code logs -> 2015 -> Sun, 22 Feb 2015< code.20150221.log - code.20150223.log >
--- Log opened Sun Feb 22 00:00:50 2015
00:01 gnolam [lenin@Nightstar-t1tbf0.cust.bahnhof.se] has joined #code
00:01 mode/#code [+o gnolam] by ChanServ
01:00 grindhold_ [quassel@Nightstar-uufabm.zebra.fastwebserver.de] has joined #code
01:03 grindhold_ [quassel@Nightstar-uufabm.zebra.fastwebserver.de] has quit [[NS] Quit: No Ping reply in 180 seconds.]
01:10 Orthia is now known as Reivthia
01:29
<&Derakon>
Ugh, I swear I had this written down.
01:29
<&Derakon>
"rsync -avvpH source dest" is how I'd go about making a backup of a directory, yes?
01:30
<&Derakon>
Or was it -h...
01:37 grindhold [quassel@Nightstar-0ona3l.cyan.servdiscount-customer.com] has quit [Ping timeout: 121 seconds]
01:49 grindhold [quassel@Nightstar-0ona3l.cyan.servdiscount-customer.com] has joined #code
02:13 Derakon [chriswei@Nightstar-5fqf0m.ca.comcast.net] has quit [[NS] Quit: Leaving]
05:12 Derakon [chriswei@Nightstar-5fqf0m.ca.comcast.net] has joined #code
05:12 mode/#code [+ao Derakon Derakon] by ChanServ
05:23 Derakon is now known as Derakon[AFK]
05:57 celticminstrel [celticminst@Nightstar-j5b.hci.93.65.IP] has quit [[NS] Quit: And lo! The computer falls into a deep sleep, to awake again some other day!]
06:43 Kindamoody[zZz] is now known as Kindamoody
06:43 Harlow [harlow@Nightstar-oteu2h.il.comcast.net] has joined #code
06:53 Vornicus [vorn@ServerAdministrator.Nightstar.Net] has quit [[NS] Quit: Leaving]
07:48
< Harlow>
anyone able to assist with pointers? (only like 2 levels deep)
07:59
< Harlow>
never mind, got it :p
08:57 Vash [Vash@Nightstar-uhn82m.ct.comcast.net] has quit [[NS] Quit: Quit]
08:58 Kindamoody is now known as Kindamoody|out
10:01 Checkmate [Z@Nightstar-484uip.cust.comxnet.dk] has joined #code
10:01 mode/#code [+o Checkmate] by ChanServ
10:14 Harlow [harlow@Nightstar-oteu2h.il.comcast.net] has quit [[NS] Quit: BED]
12:17
<@gnolam>
https://twitter.com/gardaud/status/568566643669929984/photo/1
12:32
<@TheWatcher>
I'm convinced!
15:16 Vornicus [vorn@ServerAdministrator.Nightstar.Net] has joined #code
15:16 mode/#code [+qo Vornicus Vornicus] by ChanServ
17:01 thalass [thalass@Nightstar-h1qmno.eastlink.ca] has joined #code
17:01 mode/#code [+o thalass] by ChanServ
17:10 celticminstrel [celticminst@Nightstar-j5b.hci.93.65.IP] has joined #code
17:10 mode/#code [+o celticminstrel] by ChanServ
17:35
< grindhold>
does anyone of you know something similar to distrowatch.com just for programming languages instead of *nix-distros?
17:36
<&McMartin>
For tracking things like gcc patches, or revisions to Rust/Go/etc?
17:37
<&McMartin>
Either way I am in fact unaware of such a thing but would be interested to know of one if it does exist.
17:38
< grindhold>
most interesting would be the popularity ranking. you could discover cool new uprising languages faster e.t.c
17:39
< grindhold>
i know, github does research on language popularity, but only on their own data.
17:39
<&McMartin>
Mmm. Popularity isn't really the metric I'd want for that, but others are of course harder to measure mechanically.
17:39
<&McMartin>
There's also TIOBE's list.
