Talk:Debian/Archive 8
 | This is an archive of past discussions. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page. |
| Archive 5 | Archive 6 | Archive 7 | Archive 8 | Archive 9 | Archive 10 |
Table usability
I have reverted this recent edit to Debian. "People have small screens" does not seem a valid reason because "Release date" is longer than those dates. Furthermore, the timeline image below has a width of 860 pixels, wider than the table. Could Derianus provide more details about the display device on the article's talk page? 84.127.80.114 (talk) 06:50, 20 September 2014 (UTC)
@"People have small screens" does not seem a valid reason because "Release date" is longer than those dates. - But "Release date" has a space in the middle which enables display software to insert a new line. The old format is more concise. The table has seven columns, so it takes up a lot of space already.
Wikipedia:Manual of Style/Dates and numbers#Dates and years allows this format for tables. The object containing the dates in question is a table. Derianus (talk) 09:56, 20 September 2014 (UTC)
- I advise Derianus to familiarize with the WP:BOLD, revert, discuss cycle and leave the status quo up, i.e., revert to the previous version. I would also recommend WP:Edit warring, given the activity in Regions of Ukraine.
- Derianus is not addressing my concerns. Is the editor actually seeing the alleged problem? "Release date" is inside
<span class="nowrap">. Inserting a new line means ignoring the CSS and that would mean ignoring<span style="white-space:nowrap;">for the dates too. - The editor may have noticed the "Supported until" column. There are dates which cannot follow the yyyy-mm-dd format. What is Derianus' suggestion to solve the format inconsistency in the table?
- (By the way, using the "Engineering and technology" link at the top of this page would really improve this article.) 84.127.80.114 (talk) 01:39, 21 September 2014 (UTC)
- Why handing out advises to others that you don't follow yourself? I simply reverted to the ISO 8601 format. Derianus (talk) 04:10, 23 September 2014 (UTC)
- Derianus is refusing to address my concerns. There is consensus for the mdy format: the bold edit using words was not disputed or reverted by subsequent edits; I revised that edit later in the good article review without dispute, reaching a new consensus; I have revised the date format specifically.
- There is no evidence of consensus change. The other editors are not participating or reverting, but given that they are not answering my calls to improve this article either, I do not find silence to be a valid reason. They are busy and maybe they trust the current effort, which Derianus is invited to join in.
- I repeat, what is Derianus' suggestion to solve the format inconsistency in the table? 84.127.80.114 (talk) 00:22, 24 September 2014 (UTC)
- "I do not find silence to be a valid reason" - you take silence as you like. E.g. when it was changed in the direction you prefer you take it as evidence for new consensus. Now, you do exactly the opposite. Derianus (talk) 20:21, 26 September 2014 (UTC)
- Well, at least I am disputing Derianus' edit; silence is definitely not a valid reason now. I do not remember using only silence to support the word-based date format, which is also supported by consensus through editing.
- I would like to note that Derianus' arguments are focused on my reasoning consistency exclusively. I do not mind being in question, but some words concerning the table would be appreciated. 84.127.80.114 (talk) 23:26, 26 September 2014 (UTC)
- "I do not find silence to be a valid reason" - you take silence as you like. E.g. when it was changed in the direction you prefer you take it as evidence for new consensus. Now, you do exactly the opposite. Derianus (talk) 20:21, 26 September 2014 (UTC)
- Why handing out advises to others that you don't follow yourself? I simply reverted to the ISO 8601 format. Derianus (talk) 04:10, 23 September 2014 (UTC)
- WP:SILENCE does not apply to this as an editor has opposed both sets of changes - a new consensus is only reached when a new undisputed version of the article is created.
- Table: It is inconsistent to have one style of date in one column and one in another. If there is a policy regarding the use of one particular format in a column, then IMO both should use that format otherwise it would be based on preference. MOS:DATEFORMAT shows both are acceptable in tables. It is a stylistic issue rather than a major problem and I would select the template, dts, abbreviated date - testing on a 5" screen shows that neither format results in more columns/text being displayed. To get more displayed the nowrap on "Release date" would need to be removed - then the new format would be better justified.
