Vote On A Distribution For Linux Benchmarking

[LinuxGamer]

Are you really stupid or are you just pretending to be stupid?

Me and the English language don't get a long and why do you not have any Bases when calling some one stupid

[jayrulez]

Me and the English language don't get a long and why do you not have any Bases when calling some one stupid

You claim that the Wayland API is stable and complete and awesome and anything else you want to use to describe it, then you say Mir is just a copy&paste of Wayland code then say Mir API is a mess and whatever else.

If Mir is just a copy & paste of Wayland, aren't you essentially talking about Wayland when you talk about Mir then?

<s>
Look at this beautiful piece of code in Mir:

int mir::setMaster(mir::MirLauncher *ml, mir::MsgHandler *msg, ssize_t len)
{
	int ret = -1;
	mir::MsgHandler *cmsg;
	mir::MirLauncher::SetMaster *message;
	union cmsg_data *data;

	if (len != sizeof(*message)) {
		error(0, 0, "missing value in mir::setMaster request");
		goto out;
	}

	message = msg->msg_iov->iov_base;

	cmsg = CMSG_FIRSTHDR(msg);
	if (!cmsg ||
	    cmsg->cmsg_level != SOL_SOCKET ||
	    cmsg->cmsg_type != SCM_RIGHTS) {
		error(0, 0, "invalid control message");
		goto out;
	}

	data = (union cmsg_data *) CMSG_DATA(cmsg);
	if (data->fd == -1) {
		error(0, 0, "missing drm fd in socket request");
		goto out;
	}

	if (message->set_master)
		ret = drmSetMaster(data->fd);
	else
		ret = drmDropMaster(data->fd);

	close(data->fd);
out:
	do {
		len = send(mir->sock[0], &ret, sizeof ret, 0);
	} while (len < 0 && errno == EINTR);
	if (len < 0)
		return -1;

	return 0;
}

How would that look in Wayland and Weston?
</s>

[9aus4000]

The only reason the results are varied is because it became an outside population contest. Some distros are advertising to come vote on phoronix. aka people who will never even visit the site to see the news.
I'd be willing to wager at least half the votes were accounts created within the last week and 0 posts.

Funny you say that...

I guess I just never saw a reason to create an account when I was just perfectly content with reading the news, but I just had to vote :P

I voted ArchLinux because it's the bleeding edge of Linux software, and most tests benefit from bleeding edge software.

BTW lotsa hate going around on this fourm, it's ridiculous!

[DanL]

In this thread, LinuxGamer also called Ubuntu's repo buggy and then told someone who suggested using Xubuntu that they were making sense. The rate of ignorant and/or contradicting statements coming out of that account is incredible.... and that's just in one thread!

[jrch2k8]

You claim that the Wayland API is stable and complete and awesome and anything else you want to use to describe it, then you say Mir is just a copy&paste of Wayland code then say Mir API is a mess and whatever else.

If Mir is just a copy & paste of Wayland, aren't you essentially talking about Wayland when you talk about Mir then?

<s>
Look at this beautiful piece of code in Mir:

int mir::setMaster(mir::MirLauncher *ml, mir::MsgHandler *msg, ssize_t len)
{
	int ret = -1;
	mir::MsgHandler *cmsg;
	mir::MirLauncher::SetMaster *message;
	union cmsg_data *data;

	if (len != sizeof(*message)) {
		error(0, 0, "missing value in mir::setMaster request");
		goto out;
	}

	message = msg->msg_iov->iov_base;

	cmsg = CMSG_FIRSTHDR(msg);
	if (!cmsg ||
	    cmsg->cmsg_level != SOL_SOCKET ||
	    cmsg->cmsg_type != SCM_RIGHTS) {
		error(0, 0, "invalid control message");
		goto out;
	}

	data = (union cmsg_data *) CMSG_DATA(cmsg);
	if (data->fd == -1) {
		error(0, 0, "missing drm fd in socket request");
		goto out;
	}

	if (message->set_master)
		ret = drmSetMaster(data->fd);
	else
		ret = drmDropMaster(data->fd);

	close(data->fd);
out:
	do {
		len = send(mir->sock[0], &ret, sizeof ret, 0);
	} while (len < 0 && errno == EINTR);
	if (len < 0)
		return -1;

	return 0;
}

How would that look in Wayland and Weston?
</s>

i read goto in c++? is for real a goto in c++? c++ RAII got nuked from gcc/clang? switch seems too mainstream? the label out: code is horrible!!!, the unions are fishy, you tracked the definition? you notice loops are really fat for no reason?, i don't say it won't compile i don't say it won't work but call it beautiful? just proves without a doubt you are not a developer or you used to be back when Micheal Jackson was still black.

[jayrulez]

i read goto in c++? is for real a goto in c++? c++ RAII got nuked from gcc/clang? switch seems too mainstream? the label out: code is horrible!!!, the unions are fishy, you tracked the definition? you notice loops are really fat for no reason?, i don't say it won't compile i don't say it won't work but call it beautiful? just proves without a doubt you are not a developer or you used to be back when Micheal Jackson was still black.

Did you miss the sarcasm tag?

That's actually from Wayland or rather Weston... But I knew some idiot would claim it is ugly just because I said it is from Mir. It is the handle_setmaster function in this file: http://cgit.freedesktop.org/wayland/...eston-launch.c

I just changed some names around to remove any hint that it is from Weston.

[Vim_User]

You didn't get my point about Fedora. In this case some Ubuntu haters created fake accounts to vote against Ubuntu. "LinuxGamer" must have more than 15 accounts!

Show us your evidence. What makes you think that some of the Ubu-Trolls did not have done what you accuse LinuxGamer to have done?

[Vim_User]

bla bla bla, there is absolutely no way the known statistics (distros share) legitimately change in a poll. You can like whatever you want but there is the "one person-one vote" rule and the Ubuntu "people" are more than Fedora "people" and the Fedora "people" are more than the Manjaro "people".

So the poll is not legitimate because statistics dont lie but they can be cheated.

Look at the poll now. does it reflect the real statistical numbers?
How do you explain the Fedora staying behind after a couple of days in front of Ubunu?

Think...

Maybe you should think. Why the hell should the Phoronix readers, most of them interested in technical details (some of them only here for trolling, as it seems), unlike the Linux users in general, have the same distro usage statistics as the Linux users in general?

[leonmaxx]

Who's bot voted for Arch?

[GreatEmerald]

Rather than primarily focusing just on one distribution for the majority of benchmarks, why not pick the most recent releases of the five (or, if that's too time-consuming, narrow it down to two or three) distinctly separate desktop-oriented distributions (or fully compatible derivatives thereof) with the greatest targeted vendor support? That would leave out a lot of great distributions, but seems more practical. It would provide some degree of variation, while omitting more niche distributions.

Because that's too time-consuming for Michael.

Who's bot voted for Arch?

If you missed it, it's Manjaro people advertising the poll everywhere.