SoylentNews is people
SoylentNews
from the we've-been-here-for-you;-will-you-be-here-for-us? dept.
[20200629_140251 UTC: Update 1: Encourage taking care of personal/local needs, first.]
[20200629_191024 UTC: Update 2: Added stretch goal of $1000.00 ]
[20200630_023201 UTC: Update 3: Increased stretch goal from $1000.00 to $2000.00]
[20200630_023201 UTC] What is possibly one of the worst things to hear from an editor? "I'm at a loss for words." Well, it's happened. The SoylentNews community has done it, again! We started today needing $800 to cover projected operating expenses of $3500 for the first half of the year. And you did it! So, I added a stretch goal of an additional $1000. Now you have gone and reached that goal, too! We'd run at a significant loss ($6000 so far), so that is very much appreciated! THANK-YOU!!!. Stretch goal has now been increased to $2000 [so we can continue to track your subscriptions in the Site News block]. Dare I hope? --martyb
[20200629_191024 UTC] The SoylentNews community is AMAZING! In these especially difficult circumstances, we've reached our original goal for ongoing expenses... and then some!
Thank You!!!!
We started today (Monday June 29) needing over $700.00 to cover projected operating expenses for the first half of the year.
We not only reached our original goal of $3500.00, but I added a stretch goal of $1000.00 and we are already 66% of the way to reaching *that*!
Why a stretch goal? Because we have been running at a deficit for a few years. We are are still about $6,000.00 short of having sufficient funds to pay back our benefactor's original $10,000.00 outlay. Any additional funds raised will go towards that purpose whilst giving us a larger safety cushion. --martyb
The original story (after performing Update #1) appears below:
SoylentNews could use your help.
tl;dr The first half of our fiscal year runs Wed. January 1 through Tue. June 30, inclusive. We are at 80% of the funds needed to cover our expenses for the period. If money is tight for you, take care of yourself first. But, if you can help, it would mean a lot to help us to continue to be here for you.
Please subscribe. The subscription amount provided (e.g. $20.00 for 1 year) is the minimum amount for that period; you can change that default to any larger value.
To all who started a new subscription or renewed an existing subscription: Thank You!
Times are tough. First, please take care of yourself and those close to you. But, if you do have funds to spare, we would very much appreciate your support!
Where We Stand:
So far, we have had 106 subscriptions this year which have netted us an estimated $2,794.92 (after processing fees from Stripe/Paypal) towards our goal of $3,500.00.
We run a very lean operation; $20/day keeps everything going. Staffing is all-volunteer; nobody has ever been paid anything for their work on SoylentNews. That includes the editors who get the stories out on the main page. The sysadmins who keep everything running: the servers and all the services like the MySQL databases, Apache HTTP Server, IRC (Internet Relay Chat), email... it's a long list. That we so rarely have issues is a testament to how fortunate we are to have professionals who donate their free time to keep things running. We had to incorporate to be able to accept subscriptions to pay expenses. And with that there are fees for maintaining the incorporation, calculating taxes, and paying them.
Subscriptions Breakdown:
Number of subscriptions for each subscription amount, and the totals at that level, so far in 2020:
Qty Sub Amt Total 7 $4.00 $28.00 12 $5.00 $60.00 2 $12.00 $24.00 63 $20.00 $1260.00 2 $25.00 $50.00 3 $30.00 $90.00 1 $36.60 $36.60 1 $39.39 $39.39 3 $40.00 $120.00 4 $50.00 $200.00 1 $60.00 $60.00 2 $100.00 $200.00 1 $113.00 $113.00 2 $120.00 $240.00
The Pandemic Sucks:
The world has changed in the past six months.
A lot.
The pandemic hit and with it came lock-downs, work-from-home, and social distancing. Closures of movie theaters, restaurants, and bars. Video conferencing became a norm as in-person gatherings were prohibited. And for good reason: worldwide, over 10 million are known to have been infected and over a half million have died. Untold struggles and suffering as we attempt to understand and adapt to a new reality.
We recognized that many of the community were struggling. On April 19th, we extended all subscriptions that were due to expire in April or May to the end of May. If money was tight and it was a choice of renewing your subscription or paying your bills, we'd rather you spend your money locally and so thereby help keep the money in your local community.
