Home »Forum»Open mic at the Blue Iguana »Wikiloops backstage - have a look behind the curtain

wikiloops backstage - have a look behind the curtain

posted on #1
User Avatar
Posts: 2564
Joined: 30.12.10
The idea to this thread was once more sparked by a good face-to-face conversation with Wade,
what I'd like to offer here is a refreshed overview on the multitude of things and tasks that do add up to what you know as wikiloops.com.
The idea and intention is to give insight and to hint at possible fields where volunteer help would be very welcome -
I'll be the first to admit that not all fields I'll mention below have been dealt with in a satisfying way in the past,
what you have witnessed so far has always been the best that could be done with little manpower and little budget.
But see for yourself, here are the fields of action involved in operating wikiloops:

1st: It takes someone as CEO of the wikiloops media UG - the "guy in charge"
That guy gets the naughty mails from tax offices, has to do the accounting, pay the bills for servers and external services and keep up communication with all involved parties. Of course the CEO carries the legal responsibility for anything that might go wrong, and has to take care to see the genuine ideas and motives of the project resembled in its development.

2nd: It takes someone who monitors the evolution of the internet - the "web evolution expert"
Part of the job is to permanently monitor how new devices, rising average connection speeds and new browser and operating system generations change the way people use the internet.
In the past 6 years of developing wikiloops, browsers have changed dramaticly and the use of touchscreen devices is on a rise. The initial 2011 version of wikiloops would not work on modern browsers for various reasons. The well known "flash" technique has been dropped, the visual styling has changed to a new format called CSS3, the secure https: connection is a must-have today etc etc - none of that has to do with what is unique about wikiloops, its simply keeping a website up to the expectation of "yes, that looks like a website in 2017".
The task described here is simply to notice such changes ahead of time, stay on the look-out for arising standards and new solutions.
To name one upcoming change noticed by the "evolution monitoring"-guy: Younger users are turning away from using email accounts and prefer messenger apps like whatsapp. To stick to sending sign-up emails and notification emails will not do wikiloops any good in the long run since those users will perceive that as complicated and old fashioned.

3rd: It takes people with coding skills - the "programmer & core developer"
- develop functionalities in .php and javascript
- develop data models which allow accessing the required data quickly (think: how to assemble the tracks title, listener count, the comments and alternative offers in smart directories which allow assembling that quickly for each and every track - that's a data model)

4th: It takes someone with design skills to let things look smooth - the "visual expert"
That guy needs to have some photoshop skills, CSS knowledge and a hand for sticking to some cooperate identity idea

5th: It takes some user experience expert to optimise the page - the "UX expert"
User experience (UX) is all about designing the page in a way that will let visitors feel it is self-explanatory and will reduce frustrating moments of "now, that doesn't seem to work as I thought it would".
The UX expert will permanently monitor how people browse from page to page by looking at the logfiles, and will try to understand why people have left at some point, or why they didn't take the expected action.
In a next step, the UX guy may do A-B tests to see if the UX can be improved:
One simple example was the decision what to show to people looking for "Drums" on the wikiloops search engine. We used to show any track that involved drums for a while, but after monitoring users a little, I switched to showing "drums only"-tracks instead, which seemed to be the closer expectation-match. Again, this task has nothing to do with individual user experience on a human-2-human level. Site visitors are generally unaware of UX, it is perceived rather sub-conscious.

6th: It takes someone who cares about search engines - the "SEO expert"
Now, this is the guy who knows about search engine optimization (SEO), and he might as well be considered the most important guy on the team:
You would probably not be here today, nor would there be 100k tracks on wikiloops if there had not been a great amount of work spent on making sure search engines do present wikiloops among the top 10 results for backing tracks. If we have 50k members today, one may safely assume at least 80% of them found wikiloops by search engine requests.
The things to know about the job of the SEO dude are: SEO is a wicked game with constantly changing rules and new factors coming into the game, which are not publicly announced, but need to be gathered by monitoring how rankings change (no chance to predict what's coming).
The SEO guy will monitor different tools and statistics and bug the rest of the team to fullfill his special requirements, be it to follow certain guidelines about how much content may be on a page, certain speed requirements that need to be met by the server backend engineers or certain technology changes because search engines will down-rank something (for example, pages which do not come via https).

