHP, static HTML (possibly server-side includes if
you're lucky), or CGI. So CGI is often the easiest way to get
dynamic websites written in Python (or Perl) onto cheap web
providers. And for the small sites that use such hosting, the
overhead of CGI doesn't generally cause much issue.

-tkc

 

[Joshua Landau]

If you are a beginner to a programming language, assume that anything
that doesn't work the way you expect is a bug in YOUR code, or YOUR
understanding, not in the language.

Not just beginners. Out of the hundreds of times where I've gone "this
*can't* make sense", I think only one -maybe two- was an actual bug in
Python.

 

[Steven D'Aprano]

This forum (both a Usenet newsgroup and a mailing list) is part of the
Python community. So the “Python Community Code of Conduct†applies.

<URL:http://www.python.org/psf/codeofconduct/>

It certainly does not apply. Even putting aside the practical
impossibility of enforcing it on an unmoderated mailing list/news group,
the Code of Conduct needs to be adopted by each community group that
chooses to do so. The Python Software Foundation (PSF) has approved the
CoC, but it is only *suggested* that Python-related communities adopt it,
it is not mandatory. Nor could it be mandatory -- the PSF has no power to
tell arbitrary Python groups how to run their community. (Although in
principle it could choose to withhold funding from those that do not.)

The PSF blog post introducing this makes it clear that the PSF
*encourages* community groups to adopt this CoC, but that is all:

The PSF supports and advocates for the use of the CoC throughout the
community, but without adoption by specific areas, the CoC is merely
a document that the Foundation is supportive of. The way it’s useful
is that an area of the community can adopt the CoC and use it as a
guideline for participation. IT COULD BE adopted by mailing lists,
IRC channels, the bug tracker, user groups, sprints, and more.

For example, a mailing list COULD SAY that their membership should adhere
to the CoC. [emphasis added]
[end quote]

http://pyfound.blogspot.com.au/2013/06/announcing-code-of-conduct-for-use-by.html

To my knowledge, only two groups have adopted this so far: the PSF itself
have adopted it for their internal (and private) mailing lists), and the
python-ideas mailing list, whose two moderators adopted it without
consultation with the mailing list subscribers, not that I'm bitter or
anything.

That was formed by the Python community and adopted by our gracious
hosts, the Python Software Foundation, who provide this forum for our
use.

The Code of Conduct was not formed by the Python community. It was formed
by a small but influential subset of the Python community, the PSF. The
greater Python community includes large numbers of people who are not
members of the PSF, are not on this mailing list, and indeed may not be
on any mailing list at all.

It is absurd to think that the PSF can unilaterally decide how (say)
StackOverflow users are allowed to behave when they ask Python
questions. Now, if we were talking about the Python Secret Underground,
which emphatically does not exist, then

 

[Terry Reedy]

I would revise this to "If you are have really found a bug in Python..."
How does a newbie know?

I'm not sure I agree with that one, at least not fully. It's certainly
true that you shouldn't expect anybody to do anything about a bug unless
you open an issue.

On the other hand, I often find it useful to discuss things that I
believe are bugs on c.l.p first. Sometimes people will explain to me
that I'm just doing it wrong. Sometimes the discussion will end up
with, "Yeah, that's a bug".

usually followed by "File a tracker issue" or "I opened an tracker issue
for this." (I have done that several times, though I sometimes prefer a
person learn how to do it themselves.)

In either case, it serves as a good initial
filter for whether I should file a bug or not, and the discussion is
often educational.

Ask here first.
With a subject line that says 'I think ...' or "Problem with ...'
Advantages:

1. At least half the bugs newbies report are not. The tracker does not
need the extra clutter.
2. Filing a tracker issue sometimes creates a 'mental investment' in the
mis-perception, which leads to resentment upon explanation.
3. There are lots of people here ready to help and answer questions.
Any sensible question usually gets multiple responses, usually within a
few hours or a day. (Invalid tracker reports may sit for days and get
one short response.)
4. Explanations posted here benefit lots of people, rather than just 1.
5. A question posted here may elicit essential information, like which
systems or which versions have the problem.
6. If you make an informed post to the tracker backed up by at least
opinion, at least one tracker responder be in a better mode when responding.

