*** jamest has quit IRC
*** jamest has joined #gnuenterprise
*** jamest has left #gnuenterprise
*** jamest has joined #gnuenterprise
*** jamest has left #gnuenterprise
*** jamest has joined #gnuenterprise
*** reinhard has joined #gnuenterprise
*** johannesV has joined #gnuenterprise
<reinhard> dimas: do you have any plans on working on translations for gnue again?
<reinhard> ru.po is heavily out of date for most packages
<reinhard> http://www.sysadminday.com/
<reinhard> might be of interest to some people here ;-)
*** kilo has joined #gnuenterprise
<dimas> reinhard: i do have :)
<dimas> hello all
<ajmitch> hi
<dimas> hi ajmitch
<reinhard> dimas: we plan on releasing gnue-appserver 0.1 next week
<ajmitch> great
<reinhard> which will be the official release for quite some time probably
<reinhard> and will maybe also be the version to be included in sarge
<ajmitch> if we're really lucky it'll even get into sarge ;)
<reinhard> so you might want to look at appserver's po file over the weekend or so
<reinhard> it's not *that* much fortunately
<dimas> what should be translated in case of appserver?
<reinhard> there is a ru.po
<dimas> ok
<dimas> msgid "%s: _getNewItem() not implemented!"
<dimas> msgstr ""
<dimas> should msgstr be something like "%s: _getNewItem() <translation here>"?
<reinhard> could also be something like
<reinhard> "%s: no implementation for _getNewItem()"
<reinhard> that is
<reinhard> the _getNewItem() doesn't have to be at the start of the text
<johannesV> this is a bad example
*** deke has quit IRC
*** deke has joined #gnuenterprise
<dimas> johannesV: could you give any better? :)
<johannesV> well, i just thought that _getNewItem () might be a bit confusing
<johannesV> msgid "Cannot open file %s"
<johannesV> msgstr "Die Datei %s kann nicht geoeffnet werden"
*** nickr has quit IRC
*** sjc has joined #gnuenterprise
<dimas> johannesV, reinhard thanks, now i see
*** sjc has quit IRC
*** dimas has quit IRC
<reinhard> l8r all
*** reinhard has quit IRC
*** mangomadness has joined #gnuenterprise
<SachaS> have a nice weekend
<SachaS> happy coding for those you do :)
*** SachaS has quit IRC
*** jamest has joined #gnuenterprise
*** kilo has quit IRC
* apropos is back from away after 15 hrs 7 mins 20 secs
*** kilo has joined #gnuenterprise
<chillywilly> anyone know stinking perl?
<chillywilly> I thought when you did $var =~ s/foo/bar/g; it substituted the value of he variable with the result of the regexp
*** lxf has joined #gnuenterprise
*** johannesV_ has joined #gnuenterprise
*** johannesV has quit IRC
*** dimas has joined #gnuenterprise
<derek> hmm i know activestate perl, but havent heard of stinking perl
*** jcater has joined #gnuenterprise
<chillywilly> derek: all perl == stinking
<derek> true dat
*** nickr has joined #gnuenterprise
*** nickr has joined #gnuenterprise
*** lxf has quit IRC
<chillywilly> interesting hostmask nickr 
<chillywilly> did you donate some money to the pdpc?
*** reinhard has joined #gnuenterprise
*** holycow has joined #gnuenterprise
*** gsoti has joined #gnuenterprise
*** gsoti has left #gnuenterprise
*** gsoti has joined #gnuenterprise
*** Vee2d2 has quit IRC
*** Vee2d2 has joined #gnuenterprise
<whizman> Continuing my thread on format masks...
<whizman> I have come to view this from 2 different perspectives:
<whizman> One is the powerful, flexible, standards-compliant format mask implementation that we all want.