17:39
<&McMartin>
http://www.tiobe.com/index.php/content/paperinfo/tpci/index.html
17:42
< grindhold>
McMartin: nice for starters, but their way to calculate the popularity index is apparently pretty corporate-centered
17:42
<&McMartin>
"The ratings are based on the number of skilled engineers world-wide, courses and third party vendors."
17:42
<&McMartin>
Yeah
17:43
<&McMartin>
But, um, grass-roots bandwagon languages have this unnerving tendency to be like PHP
17:44 * McMartin reads the Go and Rust blogs atm and has flipped through the Swift books, and lately that's been about it for him and newish stuff.
17:44
<&jerith>
There seem to be a lot of new languages popping up these days.
17:44
<&McMartin>
Yeah. A lot of them are borrowing strongly from the ML tradition too.
17:45
< grindhold>
jerith: exactly this is why nowadays i wish for such a site
17:45
<&McMartin>
... F# at 14? Dang.
17:45
<&McMartin>
I need an excuse to write something in that.
17:46
<&McMartin>
(Because I haven't done anything serious in an ML dialect since like 2006)
17:46
<&McMartin>
Hee hee, Inform is on the list
17:46
<&McMartin>
Right between Go and PL/I
17:47
<&McMartin>
... I have dealt with all three of these languages in the past five months.
17:47
<&jerith>
Delphi and VB are on the rise?
17:47
< grindhold>
five years ago, there was the magic scripting square made up of perl, python, ruby and php, then there was c, c++, some functional stuff noone cared about and the proprietary abominations, apple and m$ created out of C... today, there is so much more and this trend is still intesifying
17:48
<&McMartin>
jerith: Would not be surprised if VB.NET became more relevant as Win7 gets more widely deployed and it can be used to implement PowerShell extensions
17:48
<&McMartin>
You're still "supposed" to do that with C#, but bindings is bindings
17:48
<@Alek>
win7 is getting replaced this fall. so's w8/w8.1
17:48
<@Alek>
¬_¬
17:49
< grindhold>
wait what?
17:49
<@Alek>
with wx being free to upgrade for a year, a LOT of people will take advantage, I'm sure.
17:49
< grindhold>
i thought windows 8 came out a year ago or so
17:49
<&McMartin>
No, I mean "people are finally not having to restrict themselves to stuff that runs on XP"
17:49
<@Alek>
err. more like 3.
17:49
<&McMartin>
I'll be astonished if corporate uptake of Win10 is rapid.
17:49
<@Alek>
3 years this august, yes.
17:50
<@Alek>
McM: free. upgrade.
17:50
<&McMartin>
Only if your IT staff works for free.
17:50
<@Alek>
although the more hardcore will wait to test it out exhaustively, sure.
17:50
<&McMartin>
Based on how long it took Win7 to roll out, this is a process that takes approximately seven years.
17:50
<@Alek>
haha.
17:50
<&jerith>
Alek: That'll probably increase uptake, but it's unlikely to be dramatic in the corporate market.
17:51
<@Alek>
w7 released in 09.
17:51
<&McMartin>
Yeah, OK, so, four years.
17:51
<~Vornicus>
All the software on all the computers in my office is cheaper than one employee.
17:52
<&McMartin>
The corporate licensing is different anyway
17:52
<&McMartin>
I'm not sure how that's going to look.
17:52
<&McMartin>
Post-W10, I think the upgrade streams will be faster.
17:52
<@Alek>
and then there's federal, I'm pretty sure I've seen some still using 95.
17:52
<&McMartin>
But the jump *to* it...
17:52
<&jerith>
All the hardware and software together on all my computers is cheaper than one employee.
17:52
< grindhold>
Alek: lol. time flies. i quit windows when vista came out.
17:53
<&jerith>
Cheaper than two months of an employee, if that employee is me.
17:53
<&McMartin>
7 was basically the fixed version of Vista
17:53
<&McMartin>
There is some speculation that Win8/8.1 are the equivalent of Vista for whatever comes next
17:53
<~Vornicus>
I know a person running a computer that's 23 years old.
