Talk:Tz database

From Wikipedia, the free encyclopedia
Jump to: navigation, search
WikiProject Time (Rated C-class, Mid-importance)
WikiProject icon This article is within the scope of WikiProject Time, a collaborative effort to improve the coverage of Time on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.
C-Class article C  This article has been rated as C-Class on the project's quality scale.
 Mid  This article has been rated as Mid-importance on the project's importance scale.


I'm questioning AIX being listed in the "Use in software systems" section. IBM has been unable to compile correct zoneinfo files for years. We found that out while testing the US 2007 DST patches and the more recent Australian patches. They are aware of it, but it is pretty obvious that nobody is using their Olson data.

This is what IBM has to say about it:

"Some applications rely on timezone information from zoneinfo files in /usr/share/lib/zoneinfo. However we are not aware of any common applications that use these. At this time updated zoneinfo files are not available."

And here, they come right out and say that as far as the OS is concerned, DST is hard coded:

"If the Daylight Savings Time option is enabled, the default in AIX is for the system time to move forward 1 hour (to DST) at 2:00am the first Sunday in April, and move back one hour (to Standard Time) at 2:00 a.m. on the last Sunday in October. Starting in 2007 this switch occurs on the second Sunday in March and ends the first Sunday in November. The default is hard coded and is not stored in any user accessible file..."

I respectfully suggest that a parenthetical note be added such as "(includes zdump and zic utilties, but the OS itself does not use the zoneinfo data)" or AIX simply be removed from this page entirely.


--Bob (talk) 13:05, 18 March 2008 (UTC)

Yep, thanks... I added this info (as a separate bullet point with refs. --Dfred (talk) 16:57, 18 March 2008 (UTC)
MySQL uses zoneinfo under AIX. I'll add that too. One citation suffices from IBM, and one for MySQL. Eubulides (talk) 17:22, 18 March 2008 (UTC)
The "DST is hard coded" citation really means "the system default is not configurable by normal users"; they're using POSIX time zone strings, which do let you change when the time shift occurs, as well as the amount of the shift. The administrator can change the system default value from the default, and the user can change their value (but not the system value):

"However, the date and time at which the switch to DST and ST occurs can be customized by root (global environment) or by users (user environment) by setting the $TZ environment variable."

In addition, they seem to be saying that, as of AIX 6.1, they support the Olson database in the OS:

"PLEASE NOTE: The default timezone format for AIX 6.1 and AIX 7 is Olson Time, this technote does not apply to Olson Time, but standard POSIX timezone formatting. I am including APARs that are for Olson Time below for completeness."

I'll update the page. Guy Harris (talk) 23:19, 11 March 2011 (UTC)
I am debugging GLIB-2.43.3 and am proposing a small change so that TZDIR, when not defined, for AIX, defaults to /usr/share/lib/zoneinfo, rather than /usr/share/zoneinfo as is common elsewhere. The reference is the output of the command

michael@x065:[/usr/include/sys]lslpp -w /usr/share/lib/zoneinfo

 File                                        Fileset               Type
 /usr/share/lib/zoneinfo                     bos.rte               Directory
While AIX did not use "olsen" before 2007, all the (smit) dialogs push you to "Olsen" as preferred layout. Any AIX system is a reference. I do not dare comment on how current the database is at most recent service pack (SP) for a particular TL (technology level). Intermediate fixes (ifix) have been released, in particular for DST changes, in the past.
deleted a bunch of my own info - a better telling - as it answers the question I had when I found this page re: the files in /usr/share/lib/zoneinfo before AIX 6.1 (see paragraph AIX (5.3 and earlier)) [1]
Being new to wiki-edits, if I have been too talky - feel free to clean-up. MichaelAM (talk) 13:02, 2 February 2015 (UTC)

Astrolabe Lawsuit[edit]

Blog post covering the lawsuit: [1]

And a slashdot article and discussion on the lawsuit: [2] — Preceding unsigned comment added by Leewalton (talkcontribs) 10:53, 7 October 2011 (UTC)

Section discussing the lawsuit: Tz database#2011 lawsuit. Guy Harris (talk) 19:05, 7 October 2011 (UTC)

TZ Strings[edit]

I think a page (linked from this one) on TZ strings might be useful, though I'd not venture to write one. (talk) 19:17, 30 October 2008 (UTC)

Tz database#Names of time zones describes the rules for constructing time zone names for the database. Category:Tz database is a category that includes a pile of individual Wikipedia pages for individual zone names and redirects to locations. List of tz database time zones gives a list of zone names. Guy Harris (talk) 00:54, 7 October 2011 (UTC)