<whizman> Second is the immediacy of delivering to my client (in the USA) a forms app that uses comma as thousands separator
<whizman> For the latter, I have created a hard-coded, straight-to-the-point patch against GNUe-Forms-0.5.5
<whizman> Anyone who may be interested in that patch can download it from:
<whizman> http://whizman.com/Project/Gnue/forms-thousep-comma-usa.tar.gz
<whizman> Regarding the former, I have looked over, and have started tinkering with, gnue-common/src/formatting/NumberMask.py (from svn co 2004-07-28)
<whizman> I remain interested in discussing the matter with jcater at a time that is convenient for the both of us (as I understand from previous sessions on #gnuenterprise that format mask is jcater's area)
*** reinhard has quit IRC
*** johannesV_ has quit IRC
*** someon has joined #gnuenterprise
<jamest> whizman: cool
<jamest> we have several hacks in gnue atm to deal with those I need it now formatting issues
<jamest> as soon as we figure out how jcater keeps getting out of the leg shackles i'm sure we'll have the real deal in place
<someon> whizman: WRT to masks, I think that there needs to be some consideration given as to how to handle display mask vs. input masks....  an input mask could be written like a regex...  [A-z][A-z][0-9][0-9][0-9] requires 2 Alpha, 3 number....  whereas a display mask would be something like ###,###,##0.00 would display a minimum of 1 digit and decimals, seperated by ","
<jamest> someon: have you seen the stuff jcater was putting into common?
<someon> erm...  Thousands seperated by ,
<someon> CAn't :(
<jamest> ?
<jamest> there was a technote in common i stuck in there probably a year or two ago
<someon> Stupid firewalls at work, and no net-link at home.
<jamest> IIRC jcater did something even better 
<jamest> it only IIRC needs relinked into forms
<jamest> but i've not looked at the stuff since forever
<someon> I'm just meaning that you could have an input mask that requires only numbers, and then a display mask that includes the ,'s  problem is: what do you do if someone puts the , in when they are doing the data-entry?
<someon> In the raw DB, you could store the number 123456.78 but display in forms as 123,456.78
<jamest> thats what it does now
<jamest> actually
<jamest> if you look in the gfdisplayhandler code
<someon> ON a completely different note.....  Does anyone know how to program BlackBerry?  I'd love to see a GNUe Client for BlackBerry, but you would have to figure out how to deal with the BlackBerry Enterprise Server in the way
<jamest> there is a working copy and a display copy
<jcater> fwiw
<jcater> I have all next week off
<jcater> and hope to gnue
<jamest> what!??!?
<jamest> you have a floor to put in !
<jamest> i'm sure the little lady has a list all ready for you
<derek> rofl
<derek> well you see
<derek> jcater is going to write  a gnue floor estimator
<derek> where you enter dimensions of your floor
<derek> cost of different flooring
<derek> and it spits out reports on what floors are within your budget
<derek> and sends edi documents to home improvement store to purchase them
<derek> and edi to pay them
<derek> so you see technically to do his "chores" he has to gnue
<derek> yeah thats it
<someon> But I'm sure it will take all week to figure out how to deal with non-square floor-plans...
<someon> walls that are curved, walls that don't meet with 90 angles...
<someon> Just to really confuse things more...  Has any thought been given to Synch-Mod-Synch devices (eg non-connected Palm Devices or WinCE devices)
<jamest> someon: not really
<jamest> i believe i'll be looking into using some type of portable barcode system in the next or so
<jamest> and there is a pda with that feature
<jamest> but using a palm for data entry
* jamest shudders
<jamest> though with a foldable keyboard they are pretty cool
<nickr> It would seem that if the data entry conisted mostly of scant text fields interspersed with checklist or other such information, that it would be rather a good way to enter such data.
<nickr> Especially if said entry was primarily write-only.
<jcater> wtf are we talking about?
<jamest> http://www.microsoft.com/technet/security/bulletin/MS04-025.mspx
<jamest> ahhhhhh
<jamest> the smell of fresh patches
<jamest> jcater: dunno
<jamest> gnue on sub-pc portables i thought
<whizman> jamest, someon: sorry, i stepped out to lunch without leaving a note here...
<someon> All good.  I did same.
