My stackoverflow question was closed so here's a blog post about CoreWCF

https://richardcocks.github.io/2025-05-08-CoreWCF.html

by eterm • 1 year ago

123 185 1 year ago

richardcocks.github.io

eterm8• 1 year ago

A post in which I try to rubber-duck a CoreWCF issue I've been having, because stackoverflow no longer seems suitable for asking questions about programming issues.

Screaming into the void of the blogosphere is catharsis for getting my SO question closed.

And because I know you're all nosy, the SO question is here: https://stackoverflow.com/questions/79605462/high-cpu-usage-... . Please feel free to point out more ways in which I screwed up asking my SO question.

agos3• 1 year ago

for what it's worth, I submitted the question as a candidate for reopening

aflukasz• 1 year ago

FWIW seems open to me right now.

eterm8• 1 year ago

Thank you, it has now been re-opened.

matsemann• 1 year ago

Honestly, I agree with it not being a good fit for a Q&A site. It's a debugging problem, probably needing a discussion, and might even not be of any use to others being that "high cpu" is kinda vague. Seems better suited for a bug report / issue tracker of the relevant library.

jve• 1 year ago

A community site for multiplayer debugging... I like it! Some people like to tackle problems and feel rewarded when they crack the nut :)

eterm8• 1 year ago

Your'e right, "High CPU" just means more than zero. It was a symptom of the stream continuing to be written to. I've edited the title now to be better.

wokwokwok2• 1 year ago

How can a question that is:

1) clearly technical

2) reproducible

3) has a clear failure condition

Not be a suitable candidate for S/O?

Did we step into a dimension where only "How do I print('hello world')?" is a valid question while I wasn't watching, because it has a trivial one-line answer?

Hard questions doesn't mean they're bad, it just means many people aren't competent answer them. The same goes for obscure questions; there might just not be many people who care, but the question itself is entirely valid.

Does that mean they're not suitable for S/O?

I... can't believe anyone seriously believes that hard niche problems are too obscure or too hard for S/O to be bothered to grace themselves with.

It's absurd.

It just baffles me that a question that might take some effort to figure an answer out to might 'not be suitable' to S/O.

francisofascii• 1 year ago

I could see it being useful to others. If there is an internal bug that causes the issue or even a code pitfall that causes this issue.

kristianp• 1 year ago

Stackoverflow is basically a site to help people with programming problems. That includes debugging a problem. I think it's sad that many types of question are not considered suitable for the site. The question was closed as "Not suitable for this site", whatever that means.

Tangentially, one thing I think StackOverflow misses in its policy is in not allowing questions that ask for help to evaluate libraries and products in the developer space. There are many questions on SO where someone has asked what is the best library for X, and the question is closed due to site policy, but I think a discussion of what library is best for a use case is a very useful thing to have on SO. The answer to the question can change over time, and multiple alternative solutions can be presented. This is the kind of thing that Reddit has taken over for many different product recommendations, but reddit answers are just comments, they don't have the UI of SO's for commenting on questions and answers.

balls187• 1 year ago

I don't see anything wrong with your SO question (I am a long time contributor), and don't see why it would have been closed.

I will say, this is a level of question that is too sophisticated for SO, and likely will only have an answer once you figure it out and go back and answer your question.

Are you confident the code is the issue--have you repro'd it consistently with different versions of .NET? What about reproing on different machines? Locally?

rendaw3• 1 year ago

Do they have some metric whereby all questions must be answered in some time period? Why not just leave it open then? What if someone else has the same question and figures it out later, now they can't answer it either!

I had the same suspicion though, which is that if the question is too difficult it triggers some reflex...

skinkestek• 1 year ago

I will say, this is a level of question that is too sophisticated for SO

Could it be that, for years, they've been selecting for things like:

- Low friction onboarding

- Upvoting “easy wins” and thought experiments over real-world, in-the-trenches problems

- Penalizing questions that didn’t fit the format—even when those of us around since the early days knew they’d likely get good answers within a few days

- Incentivizing moderators to strictly enforce rules that, frankly, weren’t necessarily great to begin with

Just a thought.

robertlagrant3• 1 year ago

Please feel free to point out more ways in which I screwed up asking my SO question.

With pleasure! SO is definitely more of a distinct Q&A site and not a discursive, open-ended collaborate and problem-solve site.

zahlman39• 1 year ago

I fully expect nobody in this comment section to care about the CoreWCF content. (I don't even know offhand what that is.) In my experience, people love talking about Stack Overflow in places that are about programming but aren't Stack Overflow, so.

(Edit: it seems people do care about CoreWCF ITT. That's nice to see.)

Screaming into the void of the blogosphere is catharsis for getting my SO question closed.

That's fine. Almost everyone who comes to SO, in my experience, has a fundamentally wrong idea about how the site is intended to work. That includes people who don't have a question and only want to post answers. Unfortunately, it's difficult to explain because people find the model unintuitive - the UI affords using the place just like many others, even though the site was created exactly to get away from frustrations caused by older models (https://meta.stackexchange.com/questions/92107). And the real objective is a synthesis of many not-always-compatible ideas (https://meta.stackoverflow.com/questions/254770). My personal sense is that the community didn't really get a handle on "what SO is" until around the time that new question volume peaked (way back in 2014).

Even then, people can hang around for years and not really get it (e.g. https://meta.stackoverflow.com/questions/427224) - in large part because the policies have been inconsistently applied on a volunteer basis, and the people who are allowed to e.g. cast close votes are vastly outnumbered.

We generally don't care about people not liking the Stack Overflow model while discussing it off-site. There's far too much of that to worry about. But that doesn't mean we'll change to accommodate everyone else. The entire point is to provide something that isn't available everywhere you look: a polished artifact, an organized repository of commonly-needed, high-quality answers to clear, focused, practical questions.

Do we accomplish that goal? Hell no, not by a long shot. But there are some real gems in there - and a few of them have millions of views. And as the rate of new questions slows, users who put on the "curator" hat become able to keep on top of the incoming queue, filter through for what's of value (and not a duplicate), and even turn attention towards the old Q&A to improve it (incidentally, a lot of that work is rounding up old duplicates that went unnoticed).

I had forgotten that any external links are a big no-no in SO land, so my question immediately attracted 2 close votes.

The problem isn't simply including an external link (we'll happily just edit those out if they aren't necessary). The problem occurs when a question appears to depend upon the externally linked content. We can't accept that (https://meta.stackoverflow.com/questions/254428) because of link rot and licensing issues (someone who wants to answer you often needs to be able to cite the code; posting on-site automatically licenses the content appropriately, per the terms of service) but mainly because of scope - a question that's suitable for the Stack Exchange format would fit neatly within the actual question text.

We don't want to do detailed analysis of the problem you encountered, even if we're capable of it, because questions are for everyone. They need to be able to reflect a problem that other people could a) have; b) plausibly search for; and c) recognize if they found it. Answers to a question need to make sense in general to people who would ask - not just in the specific context of one person's original problem. In short, we want a question, not a problem - and extracting a proper question starts with (https://meta.stackoverflow.com/questions/261592) your own analysis.

