« first day (2093 days earlier)      last day (2894 days later) » 
07:00 - 19:0019:00 - 23:00

user15026
7:23 AM
@KitZ.Fox couldn't sleep so I watched your toast video and it was just what I needed. (although now I want toast and I am out of bread.)
 
9:09 AM
why is this room so dark ?
 
???
 
@Ixrec dark as in dead
 
people rarely talk here
 
it can get lively here
depends on when people need their sleep
 
few chatrooms are active 24/7, partly because few people want to spend literally every hour of the day chatting
though this one has been quiet-ish lately due to a number of power users sort of kind of leaving the site
 
9:25 AM
ah I see, that could be it
 
10:08 AM
I am having a trouble with complex logic, if anyone interested to have a look
 
10:28 AM
This site is meant for specific problems. Maybe it is better that you try to find a more specific discussion board for beginning programmers. — Lexib0y 6 secs ago
 
 
1 hour later…
11:44 AM
I don't get why people are so happy about the result of Oracle v. Google. Sure, it was the best possible result. But it's still pretty crappy.
 
I'm resisting the urge to try and figure out wtf the outcome of all those cases means since I know it's never going to have any impact other than getting me into more internet arguments with everyone else who thinks they know what it means
 
The only thing I'm learning is that the US court system is almost as screwed up as US intellectual property laws.
 
there's probably a causation relationship in there
 
Perhaps. I mistakenly thought that federal courts were the same. But one court said that copyright does apply to APIs. Other federal courts at the same level aren't bound by that decision and can find the exact opposite.
So unless either the Supreme Court or Congress weighs in at a federal level, we don't know for sure if APIs can be protected by copyright. But one set of courts says that it is and can follow that precedent while others can ignore it.
 
and chances are it'll go to settlement and remain unresolved...
 
11:54 AM
I hate unresolved issues. Need more Thunderdome.
 
I thought it had gotten more complicated than "Can APIs be protected by copyright? Y/N"
 
So that was a question at one point. The Federal Circuit said yes, even though Google argued no and the EFF submitted a brief from a bunch of industry people on how damaging a yes would be to science and engineering.
So then it went back and the jury had to answer the question of "Was Google's use of this copyrighted API fair use?"
 
so the full answer is apparently APIs are protected by copyright but cleanroom implementations of them are fair use?
 
that sounds like...the question that actually mattered got resolved in the obviously completely wrong and highly damaging way, but then they found a loophole so...now what?
 
Now, we wait.
 
11:58 AM
until oracle files the appeal
 
this makes me wonder if their definition of API is completely different from ours
like they're talking about the typical implementations of the API rather than the interface part or the formal specification or whatever
 
The original judge that said that APIs aren't copyrightable was probably the most appropriate judge in the US to make that decision. A BS in mathematics and someone who knows how to write code.
@Ixrec The thing in question was the method declarations. Something like "can you copyright public String toString()?"
William Haskell Alsup (born 1945) is a United States federal judge. Born in Jackson, Mississippi, Alsup received a B.S. in mathematics from Mississippi State University in 1967, a J.D. from Harvard University in 1971, and an M.P.P. from Harvard University's John F. Kennedy School of Government in 1971. He was a law clerk to Justice William O. Douglas of the Supreme Court of the United States from 1971 to 1972. Alsup was in private practice in San Francisco, California from 1972 to 1978, and was then an Assistant to the U.S. Solicitor General in the United States Department of Justice from 1978...
 
what exactly was the argument in favor of copyrighting that?
 
That dude for Supreme Court. He went and learned Java so he would know WTF he was talking about.
 
middle name Haskell?
 
12:00 PM
that's an awesome coincidence
 
@ratchetfreak o_o It cannot be unseen.
 
Judge Haskell knows programming...
 
Oh. The Supreme Court declined to hear the case. So there can now be countless trials to determine if the use of a particular API or not is fair use.
Isn't reimplementing proprietary interfaces how Stallman and company got their start, and created the GPL?
 
it's how a lot of current implementations of nearly anything got started
 
this sort of reinforces why we need the GPL
 
12:05 PM
@Ixrec This could also break the GPL. If the API is protected and it's found that my use of it isn't fair use, then I can't license my implementation of it under the GPL.
 
