Wiki Edit War on using/avoiding semicolon lists

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
124 messages Options
1 ... 34567
Reply | Threaded
Open this post in threaded view
|

Re: Wiki Edit War on using/avoiding semicolon lists

Никита
Well not perfect solution at the moment, but least I don't need to teach somebody regexes: "color:green"=yes | "color:lightgreen"=yes | "color:bluegreen"=yes | ...

"But how this is different from regexes?" you would say.

1. There no regexes at all
2. There should be pages about http://wiki.openstreetmap.org/wiki/Tag:color:bluegreen=yes redirecting to page with actual scheme or defition in user language...
3. No performance penalties from regexes

Alternatively you may always use multiple tags
color:green=yes
color:green:lightgreen=yes

Nobody restricts you with tagging precision here. Actually we don't have such problem beoynd tagging list, our tags in OSM are way more easier: http://wiki.openstreetmap.org/wiki/Key:fuel

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Wiki Edit War on using/avoiding semicolon lists

Marc Gemis
I don't know the complete list of green variations. You assume there are only three, there are many more, in order to catch them all you need color:*green*=yes, a (simple) regular expression. Any other answer you come up with is incomplete in my opinion. 

Coming back to the traffic calming example. traffic_calming:table=yes and traffic_calming:choker=yes.
When I want to know all traffic calming features I need traffic_calming:*=yes, again a regular expression.

Even with your schema you will not be able to avoid that people will need regular expressions to express some queries. I am not going to look for other examples. I close the discussion from my side now, and accept that you disagree. And that neither of us will be able to convince the other that their solution is the silver bullet. Raw data  (e.g. try to obtain all information about a bus route, and watch help.osm.org for some of the other real data needs ) will always be too difficult to cope with for some people and they will always have to rely on software that makes it easier for them to work with it. That software can always be written by people that understand, data, programming and regular expressions.

regards 

m

On Sat, Jan 24, 2015 at 8:32 AM, Никита <[hidden email]> wrote:
Well not perfect solution at the moment, but least I don't need to teach somebody regexes: "color:green"=yes | "color:lightgreen"=yes | "color:bluegreen"=yes | ...



_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Wiki Edit War on using/avoiding semicolon lists

Paul Johnson-3
In reply to this post by Никита
On Tue, Jan 20, 2015 at 7:51 PM, Никита <[hidden email]> wrote:
Wow. Quality of discussion here.

I even find the second example more difficult to visualize. It's just worse than the first in every respect

payment=efectivo;visa;mastercard;american␣express
payment=mastercard;visa;efectivo

Now try to find efectivo with your regexes.

If you want to tell me something about /.*efectivo.*/ you have no idea what OSM about and how regexes work. Say hello to

payment=mastercard;visa;efectivonotinspain

There programmers out there but I'm an idiot and cannot teach how to write REGEX to casual user. They are true heros OSM need: http://xkcd.com/208/.

How to query for "payment:efectivo"="yes"? I definitely need regex here. Please, help me!

Most people (and I don't mean just OpenStreetMap-aware types) don't care about the data model.  That's for the application to deal with.  So, this is a non-issue for anybody that isn't working directly with the raw data.

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Wiki Edit War on using/avoiding semicolon lists

Никита
In reply to this post by Marc Gemis
in order to catch them all you need color:*green*=yes, a (simple) regular expression.
No, I don't. I use tags

color:green=yes
color:green:lightgreen=yes

And will search for "color:green"=yes while I wait for native multivalue tags.

Even with your schema you will not be able to avoid that people will need regular expressions to express some queries.
Unjustified statement. Read above. There no need in regexes if you know set logic and how to use and document tags properly. 

Clueless people should not enforce their regexes to everyone in OSM. 

The end.

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Wiki Edit War on using/avoiding semicolon lists

Никита
In reply to this post by Paul Johnson-3
in order to catch them all you need color:*green*=yes, a (simple) regular expression
I don't need any regex. STOP.

color:green=yes
color:green:lightgreen=yes

"color:green"="yes" query is dead simple.

Even with your schema you will not be able to avoid that people will need regular expressions to express some queries.
What. 

So, this is a non-issue for anybody that isn't working directly with the raw data.
What. Do you ever open JOSM? 
  • we can provide newbies them with link to wiki. 
  • we don't need to teach every person how to parse "japanese" from cuisine=mexican;japanese using f#$% regexes
  • we can provide clear definition at wiki page for iD or JOSM developers with description of tag instead of guessing by taginfo stats EVERY time they want to adjust something in presets
  • custom strings in editors or JOSM presets are easier to add
  • we get benefits from taginfo stats by using xxx:yyy=yes
  • advanced set querying for users, 
  • reduced cpu load for database because there no need to compute smart regexes

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Wiki Edit War on using/avoiding semicolon lists

Martin Vonwald (Imagic)
In reply to this post by Никита


2015-01-24 14:29 GMT+01:00 Никита <[hidden email]>:
Clueless people

Once again I want to thank you for your kind words.
 

The end.

Any chance, that you will follow this rule anytime soon?