"How do I do X?" questions are usually much easier to ask in the format, and are very valuable and can end up very well regarded, even when they're on very basic topics. But "what went wrong with Y code?" is not fully refined. What we're really looking for is more like "why does Y' code construct do Z?" - where the specific, exact cause of failure (https://stackoverflow.com/help/minimal-reproducible-example) is extracted from your own debugging session (along with reproducing input and actual vs expected output).

Two days later my question got it's third vote for closure, and remains unanswered and now closed forever.

This is literally not how Stack Overflow works. The OP has at least (https://stackoverflow.com/help/auto-deleted-questions) 9 days to fix the question and nominate it for reopening until it gets "deleted"; but even then it's a soft deletion (delisting) which is still reversible - you can find the question from your personal listing (https://stackoverflow.com/users/deleted-questions/current while logged in; or replace 'current' with your user ID), edit and nominate for undeletion.

The established policy is that we intentionally close questions that don't meet standards (https://meta.stackoverflow.com/questions/417476) as quickly as possible (https://meta.stackoverflow.com/questions/260263). The main point of this is to prevent the sort of people (notice that https://meta.stackoverflow.com/questions/271684 is over 10 years old; and the original complaint https://meta.stackexchange.com/questions/9731 is from before the official launch, during the private beta) who would otherwise hang out on a traditional discussion forum 12 hours a day from trying to read the OP's mind, repost the same basic explanation of the same basic idea dozens of times, etc.

(Unfortunately, the incentive system is completely broken - https://meta.stackexchange.com/questions/387356 - and the company's interests are not aligned with the community, so this is a losing battle.)

And, in fact, your question has been reopened, as of about 3/4 of an hour after your comment that I'm replying to. Stack Overflow is not at all immune to external pressure - after all, many regulars there are also on HN and other usual-suspect sites.

It also looks like your edits have actually improved the question. In particular, adding in a definite conclusion from your profiling attempt.

(We understand that a lot of people in a situation like yours wouldn't necessarily know how to use a profiler and wouldn't necessarily be able to come up with a theory about what's wrong. That isn't our problem. We aren't offering tech support. It's a bitter pill for almost everyone, but Stack Overflow by design is not there to make your code work. It's there to answer questions that arise during your attempt. And a question like yours, properly refined, can help those other people.)

jerrac• 1 year ago

Almost everyone who comes to SO, in my experience, has a fundamentally wrong idea about how the site is intended to work.

True. I quit trying to do anything there once I realized that SO was fundamentally not useful to me. It advertised as a gamified Q&A platform, but was actually a knowledge base psudeo wiki thing structured in way that didn't lend itself to answering the questions I needed answered.

So, I think a lot of the negative reactions are deserved, because SO looks like something it isn't.

People want a place to get help. SO looks like a place to get help. But SO is a place to ask for help only if your problem fits a specific set of requirements. And since most problems will never meet said requirements, most people can never actually get help on SO.

I post this in part because I'm still saltly about how much time I wasted trying to get help only to get downvoted, but also because if SO actually wants to do what they say, they really need to restructure into something that actually looks like what they want to be.

My suggestion would be to have two sites, one that is actually a general Q&A site like what everyone is after, the other is the kind of knowledge repository that SO wants to be. Then you just promote the really good questions from the Q&A site into the other site.

I'd also recommend ending the whole "downvote" idea. I have yet to see it not result in cliques and in discriminating against viewpoints the people with downvote permissions don't like. Let a lack of upvotes cause poor content to drop to the bottom.

rendaw3• 1 year ago

Is this satire? First you argue that it's a site for "practical" questions and therefore closing the question was exactly how the site should work (how do you determine it's practical), then it gets reopened and you say oh the question actually was practical so it's obvious it'd be reopened?

The current state of SO is what it is, but saying this is how it was intended to be is 100% revisionist history.

Also "we" "our"... do you work at SO?

npodbielski• 1 year ago

Hey mate is it your post? I did glanced at it and it does not stop because server is not notified that client is not there.

Or at least that is my guess, since I stopped working with WCF about 2016 probably.

Anyway in newer version of .net you CancellationToken everywhere what would do exactly that: tell your server that client disconnected. That would be my first try on fixing it.

Use token that is sent via HTTP implementation to the endpoint, pass it to your stream and when it is cancelled, end the stream. Stream ends, endpoint finishes, not CPU load.

treebeard901• 1 year ago

Yeah the stream isnt being closed properly as it is likely still trying to fill a buffer that no longer exists. The CancellationToken is a good idea regardless because you know once youve written the result, that you no longer need to wait for results.

Using has similar problems with things like database connections remaining open.

the_clarence• 1 year ago

The same has happened on reddit a long time ago. Most users give up early because they get their forst posts (on any community) removed many times before they can manage (if they do manage) to post it. If the feedback loop was faster (you instantly get feedback on why the post doesn't go through) it would be better although you would already lose some users. The situation is so bad that I predict reddit is slowly dying already

ryandrake• 1 year ago

As a general rule, I'm not going to take the time to donate free content to a site where moderators just delete it. This goes for S.O., Wikipedia, Reddit, Social Media, OSM, even HN. If my posts ever start getting flag-killed here, I'm not going to complain--I'm just going to leave, assuming the feelings are mutual. I used to habitually post to Fark.com, and when their moderators started going out of control and deleting my (and others') posts, I just canceled my subscription and went away. Who needs that grief?

If S.O. believes that deleting everything users post there is somehow improving their site and going to make it relevant again, more power to them. It's their site. Let's see how that goes for them.

socalgal22• 1 year ago

this will no longer happen because everyone has learned to just ask the LLM who will just answer. No need for reddit or S.O. for tech questions

rendaw3• 1 year ago

Just anecdoting, a couple weeks ago I asked my first question about the definition of `width` in CSS and how it pertains to minimum size calculations.

It was immediately closed as off-topic with 3 votes. One commenter asked "can't you figure that out using developer tools?"

It was eventually reopened, but I'm never asking a question there again. FWIW one of the governing web bodies essentially says "If you have technical questions, ask them on SO and not here" - I take this to mean that they don't care if you are able to answer your question, they just want you to go away.

prinny_• 1 year ago

I have stopped asking questions there. The moderation is so heavy handed it astonished me. I asked a question about some typescript types of an external library, the very author of those types replied and answered my question and I replied that I was happy with his answer and to thank him for his work on the library. The mods deleted my reply.

Was it so important for the quality of the site to delete a reply thanking someone for their OSS contribution? Is SO a better site now? Meanwhile they have 10 year old answers still up, with no way to indicate they are outdated unless the author themselves goes through the trouble of updating their answer.

Const-me4• 1 year ago

Not sure it’s the main issue, but your infinite random stream implementation is broken due to integer overflow. Stream.Position is using int64 for a reason, but your _sequence field is int32 which wraps after 2GB. BTW C# compiler has a setting to convert arithmetic overflows into runtime exceptions, too bad Microsoft neglected to enable by default.

throwaway20372• 1 year ago

I upvoted your SO.com question Can you put a bounty on it? That will help.

palata12• 1 year ago

I used to be very active on StackOverflow, it was a great platform.

