IRC Logs

Time Nick Message
22:51 rangi yup pretty cool :)
22:50 cait nice!
22:50 rangi http://jon.geek.nz/2012/06/115/
21:15 huginn rangi: The current temperature in Wellington, New Zealand is 5.0Â°C (9:00 AM NZST on July 01, 2012). Conditions: Mostly Cloudy. Humidity: 93%. Dew Point: 4.0Â°C. Windchill: 4.0Â°C. Pressure: 30.33 in 1027 hPa (Steady).
21:15 rangi @wunder nzwn
21:10 huginn cait: The current temperature in Konstanz, Germany is 19.0Â°C (11:00 PM CEST on June 30, 2012). Conditions: Partly Cloudy. Humidity: 76%. Dew Point: 16.0Â°C. Pressure: 29.83 in 1010 hPa (Rising).
21:10 cait @wunder Konstanz
18:35 cait bislang macht es noch sinn
18:34 mbalmer_ naja, das meiste davon ist eher unrecht als recht...
18:33 cait yes, urheberrecht, patentrecht, vertragsrecht, softwarerecht... my new course for the distance study
18:32 cait drojf: laptop shut down
18:31 cait hm
18:08 drojf it law? like urheberrecht? nice weekend reading ;)
18:06 drojf how far is tagerwilen from konstanz? or is it a part of it?
18:06 cait ok, back to reading it law :)
18:06 cait now
18:06 cait yep
18:06 drojf haha
18:06 drojf just wait a few minutes
18:06 huginn cait: The current temperature in Konstanz, Germany is 24.0Â°C (8:00 PM CEST on June 30, 2012). Conditions: Thunderstorms and Rain. Humidity: 51%. Dew Point: 16.0Â°C. Pressure: 29.80 in 1009 hPa (Falling).
18:06 cait @wunder Konstanz
18:06 cait still not really true :)
18:04 drojf a drastic weather change in 15 minutes ;)
18:01 huginn cait: The current temperature in Taegerwilen, Taegerwilen, Germany is 24.6Â°C (8:00 PM CEST on June 30, 2012). Conditions: Mostly Cloudy. Humidity: 69%. Dew Point: 19.0Â°C. Pressure: 29.83 in 1010 hPa (Steady).
18:01 cait @wunder Konstanz
18:00 drojf :D
17:51 cait clear? there is storm and rain!
17:50 huginn cait: The current temperature in Taegerwilen, Taegerwilen, Germany is 28.0Â°C (7:45 PM CEST on June 30, 2012). Conditions: Clear. Humidity: 65%. Dew Point: 21.0Â°C. Pressure: 29.83 in 1010 hPa (Steady).
17:50 cait @wunder Konstanz
14:16 * mtompset waves.
14:16 * mtompset smiles.
14:15 cait autocomplete failing me :)
14:15 cait mtompset:
14:15 cait good night mt
14:15 mtompset Well, have a good day (24 hour period), folks. I need to get to bed at a regular hour.
14:13 huginn drojf: The current temperature in Prenzlauer Berg, Berlin, Germany is 27.8Â°C (3:43 PM CEST on June 30, 2012). Conditions: Partly Cloudy. Humidity: 41%. Dew Point: 13.0Â°C. Pressure: 29.92 in 1013 hPa (Steady).
14:13 drojf @wunder berlin, germany
14:13 cait ouch
14:13 mtompset I'm glad moisturizers exist!
14:13 mtompset What's worse is when my term is up... going back to Canada... my skin dries up and starts to crack, because it has adapted to the more humid climate.
14:12 cait well, I am not used to it and it's too warm for my liking
14:12 mtompset But I know what you mean. :)
14:12 cait true :)
14:11 cait they said they will be, but it sure doesn't look like it
14:11 mtompset No, the problem HERE is humidity. :)
14:11 cait hope we get some rain later tonight
14:11 cait the problem here is humidity
14:11 jcamins Yup.
14:10 huginn mtompset: The current temperature in Manila, Philippines is 24.0Â°C (10:00 PM PHT on June 30, 2012). Conditions: Light Rain. Humidity: 100%. Dew Point: 24.0Â°C. Pressure: 29.86 in 1011 hPa (Rising).
14:10 mtompset @wunder MNL
14:10 jcamins I'll check that I got my math right.