<whizman> (am catching up reading conversation above)
*** kilo has quit IRC
<someon> Jamest: the best portable barcode devices I have seen either feed out in a CSV format, but use proprietary protocols over a serial link, OR are Palm Based.  Symbol makes a ruggedized barcode/plam combination device
<holycow> someon, that would be cool actually
<holycow> the systems tend to be expensive, but i'd love to have a palm device outfitted with one of these, feeding into a gnue asset management system
<whizman> jamest: in reply to comment above: currently numerics do in fact *not* display with the comma, as the code to do so is commented out (Forms-0.5.5, Forms-svn-2004-07-28)
<whizman> My patch mentioned above uncomments that code for display, and does other temporary hacks for input (basically accept arbitrariliy many commas b/t digits, and ignore them when input complete)
<jcater> whizman: we did exactly what you're patching against
<jcater> and it was not well-received
<jcater> so we took it out
<jcater> as 1) we have a large international base
<whizman> it is only a temporary workaround, agreed
<jcater> and 2) a lot of numeric fields didn't need comified
<jcater> (zipcodes, etc)
<whizman> need flexibility and i18n of format masks
<whizman> totally agreed - i thought i had said clearly enough that what i posted earlier was only temporary, for specific cases, in North America.
<whizman> In any case, i am completely agreed that we want a "real" format mask solution.
<whizman> jcater: you mentioned that you hope to gnue (newly-coined verb?) next week...  is format mask something that you think you can most effectively work on alone, or in coordination with another developer?
<someon> The system I currently use (sorry closed-source) allows input masks and display masks on a per-field basis.
<jamest> forms would be on a per entry basis
<jamest> entry:field is an N:1 relationship in forms 
<whizman> someon:  yes, exactly, and i have noted your point above that input and display really are different issues.
<whizman> also, i would suggest '9' and '#' be synonyms in numeric masks (Oracle-think)
<someon> Problem that I see is that in North America, we tend to use , as a grouping marker, and . to indicate decimal delimer.  Europe is different IIRC....  a number of countries use , to delimit decimals.
<whizman> someon: agreed, needs to be fully internationalized/localized, per-entry; display and input are different issues
<someon> As I used it in the display mask, # would be shown, if necessary, a 0 would ALWAYS be shown.  ie in a display mask of ###,##0.00 if the DB had 3, display would be 3.00 but if the DB had .49, dispay would be 0.49, if the DB had 3000, display would be 3,000.00
<whizman> perhaps there has been a misunderstanding about the word "patch" which i maybe misused a couple hours ago...  i did *not* ever intend that patch to be merged into the codebase; rather it is available in the meantime to consultants/end-users who may have a specific need and be located in the N and W hemispheres.
<jcater> whizman: ah
<whizman> someon:  yes, i understand required-digit vs. optional-digit.  i implemented some of such for Oracle.  what i am saying is i would like to able to use Oracle notation:  9 = optional digit, 0 = required digit
<jcater> NumberMask.py, et all, are an old attempt
<jcater> look in gnue/gnue-common/src/formatting/masks
<jcater> it's a lex-based parser
<whizman> jcater: apologies for the confusion, and thanks for the opportunity to clarify.
<jcater> very fast
<someon> whizman: I'm not 100% familiar with Oracle.  We have a few Oracle installs here, but I'm not allowed to play with them :(
<jcater> hmm, my README is missing
<jcater> which explained all the tokens
<whizman> OK, that was going to be one of my first questions - i was not sure what was old and what new.
<whizman> jcater:  :( lack of readme could explain some of how i was getting lost yesterday :) first looked into doing it right, then wrote a temporary workaround for my client.
<someon> I get to play with MS-SQL (ugg), MySQL, Interbase, and a little bit of PostgreSQL
<whizman> someon: try   http://mirrors.spb.ru/oracledocs/server/sql73/toc.html
<chillywilly> anyone see the new mysql clustering support?
<chillywilly> http://dev.mysql.com/doc/mysql/en/NDBCluster.html
<jamest> anyone wonder if they did as good a job on clustering as they have on the rest of their database?
* jamest runs
<whizman> someon:  btw, if my note above was vague when i said "implemented for Oracle", i mean the implemtation in Oracle's source code (C language libs), not referring merely to using fmt masks in Oracle*Forms
<jamest> whoops, typo
<jamest> s/database/"database"
<chillywilly> jamest: I don't understand blind hatred of mysql
<jcater> jesus
<someon> I saw PostgreSQL was doing replication a while ago, but I'd be interested in an Active-Active DB cluster
<jcater> mysql and clustering
* jcater gets sick to my stomach
<chillywilly> it's not as featureful as postgres but it is damn fast
<jcater> trusting something like clustering to people who took HOW long to get decent transactional support?
