On 25/12/2018 6:01 PM, Joakim wrote:
> See my responses to Nicholas above, I don't think the Android port merits a talk. By the same standards I apply to others' talks above, I don't think my work merits a talk either. ;)

A talk covering ARM and Android development in general would be very well received in the context of D. If you want to be convinced we could do a poll on who would want to see it (but I expect quite a large number of people would be in support of).

On Tuesday, 25 December 2018 at 07:10:46 UTC, rikki cattermole wrote:
> On 25/12/2018 6:01 PM, Joakim wrote:
>> See my responses to Nicholas above, I don't think the Android port merits a talk. By the same standards I apply to others' talks above, I don't think my work merits a talk either. ;)
>
> A talk covering ARM and Android development in general would be very well received in the context of D. If you want to be convinced we could do a poll on who would want to see it (but I expect quite a large number of people would be in support of).

I don't see how it could be worthwhile: nobody has ever given such a DConf talk about a port to a specific platform because it doesn't really make sense. The whole point of a port is to abstract away the platform, so you can simply recompile most of your D source for it, as H. S. Teoh has indicated he's been able to do with the Android app he's been developing in D recently.

The way to do that talk is to abstract multiple ports into a general porting guide, which is the talk Kai already gave, or maybe talk about the details of a port to a very obscure or different platform, as Igor did this year:

https://dconf.org/2018/talks/cesi.html

While it was fascinating to hear how much work he put into it, much more than me, my interest was squelched somewhat because he couldn't reveal the platform and it's likely I would never use it anyway (not a game programmer). I mean, who really develops for non-Windows, non-Posix OS platforms? I haven't since college. For those few who do, maybe the talk was great. But the Android port wasn't that obscure: it's basically a linux/ARM distro with a different libc, Bionic.

If you really mean "ARM and Android development in general" and not the details of the port, I can't claim much knowledge of that, as I don't have a large Android codebase that I've developed and deployed. Hopefully, even if I did, there would be nothing to say: as it should be pretty similar to writing D code for a desktop platform.

My phone- on whose 5.5" screen I'm viewing the text of this forum response as I type it out on a separate, full-sized bluetooth keyboard paired with it- has 6 GBs of RAM and 128 GBs of storage (of which I have 8 GB free right now). That's about what midrange desktops and laptops come with these days (though with much larger screens ;) ), so you can't say mobile presents much of a constraint in terms of hardware. I've pointed out before that I compile code on my phone about as fast as a Macbook Air from a couple years ago:

https://forum.dlang.org/thread/sqbtgmbtrorgthsplvho@forum.dlang.org

If you see some other angle on an Android talk that I'm missing, I'd be happy to hear it, but I don't see it. Maybe someday when I have a huge, successful Android app in D, I'll write about or put up a talk online about the architecture I used, but hopefully there won't be much specific to Android there. :)

On Tuesday, 25 December 2018 at 05:01:43 UTC, Joakim wrote:
> On Monday, 24 December 2018 at 22:22:08 UTC, Steven Schveighoffer wrote:
> The 0.1% of the community that attend seem to like it, the vast majority don't, or at least don't care.

You think we have 200k users? More to the point you neglect the benefit of development and progress is shared by all users.

>>> I, for one, will not be donating to the foundation as long as they continue to waste money this way, just as others have said they won't donate as long as it doesn't put out a Vision document anymore or otherwise communicate what it's doing with their money.

I agree this does need to happen, the foundation will be having a another meeting in Feb to set the vision, which I hope will be a little more planned and productive than the last one.

>> Nobody is asking for your money for this conference (unless you want to attend), and if you feel this way, that's totally your choice.
>
> I'm not talking about the registration fee, I'm talking about contributing anything to the foundation, which Walter indicates above covers some of the expenses for DConf.

Some additional transparency would help, Mike?

>>I like the results that come from the conferences, I've
>> been to all of them since 2013, on my dime for 3, and with assistance for 3. I felt it was 100% worth it for all.
>
> Yet you cannot give a single reason _why_ you felt it was worth it, or why my suggestions wouldn't make it better.

I'll give my reasons:
I got a job out of it.
I got useful insight into various bits of the compiler.
I got connections for collaboration with stuff that I'm interested.

> If you're making a bad decision, it _should_ be questioned.

Indeed, but none of us think DConf is a bad idea or that the format doesn't work for us.

> Almost nothing that has been decided so far would stop most of my three suggestions from still being implemented.

You haven't managed to convince us that that would be an improvement.

> As for how they feel about it, I don't care. The reason most projects and companies fail is because the decision-making process stops being about putting out a good product but about "feelings" and various people "saving face," especially when higher up the hierarchy, ie politics. And don't make up some nonsense that I'm saying that it's okay if everybody starts cursing each other out like Linus did: we're talking about _questioning a decision_. That is the whole point of having a community.
>
> The day this community starts being more about saving face is the day I leave it, as that's the beginning of the end, and I don't want to be around for that end.