14:10 mtompset 36C?
14:10 jcamins Well, 8 degrees Celsius, I guess.
14:10 jcamins mtompset: yeah, but it's going up another 15 degrees.
14:09 mtompset One has a heat advisory, the other doesn't, but the Germany one is hotter.
14:09 * mtompset laughs.
14:08 huginn jcamins: The current temperature in Taegerwilen, Taegerwilen, Germany is 30.7Â°C (4:05 PM CEST on June 30, 2012). Conditions: Clear. Humidity: 56%. Dew Point: 21.0Â°C. Pressure: 29.86 in 1011 hPa (Steady).
14:08 jcamins @wunder Konstanz
14:08 huginn jcamins: The current temperature in APRSWXNET Jackson Heights NY US, Corona, New York is 28.9Â°C (9:32 AM EDT on June 30, 2012). Conditions: Scattered Clouds. Humidity: 48%. Dew Point: 17.0Â°C. Pressure: 29.75 in 1007 hPa (Rising). Heat advisory in effect from 1 PM this afternoon to 7 PM EDT this evening...
14:08 jcamins @wunder 11375
14:08 jcamins All the other languages are translated from English.
14:08 jcamins Oh, I see what you're asking about.
14:08 mtompset Well, it's not strange...
14:07 mtompset It's strange that no one really noticed this, because staff-global.css is over-ridden by login.css at the login screen, but becomes the default once you are logged in.
14:06 mtompset Thankfully, no strings to translate in the patch.
14:05 jcamins There should not be any linguistic content in the CSS files.
14:05 jcamins The inc files will be translated.
14:05 mtompset or does the installing of other languages convert the English CSS and INCLUDE files?
14:04 mtompset languages make a patch like that harder.
14:03 huginn 04Bug http://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=8339 enhancement, P5 - low, ---, oleonard, Needs Signoff , div id="login" should be id="login_controls" in header.inc
14:03 mtompset bug 8339
14:03 mtompset There generated a partial patch for a styling enhancement for the staff client. :)
13:55 jcamins Right.
13:53 mtompset Ah... Business owners get "no" holidays. :(
13:50 jcamins mtompset: I have my own business.
13:49 mtompset Never heard of them. Can't comment, sorry.
13:49 Waylon "LVM is good, there is another option for backup from r1soft.com available ." any comments about r1soft?
13:48 mtompset Long weekend?
13:40 jcamins Not as such.
13:38 mtompset Do you get July 4 off, jcamins?
13:33 mtompset There are things you miss when you are not in your home country.
13:33 mtompset Dear Timmies, please open one in the Philippines. Preferably in Santa Cruz, Laguna. :) Thanks, Mark
13:32 mtompset Though a doughtnut and coffee/tea/iced cappucino from Tim Hortons would be good too. But you probably don't have one locally there.
13:30 mtompset Pizza and coke is the breakfast of coding champions.
13:30 mtompset Good Lord! Drink some caffeine!
13:30 cait mmh brunch?
13:29 jcamins No breakfast or caffeine yet today.
13:29 jcamins Sorry. :)
13:29 jcamins I meant branch.
13:27 mtompset you said, "does not apply them to any master"
13:27 mtompset That's what I am asking you.
13:27 jcamins You should only have one "master" branch.
13:27 jcamins What do you mean any master?
13:27 mtompset any master?
13:25 * jcamins just learned this.
13:25 jcamins cait's workflow is to do a git pull on the mater branch, then rebase all the topic branches.
13:25 jcamins git pull merges in the changes.
13:24 jcamins git fetch downloads the changes and does not apply them to any master.
13:24 mtompset What is the difference between git fetch and git pull?
13:24 mtompset when I select my master, will it have changed? or do I still have to git fetch?
13:23 jcamins Right.
13:23 mtompset so if my master is out of date, and I already made a patch branch.... git fetch, git rebase before the edit whatever?
13:22 jcamins You don't need to rebase your master branch at all.
13:21 mtompset so on my master before I make a patch branch, git fetch, git rebase, then git checkout -b bug_XXXX origin, edit whatever, git add whatever, git commit, git format-patch origin -o ~
13:20 jcamins Use git fetch, then git rebase
13:20 jcamins Moreover, don't use git pull.
13:20 jcamins No.
13:19 mtompset if I git pull with my master branch selected, will that affect my other branches?
