[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: mailing list vs "the futur"

I was going to say that you and I have started going round in circles and should just agree to disagree about certain things but this is a different strand of the discussion that still seems to be advancing.

On 28/08/2018 20:01, Michael Stone wrote:
On Tue, Aug 28, 2018 at 04:46:15PM +0100, Mark Rousell wrote:
web forums, app-based, IM-style, etc.) but none of that, to my mind, lessens
NNTP's ideal applicability to getting private discussion group messages from
place to place (the front end UI/UX being a different thing again).

Ignoring the changes to user requirements for UI/UX is at least part of why NNTP is no longer a major factor in internet usage.

I agree to a considerable extent and I don't think I advocated ignoring changes to user requirements for UI/UX. When I say "the front end UI/UX being a different thing again" I mean that it's a different discussion, not that a designer should ignore it.

And making NNTP available as an access method (certainly not a sole access method) is not ignoring it.

The key advantage of NNTP over email/SMTP in terms of bandwidth efficiency is
that, with NNTP, messages are only sent to users when they explicitly ask for
them. This is more bandwidth-efficient than an email-based list since the
email-based list must send out messages to each and every user whether or not
they want them.

It's more efficient at the provider level until someone decides they want all of the messages local, either as an archive or to run local search tools, etc. At that point you're transferring all of the messages just as you would via SMTP and it's basically a wash.

That's fine. Not everyone wants that but, for those who do, it's certainly no worse than email. So I don't see it as a problem.

If you have a bunch of users on remote SMTP and NNTP servers then it's always a wash. (MUAs don't typically download the entire message body unless asked to, just as news readers don't typically download the entire message body unless asked to.) Basically, the efficiency argument is bogus.

I can only say that I disagree that the bandwidth efficiency is bogus overall.

If by "remote [...] NNTP servers" you mean other NNTP servers that are federated with your own, then this is surely a bandwidth saving compared to email. I.e. The data only needs to be sent once to the remote NNTP servers for local distribution to users who connect to those servers, thus reducing bandwidth usage overall.

the bandwidth in question is so small as to not matter anyway.(We call this "premature optimization"; there are other concerns that are far more significant--like the UI/UX.) The entire 25 year archive of debian lists is probably on the order of one or two netflix movies.

Whilst I agree that UI/UX are important to users (which of course is exactly why many prefer to receive their message via NNTP, so that they can control their UI/UX), it is not necessarily the case that bandwidth is always small. A static archive is not the bandwidth.

To my mind, a REST protocol has a different (but overlapping) use case to NNTP
or email lists. I know of no standard, open REST protocol that replaces either
NNTP or email discussion lists, for example.

But there are a heck of a lot more deployed REST clients than NNTP clients.

I know you know this but I'll say it anyway: REST isn't a single protocol, it's just a type of protocol. There are loads of REST-based protocols around. Which one do you choose? There are no standardised REST protocols for message distribution that I am aware of. There's nothing REST-based that is like SMTP, or POP3, or IMAP, or NNTP, or anything else that has broad client support across a range of device types in this context.

Or is there? Have I overlooked anything?

Furthermore, there's no point saying that there are other protocols when those other protocols do not and cannot address the use case that one particular protocol, NNTP in this case, can and does address.

I am not saying that REST does not have its place. It's just that NNTP (alongside other protocols and types of protocol, both standardised and proprietary) is something that fulfils a currently commonly unfilled use case, one that in fact does have demand.

You can shake your fist at the cloud all you want, but reality is what it is. Consequently, a good experience for HTTP consumers is going to be a higher priority than NNTP users.

There are many possible priorities and different businesses and services providers have different customer bases who have different sets of priorities. I've not said that a web browser-based UI is unimportant (in fact I think you'll note I've said that it is important for most users) but that doesn't mean that a variety of access methodologies are not also important.

The numbers are so skewed as to raise valid questions about whether the small number of NNTP users is worth committing development and operational resources to. Many of the forums I can think of that supported HTTP + NNTP 10 or 15 years ago have long since decided that the answer was no.

Cool. That's fine for them.

(Though, I might start some kind of drinking game based on how many times you say "conflation of usenet and NNTP" or equivalent--it's gotten kinda amusing, if repetitive.) You also seem to have been the one to start making other unfounded assertions about my background and why I have the opinions I have. Honestly, I'd prefer if you just stick to stuff that's somehow relevant to the discussion and avoid erroneous ad hominems.