<jcater> not losing data is a strong feature, imho =)
<chillywilly> I don't think it was implemented right away
<chillywilly> you can use innodb tables now with transactions and crap
<chillywilly> I still like having real referential integrity though so postgres will always be my favorite
<chillywilly> I was looking at pgreplication...but they don't keep up with teh latest release
<whizman> jcater: do you have a copy of above-mentioned README, or was that a casualty of  rm -rf ?
<chillywilly> r -rf /dev/jcater
<chillywilly> rm*
<jamest> chillywilly: http://sql-info.de/mysql/gotchas.html
<jamest> there is a huge list there
<jamest> some of which would have bit me in the arse had i been a mysql user
<jamest> on same thing is a link to the postgreql db with gotchas
<jamest> compare the 2 lists
<jamest> the postgresql ones don't result in my $15,000 invoice quietly, and without error, becoming a $9999.99 invoice
<jamest> and people think it's ok to write accounting stuff with it as a backend
<jamest> the accountants i work with don't trust the computer already
<chillywilly> well, we're going with postgres all the way now anyway :)
<jamest> cool
<jamest> you did ask though why the blind hatred 
<chillywilly> yes I did
<jamest> that list was an eye opener for me the first time i saw it 
<chillywilly> that's good enough I suppose
<jamest> lots of "oh crap. you're kidding me." things in there
<jamest> thus
<jamest> my hatred isn't blind 
<jamest> :)
<whizman> someon:   http://mirrors.spb.ru/oracledocs/server/sql73/ch303.html#toc034  -  I remember working with technical writer Sandy S. to correct the documentation on number format masks (I am Randall Whitman)
<whizman> jamest: thanks for the MySQL Gotchas link - wish i had that 1+ years ago when my biggest client started migration from Unify to MySQL
<jamest> lol randal, that's cool
<jamest> i recall the old oracle manuals from early 90s
<whizman> jamest: nits: 2L's in "Randall"  :)
<jamest> where the pages would (whoops, no offense, i mispell everything, ask these gusy) duplicate whole sections
<jamest> 1/2 of page 160 would also be on 161
<jamest> either that or my co-workers were screwing with me
<chillywilly> hmmmmm, these mysql gotcha are interesting
<jamest> s/interesting/scary
<whizman> hehe I personally knew Oracle technical writers who were not from one of USA, UK, Canada, Australia, NZ, Jamaica ... i
<jamest> i believe it
<chillywilly> do you ever use bigserial as the primary key type in a table?
<whizman> ah, but you were commenting on typesetting, not spelling, in Orcl manuals, sorry
<chillywilly> and don't you guys think it is a waste to use varchar(255) for every damn text field in a db? cause I know someone (who I will keep nameless) that does this all the time
*** sjc has joined #gnuenterprise
<holycow> with hard drives aproaching 1/4 terrabyte, does it matter?
<chillywilly> queries and indexes are slower?
<jamest> why slower?
<whizman> chillywilly: that depends whether your primary income center is *hardware*
<jamest> IIRC in postgresql the char() type is really a padded varchar()
<jamest> when they switch to that (again IIRC) there was discussion that it really didn't hurt performance
<jamest> around the 7.0 timeframe, kinda, maybe
<chillywilly> I was jsut wondering if anyone else though ti was odd that you would use the biggedt varchar you could instead of deciding how big the field really needs to be
<jamest> i wouldn't do it
<jamest> as my users
<jamest> sigh
<chillywilly> HEHEH
<chillywilly> LKSJdfl;kajsdl;fkjasdl;fkjsl;dkfjl;sdkfjl;skdjfl;ksjdl;fkjsdl;fkj
<jamest> they'd figure it out and use it for notes
<chillywilly> :)
<jamest> they used the damn phone # field for notes at one time
<jamest> until the beatings were handed out
<chillywilly> hehe
<jamest> sigh
<jamest> now they use part numbers for notes
<jamest> ref integ stop it?
<jamest> nope, the client didn't want that slowing down the sales staff
<jamest> about time for some more free beatings
<chillywilly> bah
<whizman> those people definitely need input validation!
<chillywilly> I really need to redesign our system's database though
<chillywilly> the guy who designed it was on crack I think ;)
<jamest> who's that? the old pysco chilly?