 

[Chris Angelico]

A popular language like Python has been around for about 20 years. It is
in daily use by tens of thousands of people around the world. What are
the chances that you, in your first week of using Python, just happened
to stumble across a bug that *nobody else in the world* has noticed?

Of course, it's possible for there to be dark corners. But if you're
working with those, you know it full well. The dark corners of Python
might be in some of its more obscure modules, or maybe in IPv6
handling, or perhaps some weird platform that misbehaves, but they're
definitely not in lists and tuples treated normally.

ChrisA

 

[rusi]

Ideally, this community is healthy enough for us to enforce the code of
conduct of our host, through social convention among us all.

Thanks Ben for that.
Lets not stymie Steve Simmons original suggestion for a CoC by making it into a formal CoC and then saying its impossible.

There is a good deal of informal enforcement already. Consider eg:
- Mark's footnote reminding about GG problems and their solutions <link>
- Steven's brusque 'cut-the-crap' type responses
- Your own prompt STOPs to racist/sexist etc posts
etc

So we may take and welcome Steve Simmons' call to expand that list.
I certainly welcome the suggestion.

 

[Roy Smith]

Of course, it's possible for there to be dark corners. But if you're
working with those, you know it full well. The dark corners of Python
might be in some of its more obscure modules, or maybe in IPv6
handling,

The sad thing about this statement is that IPv6 has been around for
about as long as Python.

 

[Roy Smith]

6. If you make an informed post to the tracker backed up by at least
opinion, at least one tracker responder be in a better mode when responding.

What I generally do is summarize the problem in the tracker, but also
include a link to the google groups archive of the mailing list thread.
Then, if anybody wants to see the whole background discussion, it's easy
to find.

The most common outcome, for me at least, is that I end up opening a bug
against the docs, because they didn't do a good job of explaining how
something is supposed to work.

 

[Chris Angelico]

The sad thing about this statement is that IPv6 has been around for
about as long as Python.

Yeah, and yet so many ISPs simply don't support it (only one of the
Australian ISPs I've worked with does - Internode). So heaps of code
just doesn't get tested in an IPv6 environment. I don't know whether
there's anything in Python that would fail in that way, but there
might be; if someone starts writing a DNS server in Python and runs
into difficulties with AAAA records, then it actually might be a
library bug.

But the concept still applies: Extraordinary claims require
extraordinary evidence.[1] Before ranting loudly about how Python is
buggy, make absolutely sure you're right, and show the evidence.

ChrisA

[1] http://www.catb.org/esr/faqs/smart-questions.html#idp5249360

 

[Steve Simmons]

Thanks Ben for that.
Lets not stymie Steve Simmons original suggestion for a CoC by making it into a formal CoC and then saying its impossible.

There is a good deal of informal enforcement already. Consider eg:
- Mark's footnote reminding about GG problems and their solutions <link>
- Steven's brusque 'cut-the-crap' type responses
- Your own prompt STOPs to racist/sexist etc posts
etc

So we may take and welcome Steve Simmons' call to expand that list.
I certainly welcome the suggestion.

Thanks Rusi - and everyone else that provided feedback.

I understand (and had already anticipated) that this type of group can't
and probably shouldn't be moderated in the more draconian sense of the
word but there's no reason why we couldn't create some behavioural and
technical guidelines. I had it in mind to move to a few (?!) paragraphs
that read well when considered as a single document but that were also
suitable to cut'n'paste into a response - much as Mark puts his GG
warning in every mail.

The aim would be to put a considered response in a first reply that
tells the OP 'this is what you need to do to get a sensible answer'. We
have this for GG & code fragments and I see no reason why we can't have
it for bad language; bad attitude; unreasonable expectations; thread
high-jacking; top posting; and all the other sins we expect Noobs (and
some long standing members!) to avoid.