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Wiki Edit War on using/avoiding semicolon lists

Nelson A. de Oliveira
In reply to this post by Никита
On Sat, Jan 24, 2015 at 11:36 AM, Никита <[hidden email]> wrote:
> reduced cpu load for database because there no need to compute smart regexes

You know that it's always a trade-off, right?
While the CPU usage *could* be lowered, disk usage/IO and network
traffic could increase.
There is no magic with your approach.

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Wiki Edit War on using/avoiding semicolon lists

Richard Welty-2
On 1/24/15 10:58 AM, Nelson A. de Oliveira wrote:
> On Sat, Jan 24, 2015 at 11:36 AM, Никита <[hidden email]> wrote:
>> reduced cpu load for database because there no need to compute smart regexes
> You know that it's always a trade-off, right?
> While the CPU usage *could* be lowered, disk usage/IO and network
> traffic could increase.
> There is no magic with your approach.
furthermore, there is no reason to assume that the computation is
being done on the same system; in fact, it is unlikely that any regex
computation is being done on the DB server.

i find the case for this unconvincing, it seems mostly motivated by
an intense dislike for regular expressions. but the regex has been
a standard technique for decades; i suggest learning to deal with
it.

richard

--
[hidden email]
  Averill Park Networking - GIS & IT Consulting
  OpenStreetMap - PostgreSQL - Linux
  Java - Web Applications - Search


_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Wiki Edit War on using/avoiding semicolon lists

Никита
You know that it's always a trade-off, right?
Exactly. Regex advocates are ponies in DB design.

disk usage/IO 
Index lookup for "color:green:lightgreen"=yes is fast.
full table scan just to compute regex for each value is not

Or wait do you have an custom DB for OSM tailored both for regexes and geo-queries?

> network traffic could increase.
1-2 more tags per object? So? Yes, it will increase traffic load.

There no option to choose between if somebody choose to use ";" in value. You have to use regexes. FYI, this is not sane thing to do with relational databases. I'd rather add more indexes that will search for answers like these: http://dba.stackexchange.com/questions/10694/pattern-matching-with-like-similar-to-or-regular-expressions-in-postgresql

Do you get my point now about technical aspect of this? We need multivalue tags, there easier way to avoid problem: avoid multiple values in value part of key=value.

 i suggest learning to deal with it.
Are you an idiot? I mean really. Try answer these points:
  • we can provide clear definition at wiki page for iD or JOSM developers with description of tag instead of guessing by taginfo stats EVERY time they want to adjust something in presets
  • custom strings in editors or JOSM presets are easier to add

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Wiki Edit War on using/avoiding semicolon lists

Richard Welty-2
On 1/24/15 11:20 AM, Никита wrote:
ltivalue tags, there easier way to avoid problem: avoid multiple values in value part of key=value.

 i suggest learning to deal with it.
Are you an idiot? I mean really.
ok, i'm done here.

richard
-- 
[hidden email]
 Averill Park Networking - GIS & IT Consulting
 OpenStreetMap - PostgreSQL - Linux
 Java - Web Applications - Search

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Wiki Edit War on using/avoiding semicolon lists

Martin Vonwald (Imagic)
In reply to this post by Никита


2015-01-24 17:20 GMT+01:00 Никита <[hidden email]>:
Are you an idiot? I mean really.

I hereby request a ban of this individual from this mailing list and I definitively support an OSM-wide ban.

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Wiki Edit War on using/avoiding semicolon lists

Jo-2
In reply to this post by Никита
>>  i suggest learning to deal with it.
>Are you an idiot? I mean really. Try answer these points:

Insulting people will get you nowhere at all. If you want to be able to perfom index searches, then import the data in tables with fields allowing you to do so. That's the great thing about open data. You can work with it any way you like.

By now, it should be clear you're not finding any support here for converting semicolon delimited lists to some more verbose format.

The only consensus I see emerging is that we should avoid them where this is practical and possible.

Jo

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Wiki Edit War on using/avoiding semicolon lists

Никита
In reply to this post by Richard Welty-2
"I suggest you to deal with it" (sic!)

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Wiki Edit War on using/avoiding semicolon lists

Ilpo Järvinen
In reply to this post by Никита
On Sat, 24 Jan 2015, Никита wrote:

> > You know that it's always a trade-off, right?
> Exactly. Regex advocates are ponies in DB design.
>
> > disk usage/IO 
> Index lookup for "color:green:lightgreen"=yes is fast.

So is the index lookup for color=...;lightgreen;... !

> > network traffic could increase.
> 1-2 more tags per object? So? Yes, it will increase traffic load.
>
> There no option to choose between if somebody choose to use ";" in value.
> You have to use regexes. FYI, this is not sane thing to do with relational
> databases.

This is a false statement. Nobody (except you) is forcing anyone to store
semicolon values as literals rather than e.g. into a postgresql array.
Your argument entirely *depends* on that stupid way of loading multivalue
into the DB. I wish you'd really drop the discussion about DBs and using
regexp there as it's not the strongpoint of your argument really.

There are good points in your dislike of semicolons but they are certainly
not on the DB side.