A literary appreciation of the Olson/Zoneinfo/tz database[edit]

mailto links[edit]

There's a load of references for the tz mailing list which are links for an email address. Is that useful? (that they are hyperlinks for mailto) KayEss | talk 15:33, 28 October 2009 (UTC)

They appear to be gone. Guy Harris (talk) 00:48, 7 October 2011 (UTC)

DST rules[edit]

Somehow, somewhere, I need a section about DST rules. -DePiep (talk) 22:40, 11 March 2011 (UTC)

What about them? If you mean "a section that describes how you specify DST rules", just find a copy of the zic(8) or zic(1m) man page somewhere on the Intertubes and link to it. Guy Harris (talk) 22:59, 11 March 2011 (UTC)
No. I asked: what are the DST procedures? Then, maybe the wiki pages you mention zic(8) or zic(1m) describe & answer it. (Now serious: where does this wiki describe the DST rule for e.g. the European Union?) -DePiep (talk) 23:08, 11 March 2011 (UTC)
Those aren't wiki pages, they're man pages on various UN*Xes. To see DST rules for various countries in Western Europe (not all of which are members of the EU, and, as it's Western Europe, not all countries in the EU are there), see Western European Summer Time. See other pages linked to by Time in Europe (in the template) for other countries. Those might not give historical DST rules for all the countries in question. Guy Harris (talk) 00:47, 7 October 2011 (UTC)
So, tz database does not define DST rules? When is starts, when it ends, and how much the shift is? (of course this is defined per year and per zone). -DePiep (talk) 10:55, 28 February 2012 (UTC)
────────────────────────────────────────────────────────────────────────────────────────────────────The tz database does contain information about DST rules. However, this article does not need to (and probably should not) duplicate the entire database. If you want to try to work in a link to Daylight saving time by country somewhere, that might be a worthwhile addition. Anomie 13:20, 28 February 2012 (UTC)
This is the first ref in the article: Eggert, Paul (2007-11-29). "Sources for time zone and daylight saving time data". Retrieved 2007-12-03.  Unknown parameter |coauthors= ignored (|author= suggested) (help) (but this might not be the true definition source, though the authors names ring a bell). The link says in paragraph 1: It [tz Database] is updated periodically to reflect changes made by political bodies to time zone boundaries, UTC offsets, and daylight-saving rules. More than just interestingly, some links are to ... WP-pages. But let's assume that DST-rules are not in the tz database. Then how (by what data source) does a Linux computer automatically and correctly adjust to and from DST? -DePiep (talk) 13:43, 28 February 2012 (UTC)
This is a lead: [3]. I did google using "-wikipedia". -DePiep (talk) 13:48, 28 February 2012 (UTC)
That page just refers to the tz database (the "DST rule name"s are the names used in the tz database source files, and the "TZ zone names" are the names given to the zones in the tz database); that page is not a data source that acts as an alternative to the tz database, it just gives a summary of the currently in effect part of the tz database specifications for various locales (not the historical part). Guy Harris (talk) 18:52, 28 February 2012 (UTC)
If the DST rules aren't in the database, then a system that uses the tz database (whether it's a Linux distribution or a Solaris system or a *BSD system or a (Mac) OS X system or an iOS system or...) will probably not correctly adjust to and from DST unless the TZ environment variable is set to a POSIX-specified string that also contains the right rules (which is not possible in the general case - the POSIX-specified strings (see the description of the "TZ" environment variable on this page - that page will probably require you to open a free account with the Open Group if you don't already have one) can't handle arbitrary changes in the rules over time). Guy Harris (talk) 18:47, 28 February 2012 (UTC)
Well, it was just a checking question and yo answered it. I don't want to read "what are the DST rules in Europe", of course. I want to read or add or have added: what DST rules does tz database provide, and how?-. DePiep (talk) 21:00, 28 February 2012 (UTC)
Hi there, DePiep... For the actual human-readable rules that make up the database, refer to the contents of tzdata-latest.tar.gz from the official IANA ftp site. Inside that gzip compressed tar file you will find several files that can be compiled with the 'zic' command (as noted above). If you want to know the method by which the zones are actually defined (including DST calculations) the information you seek is there. Since these files are in the public domain, I agree that a small, well-chosen excerpt might help to demystify how the database is actually constructed. Regards, --Dfred (talk) 21:34, 28 February 2012 (UTC)

(edit conflict) (will read after)