After a while, I stopped having to post questions about "common frameworks", either because I could do with the official docs of because there was already a StackOverflow answer for my question.

What was becoming more common was that I would have a question similar to an existing unanswered one. Or that my question would never receive an answer (presumably because my questions were becoming more tricky/niche). So what I started doing was answering my own question (or answering those existing unanswered ones) after solving it on my own. Still, it was fine and I was contributing.

And for some reason, a few years ago my questions started being closed for no apparent reason other than "those who reviewed it have no clue and think that it is invalid". Many times they closed even though I had posted both the question and the answer at the same time (as a way to help others)! The first few times, I fought to get my question reopened and guess what? They all got a few tens of votes in the following year. Not so useless, eh?

Still, that toxic moderation hasn't changed. If anything, it has gotten worse. So I stopped contributing to StackOverflow entirely. If I find information there, that's great, if not, I won't go and add it once I find a solution for myself. I am usually better off opening an issue or discussion directly with the upstream project, bypassing StackOverflow's moderation.

I heard people mentioning that LLMs were hurting StackOverflow badly. I'm here to say that what pushed me away was the toxic moderation, not LLMs.

esafak4• 1 year ago

The moderators were elected. What should StackOverflow have done, held a vote of no confidence? Given them less power; make moderation more democratic?

agos3• 1 year ago

maybe set different guidelines for moderation? have some form of meta moderation?

hobs2• 1 year ago

The core problem of SO was the the goal of it (and what made it great) is very much in tension with "I want to ask whatever questions I want."

The original idea of SO was building a knowledge repository, and that meant no duplication and pruning it endlessly to make sure it was useful and up to date (which pretty much failed until recently, until its probably far too late) - this core tenet is something the moderators take seriously, and people using the site as questioners (not searchers) absolutely hate.

You can see they are trying to experiment (again probably too late) with how to make question asking easier, more friendly, etc - but that sort of cuts against the core original goals of SO and that's why the mods and the users seemed to be always in tension.

fiskfiskfisk• 1 year ago

In my own experience it's not often the elected moderators that are the problem, but those with a golden tag in a specific tag. They're far too eager to close questions because they're the ones culling through a tag often - and then close the question as they quickly think "oh, it's that again".

But it often isn't, they just didn't spend enough time to see nuance.

And neither do they see that even if _they_ understand that the question linked to is the same thing, there is no way the asker can understand what the similarity is from their knowledge point of view (or why the linked duplicate question is the same question).

mschuster91• 1 year ago

There should always be some sort of human to appeal to. Even Wikipedia has it this way at least formally [1], although "office actions" overruling community moderation decisions are extremely rare.

[1] https://foundation.wikimedia.org/wiki/Policy:Wikimedia_Found...

zabzonk• 1 year ago

Yes, they WERE elected, by the community (i.e. those asking and answering questions) and did a good job. Then those elected were dumped by the new owners and replaced by a bunch of yes-men. The people voting on closing submissions were not mods (in general) but ordinary users with sufficient rep.

lolinder5• 1 year ago

Look, I'm all for democracy in the real world, but this is a very bad use of democratic processes for a number of reasons:

1. First and foremost, it's not a democracy if your turnout is too low. The 2024 election had voter turnout of 2%, which would be a catastrophic turnout in any real democracy. Either too few knew about the election, too few thought their vote mattered, or too few had any idea how to choose between the options. Any of these reasons requires immediate and pressing attention if a democracy is to be called that.

2. Never having re-elections makes it useless. For it to be a truly democratic process there absolutely needs to be a way to withdraw consent from a moderator who behaves differently than expected. So yeah, a no confidence vote would be an option, or better yet regular elections to hold a position.

I'm afraid that the largest problem is that democracy is really just a bad fit for this kind of site. By its nature the only people who are likely to vote in this type of election are a small dedicated core, not the enormous number of users that the site actually serves. A small core of contributors to a community resource invariably seems to develop a sense of "us against the world"—the thin blue line of police lore isn't an isolated thing, it's what happens when people view themselves as lone defenders of something they care about. And just like with police, that can result in a toxic culture that begins to actively degrade the plebian outgroup that they started out serving.

I don't have a better answer, but I don't believe democracy is a good fit for moderators on the scale of community that Stack Overflow operates. It's too big to have good turnout, and the problems caused by bad turnout have become catastrophic.

wokwokwok2• 1 year ago

How about making it a site where only people who answer questions can even be eligible to be moderators?

What if moderators had to actually have karma from recently answering questions or they lose mod privileges?

Wouldn't that be a fresh change. You'd have to actually work to be a mod.

...

It shouldn't be controversial. That mods currently make visitors unwelcome is disgrace. :(

That SO incentivizes that behavior is ridiculous.

bombcar3• 1 year ago

Slashdot of all places basically solved the moderation problem, with random moderators selected from the pool of "know users" and then others selected to meta moderate.

malfist2• 1 year ago

I think you misunderstand how stackoverflow works.

Super moderators are elected, but not your regular "moderators". In stack overflow, regular folks you have enough karma are moderators and can cast votes, or initiate voting on moderation action. Enough votes and the action happens.

The elected moderators aren't the problem, generally, it's that anyone with a bit of karma can go power tripping and if you get enough of those people on an ever growing platform, they reach a critical mass to stifle anything.

So yes, some moderators are elected, and yes moderation is very democratic.

zahlman39• 1 year ago

The moderators were elected.

The overwhelming majority of the actions people complain about in this context (never mind that they don't understand the purpose of those actions or the underlying objectives) are not performed by moderators. They are curation actions taken by members of the community.

The rights to do so are awarded based on reputation, in a very poorly thought out and fundamentally broken incentive system; but there are far more people involved than the moderators. You can query by reputation at https://data.stackexchange.com/stackoverflow/query/1834631/c... : there are about 29 million total user accounts, 3.3 million which may upvote, 1.1 million which may downvote, 150 thousand which may unilaterally edit posts, 100 thousand which may vote to close questions, 28 thousand which may vote to soft-delete posts (and view soft-deleted posts), 9300 with access to internal site analytics...

and twenty-four moderators (https://stackoverflow.com/users?tab=moderators). Who are not the highest-reputation users. (I have more reputation than over half of them, and I frequently complain about users with over ten times my reputation.)

agos3• 1 year ago

just a few months ago they closed a question of mine that I posted in 2010 (!), which in the meanwhile had gathered more than 1000 votes, nearly one million views, and 20 or so answers. I get it that it does not meet their most recent criteria, but closing a question after 15 years telling me to edit my question and read the comments on how it could be improved (there were none) sounds tone deaf and unnecessarily bureaucratic

Larrikin2• 1 year ago

What was the question?

If it is from 2010 and was a relevant question or answer then but has since become irrelevant or even wrong because the framework or language has moved on I actually support this kind of clean up.