Folding@Home (F@H):
SoylentNews is helping in the fight against SARS-CoV-2. You might not be aware, but SoylentNews has a Folding@Home team. We are currently ranked in the top 300 teama in the world (#297 out of 254150 teams)!
F@H is a distributed computing project designed to help understand how proteins fold and thereby search for cures to various diseases. It was originally focused on Parkinson's and Huntington's diseases as well as cancer. With the appearance of the SARS-CoV-2 virus, F@H has pivoted to trying to unravel the behavior of that virus. To this end, many large infrastructure companies (like AWS, Microsoft, Oracle, and Google) have joined the effort.
How it works: Install the client on your computer, instruct it what amount of resources to use, and you are ready to go. The client will periodically download work units and, when completed, upload the results to the F@H servers. The faster the results are computed, the more points are earned. We are team #230319. If you have computes to spare, we'd love to have you join us!
Stories and Discussions:
Through all this, we here at SoylentNews have persevered. People from all over joined us in discussions on the pandemic and so many other topics. We aim for news with a technological focus but will occasionally offer something a little offbeat.
So far in 2020, SoylentNews has posted over 2,100 stories. Separately, the community has posted 700 journal entries. To these 2,800 items, the community has posted 76,000 comments — over 400 comments per day! In addition, there have been over 55,000 comment moderations — that's nearly 300 per day.
Server Upgrades:
We are continuing our efforts to move services from beryllium (our only Centos server) to aluminum (Gentoo). Deucalion (on IRC; aka Juggs on SoylentNews) has been trudging along trying to get things brought over for IRC (Internet Relate Chat). He reports he had a 100-hour long week at work last week, but still managed to make some progress on this over the weekend. There are significant differences between the two, so it has been quite the challenge. Getting userids added to the correct groups; setting up ACLs; chron syntax incompatibilities; the list goes on and on.
(Score: 4, Touché) by RS3 on Monday June 29 2020, @04:01PM (18 children)
I've offered to host. Not enough?
(Score: 2) by c0lo on Tuesday June 30 2020, @02:49AM (17 children)
[Citation needed] - the initial "hosting somewhere else" doesn't seem a serious offer to host.
That being said, I can't totally fault the AC, in spite of being almost rude.
To elaborate: here's a post coming for a person that couldn't find $20 for a subscription, questioning the due diligence of the current S/N admin team.
May be a fault of in the phrasing, but you'll have to admit that this interpretation is possible.
$20 is a lot easier to pay than the effort and resources incurred by hosting.
This makes a hosting offer sound like one too good to be true. I'm not saying that's impossible or insincere, I'm saying that the context doesn't make it probable.
https://www.youtube.com/watch?v=aoFiw2jMy-0 https://soylentnews.org/~MichaelDavidCrawford
(Score: 2) by The Mighty Buzzard on Tuesday June 30 2020, @03:40AM (10 children)
We've had several offers of hosting. Currently we're (well I'm at least) pretty damned happy with Linode though. They really do offer a good service, above and beyond just space, cycles, and bandwidth. We've been considering taking someone up on their offer for offsite backups though since we parted ways with RamNode. Only problem there is we'd have to work up an encrypted backup script first, cause it would just be unnecessary temptation and all around bad practice to stick unencrypted DB backups on a user who isn't staff's servers.
My rights don't end where your fear begins.
(Score: 2) by RS3 on Tuesday June 30 2020, @04:07AM (9 children)
Thank you for responding to c0lo. He attacks me a fair amount, often AC. I couldn't come up with anything nice to write, so... He interprets too much- he thinks he has sixth sense I guess.
Anyway, of course you'd want to encrypt DB backups going anywhere outside of the direct admins' controls.
I guess you'd just pipe mysqldump output through GnuPG or bcrypt or ccrypt or what're your thoughts?
I don't encrypt my backups currently. Just cron scripts that do some piping and copying and stuff currently. It's not enough of a load to care to do anything any fancier.
(Score: 2) by DECbot on Tuesday June 30 2020, @05:00AM (6 children)
You could try putting the database on a natively encrypted ZFS filesystem. Then cron job snapshots that get ZFS sent somewhere. That way integrity of the backups can continue to be checked without decrypting on the backup server.