- mthinkcpp (talk) 12:55, 28 September 2014 (UTC)
- "It is inconsistent to have one style of date in one column and one in another." - Maybe you can play the magician and add exact dates to the table where they are missing? As long as you cannot, the inconsistency cannot be removed without replacing exact days with month values in the other column. Derianus (talk) 21:06, 28 September 2014 (UTC)
- Since Derianus cannot solve the inconsistency without discarding information, the user should revert to the mdy format until the missing exact dates are provided. The mdy and month-year formats are both word-based. 84.127.80.114 (talk) 21:39, 28 September 2014 (UTC)
- "It is inconsistent to have one style of date in one column and one in another." - Maybe you can play the magician and add exact dates to the table where they are missing? As long as you cannot, the inconsistency cannot be removed without replacing exact days with month values in the other column. Derianus (talk) 21:06, 28 September 2014 (UTC)
I love the chaotic sorting in the timeline table (mixed legend row, TBAs, "supported until" column), although it may be my browser. It is no secret that I am in favor of providing humorous content to happy readers. I had another Easter egg proposal for Debian (not mentioned), but perhaps other users may want to contribute. How about a picture of Debian headquarters? 84.127.80.114 (talk) 17:21, 7 October 2014 (UTC)
yyyy-mm-dd limitation
I still believe Dsimic made a sensible decision. I have tried alternatives to express dates with month accuracy (c. 2000-10-01; 2000-10-??; 2000-10-00; 2000-10-); there is not a clean option. yyyy-mm is a valid ISO 8601 format, but is unacceptable as per WP:BADDATEFORMAT guideline. Word-based formats should be used.
However, this is exactly the same problem I am facing with citations in another article:
| Reference | date | accessdate |
|---|---|---|
| 1. Ghost in the Shell Preview | September 1997 | |
| 2. Ghost in the Shell | 1997-12-10 | 2013-09-03 |
Access dates are exact, but source dates may be reduced to month accuracy.
No citation in the Debian article has month accuracy. If a new source about Debian appears with this accuracy, should the month information be discarded because of a style issue? Should all the other dates be converted to mdy? This consistency problem is not covered by the manual of style. A decision should be made:
- Ignore the style inconsistency (current state).
- Stick to ISO 8601 and ignore the manual of style.
I am fine with either. 84.127.80.114 (talk) 03:02, 29 September 2014 (UTC)
- Hey everyone! FWIW, I prefer MDY dates in articles, and in text in general – in such "environments" MDY dates just look better and more readable to me. On the other hand, YYYY-MM-DD format is much more usable when it's about easier sorting etc., but that usually doesn't apply to articles. Also, the argument of small screens simply doesn't make sense to me, as there's little chance that a few characters more would cause readability issues. Just my $0.02. :) — Dsimic (talk | contribs) 04:08, 29 September 2014 (UTC)
- My position, keep all the information, but use MDY. It is reasonably clear when the day/month has been omitted. Also it is more consistent to use MDY and just omit parts than to use both YYYY-MM-DD and MDY. mthinkcpp (talk) 17:09, 29 September 2014 (UTC)
Why has this ended up in an ISO 8601 versus MDY discussion? This is not the place for such a debate and I will not participate (except for one comment). I thought that I raised a simple question: in an ISO 8601 context, how should dates with month accuracy be formatted? Removing the data does not solve the problem. I guess I will have to make the first decision: in an ISO 8601 context, the ISO 8601 format should be followed, i.e., YYYY-MM.
If Derianus is really interested in ISO 8601, then the user should know that this source is not reliable because it discourages YYYYMMDD, which is the basic format according to the standard, section 4.1.2.2. So, if width is an issue, the user should definitely push for this format.
I thought Wikipedia and Debian have to do with sharing. We have 77 localizations of the Debian article. Would it be nice if all these numerical data and timeline charts were shared across all these encyclopedias? Would it be useful if tables were moved to Commons and reused as templates, just like the media files? Could the dates be in a compact format such as YYYYMMDD and later localized in each Wikipedia?
