Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Tyker]

 It would be possible to add any number of pictures, so it has not to be restricted to six.



I don't where the database is to be located but images are not disk-storage friendly and come in 57 varieties  - tif,bmf,jpg etc

[Bill the Farmer]

There's a document on the Hohner site: Hohner Akkordeon Modelle, which has a lot of possibly useful info.

I had a look at the documentation for MediaWiki. It looks complicated enough without trying to shoehorn another data model on top. Would it be possible to use something like a template for new pages, possibly partially filled in according to context, so that users can just fill in the blanks? If a particular make or model warrants additional info, it can just be added.

[Theo]

Q for Theo.  You suggested it may even be a topic for Wikipedia.  Can you see anyway in could be an integral part of Melodeon.net ?
I am undecided on which is best.  Melodeon.net members will be very careful over accuracy and it would only be members who could update whereas in the open sea of Wikipedia the wide band of knowledge may be offset by 'I think I know brigade' and the malicious.

But I still think that all this detailed discussion of database structures is putting the cart before the horse.  The first step should be to make some sensible estimates of the amount of information you need to collect, what are the sources, and who will have enough time to collect it. 

I imagine there will be quite a lot of material that will be fairly easy to get hold of, such as the stuff Hohner published in 2003, and makers websites will give some information about the bigger factories current offerings, but where will you get the huge bulk of information on instruments made over the past century and a half?  The only way I can see of doing it is to spend many hours visiting museums, libraries and private collections and searching through printed sources, and cataloguing instruments.  Its a massive task, probably several years work by a team of people.  The sources are spread around the world, Europe, N America, S America, S Africa and possibly more.

I'll say again, its a great idea, but whoever takes it on has to go in with their eyes open and realise what an enormous task it is.

[Waltham]

How about we take a turn at looking it from the other end and see how far we can get with a complete history of the Hohner Pokerwork?  Shy Talk has already provided a skeleton in the FAQ thread, how about we see how far the 'hive mind' of this forum can get at puttiing together a definitive history of the instrument by way of forum posts?  If we succeed we'll have a better idea of how the database should work.  If we can't then we'll at least have saved the effort of building a database that's unlikely to be filled.

[Theo]

Good idea, and it also points up something that would not fit easily into a flat file database, that is the history of changes and developments in a single model. 

[Martin J]

Hi Theo
I understand where you are coming from in considering the amount of work needed to plan and implement as complete a database as is possible.  I agree that a such a resource would be a monumental task but the 'hive mind' version, as Waltham puts it, is the organic route I was proposing.

I only ever envisaged an organic information source which would be updated on an as and when basis by the members.  This was promoted by reading threads in the Instrument Makes and Models section and thinking wouldn't it be a good idea if this information were recorded in an orderly, searchable repository instead of being piece meal in different threads.  I also thought it should be an integral part of Instrument Makes and Models but have no idea how that can be achieved.

I could see a flat database working for this with the history of changes being catered for in the the first few fields of Make, Model and Mark.  Each unique mark would be recorded as a separate record or as a sub section of the home record.  A search on listings would bring up these three items, together with other key elements, which would allow the researcher to choose.  A simple MS Works Database does the whole job except pictures.  A database with combo boxes would also be beneficial in speeding up both entries and searches.

The wonderful fact sheets from Hohner refereed to by Bill can be entered on an as and when basis by anyone who was interested in a particular mark.  Each member could enter the details of their own 'exotica', - see Sebastian's friend Anders Trabjerg and his mid-1930s, 8-voice, 1.3-row D/C# Baldoni, Bartoli for real exotica and I'm sure there is more.

Once a vehicle is in place, makers could be advised to check entries about their products and to update us with new offerings.  The work could and should be spread and the project should evolve over time.  You are quite right Theo, it is too much work for even a small team but open to the community it could happen.

[Theo]

Hi Theo
I understand where you are coming from in considering the amount of work needed to plan and implement as complete a database as is possible.  I agree that a such a resource would be a monumental task but the 'hive mind' version, as Waltham puts it, is the organic route I was proposing.

Ah I see, so the idea has changed then, I was still thinking of your original proposal for:

What about a History section with full details of every maker and model ?

I could see a flat database working for this with the history of changes being catered for in the the first few fields of Make, Model and Mark.  Each unique mark would be recorded as a separate record or as a sub section of the home record.

But what you are proposing there is squashing the data into a flat file, when the structure of the data is actually more complex and really would justify a relational database approach, especially as the volume of data grows.

simple MS Works Database does the whole job except pictures.  A database with combo boxes would also be beneficial in speeding up both entries and searches.