13:19 mtompset plus my master branch.
13:18 mtompset I have all these branches now.
13:18 mtompset okay... last question, and then I'll probably start shutting down for the evening.
13:18 mtompset YAY!
13:18 mtompset and that isn't inside my git repositories.
13:16 mtompset I can put it there.
13:15 mtompset I configured the default apache to run on 4080.
13:15 mtompset Nevermind... I have an idea.
13:15 jcamins That's okay.
13:15 jcamins Yes.
13:14 mtompset But it will list it as a modified file.
13:13 jcamins mtompset: you just copy the file and don't add it to git.
13:13 mtompset How would I do something similar on my DEV install without triggering that extra file to upload?
13:12 mtompset On our 3.6.3 production system I copied staff-global.css and tweaked it.
13:11 mtompset I am much happier with the styling layout of 3.9.x than 3.6.3
13:07 mtompset I think browsers will let me. :)
13:07 mtompset I'm overriding height.
13:07 mtompset Yes, it did. That's why I jumped for joy. :)
13:07 jcamins In practice there are some things that certain browsers just won't override.
13:06 jcamins mtompset: in principle if you add !important to your alternatives, that should fix it.
13:06 mtompset If they don't change, the rsync is minimal.
13:06 mtompset rsync
13:04 Waylon forget the db dump, mtompset .. try 250 gb of pdf files.
13:04 * mtompset jumps for joy, "!important overrides!"
13:03 mtompset How do I override a !important?
13:03 mtompset I think I figured out why: !important.
13:02 mtompset Particularly the DB dump. :)
13:02 Waylon yeah. going to be a bugger to back up. but i guess thats not my responsibility anymore eh?
13:02 cait too warm for me already
13:01 * cait does not want to go out there
13:01 huginn cait: The current temperature in Taegerwilen, Taegerwilen, Germany is 30.9Â°C (3:00 PM CEST on June 30, 2012). Conditions: Clear. Humidity: 56%. Dew Point: 21.0Â°C. Pressure: 29.89 in 1012 hPa (Steady).
13:01 cait @wunder Konstanz
13:01 jcamins Waylon: so? As long as it works.
13:01 Waylon ings simple..." -sighs- I guess we're going to have a Goliath.
13:01 Waylon sysadmin wants to create a single Golaith of a server. when Davids are enough. I also raised the point about server failure/boot fail "based from our exp. the only thing likely to fail is the Hardisk, if we provide redundancy in this level we have nothing much to worry. in our present XCP / VM / LB / glusterFS / sync scripts / cronbal ..... There is more chance for something to fail than in a single server.. The best way is to keep th
12:59 jcamins Make T::T URLencode it.
12:59 cait I tried recently and had a lot of problems with encoding :(
12:59 jcamins mailto: link.
12:59 cait and how do you get the informatoin into the mail?
12:59 jcamins mtompset: h1#logo > a
12:58 cait they have to be at a computer with a mail client
12:58 cait yeah but there are problems with that
12:58 cait ah
12:58 jcamins cait: for non-authenticated users to send carts.
12:58 cait jcamins hm?
12:58 Waylon Hmmmmmmmmm..
12:58 jcamins We could use a mailto: link, I suppose.
12:57 jcamins That was a guess.
12:56 mtompset Where is that in staff-global.css?!
12:56 jcamins mtompset: because the text height is set on h1#logo > span.
12:55 mtompset jcamins, why does my inline h1#logo override the background image fine, but my h1#logo a doesn't override the text height?
12:52 jcamins If you find a book you're interested in, you can add it to your cart and e-mail it to yourself.
12:52 cait or wahtever
12:52 cait or a friend
12:52 cait or you can send yourself a cart with a list of books you want to get from the library
12:52 cait one of our libraries uses it to have students mail them books they want them to get ready for them
12:52 jcamins mtompset: in the OPAC.
12:52 cait I think there are different use cases
12:51 mtompset who is the send a cart message for?
12:51 jcamins Good time of day.
12:50 * jcamins is not sure how he feels about this.
12:50 jcamins You're right. You cannot send a cart if you're not logged in in 3.6.
12:50 mtompset Greetings, jcamins.
12:50 cait that#s why inline styles are evil
12:50 cait yes, an inline style should override everything else
12:50 mtompset Shouldn't the inline over-ride the staff-global.css period?!
