Chat log from the meeting on 2024-03-12

From OpenSimulator

Jump to: navigation, search
[11:00 AM PT]  Andrew Hellershanks: Hello, everyone.
[11:01 AM PT]  Ubit Umarov: who was shyfy linden?
[11:01 AM PT]  Andrew Hellershanks: Someone needs to fix the typo in the chat logger.
[11:01 AM PT]  Joe Magarac: Hello all.
[11:01 AM PT]  Andrew Hellershanks: Hello, Joe.
[11:01 AM PT]  Ubit Umarov: hi
[11:01 AM PT]  Gavin.Hird Hi
[11:01 AM PT]  Gavin.Hird shyfly?
[11:02 AM PT]  Gavin.Hird shyfy*
[11:02 AM PT]  Ubit Umarov:
[11:02 AM PT]  Gavin.Hird dunno
[11:02 AM PT]  Ubit Umarov: not patch but that one
[11:03 AM PT]  Ubit Umarov:
[11:03 AM PT]  Gavin.Hird whoever wrote that, I will not miss Mojo
[11:03 AM PT]  Gavin.Hird VP of Marketing apparently
[11:03 AM PT]  Vincent.Sylvester Linden bashing was always low hanging fruit, but dang they manage on their own to be ten times worse
[11:04 AM PT]  Andrew Hellershanks: Hello, Jamie.
[11:04 AM PT]  Andrew Hellershanks: Hello, Kayaker.
[11:04 AM PT]  Gavin.Hird I geuss they did not meet some promised dealines, so the VCs got impatient
[11:04 AM PT]  Jamie.Jordan Hi everybody
[11:04 AM PT]  Kayaker Magic: Hi all!
[11:05 AM PT]  Gavin.Hird could explain the desperate early release of PBR viewer
[11:05 AM PT]  Vincent.Sylvester hehehe
[11:05 AM PT]  Joe Magarac: There are no VCs at this point. The current owners are long-term investors from real estate and banking.
[11:05 AM PT]  Vincent.Sylvester Not sure that's better
[11:06 AM PT]  Gavin.Hird exactly
[11:06 AM PT]  Andrew Hellershanks: Wow. I don't know what is on that wordpress page but it requires you to login as part of proving you are over 18 in order to see the posting.
[11:06 AM PT]  Joe Magarac: Me either, but the Waterfield people haven't done anything really stupid.
[11:06 AM PT]  Vincent.Sylvester That tells you all you gotta know really, it's a trainwreck that registered on the Richter scale
[11:08 AM PT]  Joe Magarac: Anyway.
[11:08 AM PT]  Vincent.Sylvester To more pleasant matters. I spent last week picking apart the migration system and after I pulled my hair out I decided it just needs something a bit less... dangerous. So I wrote a new system that migrates each table rather than making large assumptions on the state of tables based on some version info
[11:09 AM PT]  Vincent.Sylvester Extending from the already present validations to include a way to move from the old to the new system and even, if nothing is present, create the required tables directly, which should speed migrations up when first starting
[11:09 AM PT]  Gavin.Hird but there still needs to be versioning on the tables
[11:09 AM PT]  Vincent.Sylvester Eventually the plan is to have a new system storing version info in the table comments for each table and migrate and validate them at each step
[11:09 AM PT]  Andrew Hellershanks: I wonder how people wind up with tables that don't match the layout they are supposed to have based on their migration level.
[11:10 AM PT]  Vincent.Sylvester Either through making their own additions, which can be useful or through the migrations failing and nothing actually noticing that one step failed for whatever reason
[11:10 AM PT]  Gavin.Hird combination of failed migration (partial) and tinkering with the tables
[11:10 AM PT]  Vincent.Sylvester It just runs the sql and then tries to start without checking
[11:11 AM PT]  Andrew Hellershanks: Are migrations not done via a transaction?
[11:11 AM PT]  Gavin.Hird I hope they are
[11:11 AM PT]  Andrew Hellershanks: As do I.
