- From: Benjamin Goering <ben@bengo.co>
- Date: Fri, 24 Mar 2023 12:18:14 -0700
- To: Melvin Carvalho <melvincarvalho@gmail.com>
- Cc: Ryan Barrett <public@ryanb.org>, Bob Wyman <bob@wyman.us>, aaronngray@gmail.com, public-swicg@w3.org
- Message-ID: <CAN+OhBOcmNR7aEpzurRM6ijBVO+Dr49RhGdKFLQz47sZ2teF-Q@mail.gmail.com>
I'm also very sympathetic to people who want to use protocols other than
ActivityPub or HTTP to do things, where content-types just don't exist
(e.g. POSIX filesystems, IPFS)
On Fri, Mar 24, 2023 at 12:17 PM Benjamin Goering <ben@bengo.co> wrote:
> > Pretty much all the conneg ive seen is broken (e.g. on mastodon too).
> > But nobody actually cares.
>
> I care, Melvin. I care. :)
>
> On Fri, Mar 24, 2023 at 10:41 AM Melvin Carvalho <melvincarvalho@gmail.com>
> wrote:
>
>>
>>
>> pá 24. 3. 2023 v 18:34 odesílatel Melvin Carvalho <
>> melvincarvalho@gmail.com> napsal:
>>
>>>
>>>
>>> pá 24. 3. 2023 v 18:23 odesílatel Benjamin Goering <ben@bengo.co>
>>> napsal:
>>>
>>>> http content negotiation is normatively required by the activitypub
>>>> spec (3.2).
>>>> I'll also point out that some URIs from the TRs already support it.
>>>>
>>>
>>> How do you ensure that each accept header gives back the same triple?
>>>
>>> For example, do they in this case.
>>>
>>> Not a pedantic point, I've been trying to fix this for years in Solid,
>>> and its made harder by the fact that the developers dont know how it
>>> works. In fact the maintainer of the reference implementation said to me
>>> recently: "What's content negotiation". It's hard to fix these things.
>>>
>>
>> Let me add an extra point
>>
>> Pretty much all the conneg ive seen is broken (e.g. on mastodon too)
>>
>> It's generally broken on vocabs ive seen too. But nobody actually
>> cares. Because, people dont actually use vocabs. They may in future e.g.
>> for xsd:typing but they are barely noticeable today. The whole overhead of
>> the thing I would argue provides very little developer value other than a
>> documentation where they can copy and paste.
>>
>>
>>>
>>>
>>>>
>>>> ```
>>>>
>>>>> bengo@bengo ~ ⚡ curl -H "Accept: application/json"
>>>>> https://www.w3.org/ns/activitystreams | head -n 10
>>>>> % Total % Received % Xferd Average Speed Time Time Time
>>>>> Current
>>>>> Dload Upload Total Spent Left
>>>>> Speed
>>>>> 100 7985 100 7985 0 0 39142 0 --:--:-- --:--:--
>>>>> --:--:-- 40125
>>>>> {
>>>>> "@context": {
>>>>> "@vocab": "_:",
>>>>> "xsd": "http://www.w3.org/2001/XMLSchema#",
>>>>> "as": "https://www.w3.org/ns/activitystreams#",
>>>>> "ldp": "http://www.w3.org/ns/ldp#",
>>>>> "vcard": "http://www.w3.org/2006/vcard/ns#",
>>>>> "id": "@id",
>>>>> "type": "@type",
>>>>> "Accept": "as:Accept",
>>>>> bengo@bengo ~ ⚡ curl -H "Accept: text/html"
>>>>> https://www.w3.org/ns/activitystreams | head -n 10
>>>>> % Total % Received % Xferd Average Speed Time Time Time
>>>>> Current
>>>>> Dload Upload Total Spent Left
>>>>> Speed
>>>>> 0 0 0 0 0 0 0 0 --:--:-- --:--:--
>>>>> --:--:-- 0<!DOCTYPE html><html dir="ltr" typeof="bibo:Document "
>>>>> prefix="bibo: http://purl.org/ontology/bibo/ w3p:
>>>>> http://www.w3.org/2001/02pd/rec54# as:
>>>>> https://www.w3.org/ns/activitystreams#" lang="en"><head><meta
>>>>> charset="utf-8"><meta name="generator" content="ReSpec 20.7.1"><meta
>>>>> name="viewport" content="width=device-width, initial-scale=1,
>>>>> shrink-to-fit=no"><meta property="dc:language" content="en" lang="">
>>>>> <title>ActivityStreams 2.0 Terms</title>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>> <style>/*****************************************************************
>>>>> * ReSpec 3 CSS
>>>>> * Robin Berjon - http://berjon.com/
>>>>> *****************************************************************/
>>>>>
>>>>> 100 33068 0 33068 0 0 148k 0 --:--:-- --:--:--
>>>>> --:--:-- 153k
>>>>> curl: (23) Failure writing output to destination
>>>>>
>>>> ```
>>>>
>>>> On Fri, Mar 24, 2023 at 10:16 AM Ryan Barrett <public@ryanb.org> wrote:
>>>>
>>>>> Sure! And feel free to redirect to a different venue if we're getting
>>>>> off topic. Off the top of my head:
>>>>>
>>>>> 1. Most web developers are not aware that conneg exists. This is
>>>>> usually fine, since it's uncommon, but can be an unpleasant surprise when
>>>>> they first hit it and don't understand it.
>>>>> 2. The `Accept` header is notoriously complicated to parse,
>>>>> generate, and fully comply with. Imo much of that complexity is unneeded in
>>>>> practice.
>>>>> 3. It often breaks caching when it's first introduced. Internal
>>>>> and external caches generally ignore Content-Type and Accept (etc) by
>>>>> default, so when conneg is first introduced, clients often get the wrong
>>>>> response type. Developers learn the hard way that they need to add Accept
>>>>> to Vary and to their framework's cache.
>>>>> 4. URLs leak across contexts. They may start out within a single
>>>>> service or API, but they gradually proliferate into user-visible links,
>>>>> object identifiers in other systems, add-on services, third party tools,
>>>>> etc. These generally don't know the original conneg requirements, which
>>>>> causes unpleasant surprises.
>>>>>
>>>>> I think most of this boils down to: modality generally considered
>>>>> harmful. When something always behaves the same way, it's reliable and easy
>>>>> to use. When it behaves differently based on something far away in the
>>>>> environment that you may not know exists, it's unreliable and surprising.
>>>>> Add in a very large ecosystem of independent tools that all need to
>>>>> interoperate, often in fine-grained ways, and you have a recipe for failure.
>>>>>
>>>>> On Fri, Mar 24, 2023 at 9:42 AM Bob Wyman <bob@wyman.us> wrote:
>>>>>
>>>>>> Melvin, Ryan,
>>>>>> Could you provide at least a short summary of why "content
>>>>>> negotiation" has not lived up to expectations? It is good to know that it
>>>>>> is non-optimal. It would be better to know why.
>>>>>>
>>>>>> bob wyman
>>>>>>
>>>>>>
>>>>>
>>>>> --
>>>>> https://snarfed.org/
>>>>>
>>>>
Received on Friday, 24 March 2023 19:18:39 UTC