I totally agree, but again, you haven't convinced us that it is an improvement.

> Not at all, the whole reason I'm willing to debate is that other worthwhile perspectives may be out there. I think the evidence and arguments strongly favor the suggestions I'm putting forward, but I'm perfectly willing to consider other arguments.
>
> That is the same stance they should have, but don't appear to. My problem with this "debate" is that nobody was able to defend the current DConf format at all.

That reasoning is backwards: in our experience DConf, as done in the past, works, and it works well. The onus is on you to convince us that it would work better the way you describe.

> Consider some of Walter's silly arguments above: at one point he says he wants "successful instantiations of your theories," implying that these are all things I'm just talking about and nobody's doing them, though it's not clear which aspects he thinks that of since I've presented evidence for much of it.
>
> But at another point, he says that other D meetups are already doing something I suggest (I pointed out that he's wrong about that one, but let's assume he believes it), so there's no reason for DConf to do it. First of all, 95+% of D meetups appear to follow the DConf format of having a single speaker lecture to a room, so why isn't that an argument against doing that yet again at DConf?

What works at one scale doesn't necessarily work at another. To do something very different from a "traditional" conference would be a significant risk when what we have works well. As noted previously your opinions would carry more weight if you had actually attended a past DConf.

> Secondly and more importantly, he's speaking out of both sides of his mouth: do you want to do something that nobody else's doing or that somebody has done? You can't argue _both_ that you don't want to do what others are doing and what nobody else is doing. And why wouldn't the former apply much more to the outdated DConf format?

I don't knowhow many times we have to say it: we do not feel the conference format is outdated.

> It's not just because of this, this is merely the final straw. I have felt that the talks were mostly not worth my time at the last couple Dconfs, that is the main reason.
>
> I see a lot of bait-and-switch going on, where the talks advertise something interesting, then talk about something else. There doesn't appear to be any attempt at quality control for the content of the DConf talks, once the presenters have been accepted. This is a problem for almost every conference, but it only aggravates the huge waste of time that is in-person talks.

It is a pity you think that, I found sone of the talks very interesting. Yes quality of the speaker and intrigue of the topic varies but such is life.

>>You have
>> contributed a lot in terms of the android port, although I haven't really programmed in android (I have a tiny bit, with Xamarin (hated it) and a bit with Java (was OK, but crazy complicated) ). I hope at some point you reconsider, I'd love to see a presentation on it.
>
> See my responses to Nicholas above, I don't think the Android port merits a talk. By the same standards I apply to others' talks above, I don't think my work merits a talk either. ;)

More's the pity.

On Tuesday, 25 December 2018 at 18:54:25 UTC, Joakim wrote:
> Simply repeating over and over again that you're not "convinced" is not an argument, nor do your own personal reasons above argue for one format over another.

I don't mean to stoke the flames on this anymore, but I do: I've been to past conferences and now that I'm not a poor student anymore I'd pay to go to them. I happen to still have plenty of free time so watching pre-recorded talks, while not a problem for me a) does not apply to everyone and b) loses interactivity.

> I asked for a rationale above and got none from Mike and a very weak, confused one from Walter. It's fairly obvious that there was never any real deliberation on the DConf format, and that you guys have dug in and decided to cut off your nose to spite your face.

I can't speak for other but the rationale or deliberation of not doing what you have suggested is the it has worked well in the past, and I see no reason for that to cease to be the case.

> Fine with me, your loss.

On the contrary, I think it will (continue to be) a massive success.

> I see, so you're arguing that DConf shouldn't be doing in-person talks because it's larger than most D meetups? Don't answer that, scale as a reason makes no sense and there's no way you can make it.

You have that backwards again, As Iain said, watching one pre-recorded talk takes an hour, which I'm sure many enthusiasts will be able and wiling to spend, scale that up to 20 and they may not be able to even if they are willing.

>> To do something very different from a "traditional" conference would be a significant risk when what we have works well.
>
> I see no "risk" whatsoever in change when the status quo is dying, and what you're already doing isn't having much impact.

and that comes down to an apparently fundament disagreement that the status quo is not good enough, In our experience the current format work very well. I hope you agree that a major change to the way things are done will likely have a major impact on the quality of the conference: if you think it is bad enough that changing it can only improve it then you would, rightly, come to the conclusion that the expected value (in the statistical sense) of changing the format is positive. However, if you think that the change may cause the quality to degrade then the expected value of the change of format drops significantly...

>> As noted previously your opinions would carry more weight if you had actually attended a past DConf.
>
> Heh, this is the dumbest possible argument anyone can put forth and you guys repeatedly make it: "I have no arguments so 'Magic! You had to be there!'"