There are a lot of best practices that just don't apply anymore that far down the line. Even simple things like whats the best way to use a variable inside of a string in Python would have an outdated (and to most users, wrong) answer if it was from 2010.

trollbridge2• 1 year ago

And the irony here is that much of what LLMs know is from training on StackOverflow.

eru5• 1 year ago

StackOverflow's content is contributed by regular folks under an open source license.

devrandoom• 1 year ago

I stopped flagging things on SO when some of my flags were deemed unhelpful. They clearly weren't.

zahlman39• 1 year ago

The underlying system has some weird behaviours and moderators are sometimes compelled to mark flags they find helpful as unhelpful. This might happen, for example, if you report something as "rude or abusive" where they agree that the content should be removed, but disagree that the user should be penalized. But also, a lot of problems are better handled by just editing out objectionable content. When you flag a question as a duplicate or as "needs improvement", that goes to the community; but everything else is sent to a team of 24 volunteers overlooking a site that still receives thousands of questions a week and is still full of years-old comments that should be cleaned up but which nobody has gotten around to yet.

handsclean• 1 year ago

I’d appreciate if somebody more familiar with SO would verify this, but I believe there’s some low constant number of close votes required to close something, and this doesn’t adapt to how many people are voting or to positive signals. Because there’s an error rate in all things, this naturally means that things are wrongly closed all the time, especially content that’s viewed a lot and not fought for.

avereveard3• 1 year ago

There's a metric incentivizing "maintenance tasks" so the system is biased toward the side of closing and duplicating.

And because recourse is so hard and goes trough the same gatekeepers anyway, they don't get any signal about the accuracy of the maintenance.

One of the reason I've left as well was bureaucrats wrecking havoc to perfectly reasonable answers trying to rack up these points.

Peak of the fenomenon was 2014 when people started publishing their so scores on their resumes, but the platform never really recovered.

palata12• 1 year ago

In my case, the questions were closed very quickly. I opened votes to reopen a few times, one of which eventually passed, and then upvotes started to come regularly.

As I said, those were pretty specialised questions, you can't expect to have 10 upvotes in the first day for those.

zerkten• 1 year ago

I had 15k reputation score at one time on SO. It recently dropped down to due to people deleting their accounts. There are review queues which appear on the top nav incentivizing power users with enough rep to go in and take action like closing requests.

Having met many SO power users in group settings over the years, I feel that there is very little tolerance for questions that require effort to understand. If it's a simple question posed by a non-English speaker that needs some thought, then it doesn't belong (but that's why comments were introduced later.) The same goes for a deeper technical question where the author gets it all out but doesn't take the time to structure or format it. The volume of behavior like this differs based on the type of question and experts prepared to weigh in.

This gets compounded by the up and comers on the reputation scale. They get their special powers and see this BOFH close behavior and replicate it. Over time it starts to become the norm. I had the ability to vote for reopens and these same people would argue about why this was a bad idea. They weren't prepared to admit they were wrong and felt they were doing God's work by ridding the site of poor questions when some of us even had the ability to make edits to clarify them.

I just opened the site after some time away. At the top, pushing the question list below the fold, are: Reputation, Badge Progress, and Watched Tags blocks. The Interesting Posts for You question feed is below that and I have to go see how that is constructed. I only ever wanted the firehose of new questions with my tags highlighted.

EDIT: The behavior I noted above is yet example of why I always want to know how a job candidate deals with ambiguity. In my experience, this has a massive impact on the ability to work independently, not piss off colleagues/clients/customers, and make good decisions.

zahlman39• 1 year ago

Hi, I'm intimately familiar with Stack Overflow (https://meta.stackoverflow.com/users/523612).

It requires either:

* Three votes (it used to be 5) from community members with the close vote privilege (awarded at 3000 reputation)

* Unilateral closure by a moderator (there are currently 24 of these: https://stackoverflow.com/users?tab=moderators - compare to 29 million user accounts: https://data.stackexchange.com/stackoverflow/query/1877958/c... )

* Unilateral closure as a duplicate by a user with the close vote privilege who also has a gold badge for one of the tags originally used on the question (https://meta.stackoverflow.com/questions/254589)

The thresholds are deliberately fairly low, mainly because closure of new bad questions must happen promptly for the site to work as intended (https://meta.stackoverflow.com/questions/260263). This is frankly a major fault in the site design; but the new Staging Ground feature (https://meta.stackoverflow.com/questions/430404) helps a lot, on the occasions when the site software actually decides to use it.

However, "closing" content "that's viewed a lot" (this basically only ever means old questions; new questions rarely ever get a lot of views, regardless of quality, unless it's from spambots - see https://meta.stackoverflow.com/questions/431084) is emphatically not wrong. We close old, popular questions all the time, because they don't currently meet site standards (usually, because they are no longer deemed on topic). This is at least partly to discourage new questions along the same lines; but the primary effect of closing a question is to prevent answers from being contributed. These old questions generally wouldn't need new answers (although edits to existing answers may be helpful - and are not blocked) even if they were still considered suitable.

metalliqaz• 1 year ago

I left SO a long time ago when it was really at its peak. Your SO score was starting to be used as a metric to get hired for jobs in the valley. At that time moderation was totally out of control. One contributor in particular* would find a post he didn't like, then go to chat rooms to rally a downvote brigade. It was so toxic I called it quits on the spot. I still use SO quite a bit, but through DDG. I search for something, it displays an answer that it scraped from SO. Some years ago I read that they made an effort to soften the atmosphere there because they realized it was chasing away female devs and other minority groups. I guess they couldn't turn it around.

*Later he took over the Flask project and I was still bitter so I stopped using that too.

ChrisMarshallNY• 1 year ago

Same here.

I have a rep that is based almost entirely on questions, not answers. I learned to ask questions fairly well (to the point I seldom get answers -there's a price to pay, for questions that are very specific).

In some cases, the question is a basic one, and doesn't need a code listing and sample project. It's still a perfectly valid, pertinent, thoughtful question, but not very verbose.

Those questions almost always get closed.

I have found that asking LLMs doesn't always get me the best answer, but I get an answer. In some cases, I can have an iterative refinement, where I keep adjusting the question, until I get a useful answer.

I've never gotten code from SO, that I can use without modification, but I have gotten some great answers, over the years, and have expressed gratitude and respect.

I have gotten quite annoyed with the "attitude" that is often expressed. There's no doubt that folks who ask questions, are considered "lesser beings" on SO. Just look at the question-to-answer ratio of the high-score individuals. Weird attitude, for a site that is pretty much completely reliant on questions.

Basically, I have just given up on SO, and have found LLMs to give me what I used to get from it.

In my opinion, they have killed SO.

matheusmoreira• 1 year ago

The constant closing of questions isn't enough for them, very often they'll straight up delete content as well, thereby completely wasting any effort that you or others have put in. Closed questions with useful answers and comments and links? Gone.