Though you might want to run that past a professional system admin, I'm just an enthusiast sysadmin for some personal servers.
cats~$ sudo chown -R us /home/base
(Score: 2) by RS3 on Tuesday June 30 2020, @12:43PM (5 children)
That's an awesome idea. I have to think it through, but maybe you're saying to grab the DB in its ZFS encrypted state, and if so, I'm not sure how you get that ZFS image but I'm sure it can be done and I've got many other pressing things stealing most of my neuron cycles. :0
I think the concept was (is?) for sending DB to offsite backups, and they (obviously) don't want to send it out unencrypted. It's just a big ugly SQL file. You can break it up too- table by table, whatever. (mysqldump is the command).
But yeah, ZFS might be a great choice, and that'd be 100% up to SN admins.
I used to be an enthusiast! Them's were fun days. I used to compile kernels almost every day. I'm not a full-time admin. Not to brag, but it's not that much fun anymore, so I do everything I can to design/build/admin for very long-term uptime. Rock-solid dependable no baby-sitting most of the time. Have had a few hardware failures- the occasional disk die and RAID rebuild. One MB literally made a small fire- 1 U rack Dell that blackened the lid. Amazingly the disks were fine so I just popped them into another spare machine and kept rolling. It was a local-only file server. Servers are under constant attack and I've never had a break in. I do employ IP blocking software that works well. There are those here (TMB cough cough) who will scoff but several sites are WordPress and the most problem I've seen is occasional (maybe 2 a year) spam comments, but 1) they don't show until an admin okays them, and 2) I really don't know how it happens- maybe some "back door" in a plugin? The site designers / admins choose their plugins, and there's a guy who does some site design / admin, but I do all the backend, updating, theme and plugin qualification, etc.
(Score: 3, Informative) by DECbot on Tuesday June 30 2020, @04:33PM (4 children)
If I were to attempt this, I'd want to ensure the database was in a rest state during the snapshot. I'm not really a DB admin, so I don't know the best way; if possible pausing transactions might work but stopping the database would definitely work. The ZFS snapshot is instant using the "zfs snapshot" command and if the zpool is natively encrypted, the snapshot will be too. To send it to a remote machine, the "zfs send" command on the sending system to send the snapshot and the "zfs receive" on the receiving system are the builtin methods to do this. That way, you don't even have to dump the database, you just send the snapshot of the filesystem underneath the database. For a recovery, you would reverse the zfs send and receive commands and import/decrypt/mount the snapshot. What's cool about this solution is the backup server can still perform zfs scrubs on the encrypted dataset to ensure block integrity and repair blocks without having to decrypt the data.
Granted running ZFS on a system is not free. ZFS likes to use the available ram to improve read performance and ECC ram is generally recommended to ensure block checksums are not corrupted in ram. There is cpu overhead to encrypt and compress the filesystem too.
I have a samba/minecraft server for pictures/music/movies on ZFS Z2 pool and a postfix/dovecot/apache/roundcube/tt-rss server on an old Jetway MB featuring a 32-bit, low power, single thread Via CPU that I'm eventually going to move to a RockPro64 with a ZFS mirror. I have plans to setup something for home assistant and Octopi on some raspberry pis some time in the future. However, that's for a much later date as I've moved the family into a new place and there's painting, yardwork, and kitchen remodeling that will take pretty much all of my free time.
My adventures in sysadmin has seen a few hd failures on old mdadmin RAID arrays, but I've mostly had to work through power supply failures likely caused by my UPSs. They have killed every power supply with PFC I left plugged in to them for any length of time. That makes senses as their approximation of a sine wave makes the PFC go mad trying to compensate for the stair stepped input voltage.
cats~$ sudo chown -R us /home/base
(Score: 2) by RS3 on Wednesday July 01 2020, @01:26AM (3 children)
All very interesting! I wasn't aware power supplies were doing PFC. You'd think the design would have factored in the harmonics because many UPSes create them. Several mental link process led me to think of a gyrator. Any idea what's failing in the PSes?