--
 i.
_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Wiki Edit War on using/avoiding semicolon lists

Никита
Insulting people will get you nowhere at all.
Well there over hundred messages and some people don't dare to study topic, I was repeating my messages multiple times already.

> If you want to be able to perform index searches, then import the data in tables with fields allowing you to do so. That's the great thing about open data. You can work with it any way you like.
Valid point, but also should suggest good practices for people who would like to benefit from default indexes:
API performance
overpass performance

These tools are quite popular right now and we should consider how people with work with data. JOSM is editor for raw osm data, not post-processes or indexed data.

not finding any support here for converting semicolon delimited lists to some more verbose format
I don't need support. People use this more verbose format right now for information they care about.

Problem was in outdated documentation at wiki and people at this tagging list...

>The only consensus I see emerging is that we should avoid them where this is practical and possible.
It was so long before discussion at tagging list. Somehow it become [hidden email].

 Nobody (except you) is forcing anyone to store semicolon values as literals rather than e.g. into a postgresql array. Your argument entirely *depends* on that stupid way of loading multivalue into the DB.
I see your point. Actually, let's talk about it. How "smart" loading script should know if key=* is multivalued or not? Should it parse every value in database?

";;;;" - 4 symbols? 2 escaped semicolons? how do you know?

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Wiki Edit War on using/avoiding semicolon lists

John F. Eldredge
In reply to this post by Martin Vonwald (Imagic)
On 01/24/2015 10:28 AM, Martin Vonwald wrote:

>
>
> 2015-01-24 17:20 GMT+01:00 Никита <[hidden email]
> <mailto:[hidden email]>>:
>
>     Are you an idiot? I mean really.
>
>
> I hereby request a ban of this individual from this mailing list and I
> definitively support an OSM-wide ban.
>
>
> _______________________________________________
> Tagging mailing list
> [hidden email]
> https://lists.openstreetmap.org/listinfo/tagging
>

Agreed.  Ad hominem attacks aren't a suitable way to discuss OSM issues.

--
John F. Eldredge -- [hidden email]
"Darkness cannot drive out darkness; only light can do that.
Hate cannot drive out hate; only love can do that."
Dr. Martin Luther King, Jr.

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Wiki Edit War on using/avoiding semicolon lists

Ilpo Järvinen
In reply to this post by Никита
On Sat, 24 Jan 2015, Никита wrote:

> Valid point, but also should suggest good practices for people who would
> like to benefit from default indexes:
> API performance

API doesn't care. Or which API call you're refering to?

> overpass performance

Overpass could handle semicolons, if it wanted to. That is, it
could provide a query interface which can fetch you the "green", etc. even
when there are semicolons used without require the use of regexp. I think
it even should do that!

BTW, who are those "people" anyway? Are you speaking for them for real
or did you elect yourself as their spokeman without asking them?

> These tools are quite popular right now and we should consider how people
> with work with data. JOSM is editor for raw osm data, not post-processes or
> indexed data.

Here goes your CPU argument too then. Does JOSM have indexes for fast
searching with tags? And I doubt that it would be all that hard to add
support to josm to handle semicolon value searching (all without regexp
magic obviously to please you, newbies, and me :-)).

> >  Nobody (except you) is forcing anyone to store semicolon values as
> > literals rather than e.g. into a postgresql array. Your argument
> > entirely *depends* on that stupid way of loading multivalue into the
> > DB.
>
> I see your point. Actually, let's talk about it. How "smart" loading
> script should know if key=* is multivalued or not? Should it parse every
> value in database?
>
> ";;;;" - 4 symbols? 2 escaped semicolons? how do you know?
I see your point, the escape needs to be defined like has been
already discussed in this thread. But actually, let's talk about it
a bit more. How would your approach deal with this then?

tag:;;;;=yes

or

tag:=yes four times(?!?)

Does that make sense to "newbies"?!?

...I suppose you'll respond that this was a bad example and I'll
whole-heartedly agree with you :-).


--
 i.
_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Wiki Edit War on using/avoiding semicolon lists

jgpacker
Our friend Никита (user Xxzme in the wiki) put his opinion in the wiki regardless of the opposition.
Since, as far as I can see, the discussion is still ongoing, I reverted his changes.
Reply | Threaded
Open this post in threaded view
|

Re: Wiki Edit War on using/avoiding semicolon lists

jgpacker
Not five minutes later, he already reverted my changes, justifying it as a "single user opinion" and "undiscussed changed".

I also fixed some of his additions in other pages, but he is already reverting them.
It seems he is trying to "win" the discussion by Fait accompli.
Reply | Threaded
Open this post in threaded view
|

Re: Wiki Edit War on using/avoiding semicolon lists

Martin Vonwald (Imagic)
Who has admin power in the Wiki? I again request a ban of this user.

Martin


2015-01-27 11:31 GMT+01:00 jgpacker <[hidden email]>:
Not five minutes later, he already reverted my changes, justifying it as a
"single user opinion" and "undiscussed changed".

I also fixed some of his additions in other pages, but he is already
reverting them.
It seems he is trying to "win" the discussion by  Fait accompli
<https://en.wikipedia.org/wiki/Wikipedia:Fait_accompli>  .


_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
1 ... 34567