17:53
<&McMartin>
This may be characterized as "hope"
17:53
<@Alek>
there WAS speculation. before w8 released.
17:54
<@Alek>
I personally think 8.1 was the fix.
17:54 * McMartin nods
17:54
<@Alek>
well, 8.2 would have been better, but it seems to have evaporated.
17:54
<&jerith>
I believe W8 still follows the "avoid every second version" rule.
17:54
<&McMartin>
jerith: That's a pretty recent rule, unless you're slipping an extra version between 2k and XP
17:54
<@Alek>
well... it's not that hard a rule.
17:55
<&McMartin>
Also, most of my Windows experience is at the "man 2" level these days, which means that 7, 8, and 8.1 were 6.1, 6.2, and 6.3 respectively
17:55
<@Alek>
hey now, 2k was corporate, I'm sure you mean 98, then ME, then XP
17:55
<&jerith>
I don't think 2k counts, being a server OS rather than a desktop OS.
17:55
<&McMartin>
Alek: I don't, because XP is a direct code descendant of 2k and both run on the NT 5 kernel.
17:56
<@Alek>
and uh. 9x was kinda iffy. 95 was the poor version, 98 was better, 98SE was best.
17:56
<&McMartin>
This may be that bias I just stated popping up~
17:56
<&McMartin>
The 9x series, including Me, were not modern OSes in the sense we'd use now, as they did not consistently implement per-process memory protection.
17:56
<@Alek>
... you're including ME in 9x???
17:57
<&McMartin>
... yes?
17:57
<&jerith>
So 98, ME, XP, Vista, 7, 8.
17:57
<~Vornicus>
same tech
17:57 * Alek snerks.
17:57
<&McMartin>
Yeah, same tech.
17:57
<&McMartin>
I'm honestly not clear on the metric you're using.
17:57
<&McMartin>
I'm using "OS core development lines"
17:57
<&jerith>
The metric I'm using is "marketed as a new major version".
17:57
<&McMartin>
(Admittedly, by this standard, that means 95 isn't a 9x, it's MS-DOS)
17:58
<@Alek>
jerith: I repeat, it went 95-98-98SE, ME-XP, Vista-7. and I consider 8-8.1. if you go by poor-good.
17:59
<&McMartin>
Right. I keep two lines of descent, MS-DOS-3.1-95-98-98SE-ME, and NT4-2K-XP-Vista-7-8-8.1
17:59
<&McMartin>
Normally I would toss ME from the list the same way I have tossed Windows for Workgroups
17:59
< grindhold>
http://arstechnica.com/information-technology/2012/05/no-cost-desktop-software-d evelopment-is-dead-on-windows-8/ lel
18:00
<&jerith>
The only versions of Windows I've actually used for any length of time are 3.1, 95, 98, and XP.
18:00
<&jerith>
Then I discovered that I can be much more productive on unix-based systems.
18:00
<&McMartin>
There was a thing that was discussed around Win8 and Metro that doesn't seem to have materialized, which was using CSS/JS to command native widget sets for desktop apps.
18:01
<&McMartin>
That would have been interesting from an "increase your developer pool" standpoint
18:01
<&McMartin>
(I am admittedly imagining something like Qt Quick but without deploying 50 megs of supporting DLLs)
18:02
<@Alek>
McM: what about 1.0, 2.0, 2.1, and 3.0 and 3.11?
18:02
<@Alek>
oh wait, 3.11 is WfW.
18:02
<@Alek>
but I considered it better than 3.1 anyway.
18:02
<&jerith>
Alek: Those are all along the MS-DOS-3.1 line.
18:02
<~Vornicus>
...I've never actually used any non-NT Windowses, now that I think about it
18:02
<@Alek>
not that I had enough experience. :/
18:03
<&McMartin>
Apparently MS internally nicknamed it "Windows for Warehouses"
18:03
<&jerith>
3.11 was basically just 3.1 with a working network stack, IIRC.
18:03
<&McMartin>
That sounds about right, though the network stack it used was not actually what the companies at the time needed.