I don't remember if I ever had a Jetway MB, but I had a Jetway graphics card that was pretty awesome for the price at the time. Early dual-monitor one. I have a couple of Via CPUs I always wanted to use for just a simple NAS. When I first got them I thought they'd work in my existing MBs, but no, had to buy special ones, which I did, but for some reason lost interest. I've noticed most P3 / socket 370 stuff runs very cool. Not so much the Tualatin (sp?) but they're still way less watts than P4. You've got great ideas! You and TMB with the remodeling. :) I do some of that but mostly for $.
One of my "things" is easy recovery. I haven't tried ZFS yet, but anything needs to be easy to boot an optical or USB drive and easily get at filesystems. That said, I'm not a huge fan of LVM because why. But I learnt and adapted. Still not a fan- I don't get it. Okay, maybe for people who are constantly moving things around... I got other things to do, like ramble on online boards. :0
No question DB backup and recoveries are tricky. You really can't guarantee 100% because for example, if right now while I'm typing the DB is shut down, backed up, I hit "submit", my post might be lost (much to the joy of some people here! :) There are a bunch of buzzwords describing all that stuff in DB handling and I forget them. One is "atomic". I just worked on a 70-year old tractor. That's my diversion from too many bytes. :)
(Score: 2) by DECbot on Wednesday July 01 2020, @11:56AM (2 children)
I don't think it was harmonics that killed the PSU. My UPSs are bypassed until there is a power loss. Then they switch on and create a very dirty stair stepped sine wave as they are el cheapo consumer grade CyberPower AVR UPSs. PFC in the fancy power supplies aggressively fix the simulated sine wave, though since the steps are big, the circuitry in the PFC doesn't last long if I have a bunch of power outages or decide to move things without turning off the servers. It's funny that I don't have this problem with the value brand PSUs without the PFC circuit. Same UPS, but years of usage in various different servers.
cats~$ sudo chown -R us /home/base
(Score: 2) by RS3 on Wednesday July 01 2020, @03:53PM (1 child)
I'm sorry, one of my problems in communications, especially written like here, is I don't know what others know. I'm an EE, and we learned an amazing and fundamental concept of any kind of signal (anything but DC) - that any waveform is made up of a series of sinusoidal waveforms of specific frequencies and amplitudes. We call it "Fourier analysis" https://en.wikipedia.org/wiki/Fourier_analysis [wikipedia.org] So by nerdly definition, anything outside of a perfect sine wave has added harmonics.
You could do a fairly simple experiment. A bit of it will require some electronic knowledge. You can use your sound input as a signal sampler, and run software that does oscilloscope waveform display (time domain) and see the stair steps, and then you can run an FFT to see the spectral components (frequency domain).
The electronic knowledge part is where you shift the 120 VAC (or whatever your "lines" voltage is depending on where you are) down to 1 V or so, so that you don't zap the sound input and potentially fry the whole computer. Also, signal ground might be a problem, so again, it requires some electronic knowledge / stuff.
You might find this interesting (or too nerdy, or overwhelming): https://www.researchgate.net/figure/Details-of-FFT-spectrum-for-output-voltage-with-SPWM_fig6_291185126 [researchgate.net]
Or just look at the pretty pictures (who really read the articles anyway!): https://www.researchgate.net/profile/Suresh_Yellasiri/publication/291185126/figure/fig5/AS:320084447907857@1453325487824/a-Simulation-validation-of-proposed-nine-level-inverter-b-Simulation-validation-of.png [researchgate.net]
https://www.researchgate.net/profile/Suresh_Yellasiri/publication/291185126/figure/fig6/AS:320159240736771@1453343319374/Details-of-FFT-spectrum-for-output-voltage-with-SPWM.png [researchgate.net]
The stair-step problem is known and being worked on, where people care to. As an EE I'm very disappointed that the PSU mfgr. didn't account for this very well known problem.
My next step (ha ha, get it? sorry...) is to do some research on PFC compensation stuff, but not now, important work calls...