Why spend your own time and effort adding content to someone else's platform anyway? It's always a much better idea to write an article on your own website than a stackoverflow answer. Stackoverflow just takes a little less effort but that doesn't matter when your effort is likely to be invalidated anyway.

karmakaze• 1 year ago

I had stopped interacting with SO as well, though the fact didn't really cross my mind. I had a few popular answers that got be enough points to be a moderator. The experience there is similar--possibly worse because the idiocy is unveiled. In the past I've often argued to reopen questions, sometimes even making edits to make it more agreeable to other mods.

It's common for those to get shouted down based on some policy or other bureaucratic nonsense by those who have no idea what the question is actually about. The problem could be that many of those who don't do, moderate. It attracts different sorts of people than those that are actually working with the things being discussed.

PaulHoule• 1 year ago

Somehow I never found the StackOverflow game to be worth playing.

In retrospect it is a case study of a particular enshittification scenario: "benign neglect" Back when they published a data dump I had a project on my speculative list to clean up their database, take only the best answers, etc. For python, the numerous Python 2 examples

   print "something"

would get rewritten to Python 3

   print("something")

basically do the maintenance work they weren't doing. Personally I find their idea of what is a valid question to ask offensive. If you're coding in Java or Javascript for example, the question of "Guava vs. Spring" or "Vue vs. React" are probably more consequential decisions for your app as opposed to anything else but questions like that are forbidden.

zahlman39• 1 year ago

If you're coding in Java or Javascript for example, the question of "Guava vs. Spring" or "Vue vs. React" are probably more consequential decisions for your app as opposed to anything else but questions like that are forbidden.

Over time we found that hardly anyone asking questions could achieve the kind of "good subjectivity" that we wanted. Questions like this attract flame wars (which are especially obnoxious in a format with answer posts with non-threaded comments) and advertising for alternatives, add-ons etc. that result in a completely derailed discussion in a place that isn't supposed to have a discussion at all.

If you want to ask "what factors should I take into consideration when choosing..." then I would agree that can in principle fit on a Q&A site. But open-endedness again makes it hard to choose the best answers and ensure they float to the top.

The general principles are much the same at Codidact Software (https://software.codidact.com), but the scope is considerably wider than Stack Overflow's (https://software.codidact.com/posts/search?search=category%3...). You might have better luck with that kind of question there.

zahlman39• 1 year ago

After a while, I stopped having to post questions about "common frameworks", either because I could do with the official docs of because there was already a StackOverflow answer for my question.

Good. That's the site working as designed and intended.

What was becoming more common was that I would have a question similar to an existing unanswered one.

Then you should improve the existing unanswered question instead, and/or draw attention to it (https://meta.stackoverflow.com/questions/265874 ; https://meta.stackoverflow.com/questions/266338). Or, yes, answer it if you can. Thank you for doing so.

Or that my question would never receive an answer (presumably because my questions were becoming more tricky/niche).

That's a big presumption. I got an answer to https://stackoverflow.com/questions/75677825/ within hours.

for some reason, a few years ago my questions started being closed for no apparent reason other than "those who reviewed it have no clue and think that it is invalid"

This is absolutely not what happened. First off, when your question is closed, you get a banner at the top of your question indicating which of the few standard close reasons was chosen. The wording isn't always a great fit (especially in the cases where people voted for more than one close reason - please keep in mind that we neither write this explanation nor get to choose the text; it's pulled from a database following simple mapping rules, and even moderators have only very indirect influence over that database) but it does normally point you in the right direction.

Second, "I don't know the answer" is not a valid close reason. People constantly accuse (on the meta site and elsewhere) that someone else's close vote was motivated by this; there's never any real way to evidence that, and this kind of accusation is in fact what we consider toxic behaviour (an assumption of bad faith).

Many times they closed even though I had posted both the question and the answer at the same time (as a way to help others)!

The fact that you provide your own answer weighs exactly zero in the calculation of closing a question. It must meet the site standards. Part of the purpose of a question is to index the information in the answer - so no matter how brilliant your explanation of the underlying problem might be, your exposition of the problem is a limiting factor.

The first few times, I fought to get my question reopened and guess what? They all got a few tens of votes in the following year.

The community does make mistakes, in both directions. The meta site exists for a reason.

But part of "fighting to get a question reopened" is editing it. Changes you might think are trivial might be crucial according to our standards. Some questions fundamentally can't be fixed; but when they can, closing a question signals that the OP's perspective is needed to fix the problem, no matter how minor. If we could fix it (without worrying about trampling on your authorial intent), we would.

Still, that toxic moderation hasn't changed. If anything, it has gotten worse.

It's not moderation, but curation. It's overwhelmingly done by a community of volunteers - not by the two dozen or so moderators (also volunteers) looking over an accumulation of literally millions of users and questions.

And it isn't "toxic". Overwhelmingly, people aren't doing it out of any kind of vendetta or a desire to cause you or anyone else a problem. They're doing it to uphold a standard (https://meta.stackoverflow.com/questions/417476/) designed (really, developed over many years by community discussion on the meta site) to accomplish particular goals (https://stackoverflow.com/tour ; https://meta.stackoverflow.com/questions/254770).

I am usually better off opening an issue or discussion directly with the upstream project

If it's something that makes sense to handle this way, it probably doesn't also make sense in the Q&A format. We can't do anything about your bug report.

I heard people mentioning that LLMs were hurting StackOverflow badly.

A lot of people think so because the volume of questions has dropped off dramatically, and there's good evidence that people will ask an LLM instead of asking on Stack Overflow.

But this is not at all "hurting Stack Overflow", unless you're a staff member at the company and you specifically worry about the effect of this decline on ad revenue.

If asking an LLM - trained on millions of existing Stack Overflow questions, along with the rest of the Internet - gives you an actually working answer (and you're either in a position where you can deal with AI hallucinations, or are lucky enough not to experience one), then that is, almost certainly, not a question that helps improve the existing resource that is Stack Overflow. It's most likely a duplicate or near-duplicate.

Duplicate questions on Stack Overflow are not inherently bad; sometimes rephrasing a question helps by providing a "signpost" so that people who think about a problem in a different way can realize that it's still the same problem, and there's still the same fundamental question to answer about it. But we want everyone who has that question to find the same collection of answers; and we want that collection of answers to be high quality, not redundant, and categorized under a high quality version of the question. That way, when you use a search engine and find Stack Overflow Q&A, you get the best possible result, as quickly as possible.

Nowadays, there are about three times as many publicly available questions on Stack Overflow as there are articles on Wikipedia. Considering that the scope of Stack Overflow is "practical questions about programming", while the scope of Wikipedia is "literally any noteworthy real-world object or phenomenon", that's clearly too many already. So why worry about the influx of new questions slowing down?

cadamsdotcom2• 1 year ago

Looks like the framework is just going to keep reading to the end of the random number stream, but of course there isnt an end to it.

Is there some kind of `IClosableStream` you can implement? That’d give you a `Closed` method, which you can then use to let either your server or stream know that it’s time to stop reading (or the stream reached EOF) - even if it’s done with a flag that’s set when the client disconnects.

Maybe there’s already an optional `Close` method you’re not overriding?

__s• 1 year ago

Stupid idea: throttle stream by putting Sleep in Read

eterm8• 1 year ago

Thanks for trying to help.