GPL is overreaching though
 
I agree, but the license doesn't matter. There's also the question of what an interface is. Can you consider the structure of a .DOCX file format an interface that is protected? Microsoft can go after Apple and Apache Foundation for their office products, for example.
 
This question is highly theoretical and most likely more suitable for programmers.stackexchange.com rather than Stack Overflow. — kb. 40 secs ago
@kb. We have the same rules on questions being too broad as Stack Overflow does. This question asks about several things, all of which are too broad for Programmers. — Thomas Owens 49 secs ago
 
I almost want to give some money to the EFF, but I don't agree with everything they do.
 
since I'm not paying attention to most of this stuff, what do you disagree with?
 
12:20 PM
A lot of their stances on national security issues, mostly.
 
1:12 PM
@AshleyNunn I'm glad you liked it.
 
1:41 PM
@ThomasOwens The first court said that API's are copyrightable. The second court said the use of said api's falls under Fair Use laws. Both positions are compatible.
 
@RobertHarvey Judge Alsup said that "So long as the specific code used to implement a method is different, anyone is free under the Copyright Act to write his or her own code to carry out exactly the same function or specification of any methods used in the Java API. It does not matter that the declaration or method header lines are identical".
He also said that the structure was not copyrightable under section 102(b) of the Copyright Act since it is a system or method of operation.
 
Yep.
It does make you wonder what he point of copyrighting an API is.
 
That was the first court saying that APIs are not copyrightable.
Oracle appealed and they said that the APIs are copyrightable and sent it back to the first court where a jury had to consider if Google's use was fair use or not.
 
You mean the "structure," right? What is the structure, anyway?
 
@RobertHarvey The organization of the files and functions.
It also applies to things like user interface screens.
 
1:45 PM
OK.
So same as it ever was, basically.
 
The fact that you have a public String toString() in an Object class is structure.
How that's copyrightable is beyond me.
I think that a judge named William Haskell Alsup who went off and learned Java to understand the trial would know, though. And he said it wasn't.
 
while not an argument I like to invoke too often, there really is no plausible explanation for that other than everyone involved in the decision making process having absolutely no idea how code works
 
@Ixrec Which is one reason why Alsup's original ruling that an API is not copyrightable is stronger. He has a BS in mathematics and went off to learn Java to understand the language before he made his ruling.
 
it makes about as much sense as copyrighting the physical structure of a book
 
@Ixrec or how using and implementing an interface works
 
1:47 PM
@Ixrec I think Google made a similar argument regarding chapter headings in a book.
 
if you look at it from that perspective then third-party unlicensed aftermarket add-ons are fair use
 
@ratchetfreak Fair use depends on specific instances of using the copyrighted content. You can't make a blanket statement.
 
look at what from what perspective?
 
@Ixrec implementing and using interfaces
 
Which is one of the problems with this ruling. Any time anyone copies an API and makes their own implementation, the owner of the API can assert that their usage wasn't fair use.
 
1:49 PM
incidentally I still don't get why fair use is relevant, I thought that was about defending works of critique or educational materials or whatever
 
Copyright is all about the expression of an idea. You could make the claim that API is not the expression of an idea because it is not an implementation. However, few would argue that REST is not the expression of an idea.
@Ixrec That's one purpose of fair use. Possibly the most important.
 
that was the original purpose of copyright, because back when it was created that verbiage accurately captured what they wanted to do
 
@RobertHarvey This is what Alsup said, until the higher court overruled that.
 
I suppose you could argue that developing a high-quality API takes time, effort, research, money, etc, and can provide substantial commercial benefits, which is the standard argument for why other copyrightable works need some form of protection from copycats
 
I'm starting to get the feeling that the US court system is suffering from too many cooks syndrome
 
1:53 PM
but of course those substantial commercial benefits only come into existence by providing an implementation of said API
copying someone else's implementation and selling that is clearly something that should be off-limits
 
Common interfaces leads to competition.
 
exactly
most of the internet wouldn't be possible if we weren't allowed to implement our own HTTP clients and servers
 