Ultimately, this whole article could be moved to Commons, automatically translated, and only the localized featured and good content would be kept, besides other bits of regional interest. Maybe I am going too fast. Any thoughts about moving the release dates to the Wikidata page and using the #property parser function? 84.127.80.114 (talk) 17:21, 7 October 2014 (UTC)
- Hm, isn't the automated translation between different human languages still just a pipe dream? Btw, YYYY-MM format is, AFAIK, not allowed by the Wikipedia's guidelines as it's really confusing. — Dsimic (talk | contribs) 18:30, 7 October 2014 (UTC)
- No consensus in the guidelines.[1] If the context uses ISO 8601 to express a date, it should be obvious that yyyy-mm indicates an approximate date in that context, not a year range that would need an en dash. 84.127.80.114 (talk) 03:44, 19 October 2014 (UTC)
- 84.127.80.114 - personal attacks aside, I agree with moving the data to WikiData. This is much better for sharing between Wikipedias and sharing with other parties. It may make it harder for non-WMF-wikis as long as they cannot use the WD-data.Derianus (talk) 22:14, 9 October 2014 (UTC)
- It looks like #property is not ready for prime time. This is not the first time that interwiki solutions have been suggested.[2][3] However, I believe that we have the elements to implement a "Commons" for templates.
- Nevertheless, it would help to know exactly what problem Derianus is trying to solve. What table does the user maintain? What software is using Wikipedia's rendered HTML instead of wiki markup? 84.127.80.114 (talk) 00:32, 11 October 2014 (UTC)
- The problem of copying is solved (by using ISO 8601). The problem that users want to destroy the possibility to copy data in wiki code and data rendered as html remains. There are dozens?hundreds of tables in Wikipedia using ISO 8601 and that format is specifically allowed for being used in tables by MOSNUM. "yyyy-mm-dd limitation" is close to a hoax. Derianus (talk) 02:28, 11 October 2014 (UTC)
- Well, MDY isn't forbidden either. Also, how often do you need to copy the dates, or how often other users report that need? — Dsimic (talk | contribs) 02:46, 11 October 2014 (UTC)
- I already explained the section title and my initial question has been answered. I appreciate Derianus' contributions, specially this one, so I should give a warning. I do not know if Derianus is aware of the edits that has done. I do not know if the user is aware of the policies either. I do like the ISO 8601 format. However, if Derianus wants cross-article consistency, that wish might become true.
- If there is some software outside of Wikipedia that depends on rendered HTML, I may be able to help. 84.127.80.114 (talk) 22:19, 11 October 2014 (UTC)
- The problem of copying is solved (by using ISO 8601). The problem that users want to destroy the possibility to copy data in wiki code and data rendered as html remains. There are dozens?hundreds of tables in Wikipedia using ISO 8601 and that format is specifically allowed for being used in tables by MOSNUM. "yyyy-mm-dd limitation" is close to a hoax. Derianus (talk) 02:28, 11 October 2014 (UTC)
YYYY-MM-DD benefits
- better for small screens
- the only option for cross article consistency
- easier sharing of data between different Wikipedia langauge editions, since ISO 8601 is an international standard
- sortable without any extra templates or hidden javascript logic - WYSIWYG
- easier sharing of data between table data in Wikipedia and tabular data from other sources — Preceding unsigned comment added by Derianus (talk • contribs) 22:43, 29 September 2014 (UTC)
The small screen argument has been given as one benefit. It was countered by:
- "Also, the argument of small screens simply doesn't make sense to me"
Maybe that user never had looked at tables with several columns having a date format longer than ISO 8601 YYYY-MM-DD.