On the client side, randomStream.Close will get called when it's disposed.

On the server side, I'm not sure what I could put into an overriden Close that wouldn't just be base.Close()? RandomStream itself doesn't own any resources that need cleaning up.

I could force WCF to use Session mode, and then add flow-control through a side-channel, so other messages could prepare the stream to internally buffer and then rewrite in requested chunks?

But at that point I might as well just use an apprpriately sized GetRandomBlock(ValueWithSequence[]), and chunk requests that way and abandon using a stream for this at all.

I'll have an experiment with that approach to try to find the best buffer size and whether streaming the buffer actually helps vs just having it as the message and letting WCF control the sending.

cadamsdotcom2• 1 year ago

What if your Close implementation sets a flag in the stream that, if set to true, causes it to respond to Read calls with an EOF?

There'd be some bookkeeping to keep track of the stream (so you can set its flag) then replace it with a new one the next time a client connects, effectively making each stream single-use. But you seem to be optimizing for rate of reads, not number of opens and closes you can do, so it seems that shouldn't be a blocker..

wzdd• 1 year ago

(non-Core) WCF seems to have a concept of "drain on close stream". Could CoreWCF be replicating that? https://stackoverflow.com/questions/1676563/why-is-wcf-readi... . The commenter there suggests Abort() instead (which is apparently a bad idea).

It's weird behaviour and I wouldn't have expected it either, since infinite length streams are pretty common.

yakz• 1 year ago

There are some events you can handle:

https://learn.microsoft.com/en-us/dotnet/api/system.servicem...

Could you use these to cancel the stream?

junto• 1 year ago

This is the route I would go too, but quite honestly once you’re using the more non-standard WCF bindings in CoreWCF, you’re kind of playing with fire anyway. Microsoft never really intended to support WCF and SOAP in Core since it relied on so much Windows API’s under the hood in .NET Framework and they only added some support behind CoreWCF because so many enterprise customers still had SOAP and .NET services flying about that they had pressure to support something in this area.

public class StreamingService : IStreamingService { private readonly ILogger<StreamingService> _logger;

    public StreamingService(ILogger<StreamingService> logger)
    {
        _logger = logger;
    }

    public Stream GetRandomStream()
    {
        var random = Random.Shared;
        var isClientConnected = true;

        var context = OperationContext.Current;
        context.InstanceContext.Closed += (s, e) =>
        {
            _logger.LogInformation("Client closed connection.");
            isClientConnected = false;
        };
        context.InstanceContext.Faulted += (s, e) =>
        {
            _logger.LogWarning("Client connection faulted.");
            isClientConnected = false;
        };

        return new MonitorableRandomStream(random, () => isClientConnected, _logger);
    }

}

public class MonitorableRandomStream : Stream { private readonly Random _random; private readonly Func<bool> _isClientConnected; private readonly ILogger _logger; private long _sequence;

    public MonitorableRandomStream(Random random, Func<bool> isClientConnected, ILogger logger)
    {
        _random = random;
        _isClientConnected = isClientConnected;
        _logger = logger;
    }

    public override bool CanRead => true;
    public override bool CanSeek => false;
    public override bool CanWrite => false;
    public override long Length => throw new NotSupportedException();
    public override long Position { get => _sequence; set => throw new NotSupportedException(); }

    public override void Flush() {}

    public override int Read(byte[] buffer, int offset, int count)
    {
        if (!_isClientConnected())
        {
            _logger.LogInformation("Stopping stream: client disconnected.");
            return 0;
        }

        var span = new Span<byte>(buffer, offset, count);
        _random.NextBytes(span);
        _sequence += count;
        return count;
    }

    public override int Read(Span<byte> buffer)
    {
        if (!_isClientConnected())
        {
            _logger.LogInformation("Stopping stream: client disconnected.");
            return 0;
        }

        _random.NextBytes(buffer);
        _sequence += buffer.Length;
        return buffer.Length;
    }

    public override long Seek(long offset, SeekOrigin origin) => throw new NotSupportedException();
    public override void SetLength(long value) => throw new NotSupportedException();
    public override void Write(byte[] buffer, int offset, int count) => throw new NotSupportedException();
}

jgalt212• 1 year ago

This is confusing enough for humans, let alone a machine that will happily reproduce calls and configuration from something that is almost identical, yet subtly different and incompatible.

Interesting point. I'm going to see if they similarly struggle generating VBA code vs generating Visual Basic code.

phkahler• 1 year ago

Stackoverflow is no longer a Q&A site. It's trying to become a curated information source. Like wikipedia or the kind of thing you'd train AI with. As such it isn't great for answering question any more.

Another old problem was notable users. There was a guy famous for his presence and answering tons of question (I forgot his name). He was actually pretty good but... he was not an expert in all the areas he'd participate in, but his answers would sometimes win because he was articulate, not because it was the best.

nicce3• 1 year ago

Likely this summarises the shift:

https://openai.com/index/api-partnership-with-stack-overflow...

zahlman39• 1 year ago

On the contrary, the people trying to keep the site a useful, curated source of information (rather than a dumping ground for people trying to get others to debug code for them) are on balance strongly against every kind of AI involvement with the site, including the use of that information for training AI (although we can't do anything about that). We curate a repository of high quality questions and answer (which is a "Q&A site", and is specifically what the site is supposed to be, so as to distinguish it from traditional forums) exactly because we want it to be a place where you get information from humans that reflects original human insight.

Which is also the reason for the ban on GenAI content (https://meta.stackoverflow.com/questions/421831).

neonsunset2• 1 year ago

If the author reads HN comments - findings like these are probably better to be submitted at https://github.com/CoreWCF/CoreWCF instead

(although I have no idea how active CoreWCF owners are w.r.t this)

shagie7• 1 year ago

... and https://github.com/CoreWCF/CoreWCF/discussions

eterm8• 1 year ago

Thanks. In my experience the CoreWCF team are fairly active, but I didn't want to bother them until I can understand if this is a problem with CoreWCF.

This might also be a problem WCF client, which is maintained by others elsewhere in a different repo: https://github.com/dotnet/wcf for the nuget package version.

But this might just be how WCF is designed. I'll try a version of this within .NET Framework, but even that might change depending on whether it's via IIS or started via ASP.NET Core, and whether it uses the built in System.ServiceModel or the nuget version.

( You can probably tell I'm a bit frustrated with MS for making a bit of a mess in the way they hurried away from .NET Framework especially with respect to WCF. )

neonsunset2• 1 year ago

Thanks. In my experience the CoreWCF team are fairly active, but I didn't want to bother them until I can understand if this is a problem with CoreWCF.

It may not be necessarily a problem but, ideally, the less users have to care about gotchas and knowing how to exactly use the API the better. There are some constraints to this but chances are at least documentation can be improved.

Plus, especially if there are not that many issues, it signals interest and active usage.

xyst• 1 year ago

C# is not my primary language but have the misfortune of debugging C#/.Net junk.