If you were not this kind, I'd have lost my patience.
I am not here with a helpdesk question. My Q is not: "what are the EU DST rules?". I am here to ask: (now reread my bold above please). I claim: this article, called tz database for a reason, should point to and describe the DST rules tz database provides. -DePiep (talk) 22:11, 28 February 2012 (UTC)
The page points to the rules, in the sense of "the source files that are compiled into the binary files used by software", by linking to the IANA time zone database page, which has a "Latest version" section with a download link for the tzdata gzipped tarball; that tarball contains all the source files. (The page obviously should not itself contain all 13000+ lines of time zone database source code, or even all 5000+ lines of non-comment time zone database source code.)
If by "describe the DST rules" you mean "describe the syntax of the rules", whether it should contain the full syntax, as described in the zic(8) man page (which is now linked to in the "External links" section), is unclear, but a summary and examples, of the sort User:Dfred gave, might be useful Guy Harris (talk) 01:27, 29 February 2012 (UTC)
OK, I've added a brief summary of "zone lines" in the "Definition of a time zone" section and made the "Daylight Saving Time (DST) rules" section describe "rule lines". Those cover the high points of the time zone source file syntax. Guy Harris (talk) 02:07, 29 February 2012 (UTC)

Inclusion of example from human-readable definition files to show how zones are actually defined?[edit]

What do others think of including a small excerpt of the human readable data from the tzdata file? Ideally it would include enough entries to fully see how the different Rule and Zone lines relate to each other. Here's an example of what I was thinking: (perhaps too big...)

(NOTE: The following text is declared to be in the public domain according to the header of the file from which it was taken, "northamerica" in tzdata2011n.tar.gz.)

# Rule  NAME    FROM    TO      TYPE    IN      ON      AT      SAVE    LETTER/S
Rule    US      1918    1919    -       Mar     lastSun 2:00    1:00    D
Rule    US      1918    1919    -       Oct     lastSun 2:00    0       S
Rule    US      1942    only    -       Feb     9       2:00    1:00    W # War
Rule    US      1945    only    -       Aug     14      23:00u  1:00    P # Peace
Rule    US      1945    only    -       Sep     30      2:00    0       S
Rule    US      1967    2006    -       Oct     lastSun 2:00    0       S
Rule    US      1967    1973    -       Apr     lastSun 2:00    1:00    D
Rule    US      1974    only    -       Jan     6       2:00    1:00    D
Rule    US      1975    only    -       Feb     23      2:00    1:00    D
Rule    US      1976    1986    -       Apr     lastSun 2:00    1:00    D
Rule    US      1987    2006    -       Apr     Sun>=1  2:00    1:00    D
Rule    US      2007    max     -       Mar     Sun>=8  2:00    1:00    D
Rule    US      2007    max     -       Nov     Sun>=1  2:00    0       S
# Rule  NAME    FROM    TO      TYPE    IN      ON      AT      SAVE    LETTER
Rule    NYC     1920    only    -       Mar     lastSun 2:00    1:00    D
Rule    NYC     1920    only    -       Oct     lastSun 2:00    0       S
Rule    NYC     1921    1966    -       Apr     lastSun 2:00    1:00    D
Rule    NYC     1921    1954    -       Sep     lastSun 2:00    0       S
Rule    NYC     1955    1966    -       Oct     lastSun 2:00    0       S
# Zone  NAME            GMTOFF  RULES   FORMAT  [UNTIL]
Zone America/New_York   -4:56:02 -      LMT     1883 Nov 18 12:03:58
                        -5:00   US      E%sT    1920
                        -5:00   NYC     E%sT    1942
                        -5:00   US      E%sT    1946
                        -5:00   NYC     E%sT    1967
                        -5:00   US      E%sT

I've included the 'US' rules to make it clear what is being referred to in the actual Zone definition. The older historical US rules could probably be pared down a bit if desired -- or a more compact example could be used. However I think any example chosen should show at least a few changes of DST rules, etc. to give a sense of how these are handled. Thoughts? --Dfred (talk) 22:06, 28 February 2012 (UTC)

Into the article asap! I created the DST section already. Though I must say, the article does not separate file-content from tz-definitions that well yet. -DePiep (talk) 22:25, 28 February 2012 (UTC)
OK, I've put that stuff in as an example, and linked to Daylight saving time in the United States#History of DST in the United States to indicate why there are so many zone and rule lines (this stuff's a lot more complicated than just "offset from DST, date and time when DST starts, date and time when DST ends", as the UN*X APIs such as localtime() and mktime() have to handle times in the past and the future). Guy Harris (talk) 02:33, 29 February 2012 (UTC)
One of the things that people trying to understand the source files tend to find very confusing is that Rule lines specify individual transitions (which can and often do repeat from one year to the next), and not periods of DST observance (which might also repeat over time). A very frequent question on the mailing-list is "Rule set X has more rules for entering DST than for returning to standard time. What does that mean?" In this it reflects the needs and organization of the binary format at the expense of some confusion when reaing the source. (This has become more of an issue as implementors using other runtime environments have chosen to parse the source files rather than reading the architecture-independent binary files and allowing zic(1) to do the heavy lifting, which in turn has required the specification of the source language to be made more explicit than "whatever zic accepts".) 121a0012 (talk) 02:54, 12 May 2012 (UTC)
Please stop explaining what people do not understand. My point was (and is): does tz_database describe (define) DST yes or no? If yes, then describe in the article. If no: even so. Why do knowing people explain here on Talk page stuff that should be in the Article? -DePiep (talk) 23:24, 14 May 2012 (UTC)