I have made no assertion of any sort whatsoever about your background. Are you by any chance referring to this passage from [🔎] 5B853D2D.8040202@signal100.com"><https://lists.debian.org/msgid-search/[🔎] 5B853D2D.8040202@signal100.com>?

"  May I ask, did you use Usenet in this timeframe? I ask this because some of your comments remind me of training courses run for certain types of professional at that time which were taught by people who, themselves, commonly had only limited, and sometimes very skewed and confused, experience of the systems and protocols they were supposedly experts on. Thus what they taught was close to, but not quite, an accurate representation of how things really were.  "
If so then, as you can see, I made no assertion about your background. I simply asked you if you used Usenet in the timeframe we've been discussing and gave my reasons for asking the question.

When you replied (in [🔎] 89d1fa34-aabd-11e8-9b6a-00163eeb5320@msgid.mathom.us"><https://lists.debian.org/msgid-search/[🔎] 89d1fa34-aabd-11e8-9b6a-00163eeb5320@msgid.mathom.us>) saying that you certainly did use Usenet in the relevant timeframe in a number of environments and that you had never been on a Usenet training course, I then replied in [🔎] 5B8553A4.1050802@signal100.com"><https://lists.debian.org/msgid-search/[🔎] 5B8553A4.1050802@signal100.com> in a hopefully inoffensive manner.

To repeat, none of this makes any kind of assertion about you or your background at all.

FWIW, another possibility for why the discussion doesn't match your desires (one that doesn't require me to have a misguided agenda of conflating usenet and NNTP based on some shadowy usenet training 20 years in the past) is that threads wander and that people talk about what they want to talk about.

Quite so and I have been happy to wander with you. The discussion above about possible current/future NNTP usage is such an example. Nevertheless, I initially commented in this thread to make certain points in response to what you and others had written and I don't apologise for pointing out that many of my comments were mainly about that subject. This is not telling you what to talk about.

Also, more than one person may have something to say, and people can respond to things from different people. Unless someone gets to decide what's on-topic and ban everyone else it's likely the internet will remain that way.

I agree. I've said nothing to the contrary.

Really? Where's the like button in the standard NNTP protocol?

It's not there. I said that expectations are an issue and apparently you expect NNTP to have a Like button. I see no reason for NNTP itself to have a Like button. There's no need to extend NNTP to include a Like button.

However, if you want a Like button when viewing messages sent to your client software via NNTP then, as I suggested, one can certainly appear. But NNTP (or email, or RSS, or some other message transfer protocol, for that matter) doesn't need to be able to support Like buttons in the protocol as long as it can transfer a message with rich content. And NNTP can transfer such content, and real world NNTP clients can display that rich content, and users who want things like Like buttons can see them and click them.

However, remember that the kinds of users who might choose to access content via email or NNTP are commonly very likely to be the kinds of users who would happily forego stuff like Like buttons. Which is rather the point of offering them alternative access methodologies. They get the efficiency they want without being burdened by the cute features that others users do like to see.

Also, in many cases, as I previously noted, a single user might choose to access a particular discussion/collaboration resource in different ways at different times from different devices. A Like button may be present in the web view of a message but absent from the view of that same message when received via NNTP.

And so there's no need to retrofit features to NNTP that don't really belong there, not least since this would alienate the very kinds of users who want NNTP.

How do I see a poster's history and other profile information via NNTP? If the answer is "well, you don't need those things" that's a sad argument for how NNTP can replace a web forum.

I agree. It's a good thing that I haven't advocated that NNTP should "replace a web forum", isn't it!

I've said that NNTP (as well as other access methodologies) can be alternatives, not replacements. What people want from forums differs, sometimes quite markedly, between users and types of users and for this reason I think it makes more sense to think of a neutral forum/group/resource than a "web forum" specifically. What one person sees as a web forum can be entirely legitimately seen by others users through the 'lens' of NNTP, or email, or other access methodologies (e.g. RSS, some sort of REST client, or who knows what).

Even in terms of a forum that is only accessible via the web, I have noted that people's perceptions of it can differ by surprising amounts depending on whether they access it in desktop mode or mobile mode, or perhaps in a simplified web format compared to a full _javascript_-enabled web format. And yet it's still the same forum. Adding other access methodologies just extends the ways in which the same resource can be perceived, depending on user and/or owner preference.

A great deal of this seems to be about expectations. Could it be that you are
expecting something that is, to my mind, unnecessary or irrelevant (i.e. not
part of the target use case).

Ah, you also get to decide what the use cases are? Not the original poster or anyone else on the list?