Seems like an issue with not closing resources properly. Looking at your server code, seems the Close and Dispose methods are not overridden. Try that?

eterm8• 1 year ago

What resources do you think should or could be handled in an overriden close method on RandomService?

It'll be calling base.Close(), and doing what else?

incoming12112• 1 year ago

"I dont know C# but ill make a bold assumption"

bitbasher• 1 year ago

I'm curious-- did you feed this into an LLM to see what it thinks the issue is?

eterm8• 1 year ago

I didn't, but I've not had much previous success with these kinds of issues.

Just for the sake of it, I've now tried pasting this whole blog into claude.

It has some strange suggestions, including many things that don't work, such as adding to the client:

                // Important: Properly close the stream
                randomStream.Close();

I read the Stream documentation ( https://learn.microsoft.com/en-us/dotnet/api/system.io.strea... ), and it points out:

This method calls Dispose, specifying true to release all resources. You do not have to specifically call the Close method. Instead, ensure that every Stream object is properly disposed.

My stream is properly disposed.

Claude also sticks in things like:

                // Simulate some work or add a small delay to prevent CPU spinning
                Thread.Sleep(1);

I can't see how that do anything to solve my issue. I suppose I should humour the machine, go full-vibe and try everything it suggests, and if I end up with a working solution go back from there, but I fear that would just leave me more confused about the underlying mechanisms here.

On the client side it not only rewrites my reading to read multiple times, but adds in:

                    await Task.Delay(500, cts.Token); // Small delay between reads

I didn't ask it to re-introduce the loop, and a 500ms delay between reads is horribly long for reading successive bytes from a stream.

The only thing that was interesting was creaitng a linked cancellation token source on the service to pass to the underlying stream and cancelling it on server shutdown.

That's a useful thing to keep in mind to help with helping to shutdown the services, but doesn't actually address the issue for a server you want to keep running.

It does also add send/receive timeouts too, and they're also worth keeping in mind, but that doesn't seem like a good mechanism for dealing with this issue. If anything, it would just mask the issue by having it write to the stream for the duration of the timeout period instead, which if short could cause a problem like this to go unnoticed until it's actually under more load.

zahlman39• 1 year ago

All of that is very much in line with what I'd expect Claude to suggest - because it's exactly what I'd expect to see on the garbage blogs that LLMs trained off of, in turn oriented towards fixing common problems from people who don't know what they're doing.

hakunin3• 1 year ago

My most recent (nearly 3 years ago) StackOverflow story:

I posted a question[1]. Got some answers, but not quite complete. Then someone came along and provided a good detailed answer. A couple of upvotes later, that answer got deleted by Community Bot. I voted to undelete it, but it still needs another vote to undelete. So I ended up copying it into my own notes blog[2].

I'm not sure why the best answer was deleted. It would've been a loss if it wasn't preserved somewhere I think.

[1] https://stackoverflow.com/q/73954228/155351

[2] https://notes.max.engineer/creating-common-interfaces-in-gol...

ayhanfuat• 1 year ago

That user was using AI to autogenerate answers. Most of his answers posted in a certain period were deleted in bulk. If you think this was not something AI generated then you can raise a flag for it to be undeleted https://meta.stackoverflow.com/q/430072

hakunin3• 1 year ago

Yeah, looking at it today I do smell the AI. Nearly 2.5 years ago I had no experience to suspect it. I'm less annoyed with it getting deleted now, but it was still a useful and relevant explanation at the time.

zahlman39• 1 year ago

For reference: almost all deletion on Stack Overflow is "soft", and deleted Q&A can be viewed by users with at least 10,000 reputation. These users can vote for undeletion, and the author can edit and submit for undeletion.

It looks like the answer you're talking about is now undeleted, with yours among the three necessary votes. (It also looks like you accepted, and awarded a bounty to, a different answer back when you asked, and the answer you're now calling "best" was posted months later. So it goes.)

The explanation for the original deletion is exactly as ayhanfuat says. LLM and other GenAI content is forbidden on Stack Overflow (https://meta.stackoverflow.com/questions/421831/), since before that answer was posted, and the author (one of the site's most prolific contributors) got in trouble for a flood of such answers. (It took quite a while for the problem to be noticed and acted upon, in large part because of a moderation strike in June 2023 which ended up creating a large backlog of flags for AI content, and in part because custom flags are required to report this. The strike, in turn, was largely about the company/staff trying to interfere with moderators attempting to detect AI content, with the site owners being unwilling to accept false positives because it would be "unfriendly". So it goes.)

hakunin3• 1 year ago

Makes sense, I recognize the LLM output now. In fact as I was posting the story here, it kinda dawned on me. Problem is, this happened before ChatGPT went mainstream, so I wasn't 100%. Definitely sympathetic to the hard decision made here, in light of this.

eYrKEC22• 1 year ago

I've found the _real_ answer in deleted answers multiple times.

Thank goodness I can see it because I have enough reputation.

BrandoElFollito• 1 year ago

I opened my SO account 15 years ago and I was liking the dynamics very much.

Now at 200k+ rep I cannot stand it anymore. Whenever I ask a question that does not make me bow backwards or discuss how my Python code looks in assembly it gets downvoted almost immediately. Go and Typescript are tags that being frustration.

Now compare this with many other Stack Exchange sites.

A question about cooking or LaTeX gets nice, coloriant answers, even when they are basic.

I get it that they are one or two orders of magnitude less crowded but this may mean that SO passed its scale limit and is done.

busymom02• 1 year ago

I have been an SO user for 13 years with around 9k karma. I have recently started having similar negative experiences.

I posted this iOS Swift Programming question today:

https://stackoverflow.com/questions/79612931/ios-custom-oper...

and was instantly downvoted and then voted to closed. I do not think this question was worth downvoting or closing. There's no hits on the warning I was getting.

People on SO nowadays seem way too eager to downvote and close.

zahlman39• 1 year ago

and was instantly downvoted and then voted to closed.

The outstanding vote to close is on the grounds of "needs details or clarity". Based on the comment that was left, I assume that the other party doesn't understand why you are asking a question, because it comes across that your motivation for asking is that you don't understand what to do, but it also comes across that the error message is already telling you exactly what to do.

If you have a different question - for example, why you should have to do the thing being suggested, or you don't understand how to do the thing being described - then you should edit the question to clarify that. With this clarification, it may be that your question is identified as a duplicate (of a more general question regarding why such "inherited conformance" must be restated in Swift, or what that means, etc.). I don't know Swift, so I can't advise there - I can only rely on my general understanding of programming languages in this general syntax family.

Regardless, the question would be improved by a proper minimal reproducible example (https://stackoverflow.com/help/minimal-reproducible-example). Minimal is a key component of this; please proactively try to remove irrelevant parts of the code example, so that you show only what directly causes the reported problem. This helps (https://meta.stackoverflow.com/questions/333952) other people understand and recognize the question - not just people who might answer it, but other people who have the same question who are looking for an existing Q&A.

There's no hits on the warning I was getting.