12:49 mtompset but text height? Noooooo.... I don't understand cascading order rules.
12:49 mtompset overrode the logo no problem.
12:49 * mtompset grumbles about cascading.
12:15 mtompset Okay... misread the CSS and after reading W3 documentation... there's only TWO h1#logos that I don't understand. :)
12:11 mtompset Greetings, matts.
12:09 cait hm never done much tweaking of the staff side
12:08 mtompset well, tweaked a copy of, and used the syspref to point at it.
12:08 mtompset last time, I just got it working without thinking.
12:07 mtompset but I figured I'd read the staff-global.css which I tweaked before and try to understand it first.
12:07 mtompset oleonard pointed me at some sysprefs for that.
12:05 mtompset the OPAC side styling worked rather well without tweaks.
12:04 mtompset This is the staff client I'm trying to theme.
12:04 mtompset firebug, yes.
12:04 cait and firebug is your friend
12:04 cait overwrite the statements by using opacusercss or one of the other related system preferences
12:04 cait if you want to change something
12:04 cait don#t change anything in there
12:03 cait anyway
12:03 cait if the seelctors are not the same, there will be a reason I guess :)
12:03 mtompset Quite confusing for a lay person in CSS.
12:02 mtompset And some of them are the same, but different values.
12:02 mtompset No.
12:02 cait and the selectors are the same for all 4?
12:02 mtompset Because I don't ever remember needing multiple sections, but perhaps CSS has changed that there is a reason.
12:01 cait um, probably not :)
12:00 mtompset Do you understand why h1#logo is given 4 sections in staff-global.css?
12:00 cait I haven't used gradients much yet
12:00 cait sometimes
12:00 mtompset Do you understand CSS, cait?
11:58 mtompset I'm trying to understand the staff-global.css
11:58 mtompset Is anyone who is a CSS expert here?
11:56 cait heh
11:49 mtompset My styling nightmares with 3.6.3 will be lessened.
11:49 mtompset Sweet! Thank you! 3.9.X has div.gradient!
10:42 cait good night rangi
10:41 mtompset Have a good night.
10:40 mtompset computer geek of some kind. :P
10:37 rangi and now im going to bed
10:37 rangi and it's a sysadmins job to look after servers anyway, not a programmer :)
10:36 mtompset Well, yes, I agree.
10:36 mtompset no, because people are multiskilled.
10:34 cait or doesn't always work
10:33 cait I think dividing into librarians and programmers don't always work
10:33 mtompset I sense there's more than the simple dichotomy I provided.
10:32 cait well, yes
10:32 mtompset so you are more of a librarian type than a programming type.
10:31 cait my coworkers take care of that :)
10:31 mtompset then who set up your koha?
10:31 cait mtompset: can't access them - sorry
10:30 mtompset that should tell you total, used, and free memory for your system.
10:30 mtompset cait, if you have a command prompt on your koha server... free -m
10:30 mtompset I just VM'd the same data on a dual core dual threaded ATOM at 1.5GHz and gave it 1GB... it still seems like not quite enough.
10:28 mtompset We are looking to increase the ram.
10:28 mtompset But we are only using 512MB, and it is awful!
10:28 mtompset we do.
10:21 cait sorry, no idea :)
10:21 Markus2 ?
10:21 Markus2 hmm.. whats your cpu/ram/disk setup
10:20 Markus2 any IO bottleneck?
10:20 Markus2 bummer.
10:20 cait yeah, but I am not the server administrator - so can't tell you much abou tit
10:19 Markus2 you cait?
10:19 cait hm yes
10:18 Markus2 does anyone use koha in a production VM'ed enviroment?
09:53 mtompset He helped me get my DEV install up and running.
09:52 cait he is a cataloging deity at least :P
09:52 mtompset With that sort of praise, jcamins sounds like a Koha deity. :)
09:52 rangi ed summers who works at library of congress and who wrote MARC::Record could tell us all :)
09:50 * mtompset laughs, "Me too!"