...and the change becomes much more risky. The argument of "your opinions would carry more weight if  you had actually attended a past DConf" stems from the fact that we have experienced the conference as a positive well worthwhile to attend event and we see the potential for this to get worse in the face of change.

On Wednesday, 26 December 2018 at 00:06:39 UTC, Nicholas Wilson wrote:
> I don't mean to stoke the flames on this anymore, but I do: I've been to past conferences and now that I'm not a poor student anymore I'd pay to go to them.

I've been to past conferences and don't consider it worth going even if I was paid to go. The only way I'd even consider it again is if either 1) it is in a location where I have other business or 2) the format has some change, even if it is a small conservative change like what I described in the previous email, in at least *some* of the talks.

The talks are either irrelevant, trivial, or vaporware (seriously, how much of the stuff described have never come to pass?), and there's really no benefit in spending several days of my life on that. Especially when I can watch it on youtube in half the time (thanks 2x speed playback)... or less (thanks skimming around to find the interesting nugget in the sea of boredom).

Wanna know my favorite part of the ones I attended? Walking around UVU with Prof. Allison.... the one thing I did that was NOT part of the official schedule.

That's why I agree with Joakim - we should tweak the format to maximize the stuff that BASICALLY EVERYONE AGREES are better - the actual in-person interaction.

On Wednesday, 26 December 2018 at 00:51:51 UTC, Adam D. Ruppe wrote:
>
> The talks are either irrelevant, trivial, or vaporware (seriously, how much of the stuff described have never come to pass?), and there's really no benefit in spending several days of my life on that. Especially when I can watch it on youtube in half the time (thanks 2x speed playback)... or less (thanks skimming around to find the interesting nugget in the sea of boredom).
>

I think this is a bit uncharitable.

Last year I had an amazing time at Dconf.
Went to bike there for 9 days (800km), to arrive the day before DConf.
Munich was a beautiful city but Switzerland was very graphic.
Friends jokingly said it was a D pilgrimage and it was, kind of :)

The talks were honestly all interesting, probably being there puts you in the mood to really get into them. It's a shame not all of them were recorded. I remember those from Johnathan, Andrei and those talks on "open methods" more distinctly. Putting a face on people you've known from the internet is really surprising.

My only regret was not sleeping at the hotel since you don't get as many occasions to meet people in a beer settings.

I was more than happy to pay the full DConf price and expenses for a trip I will remember all my life.

Awesome! I look forward to DConf 2019. I'm planning my trip already.

Thanks you all who's working behind the scenes to make this happen. It's amazing.

On Wednesday, 26 December 2018 at 11:26:52 UTC, Guillaume Piolat wrote:
> Went to bike there for 9 days (800km), to arrive the day before DConf.
> Munich was a beautiful city but Switzerland was very graphic.
> Friends jokingly said it was a D pilgrimage and it was, kind of :)

You could still have done that if the conference format was altered to increase in-person interaction!

> Putting a face on people you've known from the internet is really surprising.

You would get MORE of that if the conference format was altered to increase in-person interaction.

Keep in mind that what I want to do is to tweak dconf, not to kill it. Change from 50 minute talk to 30 minute talk + 20 minute interaction, in at least some cases.

Or it might be fun to reconfigure a day: have all the day's speakers do 15 minute initial talks and pass out "learn more" links. Then we have a *two hour* lunch/mingling break which gives everyone a chance to digest the morning's information, look into the learn more stuff, form ad-hoc study groups, etc.

Then, after lunch, the speakers return for follow up stuff. Talk part 2, public Q&A, whatever, using the remaining 25 mins each of their time.


I'm open to a lot of ideas... I just want to spend more of the in-hours time doing in-person interaction.


> My only regret was not sleeping at the hotel since you don't get as many occasions to meet people in a beer settings.

And again, a huge point I have been trying to make is we could get MORE of that if we tweaked the format!

Most everyone cites their favorite and most productive part of in-person meetings are actually the after hours stuff. (I hear this both from dconf and my day job - this is part of why it hits me so much, but the day job has been doing some tweaks this last year, to great success - I *know* these changes are for the better.)

So why not take some of the after-hours wins into the main event?

Going backwards to this:

> The talks were honestly all interesting, probably being there puts you in the mood to really get into them.

Well, wouldn't it be fun to be able to talk about them or work with the ideas more in person?!

With my compromise proposals, you'd still get much of the same talk... just use a fraction of the remaining time to interact with everyone and their code directly.

On Wednesday, 26 December 2018 at 11:26:52 UTC, Guillaume Piolat wrote:
> Last year I had an amazing time at Dconf.
> Went to bike there for 9 days (800km), to arrive the day before DConf.

👍👍 woah dude !

On Saturday, 22 December 2018 at 12:18:25 UTC, Mike Parker wrote:
> Thanks to Symmetry Investments, DConf is heading to London!

That's a funny typo you have for BeerConf there.

BEERCONF