(Score: 2) by DECbot on Wednesday July 01 2020, @06:03PM
I found the answer to this question to have the pretty pictures articulating what I think I'm working with. [superuser.com] I haven't borrowed the oscilloscope from work to determine how terrible my sine wave is, but I know I'm somewhere around low-end given how much I paid for it. I do think your opinion is correct, a decent PSU manufacturer would be aware of these low end UPSs and should perform longevity tests to make sure their products wouldn't fail prematurely. Though, I imagine the testing they actually do is not to prove the quality of the design and components, but to set the maximum warranty length. All of mine died out of warranty. I suspect that the components and design of the active PFC PSUs are paired with a low quality filter that was cheaper to produce in mass with the PLUS 90 energy efficiency rating than the circuit design using just a high quality filter circuit and no PFC but with the energy rating. Though I don't have the time and inclination to test that theory. I mean I didn't even take the time to investigate why the 1st or 2nd PSU died, I just borrowed an old PSU from a retired server that was previously on that UPS. As far as I know, there's just a blown soldered on fuse some where in those two PSUs. My theory of the PFC circuit came after the death of the second unit where I found the old PSU worked fine and I was doing some postmortem research about UPSs and PFC as that was the only thing in common with the two failed power supplies. I intended to order a replacement, but it's been 5+ years since the last failure. This PSU has been in service with that UPS for over 10 years now and I don't want to change something that's working. Maybe my previous post spoke with too much authority on the topic as I didn't validate my theory with a proper tear down of the PSUs possibly causing me to jump to wrong conclusions--as we often say here, correlation is not causation. Let me try to set the appropriate tone, these are my suspicions that I have not taken the effort to prove. Nor do I believe them strongly enough to make it a priority to determine if they are correct.
cats~$ sudo chown -R us /home/base
(Score: 3, Informative) by The Mighty Buzzard on Tuesday June 30 2020, @02:56PM (1 child)
Just to clarify, I wasn't defending you, only imparting information. You're perfectly capable of dealing with, ignoring, or laughing at culo's trolling nonsense, same as I am.
My rights don't end where your fear begins.
(Score: 2) by RS3 on Wednesday July 01 2020, @12:42AM
You stuck your neck out; I saw a fulcrum. :)
(Score: 2) by RS3 on Tuesday June 30 2020, @04:17AM (5 children)
Citation needed. Oh, you imagined that? Well okay then.
Again, you're logic-leaps ahead. So by your decree, I'm not allowed to ask a freaking question? I'm just trying to be helpful here. You really like to troll me.
With you, any left-field interpretation is expected.
When my life gets so bad that I need you to manage me, I'll change my name and move far away first.
You're so far out of reality I don't have... just not worth my time. There's much more going on than you know. Stick to what you know and stop trolling me.
(Score: 2) by c0lo on Tuesday June 30 2020, @04:38AM (4 children)
I don't even try to manage you, I'm showing a possible interpretation of what you wrote in the provided context.
Come on, show that the interpretation is impossible. Otherwise, how the heck am I going to learn the reality if you don't even.
I don't understand what's exactly you mean by "left-field interpretation". Care to explain or rephrase?
https://www.youtube.com/watch?v=aoFiw2jMy-0 https://soylentnews.org/~MichaelDavidCrawford
(Score: 2, Informative) by RS3 on Tuesday June 30 2020, @06:13AM (3 children)
There's a saying "he's out in left field" meaning "way off base" meaning fantasy-land, not reality. You do too much reading into things, then deducing conclusions that are based on incorrect guesses and artifact. It's okay to ask questions, but you ask loaded questions- ones based on incorrect assumptions. Like I said, I think you do too much interpretation. But that said, I think a lot of people do far too much interpretation- especially these days. I think it causes some of the worst problems in the world. Read up on the "Dunning-Kruger effect". https://www.verywellmind.com/an-overview-of-the-dunning-kruger-effect-4160740 [verywellmind.com] I call it arrogance, or overly-confident, but someone got some grant money to study and document it. Possibly named "Dunning" and "Kruger"... (that's my sarcastic humor that many of my friends (real life ones) would get the humor right away. Mostly the making fun of ludicrous ideas that qualify for grant $. Read about this guy if you don't know him: https://en.wikipedia.org/wiki/Matthew_Lesko [wikipedia.org] )
You make a good point about subscribing, but a tiny subscription isn't going to pay big $ that the hosting costs. The current hosting is awesome and could handle vastly more size and traffic than SN has currently. I know the founders had high hopes for a large membership here, and I was hoping the same, but I see some problems that drive people (and their $) away, IMHO.