09:50 cait lol
09:49 drojf oh that was rangi. i read his sentence in cait's voice :)
09:49 cait i am not really a marc21 expert, jcamins could probably tell you a lot more about that :)
09:49 cait totally different fields
09:49 cait yes, rangi is right
09:49 cait sorry mtompset overread your questions
09:48 matts there's a lightweight tutorial here : http://search.cpan.org/~marcc/marc-mir-0.0/lib/MARC/MIR/Tutorial.pod
09:48 rangi mtompset: no, those are whole different fields, some of which are repeatable, some are repeatable fields with repeatable subfields even
09:47 mtompset That much is true. :)
09:47 matts cait, yeah you're right... But marc::mir is really a young project (v0.0, right ?) and i don't know how far the implementation has been done... But the concept behind it is interesting
09:47 drojf that's why we have these crazy marc things in the first place ;)
09:47 mtompset but would that be an example of a repeatable subfield?
09:46 cait and so on
09:46 cait and corporates and persons go in different fields
09:46 cait so you would have 2 different fields
09:46 cait and there could be more than one author
09:46 cait it could also be a corporate
09:46 cait an author must not be a person
09:46 cait well
09:45 mtompset I'm trying to learn. :)
09:45 mtompset could you give an example, cait?
09:45 cait hi Markus2
09:44 cait because there is more than one field for that
09:44 cait matts: but it's not as easy as that with marc
09:43 mtompset so that it would prevent invalid data from being added?
09:43 matts and then manipulate your records like : author="Author" and not XXX$X = "Author"
09:43 matts yeah, but the fun stuff about this module is that you can define a bibliographic framework
09:42 mtompset Right.
09:42 drojf its just about "there is a field, it has subfields" i guess. it does not care what you use the fields for
09:42 mtompset I can generally agree with that.
09:42 matts you *put stuff* in it
09:42 matts Hmm... It's just a way to manipulate records... What you put in it doesn't really matter, does it ? It's *how* you in it that is relevant, no ?
09:41 mtompset Because all the processing logic for something like KOHA is marc-dependent.
09:40 mtompset How can it be?
09:40 mtompset really?
09:39 matts mtompset, it's marc-agnostic
09:39 mtompset or ::UNIMARC
09:39 mtompset or ::NORMARC
09:39 mtompset Then people could optimize MARC::Record::MARC21
09:39 mtompset If I were to implement things from scratch. I'd have MARC::Record... but it would be smart enough to use MARC::Record::<type>
09:38 cait I am not sure using a new module will solve all problems
09:37 mtompset But is it mostly aimed at UNIMARC? or does it also incorporate MARC21, NORMARC, etc.?
09:37 matts MARC::MIR is meant to be simple and smart
09:36 matts MARC::Record is definitively not the more elegant way to handle MARC records, right ?
09:36 cait same question here :)
09:35 mtompset I am sure there are problems with current methods, but why a new module? (I'm asking to learn, not to criticize)
09:34 matts (and post the link here)
09:34 matts I'll ask eiro for his slides
09:33 matts It's a new module for MARC records handling
09:30 mtompset http://search.cpan.org/search?query=MARC%3A%3AMIR&mode=all
09:30 mtompset Version 0.0 is the latest on CPAN.
09:30 mtompset What is MARC::MIR?
09:29 mtompset hi, drojf.
09:28 drojf hi mtompset
09:26 matts ^^
09:26 mtompset Greetings all!
09:26 drojf matts: wow, state of the art. cpan has version 0.0 (!) of marc-mir-template :D
09:24 drojf hi cait
09:23 cait hi matts and drojf
09:22 matts And it's about Data Mining (in french)
09:22 matts eiro gave a nice conference yesterday about MARC::MIR
09:21 matts It's nice
09:21 matts hi !
09:20 drojf #koha even
09:20 drojf hi #koa
09:20 drojf bonjour matts. how is the perl workshop going?
09:01 mtompset Is this an automated process, chris_n?
07:44 mtompset Greetings, #koha.
04:26 mtompset Oh that's interesting! The || probably checks in the parameter already exists. But why push it if it got template param'd? Hmmmm....
03:26 mtompset grep "\$[a-zA-Z]*\s*\=>" `find . | grep -v blib | grep -v "/misc" | grep -v "/var" | grep -v "/t/"` | less
03:26 mtompset These gems brought to you via grep. :)
03:23 mtompset $template->param($error) || $template->param( $error => 1);
03:23 mtompset AAAAH!
03:20 mtompset well, before it is initialized?!
03:19 mtompset how can you use a variable before it is defined?!