I guess I should have said this up front but let me say it now: I
completely 'get' that this won't fix the Trolls or the dyed-in-the-wool
pedants or recidivist high-jackers but it may put some buffering between:
Newb: "I want you to do my homework and I have to give it in tomorrow
morning"
and
Pythonista: "you're a dick"

If there's enough support for the general idea, I'm happy to make a
decent sized contribution but if there isn't - Nuff said.

Steve

 

[Roy Smith]

Yeah, and yet so many ISPs simply don't support it (only one of the
Australian ISPs I've worked with does - Internode).

My comment was not meant as a flame against Python, but as a flame
against the whole frigging Internet industry for dragging their feet(*)
on IPv6 rollout.

(*) It is difficult to walk with any gait other than foot-dragging when
you are suffering from a severe case of cranio-rectal inversion.

 

[Chris Angelico]

My comment was not meant as a flame against Python, but as a flame
against the whole frigging Internet industry for dragging their feet(*)
on IPv6 rollout.

(*) It is difficult to walk with any gait other than foot-dragging when
you are suffering from a severe case of cranio-rectal inversion.

Right. My examples of dark corners were non-specific too; I haven't
actually found any such corners in Python as I don't use it enough.
It's high time the world moved to IPv6 wholesale. Start leaning on
your ISPs to offer /64s (or /56s as Internode give) to all their
customers. Sooner or later they'll have to make it a priority. IPv4
depletion was back in early 2011 - that's two years since IANA gave
out the last blocks to the RIRs. Surely that means something?

At work, we took a stand on this, for what it's worth. We rent
dedicated servers from a number of hosting companies (giving us
geographically and topologically separate servers), and will not work
with anyone who doesn't give IPv6 as well as IPv4 addressing. But
since we currently put less than $100/month into servers, that's not
much of a stand

ChrisA

 

[Steven D'Aprano]

What I generally do is summarize the problem in the tracker, but also
include a link to the google groups archive of the mailing list thread.

^^^^^^^^^^^^^^^^^^^^^^

Please don't. Link to the actual mailing list thread instead:

http://mail.python.org/mailman/listinfo/python-list

This is the canonical archive for the list, and doesn't involve the
atrocious Google Groups interface.

 

[Grant Edwards]

The sad thing about this statement is that IPv6 has been around for
about as long as Python.

I've recently added IPv6 support to some embedded firmware. We had to
add it because various purchasing people (mostly government) have
non-negotiable feature checklists and "IPv6 support" is starting to
show up on the checklists. However, so far, none of the people
requiring IPv6 support actually have any plans to use IPv6 -- ever.
I'm convinced we could put "IPv6 support" on the data sheets and
continue shipping IPv4-only stuff for many years to come, and all our
customers would be perfectly happy.

 

[Roy Smith]

I've recently added IPv6 support to some embedded firmware. We had to
add it because various purchasing people (mostly government) have
non-negotiable feature checklists and "IPv6 support" is starting to
show up on the checklists.

Yup. I did one of those. Hmmm, just found the powerpoint deck dated
March 2006, so we probably started the project around September 2005.
Same deal, we couldn't sell into government accounts unless we had IPv6.
Have they actually used it? I doubt it.

The really annoying thing about the project was I never got around to
implementing rfc1924.

However, so far, none of the people requiring IPv6 support actually
have any plans to use IPv6 -- ever.

What's known in the industry as "shelfware".

 

[Grant Edwards]

Yup. I did one of those. Hmmm, just found the powerpoint deck dated
March 2006, so we probably started the project around September 2005.
Same deal, we couldn't sell into government accounts unless we had IPv6.
Have they actually used it? I doubt it.

The really annoying thing about the project was I never got around to
implementing rfc1924.

Thankfully, the network library for our RTOS had support functions to
handle rfc1924. Otherwise I would have had to do it myself because
there's no way I could have made it through the project having to read
and type full-length IPv6 addresses the whole time.

What's known in the industry as "shelfware".

Yep. 99.9% of the time, our products are used on small, isolated
industrial networks that have little or no internet connectivity. The
product line started out when a lot of customers weren't even using
IPv4 yet. We still have customers who use our products with all IP
support disabled (though these days it's always on networks that
have other IP traffic).