[11:11 AM PT]  Vincent.Sylvester They are sort of, but even if that fails it just continues as if nothing happened
[11:11 AM PT]  Vincent.Sylvester There is nothing that stops them if they report failure, it's fire and forget
[11:11 AM PT]  Andrew Hellershanks: That should have been dealt with.
[11:12 AM PT]  Gavin.Hird but transaction is at table level and not for the total successful migration up to a known level
[11:12 AM PT]  Andrew Hellershanks: If there was a failure was anything reported to the console?
[11:12 AM PT]  Gavin.Hird I think...
[11:12 AM PT]  Vincent.Sylvester There was some indication on the console, but guess what, there is so much text there after you usually don't see that
[11:12 AM PT]  Gavin.Hird yes, console spew hard to decipher
[11:12 AM PT]  Vincent.Sylvester Migration runs and then all the module messages mask the errors within seconds
[11:13 AM PT]  Andrew Hellershanks: More so when it isn't shown in red text or you see it after by reading the log file.
[11:13 AM PT]  Ubit Umarov: no they are not done on a transaction
[11:13 AM PT]  Vincent.Sylvester No red text either
[11:13 AM PT]  Ubit Umarov: thye are slow enough without one
[11:13 AM PT]  Andrew Hellershanks: ok. That is at least three problems with migrations.
[11:13 AM PT]  Lyr Lobo: Hello *smiles*
[11:13 AM PT]  Andrew Hellershanks: Hello, Lyr.
[11:14 AM PT]  Gavin.Hird Hi Lyr
[11:14 AM PT]  Cuga.Rajal Hi Lyr
[11:14 AM PT]  Gavin.Hird the postgres ones were on transaction
[11:16 AM PT]  Andrew Hellershanks: Good. Unfortunately it isn't the DB system that is used that often.
[11:16 AM PT]  Vincent.Sylvester I still have to test the changes on existing tables and might have to just ditch the old system entirely using the new system from the start, that remains to be seen
[11:17 AM PT]  Vincent.Sylvester For now I have the old stuff in there in case anyone still needs to run them, after which the new migration takes over, which does everything per table, per version with a validation afterwards to make sure they conform to what they should look like
[11:17 AM PT]  Gavin.Hird ok
[11:17 AM PT]  Vincent.Sylvester The schema is just in a file so those that made changes to tables for their own system can edit them and make the validation succeed
[11:17 AM PT]  Andrew Hellershanks: Vincent, does your new system also verify that the connection strings are valid?
[11:17 AM PT]  Vincent.Sylvester That's the only thing that already throws an exception at least
[11:19 AM PT]  Andrew Hellershanks: Unless it has changed OS has had a tendency to attempt to use a database even when the credentials in the connection strings were invalid. The program would eventually throw lots of red text but long after the point where it tried to establish a link to the database(s).
[11:19 AM PT]  Gavin.Hird :_)
[11:20 AM PT]  Vincent.Sylvester On the grid side that tends to happen, but a hard failure on the connection still just throws an exception, some modules they will attempt to connect and it does do that when you restart the database while it is still running as well. Graceful handling of that doesn't exist
[11:20 AM PT]  Vincent.Sylvester My main sticking point is whether I should leave the old system in place or make the effort to turn the existing migrations into per-table ones and use the new system for everything
[11:21 AM PT]  Vincent.Sylvester Not sure which is the better option
[11:21 AM PT]  Andrew Hellershanks: I've been meaning to look at the code where the DB links are made and at the very least make it throw red text as soon as it fails to establish a link to a database or table.
[11:22 AM PT]  Vincent.Sylvester Most modules that directly connect do throw red text, just not terminate the whole thing
[11:22 AM PT]  Gavin.Hird I might be wrong, but in my experience both robust and the simulator already do that
[11:23 AM PT]  Gavin.Hird also sim start fails immediately the connection is failing
[11:23 AM PT]  Vincent.Sylvester Yeah they will throw exception on startup if connect fails and during runtime the console goes red and yellow when it loses connection
[11:23 AM PT]  Gavin.Hird very red
[11:24 AM PT]  Andrew Hellershanks: I'll have to try it. I never noticed red text at the time of an attempted DB connection.