As the admin for a small hosting company I can do far more for SN by negotiating hosting for far far less $. Making any sense yet? I've been in private communication, a bit, that you know nothing about and isn't your business, about possibly hosting this site. I have a great deal of unused bandwidth available. Slightly used, but very awesome (more CPUs and RAM and TB than they have now) servers are available cheap ($160). I'm not (in any way) a salesman (and can't stand pushy people). I've made offers and if the admins here want to, they can contact me to discuss, brainstorm, and negotiate.
Does that make sense yet? Most people get it. I know there's a lot of good in you and I don't mind clarifying things- because you asked. :) When I was younger I didn't comprehend a lot of things in life, like movies. If it wasn't spelled out, I didn't get it, but I didn't make assumptions. Somehow my brain grew some connections and I'm able to deduce and connect things- so much that most movies and TV shows are boringly predictable. A few aren't. But I'm off topic. I don't mean to be so verbally mean, and I'm extra unhappy that I wrote sharply. You've trolled me before and I'm probably misunderstanding you- we misunderstand each other. I know I sometimes write- well, my attempts at brevity sometimes leave some people out. You like truth and hate BS (as do I) and you want backing info and citations, and that's okay, but you ask by attacking and putting me on the defensive. I'm learning, and I hope some of this is helpful to you and that we can become friends, and I mean it. Have a good day!
(Score: 3, Interesting) by c0lo on Tuesday June 30 2020, @06:37AM (2 children)
Thanks for the explanation.
I suspect a cultural reference so very specific to US, probably coming from baseball by the terminology it uses.
Apologies, after quite a large numbers of years in engineering, one gets to look of "what can go wrong" before it does.
Well, not providing enough context is bound... nay, highly likely to create a lot of problems too - my opinion leans those problems are far worse than those caused by reading too little.
And this simply because there are so many ways the things to go wrong and so few ways for things to go right.
Oh, how much of the meaning of your original message would have been more precise if you only added this tidbit to it as context.
Don't you think?
https://www.youtube.com/watch?v=aoFiw2jMy-0 https://soylentnews.org/~MichaelDavidCrawford
(Score: 2) by RS3 on Tuesday June 30 2020, @01:01PM (1 child)
Yes, sorry for the USAian colloquialism. I don't know who's from where but iirc you're NZ or AUS maybe- no matter. Non-USAians love to point out how we USAians are not as world savvy as many esp. European and Oceanic. That said, we're getting better, I think, maybe? Some of us? Trying?
My best friend is not an engineer, but certainly does intense engineering R&D and he is stunning at telling me what and how something can go wrong with any ideas, mine included. It's one of the things I most like and revere about him. Most of life needs someone to do that- govt, engineering, medicine, business... And frankly he can be very direct, but somehow his delivery doesn't come across harshly at all. But knowing you're also an engineer helps a lot- we can be a very surly group. We're so focused on facts and truth and that's just how it ends up coming across.
All I was trying to say is: you're correct, I did not give all kinds of details. I'm big on context, and to me the context is this: if I read something that I don't comprehend because data / details / citations are missing, I move on. It wasn't intended for me. If it is for me, like a work assignment, I do my best to gather the info I need. One of the "engineers" (he had no degree but was a pretty poor code monkey) I used to work with was like that- assignments from the top boss would be met with insurmountable walls from monkey. I'm kind of the opposite- I'm the eager enthusiast who loves the challenge of making something work, and occasionally to a fault, with as little help as possible (especially from the grumpy crowd). :)
I don't mean to be demeaning, and sorry for redundancy, but again, my comments were brief and made sense for the targeted audience. If you're part of the SN admin team, or just wanted to be more "in" on the conversation, it's okay to ask. Skepticism is perfectly okay by me, but maybe a less surly approach?
I have NOTHING to do with the "Dilbert" comics, but today's is just too appropriate (and actually helped me understand you before I read your most recent post:
https://assets.amuniversal.com/e679f1a08d7f013808e1005056a9545d [amuniversal.com]
Thanks man!
(Score: 2) by c0lo on Tuesday June 30 2020, @01:15PM
Touche.
About those cooties: LOL, thanks for that, made me feel good close to the time of day change (yes, I'm hailing from Australia).
https://www.youtube.com/watch?v=aoFiw2jMy-0 https://soylentnews.org/~MichaelDavidCrawford