Haskell Brooks Curry (/ˈhæskəl ˈkɜːri/; September 12, 1900 – September 1, 1982) was an American mathematician and logician. Curry is best known for his work in combinatory logic; while the initial concept of combinatory logic was based on a single paper by Moses Schönfinkel, much of the development was done by Curry. Curry is also known for Curry's paradox and the Curry–Howard correspondence. There are three programming languages named after him, Haskell, Brook and Curry, as well as the concept of currying, a technique used for transforming functions in mathematics and computer science. == Life... ==
^^^ right Haskell to judge
 
the type of question you ask usually gets a good reception here: programmers.stackexchange.commontewhizdoh 40 secs ago
 
another good analogy might be copyrighting the QWERTY keyboard layout
 
2:02 PM
You question doesn't fit with SO. Drop your question in programmers.stackexchange.comroland 52 secs ago
 
2:12 PM
@Ixrec the layout was patented google.com/patents/US4522518
 
@Ixrec That patent appears to describe a QUERTY layout with a central rcolumn of home/end etc. keys.
 
I found the dvorak patent: google.com/patents/US2040248
 
Actually it described inserting an arbitrary column of keys. Stupid patent.
 
@RobertHarvey Yeah. QWERTY is prior art. They took QWERTY and...screwed it up and patented it.
 
the dvorak patent is explicitly about the layout
 
2:19 PM
I suppose you could patent a new keyboard layout.
That's what patents are for - to protect a useful thing.
 
I don't have any problem with patents. What I have a problem with is patent trolls taking out many specious patents (with no intention of ever using them to produce anything useful) and using them in a predatory way to extort money from other companies. That is a very serious problem that perverts and distorts the patent system and anything that touches it, and it needs to be fixed.
 
like filing a patent paying to keep it private and then adjusting it to what is on the market to start filing suits
 
@ratchetfreak You can pay to keep a patent private?
The whole point of a patent is disclosure of an invention.
 
it seems like such an easy problem to solve too, just require the patent to have been used (or some evidence of preparing for an attempt to use it in the future) when filing a patent lawsuit
 
A submarine patent is a patent whose issuance and publication are intentionally delayed by the applicant for a long time, such as several years. This strategy requires a patent system where, first, patent applications are not published, and, second, patent term is measured from grant date, not from priority/filing date. In the United States, patent applications filed before November 2000 were not published and remained secret until they were granted. Analogous to a submarine, therefore, submarine patents could stay "under water" for long periods until they "emerged" and surprised the relevant market...
 
2:25 PM
Ah. It's no longer possible in the US.
Applications are published now, so you can't do this. But that's why trade secrets exist.
 
@ThomasOwens Granted my understanding of the website scopes might be lacking, but I base it entirely on Meta answers. Both meta.stackoverflow.com/questions/254570/… and meta.programmers.stackexchange.com/questions/171/… support my stance. This is definitely more whiteboard than code/IDE, no? — kb. 2 hours ago
^^^ time to properly dupe-close mentioned reference at our meta to up-to-date FAQ. Staying open it keeps confusing SO folks
Historical note: this is a very old meta question and it has been superseded by What goes on Programmers.SE? A guide for Stack Overflow. We discussed this issue in chat for reference. — Snowman Jul 10 '15 at 17:21
 
I can do that later. I just started a code review. Can you throw up some meta flags so I remember?
 
@ThomasOwens sure will do
 
Just be specific as to what I need to do so I don't need to come back here and find this message after lunch.
Thanks.
 
2:44 PM
This code review touched so many files. :\
Anyway, random question. If a company is lame and requires a certificate in field, do they care if it's expired if it's one that expires?
 
surely that would depend on the company? if they believe in arbitrary checkboxes who knows how many of the subcheckboxes they know to look for
 
@Ixrec I was more curious if anyone had experience or if there was a trend.
It does depend on the company, but is there a trend toward requiring a current certification or just taking the test and passing it at some point to prove knowledge and understanding?
 
may depend on how long ago it expired and how the tech of the certificate changed
 
3:23 PM
Your saying you have the cojones? — user1585345 49 mins ago
 
Uh...?
I read that comment chain.
WTF.
 
Guy thinks a lot of himself. Or at least his supposed innovation.
 