It is simple physics, more characters need more space on the screen and less need less. Derianus (talk) 22:29, 29 September 2014 (UTC)
- Yeah, I've looked at Wikipedia articles on small screens. Though, if you want to deal with more complicated matter, you simply need a bigger screen, right? There's no point in trying to fit a pumpkin into something that was made for an apple, so to speak. — Dsimic (talk | contribs) 22:53, 29 September 2014 (UTC)
- I've looked at that particular table, both editions and (as I said before) neither show more information. mthinkcpp (talk) 18:29, 30 September 2014 (UTC)
- Despite of this, basic laws of physics hold - a place occupied by one object cannot be occupied by another. Adjust the screen so, that the date cells are exactly filled by YYYY-MM-DD. Now try to fit MDY in the same cell, without altering any table boundary nor the size of font. Derianus (talk) 22:39, 6 October 2014 (UTC)
- At the same time, basic principles of economics also apply – as a result, even cellphones will soon have 4K screens. I'm not saying that having 4K screens in phones makes sense, but as a result in 2014 it's simply absurd to trade a few pixels for style and readability. — Dsimic (talk | contribs) 02:29, 7 October 2014 (UTC)
- Behold the laws of physics: the time is 17:21, 7 October 2014 (UTC). We control the horizontal and the vertical. Welcome to the Wiki Limits. 84.127.80.114 (talk) 17:21, 7 October 2014 (UTC)
- Despite of this, basic laws of physics hold - a place occupied by one object cannot be occupied by another. Adjust the screen so, that the date cells are exactly filled by YYYY-MM-DD. Now try to fit MDY in the same cell, without altering any table boundary nor the size of font. Derianus (talk) 22:39, 6 October 2014 (UTC)
- I've looked at that particular table, both editions and (as I said before) neither show more information. mthinkcpp (talk) 18:29, 30 September 2014 (UTC)
- "Cross-article consistency"- The Wikipedia Manual of Style points (mostly) to MDY. mthinkcpp (talk) 18:29, 30 September 2014 (UTC)
- MOS allows DMY and MDY and it prohibits switching articles from one form to the other. Using ISO 8601 in tables therefore is the only way to achieve cross-article consistency for tables. Derianus (talk) 13:48, 2 October 2014 (UTC)
- Not quite, it prohibits changing the date format without either consensus or an accepted reason mandating the change, i.e. They cannot be just changed to ISO..., MDY or DMY without consensus - although I think there is now consensus for changing to/using MDY. mthinkcpp (talk) 17:10, 7 October 2014 (UTC)
- @part 1 - OK; @part 2 - no, there is no consensus. And people have not shown why ISO 8601 YYYY-MM-DD in tables should be changed to MDY or YMD. Several benefits of ISO 8601 have been shown, counter arguments were mostly attacking the advantages, but failed, and only few disadvantages had be mentioned, e.g. ~"format inconsistency within the table" - but that does not exist.
- For the cross-article consistency - can you demonstrate that MDY is preferred by the MOS over DMY? Derianus (talk) 22:28, 9 October 2014 (UTC)
- Not quite, it prohibits changing the date format without either consensus or an accepted reason mandating the change, i.e. They cannot be just changed to ISO..., MDY or DMY without consensus - although I think there is now consensus for changing to/using MDY. mthinkcpp (talk) 17:10, 7 October 2014 (UTC)
- MOS allows DMY and MDY and it prohibits switching articles from one form to the other. Using ISO 8601 in tables therefore is the only way to achieve cross-article consistency for tables. Derianus (talk) 13:48, 2 October 2014 (UTC)
MDY limitations
While using ISO 8601 leads to cross article consistency for tables, and values can be copied between Wikipedias in different languages, MDY values cannot be copied nor will their usage in the Debian article lead to cross article consistency for tables.
Compare
Derianus (talk) 22:18, 29 September 2014 (UTC)
Cross-article consistency
I started a list at Wikipedia talk:WikiProject Software#Date format in release history sections of OS articles. Derianus (talk) 04:11, 31 October 2014 (UTC)
Dubious information in timeline
Some versions have a supported until date before release of the next version.
This applies only to values that are presented with month-precision. I fixed Potato. Hamm and Bo still have the dubious information.
Derianus (talk) 22:05, 28 September 2014 (UTC)
I removed
- 2.0 "Feb 1999" and
- 2.1 "Oct 2000 "
neither had a reference. Derianus (talk) 21:58, 29 September 2014 (UTC)
- In the meantime, let me play the magician. 84.127.82.127 (talk) 00:47, 4 November 2014 (UTC)