<jamest> well, not pysco, over caffinated
<chillywilly> nope...can't take credit for it
<chillywilly> look at your pm
<chillywilly> :)
<whizman> not *them*  -  http://lists.microshaft.org/pipermail/dmca_discuss/2003-August/005858.html
<jamest> :)
<jamest> someon: i'll be trying out the slony-i replication in a new setup in next 60-90 days
<someon> jamest: slony-i replication?  
<jamest> ja
<jamest> http://gborg.postgresql.org/project/slony1/projdisplay.php
<someon> I don't know the first thing about it...  hmm... /me goes looking
<derek> jamest: please keep me updated on what you find
<jamest> trigger based master/slave setup
<derek> i could use this in a few places
<jamest> i've read through the docs and i'm putting together a new computer network from scratch
<jamest> hope to abuse the crap out of it for a month or so then put into productoin
<jamest> pending that it works out ok
<derek> cool
<jamest> as a few lines of script and you can promote a slave to master
<someon> Yeah, but I'm not so sure I would call that true active-active... in that if the master goes down, how do you manage the updates *AFTER* it goes down until it comes back up...
<jamest> it's not
<derek> my main concern is for reporting
<someon> sorry...  saw the promotion comment afterwords.
<jamest> but it's the only thing I found for postgresql that didn't look iffy
<derek> i would like something that every hour or night or something similar
<derek> replicates down to a slave
<derek> and then i can have all reports run off the slave
<jamest> um
<derek> and all entry, etc run off the master
<jamest> you can do that with a pg_dump and a pipe
<derek> yeah but it doesnt have nice tools/interface
<jamest> i used to for my devel copy of the production db
<derek> we do it for things now
<derek> but would like a more eloquent product
<jamest> ah
<chillywilly> jamest: cool
<chillywilly> jamest: that might be the thing I was looking for
*** VeeOffice has joined #gnuenterprise
*** Vee2d2 has quit IRC
*** gsoti has left #gnuenterprise
*** kilo has joined #gnuenterprise
*** VeeOffice has quit IRC
*** VeeOffice has joined #gnuenterprise
<derek> anyone use tcpdump here before?
<derek> i want to see what MAC address something is coming from
<derek> basically if i ping a certain ip address i get 3 replies for each single ping request
<derek> i want to find out what machines (MAC addresses) the 3 diff returns are
<derek> if i run tcpdump host server and client
<derek> and ping
<derek> tcpdump doest show traffic
<derek> but when i exit tcpdump it tells me 6 packets filtered
*** kilo has quit IRC
<someon> IIRC that's because PING is not TCP it's ICMP
<chillywilly> yep
<chillywilly> try
<chillywilly> arping
<chillywilly> arp is the address resolution protocol which maps IPs to MAC addresses
<chillywilly> or better yet run arp
<chillywilly> with no args
<chillywilly> and you can see the arp cache
<someon> arp -a should show it, but on my box here, it's not working the way I expect (it's only showing my system)
<chillywilly> shows other hosts on my lan
<chillywilly> root@gandalf:/home/danielb# arp -a
<chillywilly> phone.libertyetech.com (192.168.1.5) at 00:0C:E5:C5:54:0A [ether] on eth0
<chillywilly> CPE-69-23-200-1.wi.rr.com (69.23.200.1) at 00:08:E2:32:44:54 [ether] on eth1
<chillywilly> my phone is the only other host
<chillywilly> ;P
<chillywilly> anywho...back to work...blah
<someon> ethereal will show it for you :P
*** kilo has joined #gnuenterprise
*** someon has quit IRC
*** someon has joined #gnuenterprise
*** someon has joined #gnuenterprise
<derek> hmmm
<derek> arp -a doesnt show me
<someon> doesn't show you the MAC's?
<derek> it only shows for MY machine
<derek> i want to be able to do
<derek> arp -a someserver
<someon> Yeah, that's what arp was doing for me too
<someon> I'm on a fully switched network though
<derek> so am i
<derek> now i can LITERALLY say i am sitting on the hub for phoenix
<jamest> doesn't matter if you're on a switch
<jamest> you should still see every machine you're recently communicated with
<jamest> derek try doing something like
<jamest> telnet machineName
<jamest> doesn't matter if telnet is there
<jamest> then immediately 
<jamest> /usr/sbin/arp
<someon> derek: are the devices you are looking for on the same subnet?