It does sound kind of cool.
I'm not familiar with the .NET ecosystem. Is this something useful and relevant?
 
I'm not seeing it. We use EF at my shop, but have been gravitating to a model that uses pure SQL. It's not that difficult.
 
It sounds like he made an IDE plugin to make it easy to support pure SQL?
Or am I misunderstanding?
 
3:27 PM
It apparently generates wrapper classes for your .SQL files. I guess that's kinda cool. But I need to look at it more closely.
 
chance of spam?
 
That's exactly what it is.
Really, I just poked him because he apparently thought his "innovation" was an answer to the question that was asked.
 
Is it not an answer?
 
I mean, how could it be? "My thing fixes whatever problem you think you have, and I've thrown in a set of steak knives for good measure."
 
The cafe just opened. Let me read the question and his answer again. I haven't read the question yet.
 
3:30 PM
@RobertHarvey If I had a nickel for every time a guy said that to me.
 
 
1 hour later…
4:46 PM
I like this room.
 
5:00 PM
I think you should either move it the webmasters.se or programmers.stackexchange.com. — Jerrybibo 27 secs ago
 
5:12 PM
@Jerrybibo do not encourage cross-posting. Rahul, if this question should be migrated, you'll need to flag it for a mod to migrate to another site. — Aaron Hall 1 min ago
 
5:42 PM
man...
ok, preemptive strike: have a nice long weekend!
 
I hope this is OK, but I posted another answer to the thread for changing site proposals that summarizes all changes.
1
A: New Site Name and Scope Proposals

RachelThis answer is to summarize the requests for Stack Exchange in a single post. It is a community wiki, so if something else seems to be a consensus in this thread, feel free to add or edit it. Name Change to Software Engineering Tour / Tag Line Software Engineering Stack Exchange is a qu...

Does anyone have any concerns about the initial writeup that they'd like to see changed?
 
What happened to 'Data Structures and Algorithms?'
 
It is covered under one of the other bullet points
 
There's a very specific reason I put that in there. I want the community to know that we're still about nuts and bolts (and not just management principles), even though we don't troubleshoot code.
 
I forget who posted the answer stating that... Thomas I think
 
5:52 PM
The "four bullet rule" is misguided anyway. If we need seven bullets to do it properly, then we need seven bullets.
 
Yeah, it's from Thomas' answer - "I removed "algorithm and data structure concepts", since I would consider that part of "software architecture and design"."
We can add it back in though if we think it is important
(personally algorithm and data structure questions are some of my favorites)
I am also not as familiar with the "official" terms for everything... would it make sense to merge or csv any of the existing items listed into a single bullet?
 
@Rachel Data Structures and Algorithms is not mentioned anywhere in either of those Wikipedia articles ("Software Architecture" and "Software Design"). I'm OK with removing it if that's the right thing to do, but not just because Ana says we only get four bullets.
 
"You only get four bullets" is totally the wrong way to go about this.
 
I agree with that
 
5:58 PM
@Rachel not a dupe question, then?
 
I am trying to shrink the on topic page to something simpler, like they requested
 
I think you make a good point though that they should be listed separately
added back in
 
oh no, did I miss a scope discussion?
 
Just algorithms and data structures.
It seems like the odd man out, but there's a very specific reason it is there and, say, "GOF patterns," is not.
Algorithms and Data Structures are the bread and butter of what we do. You can't swing a cat without hitting one or the other.
 
6:10 PM
you can't talk about software at all without hitting one or the other
 
Exactly.
Remove it from the scope spec, and all you have left are bullet points that scream "management."
 
I don't think software design/architecture screams "management"
that's exactly what management is incapable of doing
some managers and designers can code a little, but they can't work effectively on >10k line codebases
 
@ThomasOwens Aww you added back in the huge list of off topic items to the on-topic page text... is there any way we could reduce that? I really think once the name changes we won't need to explicitly state everything listed there
 
Sure, but if I'm not a Software Architect, can I still ask a question here?
Of course I can. Just so long as I don't ask you to fix my broken code for me.
 