But for the cooperative approach you are proposing, and working on the web you would be unable to use MS  Works.  You would need to used a multi-user, web capable database, and the one which I see used for the majority of web based databases is MySQL.  This forum runs on a MySQL database, and the information part runs on another.  If you want to start with a bare MySQL database you would end up doing a lot of scripting to  create a front end with access rights, searches, and various outputs.  It might appeal to any PHP writers on the forum, but its also not a trivial job.  That is really why I suggested using a Wiki, or there may be other public domain/open source scripts we could use.

I also thought it should be an integral part of Instrument Makes and Models but have no idea how that can be achieved.

would not be a problem if the thing is built along the lines described above.

[Martin J]

As you can tell from my ramblings I am no programmer Theo and I agree that a relational database would be ideal.  I also know the cost of commissioning such wonderful items.  The only reason I suggest MS Works 'type' databasing is that even I can programme that.  I have heard of MySQL and know it is the thing for the net but there my knowledge ends.  I am in the hands of others to guide me in the right direction to create the needed tool.  To answer one of your earlier questions, yes I am only too willing to help within my knowledge and abilities.

Sorry I was unclear in the first instance.  I see now that the original bar was much too high but it was a concept to aspire to and at that time I had no idea as to the volume of makes, models and marks which are obviously out there.

To echo your remarks,

using a Wiki, or there may be other public domain/open source scripts we could use.

does anyone know of a tool which will do the job without extensive work in making it fit ?

[Sebastian]

I don't where the database is to be located but images are not disk-storage friendly and come in 57 varieties  - tif,bmf,jpg etc

You have to store only references to the picture-files, e. g. "http://your.server.here/pictureOne.jpg; http://your.second.server/secondPicture.png". The picture-files could be located on different servers, this shouldn't be a big problem.

History of changes: This could be dealt with in the remarks-field. Maybe "remarks" isn't a really good english translation. But I would want to put everything not easily classified into this "remarks-field", that could be several paragraphs of text, if appropriated.

[TomB-R]

Y'all have seen this, no doubt, but I don't think it has been mentioned so far in this thread.

http://www.accordion-wiki.com/index.php?title=Main_Page

Tom

[GbH]

Hello All,

I've been lurking on this site for a few weeks now and have found it to be most interesting, having been a somewhat soliatary melodeon player (if I even qualify as that) for a number of years now.  Having never really met or played with any other box players, it's been quite a revelation to discover what goes on.

Anyhow, upon reading this thread, I decided it was time to break my silence, as I think I know what you need.  If you're wanting a community based wiki system, but with the structure and reporting of a regular database, then you should be looking at something called Semantic MediaWiki (http://semantic-mediawiki.org/wiki/Semantic_MediaWiki).  It's an extension to (i.e. it sits on top of) MediaWiki that allows users to 'mark-up' pages with information of a specific property.  So, for instance, on an melodeon page, information such as maker, number of rows, key, voices, weight could all be recorded specifically and subsequently queried (a regular wiki can't do this).  What's more, using the additional Semantic Forms extension, an interface can be created that guides users through the page creation process, meaning that average users don't have to know how the system works or what they're supposed to be creating - they just fill in the boxes.

I've been using and administering such a system, in our office, for 6 months now and, on the whole, it's worked very well.  When setting it up initially, it does take more time and knowledge that a regular wiki would.  However, the eventual results are worth the effort.  Bearing in mind what's been requested above, I think it would be worth consideration.

To set up and host such a system locally, you'd need a web server, MySQL, PHP support and a regular Mediawiki install, along with the Semantic extensions (all this is freely available for no charge).  Alternatively, there's a site that publicly hosts this sort of wiki.  It's at http://www.referata.com/wiki/Main_Page (I've no experience of using this service, though).

Anyhow, hope this is of some use.


Guy

[Theo]

To set up and host such a system locally, you'd need a web server, MySQL, PHP support and a regular Mediawiki install, along with the Semantic extensions (all this is freely available for no charge).  Alternatively, there's a site that publicly hosts this sort of wiki.  It's at http://www.referata.com/wiki/Main_Page (I've no experience of using this service, though).

Anyhow, hope this is of some use.


Guy

Thanks for the suggestions Guy.  All the server facilities you list are available on our server.

[Martin J]

Working at the moment, just dropped in to see how things are going on only to find we've struck gold.  Thanks Tom & Guy.  I've had a quick look and will look properly this evening but first impression is that this is what we need.  Theo says it will work on melodeon net - great.

Back later

[Sebastian]

we've struck gold.

It looks very fine indeed.

Amicalmen,

Sebastian

[Theo]

Working at the moment, just dropped in to see how things are going on only to find we've struck gold.  Thanks Tom & Guy.  I've had a quick look and will look properly this evening but first impression is that this is what we need.  Theo says it will work on melodeon net - great.

Back later

Mind you there is a bit of work setting up the structure.  I'm happy to install a copy of Semantic MediaWiki on out server, but I don't wish to do the development work that will be needed.  Volunteers???

