Originally posted by tarceri
View Post
Announcement
Collapse
No announcement yet.
What Would You Like Next On Phoronix In 2014?
Collapse
X
-
Michael Larabel
https://www.michaellarabel.com/
-
Originally posted by Michael View PostI thought I responded? Otherwise email me at michael michaellarabel.com.
Comment
-
Most filesystem benchmark articles seem to generate a lot of debate about how using default mount options, allowing time for caches to populate etc affects the measured performance.
It would be good to collect feedback from the developers and people with experience tuning filesystems and perform some comparative tests to quantify the effects of different tuning options.
Comment
-
Originally posted by Oxmyx View PostBenchmarks covering sysctl, proc/sys, module parameter & similar settings under different loads. I'm thinking mostly disk and networking for enterprise environments. There are many articles that tweak specific sysctl buffers, etc, but rarely any data to show what happens.
Thinking about tweaks, another interesting thing would be to benchmark file systems created with different mkfs options (block size etc.) to see what actual difference they make.
Originally posted by wizard69 View PostI have to also agree that coverage of how the ""BIOS"" issue is solved on ARM hardware is vital. However I'm not convinced that CoreBoot is the final answer here. We certainly need to see how things unfold and see what the industry adopts. So covering how this all rolls out in the real world is very important.
A question though, what is so bad with EFI? Calling it a train wreck seems to be going overboard.
Meanwhile UEFI answered the problem by adding additional complexity: UEFI is an OS in and of itself. That means it has fancy features (mouse, touchscreen etc. support, nice graphics and so on), but it's bloated (so much that there are issues fitting the thing on regular flash memory), slow (although UEFI also says screw the safety checks for faster boot), and worst of all, the UEFI specification is pretty much book-length. Couple that with the traditional lack of proper firmware testing, and you have a recipe for disaster. The bug surface is huge and nobody ever tests most of the features ("Windows boots? It's production-ready!"). For descriptions of some more interesting firmware-related issues, check out this thread: http://phoronix.com/forums/showthrea...-issue-stories
Comment
-
I think I speak for many when I say we want less unnecessary linking back to your own articles. It's warranted in many cases but you really overdo it sometimes. My worst experience was trying to find a link to a kickstarter page in a follow up article(I think it was the failed open source GPU). I lost count of how many links I followed to phoronix articles trying to find the link to the page. It's really unprofessional, so please work on that in 2014.
Comment
-
I would actually just like to see more of the same. Let me take this opportunity to extend my gratitude, phoronix is the best site out there for information hungry open source freaks like me. My hope is that you can keep up the great work and give us many more years of excellent coverage. Just make sure to take the time off needed to keep up motivation. I don't share any of the criticism given in this thread. Yes, of course you could be me more neutral, but then again many others around the net are that already. The more unpolished emotional approach is part of your style, I like it. Of course there are topics where I flat out disagree with your angle, but still providing that angle is your prerogative and I hope you continue to do it. The forums could be better policed, but then again there are other forums that are well policed. Having almost anarchy does give interesting dynamics, and brings out some threads that are actually interesting.
To sum it up, just keep up the good work, we love it! aaaand, please with lots of sugar and cream on top, don't let the overly critical elements in the community get to you. Some people are unable to appreciate gold even if it is straight in front of their nose.
Comment
-
Originally posted by wizard69 View PostAdblock users will not go away! In fact I find the pop up ads especially frustrating on your site, it actually reduces the amount of time I spend on your site. This mainly due to most of my access being via an iPad which doesn't support ad blocking. In any event you are actually encouraging the use of ad blockers when every page load results in the same damn ad being popped up again and again.
As for article length, sometimes they are too short sometimes not. We can debate length all we want but I think the bigger issue is too many articles at times. That is should some article have even been written? I would argue that if you can't take the time to flesh out an article maybe it shouldn't be on the site in the first place.
---
For the site itself, I've got some stuff:
Does extra video memory make any difference with either the open- or closed-source drivers, and at what levels? I suspect the open-source drivers might be helped by the extra RAM, because it would mean less thrashing of textures across PCI-E/AGP.
In the absence of ntp, how much time does a modern Linux machine lose in 24 hours? This might be a tiebreaker in some benchmarks where things are really close: if an RTC is slow enough, it will inflate the benchmark score.
Benchmarking Wine versions of apps against native Linux.
Comment
Comment