[11:24 AM PT]  Vincent.Sylvester Think it is only the modules that can have different connection strings that might not throw exceptions
[11:25 AM PT]  Vincent.Sylvester Generally what's happening more is that it starts and optional modules fail to load and you get errors in main modules
[11:25 AM PT]  Vincent.Sylvester It'll happily work without inventory module working, but that creates issues with assets pretty quickly
[11:26 AM PT]  Andrew Hellershanks: In the past when people have reported problems getting OS running, or seeing lots of red text, I used to tell them to check the connection strings.
[11:27 AM PT]  Vincent.Sylvester It is either that or just bad config in the modules. If you enable a module in one ini file you gotta also uncomment some things in another, some miss that
[11:27 AM PT]  Vincent.Sylvester Technically you could just combine it all into one config, it's what the system does anyways
[11:27 AM PT]  Vincent.Sylvester With includes you can basically include any text and it'll load that in and override based on what's loaded last
[11:28 AM PT]  Vincent.Sylvester Doesn't even check if it is a .ini heh
[11:28 AM PT]  Gavin.Hird wasted cpu cycles
[11:29 AM PT]  Ubit Umarov: ini files are merged into a big config thing
[11:29 AM PT]  Ubit Umarov: well most of them, except regions.ini
[11:29 AM PT]  Gavin.Hird is the merge logged somewhere?
[11:30 AM PT]  Ubit Umarov: nahh
[11:30 AM PT]  Gavin.Hird or the result rather
[11:30 AM PT]  Vincent.Sylvester You can do "config get" on console to see the current running config
[11:30 AM PT]  Ubit Umarov: nahh
[11:30 AM PT]  Ubit Umarov: :)
[11:30 AM PT]  Gavin.Hird :-)
[11:30 AM PT]  Kayaker Magic: config show  or config show section_name
[11:30 AM PT]  Gavin.Hird might be easier to debug the resulting one than individial files
[11:31 AM PT]  Ubit Umarov: well you can see the result
[11:31 AM PT]  Gavin.Hird right
[11:31 AM PT]  Ubit Umarov: would be a bit large log
[11:32 AM PT]  Ubit Umarov: well and since we do use a tool for it, possible not that visible
[11:32 AM PT]  Andrew Hellershanks: It isn't just a module problem. The main system has the problem with connection strings if the credentials aren't valid.
[11:32 AM PT]  Vincent.Sylvester Thing that I would like some comments on. I currently have the code to do the old migrations and only after they are done using the new system. I wonder if I should change this and just use the new system directly, converting the existing migrations to the new style instead. I'm not sure what's going to work better for most people
[11:34 AM PT]  Gavin.Hird I have never experienced the existing system as particularly catastrophic, but it would be an improvement to run each statement as a transaction
[11:34 AM PT]  Cuga.Rajal Maybe switch to the new system after some testing? Use a test DB, mess up a table intentionally by hand, then start up OS and see new system work.
[11:35 AM PT]  Gavin.Hird drop table if exist;  -woops there goes your data
[11:36 AM PT]  Vincent.Sylvester I'll add proper transaction to the handling. Right now I just run the sql and check if the query returned an error code or not. After all migrations the schema checks if the table looks as it should
[11:36 AM PT]  Ubit Umarov: what would be the help of a trans on that?
[11:36 AM PT]  Andrew Hellershanks: vincent, How do you determine whether tables have been migrated to the new system? Are you using a new table to track the changes?
[11:36 AM PT]  Ubit Umarov: data would be gone anyway
[11:36 AM PT]  Ubit Umarov: transastcion only assures that all happens..