I guess you mean that you tried searching for the warning and came up empty handed. I'm not sure why this happened for you. If I try copying and pasting into a search engine - e.g. https://duckduckgo.com/?q=Class+must+restate+inherited+%27Cl... - I get seemingly useful results off the top - including issues in the GitHub repository for Swift itself, such as one proposing and automatic "fix-it" for this error (I assume this concept is meaningful to Swift developers generally). I also see official Apple documentation at https://developer.apple.com/documentation/swift/sendable , several blog posts talking about the use of `@unchecked Sendable` with fully worked examples, etc.

I do not think this question was worth downvoting or closing. People on SO nowadays seem way too eager to downvote and close.

Nobody ever thinks their own questions are worth downvoting or closure - they wouldn't ask if they did. But the standards aren't set by the OP; they're set by policy and surrounding discussion such as https://meta.stackoverflow.com/questions/417476 and https://meta.stackoverflow.com/questions/252677 and https://stackoverflow.com/help/closed-questions and https://stackoverflow.com/help/why-vote and https://meta.stackoverflow.com/questions/429808. And they're set this way because the site isn't about answering your question in the moment; it's about building up a searchable, useful repository of high quality questions and answers.

busymom02• 1 year ago

That's very interesting. When I Googled that error, the only relevant result I got was this revenuecat one:

https://community.revenuecat.com/sdks-51/many-rc-related-war...

However, when I searched for it now (after seeing you were getting that Swift GitHub result), I got better results. No idea why I didn't get that before.

Thanks for the detailed answer. I appreciate it.

eYrKEC22• 1 year ago

Stack Overflow is like online gaming -- lots of toxic people, but I still get value out of it.

Ask a question.

Ignore the trolls.

Get a question answered that would have taken you another day to solve.

Make money.

digianarchist• 1 year ago

Back in the day, some of the most toxic people were community moderators.

I barely use it anymore. GitHub issues are far more useful these days but lacks a place to ask generic questions.

GitHub could strike a knockout blow here if they wanted to.

xeromal2• 1 year ago

Can't do that if they delete the post before it gets answered

dkackman11• 1 year ago

SO community def not reading the room

incoming12112• 1 year ago

StackOverflows days are numbered, the site isn't like it used to be, it used to be useful, but over time people spend more time trying to game the system or enforce their own interpretation of rules.

SO is trash now-a-days.

bytefish• 1 year ago

For all reasons mentioned in the article, I wouldn’t post such a question on Stackoverflow. It doesn’t fit. And I see their reasoning. Just raise an issue in the CoreWCF issue tracker. What’s the worst thing to happen?

The CoreWCF folks would be delighted to have a reproducible example, and they are best suited for answering the question. They have written CoreWCF after all! And if you don’t feel comfortable with raising issues? There’s also a Discussion feature on GitHub, where you can ask more free form questions.

And once you fixed or analyzed the issue with the CoreWCF team, you could write a blog post about it and share it with anyone having a similar issue.

That’s how I do these things.

jiggawatts2• 1 year ago

I think you found a bug, but you were doing something that was never going to work, and you reported it in the wrong place.

The right place for this kind of question is the GitHub Issue tracker for CoreWCF, where you'll find several related issues. You're not the first one to hit bugs in streaming mode, which is... fragile. Just turning HTTPS on or off can break it: https://github.com/CoreWCF/CoreWCF/issues/1391

Unbounded streaming is something WCF was not designed to support, and hence CoreWCF doesn't support it either. CancellationTokens, etc... might improve things in case of accidental disconnection of a client session, but in general WCF expects streamed responses to have a finite length and be explicitly terminated with an end-of-stream by the server.

E.g.: read the end of this section: https://learn.microsoft.com/en-us/dotnet/framework/wcf/featu...

"Streamed transfers end and the message is closed when the stream reaches the end of file (EOF). When sending a message (returning a value or invoking an operation), you can pass a FileStream and the WCF infrastructure subsequently pulls all the data from that stream until the stream has been completely read and reached EOF."

Fundamentally, WCF is a message-oriented RPC system. It has many issues with streaming: it has to be explicitly enabled, it is incompatible with many features, etc...

Just "read between the lines" in your stack trace! It mentions XML encoding in several places, but XML as used in RPC uses a "tree" structure which must be terminated with closing tags.

You might be able to shoehorn unbounded streams into CoreWCF by implementing custom channels and encodings, but at that point it isn't really WCF any more!

The bug is that CoreWCF really should give up and stop writing if the client is disconnected. Stepping through its code, it does look like it loses track of the CancellationToken at some point, probably riiiight here: https://github.com/CoreWCF/CoreWCF/blob/067f83b490332120f45e...

Note the comments surrounding it talking about things like "TODO: Verify that the cancellation token is honored for timeout" and "Since HTTP streams don't support timeouts, we can't just use TimeoutStream here." That sounds unfinished to me.

I can't be sure, but there seem to be some design flaws that also contribute to this bug. E.g.: CoreWCF uses synchronous writes to ASP.NET Core's pipeline, which is... also a buggy mess because it is so rarely used.

Then, somehow, it is still writing to a socket that is long gone, which is really odd. It really does feel like there's at least a DoS security bug here, either in ASP.NET when operating in sync mode or in CoreWCF. The issue is that malicious clients could easily request many large responses, disconnect, and the server would keep processing the requests for hours and hours.

eterm8• 1 year ago

Thank you, if you write that up as an answer on SO I would be happy to accept this there too. I'll write up a bug on the issue coreWCF issue tracker over the weekend.

The more I looked into the issue, the more it felt like I was trying to do something WCF isn't designed to do.

jiggawatts2• 1 year ago

I don't have an SO account, but I welcome the kudos here!

My experience with Microsoft-related stuff now being done "in the open" on GitHub has been great, because it means that users like us can directly open an issue ticket right where the developers are day-to-day. The tickets get attention and/or fixes more more often than screaming into the void on some generic forum full of only vaguely interested people.

darrylb42• 1 year ago

This does not appear to be support by SOAP or WCF. From the SOAP specification:

"Bindings MAY provide for streaming when processing messages. That is, SOAP nodes MAY begin processing a received SOAP message as soon as the necessary information is available. SOAP processing is specified in terms of SOAP message infosets (see 5. SOAP Message Construct). Although streaming SOAP receivers will acquire such XML infosets incrementally, SOAP processing MUST yield results identical to those that would have been achieved if the entire SOAP envelope were available prior to the start of processing. For example, as provided in 2.6 Processing SOAP Messages, identification of targeted SOAP header blocks, and checking of all mustUnderstand attributes is to be done before successful processing can proceed. Depending on the representation used for the XML infoset, and the order in which it is transmitted, this rule might limit the degree to which streaming can be achieved."

https://www.w3.org/TR/soap12-part1/

The critical note is the output for streamed, or buffered must the the same so it does not look like infinite streams can be supported.

I also played with your sample application to see if there was any indication of the connection dropping.

I had an interesting result switching to the NetTcp binding. On the client it is unable to exit until the max size of the incoming message is hit. The server continues running as there are no sent message quotas.

My stackoverflow question was closed so here's a blog post about CoreWCF

Related Stories