[GbH]

Working at the moment, just dropped in to see how things are going on only to find we've struck gold.  Thanks Tom & Guy.  I've had a quick look and will look properly this evening but first impression is that this is what we need.  Theo says it will work on melodeon net - great.

Back later

Mind you there is a bit of work setting up the structure.  I'm happy to install a copy of Semantic MediaWiki on out server, but I don't wish to do the development work that will be needed.  Volunteers???

Well, as I already have experience in setting up and running such a system, I guess it would be a bit lame if I didn't offer at least some sort of help.  I would recommend, thought, that there should be at least a couple of other that are prepared to get their hands dirty and learn how it all works. 

In the first instance, though, the following would need to be done:

-  On the server, you'd need to install the Semantic Mediawiki, Semantic Forms, Semantic Drilldown (preferably) and, if it's not already installed, the Parser Function extensions.  If Mediawiki is already in place, these are all easy to find and install - they're all on the official extensions matrix.  There are a few others that could be installed too, but I think you're better off starting simply.

-  On your base Mediawiki install, ensure image support is turned on (I think it's off by default).  Photos and diagrams are cool! 

-  People need to start thinking seriously about what types of standard page are needed, so appropriate properties, templates and forms can be created.  To start with you need to think of wiki pages in terms of simple categories.  These might be 'Instruments', 'Manufacturers', 'Players', 'Repairers', 'Events' etc - it depends what content you'd want to cover.  To begin with, you'd want a list, however small, that makes a cohesive working system.  Once you've got these categories defined, you'll need to figure out what common properties you'd need in each (and also, eventually, what type of data they'd hold).  So, for an 'Instruments' category, you might have name, number of voices, number of rows, key, date first sold, manufacturer, web link, summary, notes - really, any fields you think might be necessary, even if they're not a relevant for every entry.  Once a few of the categories are sorted out and agreed upon, it's then fairly straightforward (once you know what you're doing) to build them into the system.  Other pages, that report and use the data, follow on naturally after that (no need to worry about that right now).


Guy

[Martin J]

Earlier in the post I submitted a proposed field list which I have been updating at home as people add ideas and wish items.  Once I have tidied it up in the next day or two I will post it.  I am willing to be one of the helpers, data, structure, analysis etc but I have no programming skills.

Once the tool is in place I would be willing to write to every manufacturer (that I know of or I am given the email address of) advising them of the facility and asking them to submit details of their products (by filling in the entry form themselves) and to check the details we already have from 'the hive'.

The other massive source of info is within this forum under the new remit of Lester's FAQ section.  Shy Talker and Waltham have also supplied solid information, indeed Watham has already begun with the history of the Polka Work.  We should also request the assistance of our overseas members to enlighten us with offerings from other countries.

[GbH]

Earlier in the post I submitted a proposed field list which I have been updating at home as people add ideas and wish items.  Once I have tidied it up in the next day or two I will post it.  I am willing to be one of the helpers, data, structure, analysis etc but I have no programming skills.

Unless you're thinking of doing something well out of the ordinary (unlikely, given what's been posted so far), no conventional programming is required.  Data, structure and analysis skills are very useful, though, as is the ability to think up unambiguous terms for the properties and categories.

Another useful skill, albeit one that isn't necessarily required right away, is a bit of graphics and design finesse. 

Guy

[Sebastian]

Hi.

Sorry for ducking away for some days, but we had a little bit more to do at work due to the Asse-problems becoming more urgent. 

If there is some help, I could provide, in setting up the system, I would be glad to do so. For programming skills: I scripted together with other volunteers a member administration system for an organisation with 4000+ members or a web shop for books for a friend, I fiddled with some AJAX-applications and set up a simple Wiki for some small organisation. So I have a basic knowledge of PHP and MySQL. However, until now I know nothing about the internals of the Wikimedia-system.

Sebastian

[Viljo]

Hi!

I have been writing my accordion web pages since 2006 (release Jan. 1st!). It includes mostly history and technical issues of accordions. In my site there is accordion gallery with over 700 pictures (and going up) and this gallery is only one part of the site. Today I can say that my web page is more complete than any wikipedia today (in mount of information, not languages). Acc.gallery is not such database you have been thinking here but for me it have been very good tool to understand the historical issues more correct. My gallery includes picture + name + tuning (if a diatonic) + manufacture and place + other important issues like history of that accordion shown in the picture.

As my job has gone further, I have understood that it will never stop. Never. Only in Italy there have been 419 manufactures between years 1850 - 2003. Then there is Russian industry, European, Asian, American... concertinas, bandoneons, diatonics, chromatics... My field includes every single free-reeded European origin instrument and some very early mouth organs (i.e. everything about free-reed instruments). My project is once in a lifetime project and a lifetime project...

My site is released here www.vapaalehdykka.net in Finnish. Later on I am going to take more domains. Then there is translation coming for English, Swedish and perhaps German too.