18:03
<@Alek>
it was called Windows for Workgroups, and McM said he was discarding it. XD
18:04
<&McMartin>
Right. I suppose the filter I'm applying here is "was it aggressively marketed to consumers" and ANAICT all versions of Windows pre-3.1, including 3.0 itself, were not.
18:04
<&McMartin>
And 2K saw some consumer use, ANAICT.
18:04
<&jerith>
I thought ME was marketed fairly aggressively.
18:04
<&McMartin>
Note I'm also not listing the three versions of NT prior to NT4
18:04
<&jerith>
But I may be misremembering.
18:05
<&McMartin>
It kind of was, but I also got this impression that ME was being released under duress
18:05
<&jerith>
I think 2k saw consumer use because it was so vastly superior to 9x/ME.
18:05
<&McMartin>
That they really wanted people to go to an OS that didn't have every process share half its address space with every other process and the kernel.
18:05
<~Vornicus>
My first windows machine was 2k
18:05
<&McMartin>
viz. 2K
18:06
<@Alek>
fair enough.
18:06
<&McMartin>
But 2K being the a major revision to a kernel that wasn't actually compatible with the software everyone actually ran kind of complicated that
18:06 Turaiel[Offline] is now known as Turaiel
18:06
<&McMartin>
So XP had a minor revision that drastically boosted compatibility and then it stayed entrenched for 13 years.
18:06
<@Alek>
ME was marketed, sure, but it was also utter crap and most customers saw that fairly soon. as I recall.
18:06
<@Alek>
my first windows machine was 98se.
18:07
<&jerith>
I also recall XP being actively worse than 98 in a number of ways rather than just increased bloat while failing to be better.
18:07
<&jerith>
Err, ME.
18:07
<&McMartin>
(Note that this story also continues - Vista was NT 6.0, which broke everything so nothing ran, even though all their tests showed that on the whole the populace found the user experience better. Windows 7, 8, and 8.1 are all minor revisions that don't break earlier drivers and refine or rewrite the user level more)
18:07
<&jerith>
Not XP.
18:08
<@Alek>
oh whew. I was about to say.
18:08
<@Alek>
8 is NOT a minor revision, I'd say... but eh.
18:08
<&jerith>
Alek: 8 is a minor revision under the hood.
18:08
<@Alek>
hah
18:08
<&McMartin>
It's a minor revision at the kernel level. This is the semver definition on "minor revision"
18:08
<@Alek>
ok.
18:08
<&jerith>
It had a whole new UI layer.
18:09
<&McMartin>
Major revisions at the kernel level mean all manufacturers have to rewrite all their drivers
18:09
<&jerith>
Like swapping Gnome for Unity on Linux.
18:09 * Alek learns.
18:09
<&McMartin>
That's why Nothing Fucking Worked on 2K or Vista.
18:09
<&McMartin>
semver is great. I hear MS is abandoning in Win10, this makes me sad.
18:09
<&McMartin>
Then again, no other OS kernel anywhere else at all uses semver.
18:09
<@Alek>
so what's semver?
18:09
<&McMartin>
"Semantic Versioning"
18:10
<&McMartin>
An informal standard for what changes in x, y, z should mean in version numbers that are of the form x.y.z
18:10
<@Alek>
so I see. thanks.
18:10
<&McMartin>
Basically, a change in x ("major revision") means that it's not backwards compatible, so old stuff breaks
18:10
<&McMartin>
Change in y ("minor revision") adds new functionality but doesn't alter behavior of old code in any way
18:11
<&McMartin>
Change in z ("point revision" or "patch revision") doesn't alter the API at all and is just bugfixes, usually
18:11
<&jerith>
Alek: X.Y.Z, where changes to X indicate "lots of stuff breaks", changes to Y indicate "new stuff, but old code should still work", and changes to Z indicate "bugfixes and things".
18:11
<@Alek>
yet windows has Compatibility Mode, so doesn't that bypass semver, sorta?