tz into IANA[edit]

What happened? Can the encyclopedia describe the issue & the process? -DePiep (talk) 00:22, 14 May 2012 (UTC)

The history section summarizes the situation: there was the Astrolabe lawsuit, Olson et al weren't in a position to fight it themselves, so they handed it over to the IETF/ICANN/IANA (which may have been in the plans anyway). I have no idea where the move of the article to "IANA time zone database" might have come from, though. The IANA page just calls it "Time Zone Database". I'd support a move back pending discussion. Anomie 02:44, 14 May 2012 (UTC)
I would also support a move back to the previous article title. IANA's current stewardship of the TZ Database does not seem to constitute a change of name. In fact RFC 6557, referred to in the edit summary of the move, contains the following text in its Terminology section:
TZ Database: The Time Zone Database, sometimes referred to as the "Olson Database". This database consists of information about offsets from UTC for different ocalities, including daylight saving time (DST) transition information.
RFC 6557, 1.1. Terminology
--Dfred (talk) 12:27, 14 May 2012 (UTC)

Internal names in some Terminology section are not specifically made for being used as Wikipedia article names. They can of course be used as such, but this is by no means mandatory. TZ is cryptic, fine for insiders, but not for general audience. WIPO -> World Intellectual Property Organization, USA -> United States. Time zone database is ambiguous, since it is a generic name. "The first choice below connects to the copy of the IANA Time Zone Database maintained by the IETF Secretariat." But of course in Wikipedia the uppercase spelling is unwanted, it is not Wikipedia style to capitalize. The RFC draft is name IANA timezone database.

It goes along the lines of:

"IANA time zone database" is documented, perfectly describes the content of the page, is unambiguous, and fits well with similar terms, has no issue with capitalization (tz in URL converted to Tz). Royaume du Maroc (talk) 15:22, 14 May 2012 (UTC)

Thanks for the reply. I do not think anyone is suggesting the new article title is incorrect. However for the purposes of an article title, I would argue that the common names of the subject of this article have been, and still are, either "tz database" or "zoneinfo". The continued use of the "tz database" name as official terminology in current (post-IANA) RFCs, as well as within the code and database itself, seems like more than a mere technicality. While the generic term "Time Zone Database" could possibly be ambiguous, I wouldn't support that as an article title. The new article title seems like a case of over-precision. I don't believe there's any ambiguity in using "tz database". I do agree that capitalization is an issue with the "tz" name, but a limitation of Wikipedia doesn't change the common name of the subject.
I certainly support an alt/long name redirect from IANA time zone database to tz database for cases where using it for WLs is helpful or makes things more clear. For instance, perhaps in the individual tz timezone articles (e.g. Africa/Johannesburg, etc.) it makes sense to use the longer form to give context. I'd be interested to hear some other opinions on the matter, though... --Dfred (talk) 22:44, 14 May 2012 (UTC)
To determine an article name it is more important to cite Wikipedia conventions than to cite some technical third party paper.
Natural disambiguation: If it exists, choose a different, alternative name that the subject is also commonly called in English, albeit, not as commonly as the preferred but ambiguous title (do not, however, use obscure or made up names).
The current official name is "Time Zone Database" and this name is ambiguous. Prefixing with "IANA" and ambiguity has gone. It's the same with "airport code". One says "ICAO airport code" or "IATA airport code".
On your user page it says you are using Linux, but Wikipedia is not only for tech people. "tz" is not a string that is understandable to most readers. But "time zone" is understandable. "tz" can refer to things in other domains, e.g.: .
There is also no over-precision. Over-precision would mean one can drop a part of the name and the topic would still be the same. One cannot drop "IANA" since then the title would be generic. One cannot drop "time", since then other IANA zone data might be referred to. One cannot drop database, since the article is not only about what a IANA time zone is. Maybe, one can drop "zone" -> "IANA time database". Anyone who would want that?
Since external terminology documents don't mandate what is the common name, nor do topic insiders, it leaves the subject to a broader audience:
Expand "tz" to "time zone" and use natural disambiguation!
Royaume du Maroc (talk) 01:12, 15 May 2012 (UTC)

Requested move - back to zoneinfo - name before Tobias Conradi first touched the page[edit]

The following discussion is an archived discussion of the proposal. Please do not modify it. Subsequent comments should be made in a new section on the talk page. No further edits should be made to this section.