<derek> nope
<derek> to jamest, not someon
<jamest> what the
<derek> soleri:/home/neighborsd# telnet proxy.maricopa.gov
<derek> Trying 156.42.249.200...
<derek> telnet: Unable to connect to remote host: Connection refused
<derek> soleri:/home/neighborsd# arp
<derek> Address                  HWtype  HWaddress           Flags Mask            Ifacemongo.maricopa.gov       ether   00:50:DA:0D:48:FE   C                     eth0
<derek> 156.42.48.252            ether   00:D0:02:52:43:FC   C                     eth0
<derek> soleri:/home/neighborsd#
* someon was forgetting that pretty much everything that I was on a different subnet...  so I could only see the router's if
<jamest> what distro?
<jamest> even then he should see the router as he's talking to us
<someon> try arp -n instead (shows the IP rather than ifnames..)
<someon> He has two entries...  neither of which should be his own IP...
<derek> hmm ethereal is nice
<derek> jamest: debian
<derek> unstable
<derek> wonder what would happen if started blocking packets from cisco group management protocol :)
<chillywilly> wooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!11
<chillywilly> do it!
<chillywilly> perl makes me crazy
<derek> anyone know if it is reaosnable for a load balancer to respond in triplicate to ping requests?
<derek> specificially Windows Load Balancing Server
<apropos> are you pinging a broadcast address?
<derek> it is a proxy server
<apropos> you should only get multiple responses if you are pinging a subnet address or a subnet broadcast address. Otherwise you have an IP address collision.
<derek> i tend to agree
<derek> but dont claim to be a network stud
<apropos> you are pinging an IP address or a name?
<apropos> try a traceroute maybe
* apropos is gone: away
<derek> name
<derek> but happens for ip too (i think)
<derek> yip for ip address as well
*** kilo has quit IRC
*** jamest has quit IRC
*** kgnet has joined #gnuenterprise
<kgnet> hi
<kgnet> I installed gnue-designer.exe and I've got such error:
<kgnet> Traceback (most recent call last):
<kgnet>   File "gnue-designer", line 30, in ?
<kgnet> ImportError: No module named gnue.designer
<kgnet> What can I do ?
<kgnet> Any one ?
<kgnet> Maybe I will back tomorrow.....
*** kgnet has quit IRC
*** jcater has quit IRC
*** insanekane has joined #gnuenterprise
<insanekane> hello sjc and all :)
<sjc> hehe
<sjc> hi insanekane
<insanekane> man i didnt know this channel was there ...
<insanekane> sjc, have u tried using GNUe app server ?
<sjc> no. not yet
<insanekane> okie
*** dimas_ has joined #gnuenterprise
*** havoc_ has joined #gnuenterprise
* dimas_ is first time testing gprs
*** dimas_ has quit IRC
*** havoc has quit IRC
*** sjc has quit IRC
*** insanekane has quit IRC
<someon> Have a good weekend all!
*** someon has quit IRC
*** dcmwai has joined #gnuenterprise
*** jamest has joined #gnuenterprise
*** jamest has left #gnuenterprise
*** tocer has joined #gnuenterprise
<tocer> hello all
<tocer> any one else?
<tocer> when i install gnue cvs verion, i have a problem . setup program need linkd.exe , but where is it?
<whizman> tocer: i'm new here and i really do not know about linkd.exe, but with a little help from google i found the following web page, in case this helps any:
<whizman> http://lists.gnu.org/archive/html/commit-gnue/2002-11/msg00058.html
<tocer> thanks whizman
<tocer> i'll read it
<whizman> you're welcome
<whizman> (Google is your friend :))
<tocer> i'm searching in Goolge,but too many many result:( i'll try it
*** dcmwai has quit IRC
<whizman> tocer:  Google search     linkd.exe gnue      was a charm
<tocer> OH,I think i find a bug in \\gnue-common\\setup-cvs-win.py. and thanks whizman, I find linkd.exe in a hacker web site:)
<tocer> waiting, I'm tring it.
*** dcmwai has joined #gnuenterprise
<mangomadness> hola