[11:37 AM PT]  Vincent.Sylvester I run old migrations up till they report no pending migration and then set the table version to zero, after that I validate the table schema
[11:37 AM PT]  Ubit Umarov: ans on some operations, to keep track of all operations to revert, would possible be a bit big thing..
[11:37 AM PT]  Vincent.Sylvester I could add a validation in between to check if the tables are of the design they should be after running the old migrations
[11:38 AM PT]  Gavin.Hird version to zero?
[11:38 AM PT]  Vincent.Sylvester set table comment Version:0
[11:38 AM PT]  Vincent.Sylvester Each new migration after just increments that, one sql file per version per table, so any failure can specifically state what went wrong
[11:38 AM PT]  Andrew Hellershanks: ok
[11:39 AM PT]  Gavin.Hird but there are dependecies on table versions and actual code the grid or simulator runs
[11:39 AM PT]  Ubit Umarov: current migration thing, does make sense...  but ofc, in time juat gets 2 large and heavy
[11:39 AM PT]  Ubit Umarov: you cna't jsut compare  the columns you want to the ones present
[11:40 AM PT]  Ubit Umarov: in several cases you do need the full path..
[11:40 AM PT]  Ubit Umarov: how you got there..
[11:40 AM PT]  Gavin.Hird agreed Ubit
[11:40 AM PT]  Ubit Umarov: that is the cost of soem flexiblility
[11:40 AM PT]  Ubit Umarov: some--
[11:41 AM PT]  Vincent.Sylvester full path?
[11:41 AM PT]  Ubit Umarov: yes.. history path..
[11:41 AM PT]  Gavin.Hird table creation history
[11:41 AM PT]  Lyr Lobo: Heya Orbert!
[11:41 AM PT]  Ubit Umarov: ot that yes :)
[11:41 AM PT]  Vincent.Sylvester Ah yes, that's why so far I kept the old system and only switch to new after those finish
[11:42 AM PT]  Ubit Umarov: ofc we do have thigns like the need to install a sequence of opensim versions to updated the dbs
[11:42 AM PT]  Ubit Umarov: like neeed 0.8x then 0.9x..
[11:42 AM PT]  Ubit Umarov: to update a older one
[11:42 AM PT]  Vincent.Sylvester The new system checks if a migration file exists to change a table to a new version. Checks first if the table might not already be up to date and then applies the migration. Should it throw an error code it'll terminate telling the user to fix the table manually
[11:42 AM PT]  Gavin.Hird yes, that too
[11:43 AM PT]  Ubit Umarov: wel and same db backends may not even have tools do compare schemas
[11:43 AM PT]  Ubit Umarov: to...
[11:43 AM PT]  Vincent.Sylvester I could write a schema file for each version to check as well
[11:44 AM PT]  Vincent.Sylvester I check for columns via information schema, grab the indexes for the keys and indexes and check charset
[11:44 AM PT]  Ubit Umarov: you need to get and compare a lot..
[11:44 AM PT]  Vincent.Sylvester Enough information to create the table directly in the most up to date version
[11:44 AM PT]  Ubit Umarov: for example data types..  dimensions.. etc
[11:44 AM PT]  Ubit Umarov: all their props..
[11:45 AM PT]  Vincent.Sylvester column_name,data_type,is_nullable,column_default,column_key,character_set_name,collation_name
[11:45 AM PT]  Ubit Umarov: yeap
[11:45 AM PT]  Ubit Umarov: and when there is a change.. what is the proper update comand?
[11:46 AM PT]  Ubit Umarov: think we stil have one case where a complex sql "program" was run...