18:11
<&McMartin>
Well, they're orthogonal
18:11 Kindamoody|out is now known as Kindamoody
18:11
<&McMartin>
Win8.1 included I'm Going To Turn This Car Right Around mode, which definitely complicates it
18:11
<@Alek>
hwa
18:12
<&McMartin>
Win8.1 and later actively lie in calls to GetVersionEx and similar unless your app manifest proves that you are aware of the later OSes existing.
18:12
<&McMartin>
semver would still be useful in the presence of this because it answers the question of "how much effort do we expect it to be to start being aware of this new OS?"
18:12
<@Alek>
o_o
18:12
<&McMartin>
Yeah, that was my reaction too
18:13
<&McMartin>
The pages on MSDN have been toned down over later releases now that this is Just A Thing
18:13
<&McMartin>
But the docs for GetVersionEx from pretty much XP on were full of "god damn it, programmers, check for the version being AT LEAST what you need, not EXACTLY what you need."
18:13
<&McMartin>
And apparently nobody ever stopped
18:13
<@Alek>
... no wonder it (mostly) seems to be easier to run older games than even on 7.
18:13
<&McMartin>
Ergo, they Turned This Car Right Around and introduced Lying To Your Face mode.
18:14
<&McMartin>
Oh, no, that's just refinements.
18:14
<&McMartin>
The lying starts at 8.1, and 8.1 falsely reports as 8 without the app manifest
18:14
<&McMartin>
If you set compat mode in any version, it should say "I'm the version I'm pretending to be now"
18:14
<&McMartin>
That would be an important part of compat!
18:15
<@Alek>
so far I've had one game fail to run, even with modernization by gog and compatibility mode. Septerra Core, which I'm sad about.
18:15
<&McMartin>
But IIRC the early compat modes basically just shimmed the syscalls a bit, a la DSfix
18:15
<&McMartin>
While some versions of Win8 would run A Full Copy of XP under Hyper-V virtualization, but I don't think that happens all the time because that's, uh, an expensive proposition
18:16
<@Alek>
dang.
18:17
<&McMartin>
... huh, now that I check, the first released NT was NT3.1
18:18
<&McMartin>
And apparently that line would start OS/2-NT3.1-NT4-2K-etc
18:19
<&jerith>
VMS-OS/2-NT3.1-...?
18:20
<&jerith>
Or was VMS just a spiritual ancestor rather than sharing any implementation?
18:20
<&McMartin>
I believe the latter.
18:20
<&McMartin>
... actually, it better be the latter or IBM was commiting super-flagrant corporate espionage~
18:20
<&McMartin>
Wiki also notes that NT started out as NT OS/2 3.0 and thus the legend about rot1'ing VMS to get the name WNT cannot be true
18:21
<&McMartin>
=(
18:23 Vash [Vash@Nightstar-uhn82m.ct.comcast.net] has joined #code
18:23 mode/#code [+o Vash] by ChanServ
18:25
<&McMartin>
Re: Septerra Core specifically, though, the LP notes some stuff about crashiness
18:25
<&McMartin>
http://lparchive.org/Septerra-Core/
18:25
<&McMartin>
And by "the LP" I apparently mean "ToxicFrog's LP"
18:28
<&ToxicFrog>
The crashes I'm aware of relate to the Sonic Emitter and certain special attacks, though
18:28
<&ToxicFrog>
I have no experience with it not working at all, and I'm pretty sure the GOG version already comes with 1.04 installed
18:31
<&ToxicFrog>
McMartin: that was my first -- and to date only -- non-IF LP.
18:31
<&ToxicFrog>
I have a soft spot for Septerra Core despite the glacial pace of combat and the way-too-long dungeons.
18:31
<&ToxicFrog>
Although I did use the cheat codes heavily for that LP~
18:44
< [R]>
I got stuck on that stupid wolf trail.
18:44
< [R]>
Couldn't back up, and my save was for the start of it, and I was already in a bad position.
18:45
< [R]>
IIRC fleeing the fights was difficult too.