@ThomasOwens I think I understand your reasoning, you wanted to include links to the related SE sites. But I was hoping we could accomplish that some other way, such as a single link to a Meta FAQ post about related SE sites
 
6:14 PM
@RobertHarvey if algorithms and data structures are not clearly implied by architecture and design then I also believe that these should be explicitly stated, I would want them to be firmly in site scope
 
we should not make it our job to redirect people to every other SE site that might be useful to someone with a programming job
 
@Rachel Most of them exist because the questions there could be considered Software Engineering, but we don't want them.
 
that job is imposed upon us by the bazillion off-topic questions coming our way, but it's not a job we ever should have had nor should continue to encourage having
 
For example, licensing questions is part of software engineering. We don't want those. Professionalism is part of Software Engineering, too. Those go to The Workplace.
 
Licensing questions are not part of software engineering, any more than marketing questions are.
 
6:15 PM
I don't see how licensing is part of "software <bikeshed word>"
 
@RobertHarvey If you go by SWEBOK, they are.
 
@ThomasOwens for a site titled "Software EngineerING" though, I would not expect people to assume workplace questions are on topic. Maybe if it was titled "Software EngineerERS"
 
@ThomasOwens Go by SDLC instead. Licensing is not there.
 
@RobertHarvey That is true.
 
if you count choosing libraries/dependencies as part of the SDLC then licensing shows up, but we want to avoid any questions about specific libraries/dependencies anyway
 
6:17 PM
I don't care much for official naming buckets, if a licensing question is more about software engineering than legal matters, I'd say it's fine here. We should use our judgement in those cases, not have a black/white line drawn
 
@Rachel Licensing is purely a legal thing.
 
the argument for getting rid of licensing has nothing to do with whether it's part of Software <Bikeshed Word>
 
They would be best served at Open Source or Law. Their answers are infinitely better than ours on licensing questions.
 
@Rachel The percentage of such questions is vanishingly small. It's easier to just categorically disallow them, especially since Open Source answers such questions much better than we do.
 
the argument is that 1) we suck at licensing questions compared to OpenSource.SE and Law.SE, and 2) most of the licensing questions we get are off-topic/low quality anyway by current standards, and 3) they're one of the very few things that if considered on-topic makes it far more difficult to concisely describe our complete scope
and frankly, I find all of that combined to be pretty convincing
most of the licensing questions I've personally answered here amount to copy-pasting stuff from the FSF FAQ, so arguably none of them were particularly valuable to begin with
 
6:20 PM
But seriously, can we reduce the wording of the off-topic list? Assuming we had a meta-faq post containing other related sites, how could alternatively word it?
 
imo the off-topic list should only bother listing things that people ask regularly, which we can't stop them from asking by making some other change
 
I don't like edit wars, so i'm looking at you @ThomasOwens :)
 
but we won't know what those things are post-name change until we actually do the name change
I would prefer to err towards a short list for the following reason:
even if the name change works, 99% of off-topic questions are going to come from people who don't read that list
 
Personally I think code review and open source don't need an explicit mention. Law is good, Workplace is a maybe (personally I would remove it, I think name change is good enough), and last two bullets can be considered covered by line of text below the bullet pointed list
Which would leave us bullet points in the off-topic list of just Stack Overflow, Law, and maybe one about lifestyle/workplace/career advice
 
code review might actually be the one most worth mentioning, since they come the closest to overlapping with us in a non-obvious way
it's fairly obvious that legal advice is not part of software engineering
 
6:24 PM
From what Ana said, they want this to be very simple so someone can tell at-a-glance if a question would be OK here or not. No big long lists or walls of text to make your eyes glaze over
 
@Rachel We're never going to be 4 bullet points. Ever. Give me like 15 minutes and I'll have something to show everyone.
 