You're being sarcastic but I also think you might be misinterpreting my words here. I certainly don't get to decide what use cases exist in general but I do get to decide what the target use cases are where my own comments are concerned or where we are talking about work in which I am involved. In this part of the discussion we are discussing work in which I am involved, aren't we.

To re-phrase what I said in the above quoted passage, you clearly have certain specific expectations about what NNTP should preferably, for you, mean when added as an access methodology to forum/discussion software. Your expectations in this regard certainly constitute a valid use case for your purposes but it's not one that anything I am working on is targetting in exactly the way you'd like it. Perhaps I was being too blunt when I said that your expectations were "unnecessary or irrelevant" to my mind but... oh well. Sorry, I'm not going to add a Like button to the NNTP protocol for you (even though a message transmitted by NNTP can still include a Like button or links to other stuff).

has to be used to the exclusion of all others. Content posted to a web forum
(via web browser or via app front end or via some REST client) can perfectly
well be distributed (complete with graphics, text effects, images, even
binaries (!!) ) to participants via email or NNTP.

Sure, and for the third time: the theory doesn't match the practice. Why?

So you say and yet, from what I can see, practice often does match what I described in various ways and to various extents. There are opportunities to do more and I am involved in doing more.

Your point appears to be that no one is doing it to the extent I would ideally like to see and so there can't be any demand. That's fine. Our opinions about opportunity and demand differ.

I don't see it as either/or.  REST and other access methodologies (for reading
and/or writing) can and should co-exist.

Sure, and for the fourth time: the theory doesn't match the practice. Why?

See above. REST and other access methodologies in fact can and do co-exist, right now, today. I'd like to see more of it but it's hardly unknown. (You might well say that the preceding statement is an assertion and you'd be right in this case. But I'm content to let it stand as an assertion for the sake of brevity).

Despite this, there are still opportunities to do more. If you don't feel they are worth exploiting or fulfilling then so be it. As I say above, we differ in our interpretations and expectations.

I think web forums have become the de facto replacement for groups accessed by
NNTP because they were, more than anything else, simpler to use and access.

Uh, yeah. That isn't a meaningful feature in your mind?

Well obviously it's a meaningful feature to my mind. You can tell that I think that it's meaningful because I said that I thought that it was the key reason that web forums have become so popular! I think that's "meaningful", don't you?

However, it is unwise to presume that, just because one product has a particular unique selling point, other (partly competing) products do not have their own unique selling points that are desirable to other users or types of user. To put it another way, each product can have its own target use case.

So the NNTP client has to implement an HTTP client anyway?

No. Why would you assume that? Don't essentially all client devices have a HTTP client anyway?

Again, your expectation appears to be that the client software should do everything, even stuff that is outside the target use case.

Why not just use HTTP?

Because it doesn't suit the NNTP target use case. That is to say, efficient, standardised,  access to message content. Clicking on a link in a message and seeing it open in a web browser does not mean that the user wanted to get the message content by HTTP in the first place.

Consider the parallel scenario in email. One can access email via a web browser but many people still prefer to access email via a client program or app. It's a lot of people. Local client usage has actually increased compared to web access due to the growth of mobile devices. When they click on a link in an email, they don't necessarily expect their email client to have a built in web browser. They usually expect that their preferred web browser will open to visit the link.

Similarly, there is no reason to assume that a user who prefers to receive discussion messages by email or NNTP wants to have a web browser in his or her email/NNTP client. Most users are happy to have links open in their web browser.

Whilst a Like button is not a link to a website in semantic terms, I nevertheless think having a Like button in this context open a web browser instance is likely to be entirely acceptable (if users want to see Like buttons at all in their NNTP feed, that is).

If this is not an acceptable method of access to you then so be it, you're not a likely user of anything I'm involved in. But I should point out that you are having this conversation on an email list and it doesn't provide Like buttons at all. There's a Debian web forum you could use instead but I note that it seems to be pretty plain phpBB with no Like buttons. Note also that phpBB can send out email notifications of new messages, and those don't have any attempt at Like buttons either.

Indeed, but our discussion has ranged in various directions and are beginning
to conflate somewhat issues. I'm not pitching NNTP as a replacement for this
sort of scenario

So you want the servers to implement two or three protocols? That's got to be more complex than implementing just one.

Yes. It's also quite common in practice. Implementing more than one protocol is just fine.

There'd have to be one heck of a compelling reason to do that, more than just "I personally like NNTP".

Much like the same widely varying compelling reasons that people write mail list software, or web forum software, or Tapatalk, or plugins for these, or any other kind of software. There are many possible reasons.

Mark Rousell

Reply to: