It’s clear the bugs introduced in the server upgrade have NOT been fixed. New posts are scrambled up with old comments, and I have no idea where submitted comments are going to end up.
I recommend leaving Pharyngula for a day or two until these severe problems are actually fucking fixed. That’s what I’m going to do, anyway. Yeah, BOYCOTT PHARYNGULA.
Jesus Christ, but this thing is a mess.
Is it time to Occupy FtB?
See you later…
*Clenched tentacle salute*
If you’re going to say that, I’m leaving.
Glen Davidson
Seems better, PZ. Maybe a nice walk around the park…
*transports some bacon flavored veggie popcornz and swill disguised as beer to PZ*
Will the Occupy FtB be outdoors or in a nice cold air-conditioned server room? None of this wimpy feet-up-by-the-fireplace stuff for me! Cold air and hot rum toddies rule!
I just managed to post a few comments on the Hitchens Mormon thread and they ended up in the right place.
I can see that the earlier tranche of comments belongs to another thread.
Testing
David B
Boycott? Count me in.
As a symbol of my solidarity with your noble endeavor, I shall refuse to comment in this thread.
In the meantime, everyone is invited to my place to commiserate with my friend, who is expected to adhere strictly to a nonexistent student disciplinary procedure. The syllabus is sacred! Any advice?
Hmmm. My server is with Codero as well, and the few times I’ve needed assistance their tech guys were on the ball, up-front, bright eyed and bushy tailed and went a couple steps beyond what the contract called for with no hesitation and nary a word. I’m very happy with them.
Of course I do the huge bulk of my own low level tech admin and only needed help on very low level things like an OS upgrade where the sshd service didn’t come up…. the problems here look like database restore and web application maintenance. I don’t know if you’ve contracted those to Codero or through someone else.
Those kinds of things can get funky, particularly keeping in sync if you try to keep the service open (bad idea). You need somebody who really knows what they’re doing and ideally (but difficult in practice) finding someone with a lot of documented experience with the particular software packages you’re using.
Other than that you need to find people who are very skilled low level (where the action is) techies who can figure it out. As a low level techie I can, after a bit of tech discussion, tell whether I’m talking to someone of similar (or rarely better) skill than I at that low level stuff. But if I could tell a non-techie how to identify that kind of techie, I’d write *the* book that corporate suits in tech dependent industries (all of them now) have wet dreams about and retire to a Caribbean beach.
— TWZ
I’m swiping the picture of angry, typing Squid to use at some point!
~wonders where this Schroedingers comment will end up~
More seriously, good luck with the server -.o, hopefully the squid will be less angry at some point!
This is why I’m not pestering you guys to post anything to amuse me. It looks like bollocks all the way down.
Not to worry.
The best laid plans of……….
So you might say the server is still…kraken up? *runs*
Simple.
Step 1. Restore your backup.
(They DID make backups before doing an ‘upgrade’, right?)
Ain’t technology grand?
tap, tap, tap, can anyone hear this? Can you hear me now?
Stephanie Zvan said (in a post which seems not to have survived the transition) that the plan was to allow the techie managing the transfer to get a little sleep over the weekend.
NOW you know why industry-standard practice forbids such slackness!
Oddly enough, this is the first article at Pharyngula I’ve seen that appears to be working correctly for me…
Pack up the kids Emma, we’re staying at scienceblogs for a few days.
You should really move to The Cloud!
I’m posting a comment on “I give up” just to see where it goes.
hadron @24 says
As said low level techie, let me ask: does anyone know of a visual of a middle finger salute embedded in a broken and bleeding brain?
— TWZ
Uh…yes things do seem to be still bollixed. Some afternoon comments above, followed by some morning comments. Time is flowing backward.
Ah well.
Somebody give me a call when we reach early 2000? I want to have a word with the folks putting together the Palm Beach ballot.
Like Randomfactor, this is the first time since the upgrade started that everything appears to be normal here. At least I can actually SEE comments for the first time since the upgrade.
If you added an entry to your hosts file the last time the “new” FTB server was set up, remove it now. It may not clear all problems, but at least Facebook links work. Google “hosts file edit” for your operating system to fix this.
Another shutdown might be in order.
Will we ever get the comments back where they belong?
Well, the advertisements on the sidebar, the ones exhorting me to become an ordained pastor and answer God’s call, are up and running while your comments falter…. PZ, you know this is God speaking to you, and that this might be the last day, the last chance… become and ordained pastor and all will blog well…
I just tried to make a post as a test. It vanished. Tried twice more: gone.
Maybe I’m just going to have to move back to scienceblogs, as poorly maintained as that was.
Stangely enough, everything is working fine for me. Nothing has vanished.
BTW, the new servers are quite a bit faster than the old ones.
Wow. It looks like Obama will actually win the Democratic Nomination. Is the US brave enough to elect a middle-of-the-road black man as President? Should be an interesting election.
— (((Billy))) the Atheist
If you’re having problems, (I’m not) you may try doing a forced reset of your internet connection. If I remember the windows version correctly, first you want to close your browser(s), open a command window and type something like ipconfig /renew
If it is a move to a new box (most likely) the most likely issue with seeing confusion is host not resolving correctly. This is usually a DNS issue. The IP I’m seeing is 68.87.64.150. You could go into /etc/hosts (again, if I recall this correctly … c:\windows\system32\drivers\etc\hosts and set freethoughtblogs.com to that address – on linux and most other unix systems just /etc/hosts). If you do this manually, at some point next week you want to go back in and remove it.
In any event a DNS change should propogate through all properly configured DNS servers in (less than) 24 hours, and all servers should reset (much less than) 24 hours to check upstream changes.
The point being that everything should shake out in (theoretically, for everyone) less than two days from the moment of the change (and almost always less than a day real time). Some less than smart systems don’t always do the right thing, however.
— TWZ
This is a test of my boycott of Pharyngula. This is only a test. (Did I pass?)
Everything seems fine on this end.
This post is supposed to appear on the “I give up” thread.
Another annoying thing? I think all our backchannel communications are hosted on the same server, so I don’t get any of the status updates and all attempts to communicate my complaints are met by silence.
Hey, if I wanted unresponsiveness and intolerably slow tech support, I would be back on scienceblogs!
Just like talking to the lard our dog, really.
@ColonelZen
Best I can do for now…
http://i674.photobucket.com/albums/vv101/Konradius5/Gary%20Larson%20Comics/BrainSurgery.jpg
You are a cat person after all!
PZ, I have some experience running Linux on a Linode for a real estate firm. I can get you set up nicely with a LAMP stack (though I need to learn how to get this specific forum software running…it’s just a big PHP/MySQL webapp right? What ISN’T these days…?).
You could probably run this blog off the minimal Linode plan, though i expect you’d need to purchase additional bandwidth. Linode is very very customizable, quite obviously run by geeks like me, and I can’t recommend them highly enough.
:)
You’re funnier than you give yourself credit for, (((Billy))).
Dear PZ @39: Ah, good. I thought I was the only one blacklisted. All I’ve got is silence, too. Silence is not golden when your blog is hosed and your readers upset.
Tell you what. We should gather a few of the others and start our own enterprise, Free Freethought Blogs. And we will start out with the fastest, biggest, best server available so that we never ever have to do an upgrade ever in our entire careers.
Or until next year, when changing technology will force us to upgrade anyway. Sigh.
That would be a refreshing change from all the server software traditionally administrated by jocks.
http://www.etymonline.com/index.php?allowed_in_frame=0&search=sabotage&searchmode=none
sabotage (n.) Look up sabotage at Dictionary.com
1910, from Fr. sabotage, from saboter “to sabotage, bungle,” lit. “walk noisily,” from sabot “wooden shoe” (13c.), altered (by association with O.Fr. bot “boot”) from M.Fr. savate “old shoe,” from an unidentified source that also produced similar words in Old Provençal, Portuguese, Spanish, Italian, Arabic and Basque. In French, the sense of “deliberately and maliciously destroying property” originally was in reference to labor disputes, but the oft-repeated story that the modern meaning derives from strikers’ supposed tactic of throwing old shoes into machinery is not supported by the etymology. Likely it was not meant as a literal image; the word was used in French in a variety of “bungling” senses, such as “to play a piece of music badly.”
Are you kidding? I’m not funny. That Obama comment was from about 3 1/2 years ago. Not sure how it got here.
Or did you mean the cat comment?
@ColonelZen:
Depending on the “non-techie” and how willing they are to do a bit of research, that could be easy – just ask the prospective candidate a few questions:
1) Who was the first computer programmer, and what was she “famous” for (ie – importance to computer history)? (two possible answers here, only one of which is truly correct IMHO)
2) Who was Claude Shannon, and why are his theories important today?
3) What kind of minimal logic feature is necessary for the development of a Turing-complete machine?
4) Can you describe (schematic) or build that feature, using simple mechanical components?
5) Have you ever dreamed in code – and did your code compile after you awoke and implemented it?
6) What’s the lowest level form of programming you have done, and why did you need to do it?
7) What are your hobbies? What do you think about when not thinking about computer problems?
Items 1 and 2 are history questions; IMHO, you aren’t worth your salt if you don’t know your history (especially the answer for number 2; very relevant for the internet and networking). Number 3 shows how much said techie really knows about the underlying hardware of computation – or whether it is a “magic black-box” to them. They should be able to answer in affirmative number 4 if they got number 3 right. Numbers 5 and 6 show how dedicated they are to software development; there isn’t necessarily a “correct” answer to these, but the answers might reveal interesting insights into their person.
Finally, for number 7, if they answer anything but “I am always thinking about computers and/or computer problems, and how to solve them”, you might want to find another candidate. Maybe it’s a bit unfair, but typically the best in the tech world tend to be so dedicated to the craft, that to not be thinking about such is an anathema – it is almost completely unthinkable.
I am sure more than a few here (perhaps even yourself) could rattle off all the answers to these questions without needing to google any of them; I don’t consider any of them to be particularly difficult.
Note: I posted this answer as an article on my website, which I am not linking here; btw, @ColonelZen, your profile link appears to be broken…
crOsh:
As a Comp Sci grad student, these should be cake… on my honor, I didn’t use Google.
1. either Ada Lovelace or Grace Hopper, depending who you ask. Lovelace worked with (a simulation of) the Difference Engine, while Hopper created (ugh) COBOL.
2. Claude Shannon spearheaded research into information theory, defining such concepts as entropy in communication. His theories are important today because we’re still trying to compress data as much as we can.
3. I’m not sure what you’re asking, but I’ll take a stab at it: NAND gates? (or NOR gates, which are equivalent, but costlier to implement)?
4. Probably, depending on the technology. It’s been a long time since I studied CMOS, and I’m still working on the redstone implementation.
5. Dreamed in code, yes. Compiled, not quite. I didn’t remember the details completely, but the structure worked.
6. MIPS assembly, because it was required for undergrad Architecture.
7. Photography and Minecraft right now.
@cr0sh: Congratulations. You’ve devised a test to find obsessives who will never tell you what they’re doing to their — whoops, I mean your — computers. Good luck when they get bored and leave!
Huh. This is WordPress, fercrissakes. It isn’t like it is some sophisticated software doing complex data manipulation.
Oh well.
Dude. View. Source.
And what “isn’t” is software that does stuff that, for example, requires data integrity.
In other words, you will never-ever-ever find the “MP” part of LAMP on the floor of the NYSE (or the cloud that is NASDAQ), or running on any NASA mission hardware.
I’ve now heard from the FtB techie — work proceeds apace, it will be fixed.
It sounds like our side did everything right: we paid for the expert pros at our hosting company to do the migration, assuming that they would be the experts at their hardware and have this process down cold. Strangely, we got professionals who committed to starting this process late on a Friday night, who then announced that they wouldn’t be working on it over the weekend…so when it entered strange mutant alien territory, they just dropped it and left it all dangling.
Not impressed, Codero, not impressed at all.
Bunny:
Way to stereotype. Good luck finding somebody competent who doesn’t get labeled ‘obsessive’ by your standard.
Would you level the same claim against an auto mechanic who eats and breathes engines?
Oh gee.
Should have started it on Tuesday. Mondays are usually out, for obvious reasons. Fridays are usually out, for obvious reasons.
Offer them free pizza as a bonus if it only takes them twice as long as they said it would. LOL.
@50 – Benjamin: You win. I don’t even have a cookie to give you, though – sorry.
Regarding your answers:
1. Perfectly correct and acceptable; Lovelace also foresaw uses and applications of computers that wouldn’t be done for almost 100 years (automated music composition, for instance), Hopper is also arguably the inventor of the term “debugging”.
2. 100% there.
3. Correct – on both – nice.
4. Interesting that you thought about doing so via fabrication tech – but could you do it with Lego? Rope and pulleys? Matchboxes? You don’t have to answer; I’m sure you could.
5. Heh – glad it worked out!
6. Interesting – I once had to do some low level hex editing in the ROM monitor of an Apple IIe for some graphics coding I was working on. No assembler available, and I was too lazy to write one.
7. Aww, man – where’s the love? Just kidding…
I’d probably hire you – if I had a company and a reason to hire someone, I guess. Some code review and other would be necessary of course, but this “geek challenge” would put you high on the list. Many consider these questions too deep, too challenging; in computer science, they are considered very basic (as I am sure you are aware) – 101 stuff at best. The number of people though, who claim to be computer geeks, who don’t know the answers, though, is astounding to me sometimes. It’d be like an automobile mechanic who couldn’t explain the Otto-cycle.
@Naked Bunny with a Whip: Perhaps – but I only consider the questions to be prelim; a review of work history, actually talking to them, etc – would all be necessary to weed out the best candidate for the position.
@Usernames are stupid: Regardless of what the software is doing, having the knowledge of the answers for these questions would tell a lot about the individual, and whether they take their job/career/interest in computing seriously or not. Would you have a mechanic work on your car who couldn’t tell you the difference between fuel injection and carburetion, and how each works?
To be fair, the tech only told me in retrospect that he didn’t work weekends. He apparently did the first attempt at the migration Friday night, but it was borked. So I called Saturday morning and they called him back in for the second try. Which did get the new server up and running, but they never told us that or switched over the DNS. Then when they did switch over the DNS, they screwed it up and we had to fix it. And this morning, the tech tells me that they didn’t tell us or switch over the DNS because they had no confidence it was going to work — so apparently they just stopped and were going to let it sit there not working rather than getting other techs on the job. It was only after I sent a message complaining about the whole thing that the tech who did the migration told me he didn’t work weekends. That pretty much set me off. They started a migration at 10 pm on a Friday night and screwed it up and apparently weren’t going to do anything about it until Monday, until my phone calls pushed them about it. The other excuse was that they were waiting for information, but we had answered their questions a full 18 hours before they said they were waiting for information (and it’s all there in the trouble ticket thread, with time stamps, to prove it). Bad service, almost no communication and then lame excuses. I kind of lost my cool with the last guy I talked to on the phone and in the last message to the tech, after which he just shut up.
Bear with us, folks. We’re working to fix this mess as fast as we can.
@54 PZMyers
“Strangely, we got professionals who committed to starting this process late on a Friday night, who then announced that they wouldn’t be working on it over the weekend…”
Gosh Professor, it doesn’t seem strange to me. It seem like sabotage. It could happen!
Anyway, looks like things are going ok now.
Thanks Professor for the fine site.
It is deeply appreciated. Now if only I could get the techs at the various services that I actually pay for to be as dedicated and hard-working on my behalf.
Fuck techs.
What we had there was a failure to communicate…
and what time zone are we in now?
I hope I am not disregarding the “take a break from Pharyngula” clause that was recently postuled. Although it is probably encouraged to test things here. *but this not a test*, so don’t worry.
For 12 ColoneZen
You need some help? I am available & work well with animals?
To be me I like the observation I just made in post 48 of a 1 and a 2 and a 3 and not in that particular order *pause* *chuckle*, *sorry* that Father Ogvorbis, OM: Delightfully Machiavellian coincidentally while undoubtedly probabilistically so placed them there. Indeed Father you made my day. SBT
Now if the servers or the IP’s or the VP’s or the posts are fried or scrambled or toasted, roasted or still just plain boastin’ there are ultimately positives and negatives. Flying Spaghetti Monster I love electricity. And I do mean electricity, not emotional related human perceptions of a set of circumstances as being “a positive thing” or a “negative” thing, unless they are withstanding a massive Ion cloud than set language is deemed appropriate but anyways….iam delaying.
Have not checked that one post but I am pretty sure that something is bothering me. Oh yes; and Noo it is not my stupidity oh Mighty Oaken Minded Dendritic Giants, I feel encouraged internally to offer objectional interpretations of my . . .my . . . *hrumpf* stupid post. However, my claim is that because Iam a massive tone troll (if I understand it correctly,cuz last I checked my briefing on the lingo was Deleted (pz if ya heard), I learned more in the 69 comments or so that vanished than. . . well anyway) you shouldn’t be so mean.
Now though any semblamce of my ressurrection was grounded out and lost. So I must try. Though, it will be hard to stupidly redo a stupid post so stupid is no longer stupid through stupids own words and so that my version of reality is not relegated to that of methane resting in a cow on a farm where they kill shit.
Hopefully I won’t struggle too bad. I had asked to “take it all back” but that wasn’t going over to well *chuckle**confused* I guess that’s just; “how it be”. Since my pathetic apology was met with more savage verbal attacks I have no choice but to look at my comments & seperate my questions from my drug induced unintentional rambling long winded bullshit. Damn! Did I just do it again. Shit. Mania, what are you gonna do?! Harness it!
I believe what iam saying about that post is
“You missed the meaning of my most appropriate argument for the reasons for why PZ posted it in the first place. Dissecting the nature of the long term and very relative practice of being a citizen of the public majority is obviously where all of the reasoning was going and that’s where without reading the comments based on the questions asked in the origal post by PZ I stated questions & brief facts…& thus began my what would only be described as a journey of awareness to the Valley of the Tone Troll; where has been described that I own a nice piece of property
And shorter
“You just don’t get it.”
Does that work :) iam kidding, but don’t banish me, anything but baaaaaaaannnnn iiiissssss mmmmmeeeee nnn n t t t
Thanks for the update, Ed.
“We’re going to do the upgrade at 10pm on a Friday and then not check on it for a couple of days.”
What could possibly go wrong?
Eek.
No no no no no not “fuck techs” – techs are good.
There was me biting back all the complaints yesterday and this morning, thinking FTB’s tech had done the transfer, not realizing it was Codero. I could have been complaining loudly all this time!
Ed Brayton:
That’s okay, you were losing your cool for a whole lot of us. Thanks for pushing them to get things done and done right.
They’re busy distracting themselves on their favorite techie blogs (and football on tv over the wend, natch). The country hasn’t got a current with a sensible direction anymore. It’s just a dizzy whorl going round ‘n round before going down the drain.
Productivity? Progressivity? Professionalism?
HAHAHAHAH.
Too hard. Besides, who has the time to bother with that stuff? Their leisure is everything.
It’s not terrorists. It’s ‘stimulation’, ‘sensation’ and ‘entertainment pastimes’ – any excuse not to work. Diddle with that new gizmo. It’s awesome.
It’s epidemic procrastination.
*flushing noises*
What is this, Greece ?
@67 Anchor
interesting TED video,
http://www.ted.com/talks/zimchallenge.html
THE DEMISE OF GUYS?
There’s nothing wrong with working weekends; there are many things wrong with doing a major change right before you leave for your weekend off. They should have saved it for Monday morning.
ARGH Nothing wrong with NOT working weekends
What’s the unemployment rate there again ? Especially for an IT company or webhoster, where the problems don’t stick to a Monday-Friday schedule, support should be available 24/7.
crosh @49
A good talker who’s gone through just a couple CS courses could probably answer some of these and BS their way through the rest to at least dazzle a non techie or non-deep techie manager.
#5 isn’t much use to those of us who rarely dream (that we remember, anyway. Of those few I remember I never had one about code to compile to do anything … but once I had one where everybody’s talk (including mine) was code. And I was frustrated that nobody else’s would compile. (BAL … mainframe assembler).
#4 as mechanical isn’t as easy as it looks either … a tape which is indefinitely erasable and rewritable is not something that is easy to come by. That said, it could be beautiful.
As for my web site, sloppy, an extra e (should be fixed in this link). There’s some public face if you wish to look, but not much, and probably of little interest, and I rarely update or add to the public pages. I occasionally use things you can’t see (hopefully) to testbed things for clients (its too easy to write things with dependencies on your “home”machines software stack and configurations, hence the value of a remote server).
In fact I have little sympathy for the suits and their inability to distinguish good techies (both admin and prog’rs … I’ve been both). The only real way is the way to know good human beings. Keep them around, work with them and help them grow for a while. Mediocre can properly encouraged be made good, and good, properly encouraged can become excellent. And those who only want super performers, no matter what they’re willing to pay, are looking for whores for a one night stand, not people.
— TWZ
Because people don’t work weekends? I suppose it’s our fault for letting them out of the mines to die under open sky in the first place.
For Benjamin Geiger, the non-obsessive tech who has non-computer-related hobbies: 16-bit ALU built entirely in minecraft.
It ain’t broke so let’s fix it… stupid upgrades.
No way. If Pharyngula goes down for more than a day, I’m gonna need methadone.
Ed Brayton @ 58
You have every right to lose your cool. I’ll grant that a hosting company is a different kind of operation than dedicated staff that I’ve always been, but …
An admin/techie who “doesn’t work weekends” (at least in an emergency) is not a techie.
Ad admin/techie who walks away from a down production system (without handing it over to a tech equally/more competent, and with notice of responsibility “officer of the day” style) is not a techie.
And a support organization that requires or allows such behavior of a “system administrator” level tech is a bad organization.
The “military style” responsibility may sound overbearing to those who’ve never been a real tech at the mainframe level but that’s very much the culture I learned the craft in. The systems programmer/system administrator of a system is *responsible* for it … in extreme cases that might mean going to his boss and saying “I can’t handle this we need to contract out (specific expertise at 400/hr or more) but the admin is *always* responsible and cannot walk away from it without de-facto walking away from his job.
Of course the systems I learned this on cost half a million and rapidly up, and where deployed a day of downtime could easily cost the company tens to hundreds of thousands of dollars, but even in the later smaller web service companies I worked at, if not quite so ramrod about it, it was still the prevalent ethic of the tech support staff (and I was partly hired to re-enforce that ethic).
A service company which doesn’t foster that kind of attitude and responsibility – a roster naming the responsible admin for every minute of a system project/event who is required to be responsible for every decision and action regarding the system under his “watch” – is not doing it right.
— TWZ
I wouldn’t say that any single person ought to be on call at all times no matter what. Either the company is big enough to have people on staggered shifts and cover all the time periods needed, or the company is too small to have that many employees and they don’t do something like a major server upgrade right before going on a break.
To clarify, FtB has only one official tech. I volunteer some second-pair-of-eyes work and ad hoc migrations together now and again, and Ophelia’s and Dan Fincke’s techs are on the backend mailing lists, but it’s a one man show ultimately.
@69 joed: Heh, and the segment is immediately followed by an ad for all the awesome things you can do in a smart-car that, among other things, let’s you know what entertainment is available exactly when you need that information most, while operating a motor vehicle. That isn’t innovative. That’s idiocy.
We’re screwed.
@74 brownian: [Country going down the drain, etc] “Because people don’t work weekends? [etc]” Oh, of course not. I was being a tad facetious and perhaps you exaggerate a tad. Not working weekends is dandy. What sucks is a person who takes on a job, schedules it poorly fully knowing the weekend looms, badly misjudges the time required to complete it YET doesn’t respond to or follow through to the client’s satisfaction. Rorschach #72 is right: support should be available 24/7.
That’s not professional behavior. That’s just another schmuck who doesn’t give a phlying phuck. Bottom line, his precious weekend is more important than his client, who hired him in good faith and pays him good money for services rendered successfully, if not in the originally estimated time, in a diligent and perservering manner asap. Does anyone imagine this guy spends his wends sleeping, say, because he’s so overworked all week? Poor dear…
Oh, and Ben Geiger’s “hobby” isn’t contingent on client satisfaction either. All HE requires of his hobby is that it sufficiently pleases and amuses himself. Nice work, such as it is, even if it is “somewhat slow, big and unwieldy”. Wonderful way to spend his weekends, I’m sure. I can appreciate that kind of dedication on the justification just to see if it could be done. And if he derives pleasure from the diversion, if it entertains him, that’s all great. Too bad such diligence and discipline (dare I say it?) – passion – does not characterize the techie’s attitude toward his job.
Well, I guess it’s a bit different if you have just one guy looking after your business, as opposed to multiple staff on some kind of roster that covers 24/7 support. You can’t expect one person to be available 24/7.
But that leads me to another question, if FtB has one tech, and they are not with some professional hoster that has proper support, what was this upgrade about ? Moving FtB from Ed Brayton’s IPhone to his laptop ?
@ColonelZen:
You’re right that someone could BS their way through the questions; but then, such people might be able to do so no matter what the “test”. Nothing is perfect, and such a test as I proposed should only be one component of such an interview, and not the only requirement.
Re: your comments on #5 and #4 – I suppose #5 could be re-worded such that “non-dreamers” could participate; as far as #4 – I wasn’t interested in a complete Turing machine, but rather the individual logic element; a completely mechanical Turing machine, while possible (though not the “infinite tape” part, of course), would likely leave someone sketching or building for a while (LOL), though I have seen more than a few similar mechanical implementations out there (I saw one made of Lego, once).