18:46
<&ToxicFrog>
I'm not sure how fleeing battle works. It's a roll against something. Probably Speed.
18:47
<&ToxicFrog>
IIRC, the game automatically saves whenever you leave the world map, so it should be basically impossible to get stuck like that -- you can always rewind to where you entered the dungeon.
18:57 Turaiel is now known as Turaiel[Offline]
19:03 Derakon[AFK] is now known as Derakon
19:11 Datawalke [Datawalke@Nightstar-kt8k7i.com] has quit [Connection closed]
19:54 himi [fow035@Nightstar-v37cpe.internode.on.net] has quit [Ping timeout: 121 seconds]
20:24 Vash [Vash@Nightstar-uhn82m.ct.comcast.net] has quit [[NS] Quit: Quit]
20:35
<@Alek>
Re: Septerra Core, I tried all kinds of fixes from Google, none of them worked. it failed either before the main menu, or on it, I forget which precisely. the intro movies would play, but that was all.
20:41 Checkmate [Z@Nightstar-484uip.cust.comxnet.dk] has quit [Connection reset by peer]
20:41 Checkmate [Z@Nightstar-484uip.cust.comxnet.dk] has joined #code
20:41 mode/#code [+o Checkmate] by ChanServ
21:17
<&ToxicFrog>
Alek: yeah, I've got nothing.
21:56
< RchrdB>
Derakon: with rsync, "-H" preserves hardlinks.
21:58
< RchrdB>
On the "free upgrade" front: we sell web apps to UK government and have tons of users on I
21:59
< RchrdB>
on IE 7 because that's the version that came with Vista, which aiui is still officially supported
22:00
< RchrdB>
(Never mind that you CAN upgrade that to at last 9, maybe 10. Local govt IT are fucking addicted to using the shittest software that isn't actually forbidden :p)
22:01 Derakon [chriswei@Nightstar-5fqf0m.ca.comcast.net] has quit [[NS] Quit: Leaving]
22:02
<@TheWatcher>
RchrdB: not just local govt
22:02
<@TheWatcher>
Universities are as bad.
22:02
< RchrdB>
I really don't get it
22:03
< RchrdB>
By the time you've decided to upgrade to vista, ie9 is out for it already
22:03
< RchrdB>
Standardise on and test your apps against that! Fucking dickheads.
22:18
<&jerith>
RchrdB: At the time, the apps were probably written for IE6 and didn't work in standards-compliant browsers.
22:19
< RchrdB>
They should've broken when the mandatory upgrade past ie6 came, then
22:20
<&jerith>
Maybe they broke less in IE7?
22:21
<~Vornicus>
IE9 is when IE started getting standards compliant in a lot of ways
22:21
<&jerith>
Maybe someone said "I have IE7 here, so that's what we're all going to use".
22:21
<~Vornicus>
before then they used a different way of attaching events, and lots of the CSS didn't work.
22:24
<~Vornicus>
It is a hell of a lot easier to support "chrome, ff, ie9+" than "chrome, ff, ie8+"
22:39
<@Alek>
yup.
22:40
<@Alek>
ie9+ is actually more stable, ofttimes, than current FF or Chrome.
22:46 Vash [Vash@Nightstar-uhn82m.ct.comcast.net] has joined #code
22:46 mode/#code [+o Vash] by ChanServ
23:03
< RchrdB>
Vornicus: I am intimately aware of this. It is why I despise ours users' IT depts so often.
23:43
<@TheWatcher>
I'm lucky, I was able to say IE9+ and have it stick, but I think that was only because im the School of CS
23:44
<@TheWatcher>
When CODICIL WINTER HILL starts getting used outside it later this spring, I expect I'm going to have to deal with a lot of "Why won't this work in IE7?!"
23:47 thalass [thalass@Nightstar-h1qmno.eastlink.ca] has quit [Ping timeout: 121 seconds]
--- Log closed Mon Feb 23 00:00:06 2015
code logs -> 2015 -> Sun, 22 Feb 2015< code.20150221.log - code.20150223.log >

[ Latest log file ]