03:19 mtompset my ( $title, $LINES ) = TransformPrefsToHTML( { $title => $matched_groups }, $searchfield );
03:19 mtompset okay, this is weird:
03:13 mtompset Nothing worse than harddrive crashes causing downtime, when scheduled downtime could have prevented it.
03:12 mtompset Even quarterly probably merits the downtime.
03:11 mtompset How often do you have to reboot and fsck? If it happens monthly, I think it a scheduled downtime is in order.
03:09 jcamins_away mtompset: red bugs are higher priority.
03:08 Waylon university students use it from uni, and from home, via their uni website.
03:07 Waylon its 24x7
03:04 mtompset Or do you have users 24x7?
03:03 mtompset That should be less than an 8 hour day. Do it on a weekend, users don't notice.
03:03 mtompset So, 2 boot time down times, and the time it takes to clone.
03:02 mtompset Then the drive gets swapped in with a boot time down time.
03:02 mtompset Even a clone to a newer, larger newer drive should take several hours at worst.
03:00 mtompset you mean it is being used even on weekends?
02:59 Waylon can't really afford the extended downtime.
02:59 mtompset If you had to reboot and fsck several times, shouldn't you replace the drive?
02:59 Waylon so the hostgator management contact datacenter staff, who use the IPMI tools, i imagine.
02:58 mtompset oh, the joys of company acquisitions
02:58 Waylon so the hdd we're using is ancient and has failed a few times, requiring reboot and fscking.
02:58 Waylon their hardware used to belong to theplanet, which got brought out by softlayer.
02:57 Waylon yeah.. stupid server management company.
02:57 mtompset hostgator?
02:57 Waylon ..... hostgator can't install debian lenny it seems... not going to be fun installing koha 3.0 on squeeze. not even sure if its possible, last time i tried that it did not work i think. libXML version was too high. Moving from one hostgator server to another, to avoid possibly bad hdd, while we sort out the softlayer/3.8 issue.
02:38 mtompset Anyone know why some bug listings are in RED?
02:34 mtompset Giddiness declining... so many bugs. So much uncommon to me.
02:25 jcamins_away It's not very commonly used.
02:25 jcamins_away It means no one has tested to confirm the bug exists.
02:25 mtompset (as a bug status)
02:24 mtompset What does UNCONFIRMED mean?
02:24 jcamins_away It's just Zebra that I put on the wrong filesystem.
02:23 jcamins_away Waylon: I planned ahead and didn't have that in /var/lib.
02:23 Waylon how did you manage to move the database without downtime?
02:22 jcamins_away mtompset: yeah, but no one hit the server in that time.
02:22 mtompset There was probably a few milliseconds, jcamins_away.
02:21 * jcamins_away successfully moves /var/lib on a production server onto a different filesystem without any downtime whatsoever.
02:21 mtompset under promise, over deliver.
02:21 jcamins_away Woohoo!
02:20 jcamins_away Waylon: did Scotty teach you *nothing*?
02:20 mtompset not in the conventional sense of the word.
02:20 * jcamins_away waves.
02:20 mtompset I don't get paid.
02:20 Waylon depending on what your boss wants you to be doing.
02:20 * mtompset feels a wave of coding giddiness coming on.
02:19 Waylon and possibly even get paid for it.
02:19 Waylon of course
02:19 mtompset Hey, with this DEV install... I could actually go looking for NEW reports which need patching.
02:18 mtompset By the way, greetings, jcamins_away but not really. ;)
02:17 Waylon well.. its actually not many features.. and i did code some of them myself. I don't think it will be too difficult to recode.
02:16 mtompset Depends on your coding stamina. ;)
02:16 jcamins_away That's actually probably an optimistic estimate.
02:16 jcamins_away Waylon: you have a lot of weird custom code.
02:15 Waylon six months...
02:15 jcamins_away Right.
02:15 Waylon yeah.. not much point in getting the new hardware until then eh?
02:15 jcamins_away "Sounds great, it'll take about six months to reconcile the code enough to move to 3.8."
02:15 Waylon but yea, the lease cost will be higher of course.
02:14 jcamins_away So there you go.
02:14 * mtompset nods.
02:14 Waylon yeah.. im not paying for it...
02:13 jcamins_away Waylon: to be blunt, who cares if they're being spendthrift?