[11:46 AM PT]  Vincent.Sylvester You can technically use sql directly to only insert a field if it doesn't already exist for example
[11:47 AM PT]  Vincent.Sylvester Though I check before doing migrations if the table is not already up to date according to latest schema to avoid that
[11:47 AM PT]  Vincent.Sylvester Db should throw error codes if the sql command fails though, so if it returns one I bail immediately
[11:47 AM PT]  Gavin.Hird yes but there might be cases where you have to do actual data migration, even to other tables or fields and format conversion in the ugrade
[11:47 AM PT]  Gavin.Hird upgrade
[11:48 AM PT]  Ubit Umarov: at a point some tables will need to be split
[11:48 AM PT]  Cuga.Rajal assets, LOL
[11:48 AM PT]  Ubit Umarov: i avoid one case, by converting nonsense double to float
[11:48 AM PT]  Ubit Umarov: otherwise the table was overloaded
[11:49 AM PT]  Gavin.Hird so even if just updating or adding single fields will work most of the time, there is a need to to have the flexibility to do complext migration
[11:49 AM PT]  Joe Magarac: While we're discussing version issues, I brought up last time that those of us on the viewer dev side need more clarity about OS versions, so we can figure out what to support. Any progress on that?
[11:49 AM PT]  Ubit Umarov: there are limits to number of fileds per row :)
[11:49 AM PT]  Vincent.Sylvester It still just runs any sql you throw at it, so it could still do that just fine
[11:50 AM PT]  Ubit Umarov: joe we do have release versions..
[11:50 AM PT]  Ubit Umarov: those are identified by version number alone
[11:50 AM PT]  Ubit Umarov: like release
[11:50 AM PT]  Vincent.Sylvester Handling new tables, hm may have to look at that
[11:50 AM PT]  Ubit Umarov: btu we have devel versions...
[11:50 AM PT]  Ubit Umarov: those also need the commit hash for full id
[11:51 AM PT]  Ubit Umarov: liek you can see on the green overtext at ly right
[11:51 AM PT]  Ubit Umarov: my..
[11:51 AM PT]  Cuga.Rajal Joe, for the short term you could build-in git in the app and then do git command on the version hash to get the commit date
[11:52 AM PT]  Ubit Umarov: osgrid also adds the date yes
[11:52 AM PT]  Joe Magarac: That's ridiculous..
[11:52 AM PT]  Ubit Umarov: Version: OpenSim Nessie Dev b70b5c07e3: 2024-03-11 00:18:05  0000 (SIM-0.3/0.8)
[11:52 AM PT]  Joe Magarac: Everything is
[11:52 AM PT]  Cuga.Rajal sequential beta numbers does have some merit....
[11:53 AM PT]  Joe Magarac: That's the problem.
[11:53 AM PT]  Ubit Umarov: it is Dev..
[11:53 AM PT]  Joe Magarac: So what's Production?
[11:53 AM PT]  Ubit Umarov:
[11:53 AM PT]  Ubit Umarov: only one..
[11:53 AM PT]  Joe Magarac: OK.
[11:53 AM PT]  Joe Magarac: When will there be a new Production version?
[11:54 AM PT]  Ubit Umarov: while Dev is still "a oprn basket"
[11:54 AM PT]  Ubit Umarov: while Dev is still "a open basket"
[11:54 AM PT]  Ubit Umarov: ofc osgrid only runs dev versions :p
[11:54 AM PT]  Vincent.Sylvester Sql would create the new table and move the data, after which it just runs the validation as normal. May have to play with the mapping a bit and version that too. Could move that into a file also or simply use a different one if the new table is present. It's always a bit of a balance between assumptions and blindly running things. That's always going to be an issue, but it currently is even more so
[11:54 AM PT]  Ubit Umarov: so osgrid does list the commit number ( and date)
[11:54 AM PT]  Gavin.Hird I do too. :-)
[11:55 AM PT]  Ubit Umarov: i mesn osgrid own releases
[11:55 AM PT]  Cuga.Rajal Can it be found programatically with lsl?
[11:55 AM PT]  Ubit Umarov:
[11:55 AM PT]  Cuga.Rajal I mean, not necc lsl, programatically,
[11:56 AM PT]  Joe Magarac: Perhaps more thought needs to be given to deployment strategy. As a viewer dev, I have no idea what to target.
[11:56 AM PT]  Vincent.Sylvester Last release and master dev
[11:56 AM PT]  Joe Magarac: That's reasonable.