I know it won't be 4 bullet points in on-topic, but I'd like to reduce off-topic list to only things that deserve an explicit mention
 
I need to go look at all of the proposed lists again because I really don't get why 4 bullet points is hard to achieve
 
@ThomasOwens but I trust your judgement, and don't mind waiting to see what you have planned :)
@Ixrec It's more about trying to be very clear about our scope, instead of fitting some kind of arbitrary number of bullet points
some items offer better clarity on two or three lines than mashed into a csv of one line
 
I mean from my point of view I'd struggle to even get to 4 bullet points if I was just trying to describe the questions I want on this site
"software design and architecture" kinda covers everything I care about imo
finally skimmed all the answers, of the proposed on-topic bullet point lists, Rachel's wins hands down imo (and I am surprised to be saying that)
 
6:29 PM
I'm going overbroad here. Edits accepted. Link forthcoming still.
 
what exactly is a question that fits under "software requirements" but not design/architecture/etc?
the only thing I can think of are those godawful "I've been asked to make an X, but what exactly should it do?" questions
 
@Ixrec Don't ask me, I initially read "software requirements" as "what's the min RAM, CPU, video card, etc do I need to run X" :)
 
I was implicitly asking Thomas because that's his bullet point =)
 
Although second thought was about the requirements gathering and documentation stage of the SLDC
 
personally I've never heard any advice about requirements gathering that didn't boil down to "be good at talking and listening" and "don't assume you know what the user wants", which doesn't require much clarification
the only reason it gets any more complicated than that is because there's a lot of people who have to talk to each other about a lot of things, which creates a need to plan meetings carefully and document the results of said meetings, but that's no different from any other part of general workplace management
 
6:45 PM
@ThomasOwens so far I have seen rather skeptical assessments of the interstitial efficiency. Though these were coming from SE team folks who generally disliked this feature at smaller sites in the past. Probably works checking with SF moderators are they satisfied with how it works for them. (I recently noticed that unlike Programmers SF doesn't require registration to ask. But still, their pages look so much cleaner than ours, what kind of magic do they use?) — gnat 26 mins ago
@gnat They have the magic of not being named "TechnicalSupport.SE" with a tag line of "Server Fault is a question and answer site for tech support people", which is essentially what our site is right now to outsiders :)
 
0
Q: I have a question, but it may not fit best on Programmers. Where else can I ask it?

Thomas OwensWhat other sites are there that may be able to answer my questions about developing software?

Does that allow us to eliminate the list, perhaps?
 
I like that
 
Yesssss but it is very TLDR right now. Can we have a bullet-point single-line list of all the sites?
Maybe have the question contain the bulletpoint list, and post separate answers for each of the sites and what they're about?
 
surely it should be listing types of question rather than sites
 
Perhaps. Do we want to reorganize it?
 
6:49 PM
like a lot of Meta FAQ posts are
 
@Rachel No. I hate that.
I'd be willing to accept an answer that is a bulleted list, though.
 
and let's be honest, the users creating the problem will never read that
 
At least it exists.
 
I like the descriptions though. Too bad we don't have expanders in posts...
 
the only thing that matters is can we reduce the torrent of zero effort off-topic junk by a few orders of magnitude
 
6:50 PM
We can point to that question (in any format) in the help/on-topic page as a place to go.
 
if that happens, then things like this meta post might actually make a dent in what's left
 
We can talk about grouping the sites somehow later. By topic, perhaps? Or one site per answer? I'm not sure.
 
@ThomasOwens What part do you hate btw? the bullet point list in the question, or having multiple answers for each subject?
 
@Rachel A list in the question. The question should be a question and that's it.
I'd be OK with accepting an answer that just links to other answers, though.
 
@ThomasOwens Ok, so a TLDR at the top of the post with a bullet-point list of sites, and a "read below for more details" kind of thing would be OK?
oh ok, i see what you're saying
 
6:51 PM
There's different options.
But yes/no to the general approach?
 
I like the idea of this meta post though :)
 
user image
3
I should've tried this ages ago
I wonder how much that's affected by google knowing it's me
 
@ThomasOwens Do you care if I edit and add TLDR to the top now, or are you doing it? (don't want to overlap if you're alraedy doing it)
 
I was just adding links.
I'm a little concerned with the length of the post running up against SE limits if everything is in one post.
 
are there not MSE questions that already cover this ground?
 
6:55 PM
@Ixrec There are questions about linking to particular "where do I put X question", but no general listing for other sites of interest to users of a particualr site.
Now, the question is can we link to this instead of a big long list of places to go? Can we whittle down the "is not about" list and then throw a link to this meta-question somewhere?
 
07:00 - 19:0019:00 - 23:00

« first day (2093 days earlier)      last day (2894 days later) »