02:12 mtompset If there are costs to change, that would be used in an amortization argument.
02:10 mtompset And it will rock your Koha world.
02:10 mtompset Okay, then there's no reason to not go with the sysadmin's suggestion, because it won't cost extra.
02:09 Waylon i don't think the actual setup costs anything.
02:08 mtompset So as you can see.... that armortization costs 5x as much in my example.
02:07 mtompset You'll want to upgrade a machine between every 3-5 years to keep current and hardware defect free...
02:07 mtompset That's only $33.34 every month.
02:06 mtompset But if I buy a new computer every 5 years...
02:06 mtompset That is... $166.67 per month.
02:06 mtompset I must be amortizing $2000 over 12 months.
02:06 mtompset And buy a new computer every year.
02:05 mtompset If I pay $2000 for a brand new computer.
02:05 mtompset It's like buy a computer.
02:04 mtompset because of the whim of a sysadmin.
02:04 mtompset to spend $X again every 6 months.
02:04 mtompset if you only have it for 6 months out of Y years... that's crazy...
02:03 mtompset That was intended to run for Y years.
02:03 Waylon its kinda been sitting for 200ish days not doing much while i worked on it.
02:03 mtompset The company has paid $X so far.
02:03 Waylon amortization costs?
02:03 Waylon though, i can setup a vm with debian squeeze and koha 3.8 and see how it goes.
02:02 mtompset Let the boss put 2+2 together.
02:02 mtompset You can point out amortization costs.
02:02 Waylon meh.i can't say anything really.. im only the perl coder.
02:00 mtompset If you have only had them for 200ish days... then perhaps another 165 days? so 6 months minimum?
01:59 mtompset not a test program.
01:59 mtompset by seiging, I mean actual use.
01:58 mtompset If seiging proves it sufficient, upgrade later.
01:58 mtompset If seiging proves it insufficient, then upgrade.
01:58 mtompset After the upgrade.
01:58 mtompset Suggest that the initial setup be tested for X months.
01:57 mtompset I would argue that the initial setup costs haven't been amortized sufficiently and that there is insufficient data to prove the initial setup inadequate for an upgraded version.
01:54 Waylon the servers themselves, data center, the server setup and lease, company.
01:52 mtompset Who paid for the servers? your company or the data center?
01:51 mtompset Though, frankly, I like his suggestion.
01:51 mtompset I still think the costs haven't been amortized sufficiently is probably a good reason not to go with the sysadmin's suggestions.
01:48 mtompset Sounds like you need to set up a 3.8 system and seige it to prove that the IO is good enough.
01:47 Waylon 3.0 is a dead horse
01:47 Waylon most of which is only usable on koha 3.8
01:47 mtompset Did you look into some of the suggestions here?
01:46 mtompset http://wiki.koha-community.org/wiki/Koha_Tuning
01:46 mtompset Ah.
01:46 mtompset tuning?
01:46 Waylon in other words, if 10 users at same time, did searches, one search each 5 seconds, it would only work 94% of the time.
01:45 mtompset AH.
01:45 Waylon simulates users using hte website.
01:45 Waylon the program that automatedly requests urls, times the responses, tells if a request doesn't get a response.
01:45 * mtompset is trying to understand the definition.
01:45 mtompset slowed down?
01:44 Waylon siege?
01:44 mtompset sieged?
01:43 Waylon old koha 3.0 sieged at 10 users, 5 second delay each batch... came up with 94% availablity.
01:43 Waylon god, i wish i never sieged it now.
01:42 Waylon he doesn't think that the IO is up to it. based on my sieging of it.
01:42 mtompset all the more reason to not change.
01:42 mtompset amortized sufficiently.
01:42 Waylon well.. we haven't even used it yet.
01:41 mtompset The costs haven't been amortized yet.
01:41 mtompset His suggestions are probably good, but it's too soon after initial setup.
01:40 mtompset Just for the record, I'm no expert. :)
01:39 mtompset It probably is. He's a sysadmin, suck it up. :P
01:39 Waylon and thinks that xapi export of the lvm partitions not raw enough.
01:38 Waylon considers the dual server system, on XCP, abit of a too complicated.
01:38 mtompset That's my gut feeling. I claim no expertise.
01:38 mtompset It's not cost effective to take his suggestions so soon after initial setup.