[11:56 AM PT]  Vincent.Sylvester Everyone running ancient software, tough luck, can't backwards compatible into the stone age
[11:57 AM PT]  Gavin.Hird At the development stage you are in, possibly the release version ans that is a known entity
[11:57 AM PT]  Ubit Umarov: sadly git hash is not that easy to glue into runcode
[11:57 AM PT]  Jagga Meredith: Maybe test for ancient then say it can't be done
[11:57 AM PT]  Ubit Umarov: currently we do look for a file .version   or the git repo data as present on a normal clone
[11:58 AM PT]  Ubit Umarov: osgrid uses .version file
[11:58 AM PT]  Cuga.Rajal I may have made  a bad assumption about the hash being accessible to viewer and not date
[11:58 AM PT]  Andrew Hellershanks: It isn't that hard to do. Save it to a file when a build is run and then include that file in the source.
[11:58 AM PT]  Joe Magarac: Right now, I get good results with the version at Ubittest, and less good results with other dev versions elsewhere on OSGrid. That's why I bring this up.
[11:59 AM PT]  Andrew Hellershanks: IIRC, The dev versions of GIMP have a way to get that info and build it in to the info shown on the About box.
[11:59 AM PT]  Gavin.Hird other dev versions?
[11:59 AM PT]  Ubit Umarov: on crossings you should get similar thins on all 0.9x running ubode
[11:59 AM PT]  Vincent.Sylvester I haven't updated your Sharpview regions since I set those up a couple weeks ago, if they need an update please poke me
[11:59 AM PT]  Cuga.Rajal A viewer needs to query the version without access to files on the server
[11:59 AM PT]  Ubit Umarov: on otther things.. well version alone is just impossible
[11:59 AM PT]  Ubit Umarov: ll protocol did add soem flags to let viewers know regions features
[12:00 PM PT]  Ubit Umarov: but ll does not keep them
[12:00 PM PT]  Ubit Umarov: there should be a flag for pbr for example
[12:00 PM PT]  Ubit Umarov: but ll just changes the viewers at the same time :)
[12:01 PM PT]  Cuga.Rajal Joe are there specific features you could look for, perhaps added to capabilities, without knowing version?
[12:01 PM PT]  Ubit Umarov: also uses presence of some capabilities
[12:01 PM PT]  Vincent.Sylvester Cuga, yeah that's usually one way to get a bit better idea, but not 100% either
[12:01 PM PT]  Ubit Umarov: as region, i also use the caps viewer asks for, for example to detect eep viewers versus windlight
[12:02 PM PT]  Ubit Umarov: there is also the opensimextras
[12:02 PM PT]  Ubit Umarov: on simulator features
[12:03 PM PT]  Vincent.Sylvester I think release should have an additional one in there for the alias urls, not sure if they are reported if empty, but that would be one way, there are a few recent additions to simulator features
[12:03 PM PT]  Andrew Hellershanks: Let's pause the talk on migrations for just a moment.
[12:03 PM PT]  Gavin.Hird My viewer in About displays:
[12:03 PM PT]  Gavin.Hird You are at 484.9, 531.9, 24.7 in Dev Outreach located at  (
SLURL: hop:// Outreach/255/255/25
(global coordinates 2839270.0, 2560530.0, 24.7)
In region Dev Outreach at (11089, 10000) 
OpenSim Nessie Dev b70b5c0 (X64/Unix/DotNet)
[12:03 PM PT]  Andrew Hellershanks: We are at the top of the hour. Does anyone have a question to ask that is not about databases and/or migrations?
[12:04 PM PT]  Gavin.Hird so the last line is not that waht you are looking for Joe?
[12:04 PM PT]  Andrew Hellershanks: I'm not seeing anyone typing so I will take that as no other questions for today. You may continue. :)
[12:04 PM PT]  Joe Magarac: I need something that is ordered.
[12:05 PM PT]  Joe Magarac: Can't tell from the commit hash alone.