The result of the proposal was no consensus to move. While this form was the one preferred by a banned editor, there's no consensus that the other title is inherently superior. Cúchullain t/c 20:45, 12 June 2012 (UTC)

Tz databasezoneinfo – This was the name before User:Tobias Conradi (see sock puppets, bans etc.) first touched the page. TimeZoneEditor (talk) 04:14, 31 May 2012 (UTC)

  • Oppose. This is an established name, and most links until the latest sockpuppet attack were to "tz database". It's often good to revert banned editors, but when their work has been the norm for a long time, reverting them would cause even more disruption and confusion. Therefore, the encyclopedia will be better off going with the established name, which it has right now. Nyttend (talk) 04:34, 31 May 2012 (UTC)
    But then Tobias Conradi may have got his way - despite being banned. How can it be allowed that any contribution by one of his socks remains? Especially see, how he apparently changed the perceived name
    Maybe he just performed a second page renaming to cover the first one, so that everybody is mislead into thinking the correct widely used name is "tz database". TimeZoneEditor (talk) 04:45, 31 May 2012 (UTC)
Let's do a simple cost–benefit analysis — does it hurt more to permit this to stand or to move everything around (if we move this, we need to change the way that links are displayed in tons of articles) throughout the encyclopedia? As well, a simple Google search isn't a good argument because of (1) the extent to which results can vary from the estimated totals that you quoted [not a complaint at you; Google simply isn't very rigourous when it comes to estimating numbers of links], and (2) many of those pages aren't reliable sources. What counts is the predominant usage in reliable sources, overlooking the tons of non-reliable pages. TC is banned, but that doesn't mean that we must get rid of his edits: it means that we may get rid of them if they cause problems, but in situations like this where there aren't substantial problems, it sometimes helps the encyclopedia more to be able to keep them. Nyttend (talk) 05:18, 31 May 2012 (UTC)
  • Oppose. The policy wrt reverting banned user edits is described in Wikipedia:Banning_policy#Edits by and on behalf of banned editors (and rephrased in the tableat Difference between bans and blocks). I read this as follows: since the name tz database is verifyably correct, there is no must to undo that edit. It is not a personal creation any more. this also prevents that we would be forced to degrade the page by removing correct but banned text. The disruptor could have their way by tricking us into a wrong title? That said, there may be other reasons to change the title, which we are free to discuss. -DePiep (talk) 07:02, 31 May 2012 (UTC)
  • Support move to anything away from this crappy, meaningless, miscapitalized title that does not convey the meaning of the content of the article itself.--Kintetsubuffalo (talk) 06:11, 2 June 2012 (UTC)
The above discussion is preserved as an archive of the proposal. Please do not modify it. Subsequent comments should be made in a new section on this talk page. No further edits should be made to this section.

Time Zone[edit]

The Article, and Olson, should use Time Zone (and similar) only in accordance with its normal definition, in which only winter Time is considered. I suggest that Olson, and then the Article, should use "Time Region" or similar to mean a set of places with a matching winter offset from UTC/GMT and matching Summer Time Rules. (talk) 18:36, 18 December 2012 (UTC)

"Olson" isn't in charge any more - he still participates, but has retired from the position of chief maintainer; Paul Eggert now has that job. If you want the maintainers, and others on the mailing list, to choose a different name, feel free to ask the mailing list; whether a name change will happen is, of course, another matter. (I personally tend to refer to "tzdb zones", but that's just me.) The article itself shouldn't invent a name for the zones. Guy Harris (talk) 00:36, 6 July 2016 (UTC)

"Use in software systems" section[edit]

Today I added luatz into the listing of libraries that use the tz database, but Alexf reverted it on grounds of Wikipedia is not a collection of links. If that is the case, should the section be removed entirely? — Preceding unsigned comment added by Daurnimator (talkcontribs) 18:25, 18 August 2014 (UTC)

History section[edit]

It seems strange to me that one of the first things you read is about the lawsuit. Can we move the history section to the bottom since it is less important? — Preceding unsigned comment added by Mj1856 (talkcontribs) 00:58, 12 December 2014 (UTC)

I concur. Moved - David Gerard (talk) 14:32, 13 December 2014 (UTC)