01:37 Waylon new sysadmin
01:37 mtompset less than a year?!
01:37 Waylon 200 days?
01:37 Waylon oh..no...
01:36 mtompset Ages? Like >5 years?
01:36 Waylon ages ago. back when XCP 1.1 came out.
01:33 mtompset When did the two machines get initially get set up?
01:32 Waylon or just get axed, and get a dev/staging server in one of softlayers other data centers for safety.
01:31 Waylon so one will just get upgraded.. the other will possibly become its dev/staging server.
01:31 Waylon these two servers are already in the data center.
01:30 mtompset What will become of the two machines if he gets his data center-based server?
01:29 Waylon yeah..... that too.
01:29 mtompset Ah... he's killing your babies.
01:29 * Waylon did the setup
01:27 mtompset He's probably had to do some kludgey stuff to set the two machines up, and really would prefer a RAID setup.
01:27 mtompset consider the cost analysis of running two owned servers vs. one leased one.
01:26 Waylon two machines.
01:26 mtompset two servers?
01:26 Waylon 8 gb ram.
01:26 Waylon just simple hdds.
01:25 mtompset What is the current set up?
01:23 mtompset And if you do a cost analysis, sometimes leasing the car and upgrading and continuing a lease later is cheaper than buying it and trying to keep it going for years.
01:23 Waylon so, this single dedicated server, RAID setup will be preferable to two dedicated servers, simple storage?
01:22 mtompset So, he's trying to outsource the hardware side of his responsibilities.
01:22 mtompset But he likes dealing with system setup and updating, etc.
01:22 Waylon well.. we do have two servers, we were planning to loadbalance across.
01:22 Waylon right.
01:21 Waylon s/ha/ah
01:21 Waylon ha
01:21 mtompset He probably doesn't want to have to deal with hardware failures.
01:21 Waylon simplifiy?
01:21 mtompset He's trying to simplify his life.
01:20 mtompset But the sysadmin had to maintain it.
01:20 Waylon straight 2 hdds, 500 gb each, and 8 gb ram.
01:20 Waylon ah actually we had a simpiler setup
01:20 mtompset If it is at a data center, you probably can't get a simpler set up.
01:20 Waylon sysadmin is ours though, not the data centers
01:20 Waylon yea
01:19 mtompset Well, you said it was being hosted at a data center.
01:19 Waylon new zealand, Saturday too, 1:18 pm
01:18 mtompset Where are you, Waylon? It's Saturday here.
01:18 Waylon yeah... so i should let the sysadmin do what he wants?
01:12 mtompset 1000 searches means 1GB... it will get freed eventually though.
01:12 mtompset Now, add a "leak" factor of 1MB...
01:11 mtompset 20 searches all at once means 1GB roughly.
01:10 mtompset I noticed that each request was around 40-50MB.
01:10 mtompset yes, 512MB...
01:10 mtompset While watching memory use for our 512MB system...
01:09 mtompset Depends on how popular your OPAC and Staff Clients are?
01:09 Waylon but 16?!
01:09 Waylon yes.. 1gb is not enough for koha.
01:08 mtompset more ram = good. :) 1GB is still small, it seems.
01:08 Waylon 2 quad processors... hell...
01:08 Waylon 2 dual/quad processors / 12 or 16 GB RAM
01:07 Waylon hmm.. its in a data center. probably is hot swappable but we'd not be doing the swapping.
01:07 Waylon koha is koha app and pdf files that are currently about 240 gb total.
01:06 Waylon DB i imagine is just mysql db.
01:06 mtompset Will it be hot swappable?
01:06 Waylon RAID 10 -- 4 * 600 GB 10k disks -- usable 1.1TB for KOHA/Zebra
01:06 Waylon RAID 10 -- 4 * 300 GB 10k disks -- usable 600 GB for DB
01:06 Waylon RAID 1 -- 2 * 75 GB 10K disks -- usable 75 GB for OS
01:05 mtompset In reality, probably not.
01:05 mtompset Technically, yes.
01:05 Waylon sysadmin wants to setup a rather comprehensive RAID 1 and RAID 10 system
01:04 Waylon is there such a thing as over engineering?
01:04 Waylon hiya all.
01:03 mtompset Someone should change the topic, nengard's birthday is over now. :)
01:03 mtompset Good morning, #koha folks!