[12:05 PM PT]  Vincent.Sylvester Also commit hash on a fork you get nothing out of that either
[12:06 PM PT]  Vincent.Sylvester Just targeting release and master dev and any issues that come up through old OpenSim versions just tell them to update their stuff
[12:06 PM PT]  Ubit Umarov: actually viewers should pay attention to simualtor features on HG
[12:06 PM PT]  Vincent.Sylvester Sounds harsh, but yeah, can't live in the past forever
[12:06 PM PT]  Ubit Umarov: to findout grids changes etc
[12:07 PM PT]  Joe Magarac: I've said enough for today. Next meeting. Bye.
[12:08 PM PT]  Joe Magarac: Thanks.
[12:08 PM PT]  Vincent.Sylvester FS will tell you you cannot go somewhere when that doesn't support bento... or was it bom... I don't remember anymore
[12:08 PM PT]  Cuga.Rajal Take care Joe
[12:08 PM PT]  Jamie.Jordan thanks yall
[12:09 PM PT]  Andrew Hellershanks: ok, Joe. Thanks for dropping by.
[12:09 PM PT]  Lyr Lobo: bye Jamie!
[12:09 PM PT]  Jagga Meredith: gotta run
[12:09 PM PT]  Andrew Hellershanks: ok, Jagga. See you next week.
[12:09 PM PT]  Orbert.Tatham Talk to you all later
[12:10 PM PT]  Kayaker Magic: Buy all!
[12:10 PM PT]  Cuga.Rajal Gavin, I;m trying to get a date for the server version hash you posted, b70b5c0. Got not happy with that as short hash
[12:10 PM PT]  Cuga.Rajal Git*
[12:10 PM PT]  Andrew Hellershanks: With so many people needing to head out it is about time to close the meeting for today.
[12:10 PM PT]  Lyr Lobo: Thank  you!  Take care!
[12:11 PM PT]  Andrew Hellershanks: Any last comments before I officially draw this gathering to a close?
[12:11 PM PT]  Gavin.Hird Cuga, the info I have is the same as you see on the Sim ver script
[12:11 PM PT]  Vincent.Sylvester Remember next week and the week after will be ealier for the EU folks, then it's back to normal times after the 30th
[12:11 PM PT]  Lyr Lobo: *smiles*
[12:11 PM PT]  Cuga.Rajal I believe you, just wondering why got can;t resolve it
[12:12 PM PT]  Ubit Umarov: 31 not 30
[12:12 PM PT]  Andrew Hellershanks: Ah, yes. The twice yearly time change.
[12:12 PM PT]  Ubit Umarov: we wil change at 31
[12:12 PM PT]  Lyr Lobo: we went forward March 10
[12:12 PM PT]  Andrew Hellershanks: North America changed this past weekend.
[12:13 PM PT]  Lyr Lobo: Have a great week!
[12:13 PM PT]  Ubit Umarov: you 2 :)
[12:13 PM PT]  Gavin.Hird IDK how the code was deployed to this sim, but it is probalby built on it, so it is most likely a local hash
[12:13 PM PT]  Andrew Hellershanks: The people attending this meeting have gotten good about handling the time change.
[12:13 PM PT]  Andrew Hellershanks: I'll call this meeting to a close. Thank you all for coming. See you again next week.
[12:13 PM PT]  Cuga.Rajal ah ok, so that would confirm that the version can't be derived directly from that...
[12:13 PM PT]  Gavin.Hird I will not be here next week, and the week after I should have internet again
[12:14 PM PT]  Andrew Hellershanks: ok, Gavin.
[12:14 PM PT]  Gavin.Hird no the hash could be anything really
[12:14 PM PT]  Lyr Lobo: Thank goodness on the return of Internet, Gavin
[12:14 PM PT]  Andrew Hellershanks: Hm... can you handle time without Internet access? You might have to go and spend some time outside. ;)
[12:14 PM PT]  Lyr Lobo: I cannot imagine life without it
Personal tools
About This Wiki