Ultimately – as said before, and as you pointed out – you have to do more than test; you have to work with the person, talk to them, etc; there aren’t any easy solutions, and I hope my proposal wasn’t taken to mean that (?). I do think that the answers given would give a fascinating array of insights about the applicant, though…
lol (truly)
PZ and Ed Brayton: I question the wisdom of moving a lot of atheist blogs to FTB when a server upgrade can take out 2 dozen blogs for a day or more. There is value in diversity, and having Blag Hag (for example) on FTB doesn’t make it any easier to find or read, but it does mean that if FTB is down for any reason, not just Pharyngula but a significant part of the atheist blogosphere goes down with it.
As operators of a reasonably successful blogging platform, you should be mindful of system availability, particularly during upgrades. I read the comments; you hired what you thought were skilled, professional techs. What I gather from the comments is that their job was to move FTB to a larger server on a new IP address. There are some rules to follow to minimise pain on such a move, specifically reduce the DNS time-to-live records well in advance, copy data to the new server and configure it, and only when the new server has been tested and found working, do you switch the DNS address. The problems I read about with posts disappearing and comments appearing on the wrong post sounds like two things – firstly, sometimes accessing the new server and sometimes accessing the old. Secondly, incomplete copying of the database. It could be that FTB users were reading articles on one server and their comments were posted to another.
You need to start thinking about architecture. FTB sounds like a big Single Point Of Failure (SPOF) – one server does everything, and when it becomes overloaded you have to build a bigger server and migrate your application. That’s not the right way to do things. At least you can separate your database from the web server. That lets you run 2 or more web server machines, maybe with a load balancer in front of them, so if one of those web servers is down for upgrade or outage, all the users will go via the other server. Since the database will be separate, upgrading your web servers won’t affect the database and you won’t have mysteriously moving comments.
Also MySQL can be clustered over 2 machines, so not only will database reads be faster (or you can do twice as many per second), but there’ll be redundancy if one of the database servers goes down.
Do a google search on “scaling wordpress”. Here’s the first hit:
http://wpmu.org/scaling-wordpress-wpmu-buddypress-like-edublogs/
Being able to scale out your site is even more important if you want to bring on more atheist/freethinking bloggers.
The wpmu.org advice is pretty good – enable database caching; run nginx to distribute PHP requests to a “PHP worker pool” (you are presently running Apache); move your uploaded media to S3 or a CDN.
You will need to hire techs to make this happen, people with architectural skills whose job will not be to move your site to a bigger server, but re-engineer things so the site will be more robust and you can grow it without painful outages.
Why is a major upgrade being done starting Friday night, a high traffic time? Someone was not thinking about this? And then the system becomes broken and the tech walks away for a couple of days. Doesn’t customer service mean anything to these folks?
Thank you PZ & Ed for an excellent explanation.
Thank you elronxenu for excellent future solutions, tho I’m not sure how economically viable they are. And you could have edited the 1st couple of p-graphs out and not lost your substance.
Part of the problem is a lot of us had hosts files that we edited to get over the last migration and NO ONE MENTIONED THE IP NUMBER WOULD CHANGE. Sorry for shouting. I also wonder how many people are aware that this is why they still can’t get on the new FTB. And they’re never going to figure it out by posts on this FTB.
Perhaps a notice on the old FTB would be in order. Along with directions of how to change it back.
This wasn’t a FTB tech problem, it was a hosting tech problem. Especially the person who thought it was a good idea to do the changeover on a Friday night when he doesn’t work weekends.
Whoever you are, you would have done better to look at the traffic stats and pick a dead time … like Monday at 9 am EST.
Oh, and Ed, you were right to be upset and communicate that vehemently. Sometimes it takes a clue x 4.
Oh, FFS, whoever said FTB was loading faster is on crack.
Srlsy, the latest edition of TET containing one video (that doesn’t autoload) and 700 comments is taking 25 minutes to load.
No other heavily trafficked web site has these problems.
I’ll hit return when the site finally stops …
[10 minutes later]
Still waiting …
I’m getting very frustrated. I can see the new, unimproved FtB on my computer at work.* I can even post here, obviously. My home computer doesn’t see anything more recent than the posts from Saturday morning and there’s zero comments.
Is there anything I can do about this? I’ve dumped the cache and cookies, I’ve tried both Foxfire and Bling, I’ve tried ipconfig\renew, nothing works, including FtB.
*Some of us go to work way too early.
‘Tis, if you’re still at work, check your hosts file. FTB changed IP addresses.