7th: It takes someone to rock the social media world - the "share and like expert"
"If it doesn't happen on Facebook, it has not happened" is something marketing people say. My personal first idea of "the internet" consisted of having an email account one could log into using the netscape browser, and knowing the search engine Alta Vista (that was back in the 90s). Modern users idea of the internet consists of "Google, Amazon, Youtube and Facebook", while a yet younger generation accesses the internet via those websites' apps instead of using a browser. That is a completely different way of finding things, and I personally do believe there are a zillion potential wikiloops users out there who simply never thought of googling for backing tracks or online music collaboration - they have their facebook groups and instrument specific forums, and the only way to reach out to them is to "happen" on those networks.
But (even tho it was presented like that in the beginning of social networks) it turns out these networks are not really an easy way to reach out to people: Building a followership for some social network page is a science of its own, if you do not take the easy way and purchase paid "boosts" or buy likes.
It takes someone with a good hand to writing effective posts and news items which will be liked and shared, to keep some constant interaction going. I have been dreaming of doing that for a long time, but its one of the tasks I rarely get around to.
Besides building a followership and increasing attention, the social media representation also has the effect that it is taken to be a monitor of success when a project is evaluated. Potential sponsors will look at your Facebook followership and judge you by that, there is no way to avoid that, and that responibility rests with the social media rockers guy.

8th: It takes someone to get the word out - the "reputation manager"
This guy is looking to see wikiloops getting reviewed, noticed, spoken about and recommended in places where musicians might come across, the task mostly consists of "polishing doorknobs" by contacting musicians magazines, writing recommendations on musicians forums, asking someone with a huge social media followership to recommend wikiloops to his or her audience, contacting known online music teachers like Scott Devine or Adam Rafferty and ask them to review the project... if you belong to the 20% of members who did not cme in via a google search, you probably noticed wikiloops recommended by someone playing the "get the word out"-guys job. Unseen by anyone on wikiloops, these activities are really making a big difference, and the SEO expert is happy about any link to wikiloops as well.

9th: It takes someone to monitor the community - "big brother of the loops"
This guy makes sure new members stick to the "no covers on wikiloops"-rule and learn how to submit tracks correctly. We are trying to stay ahead of hate speech, monitor use of inapropriate images and language and fight the fire whenever a troll appears. Support emails need to be answered, forum posts need to be answered, the shoutbox needs a carefull eye and intervention every now and then. Since the show is running 24/7, one may imagine that this is quite a lot of task just to ensure a very basic level of quality & safety.

10th: It takes someone to round up funds - the "budgeteer"
This involves reaching out to potential sponsors and advertising partners, supplying them with the relevant information on "how many visitors", "what kind of people", "where from" etc etc and trying to negotiate some kind of deal. Since such negotiations are confidential, you will not see any comment on that task in the public wikiloops developers blog, but I can assure you the "budgeteer" has been working hard, sending a lot of emails, visiting the frankfurt music fair, contacting music instrument retailers around the globe etc etc. It's one of those "hidden" tasks that rarely make it to peoples notion, even tho there is a very high relevance and potential in this field.
The second and even more time-taking task to tackle by the "budgeteer" is to create calls for support towards the community, keep the permanent crowdfunding rolling, come up with incentives to support the project and witty ways of presenting them.
Out of the +2k forum posts I've written in the past 6 years, probably 25% fall into this category. The amount of work and time invested in designing functionality which is necessary to collect member support (as the donation campaign 2016 or the whole supporting membership presentation & handling) is enormous.
I read that the wikipedia spent almost half a million euros on creating their annual support campaign 2016, wished I could have gathered 10% of that to fund the whole of wikiloops for one year.

11th: It takes someone to coordinate the volunteers - the "helpers helper"
Many tasks like most prominently the translation and moderation jobs are handled by volunteers. That is really great, but as anyone working with volunteers will know, having a volunteer do something always requires some kind of initial time investment to make sure he or she understands the task and stays motivated on the job. Once you have a pool of 10 volunteers, you need to realize that you will need someone who coordinates & motivates them (otherwise they'll leave), so there's the 11th job for you.

12th: It takes someone to translate the interfaces - call em "translators"
A very rough guess on how much text needs to be translated per interface language version would land somewhere around 60-70 standard paper pages of text, and we currently have 7 languages available. A new interface version like the wikiloops 10 update that has been rolled out on the english domain in august 2017 has a follow-up translation need of roundabout 360 pages of text, requiring at least 6 volunteers to work on that (I do the german translation myself).

13th: It takes someone to adapt the Android app - the "app dev nerd"
...and ideally someone to eventually finish the iOS app. the app development is a field of its own, but of course it requires working hand in hand with the core developer, since the apps access the same data as the website. App coding is quite different to website coding, so its a field of its own to work on.

14th: It takes someone to address the community - the "community cheer-up-expert"
wikiloops wouldn't be what it is today if there had not been many addresses towards the community, be it the "site news" that eventually pop up on your newsfeeds, official blog posts, invitations to member meetings or coordinating the small nice things like sending out stickers. Opening interesting or funny forum threads that gain interest, making songs for a certain purpose or on a special topic or asking members to submit video snippets to create a fun video (we did that in 2012, that was kinda cool) are things that give wikiloops that "alive" feel. wikiloops would be valuable as a faceless collection of tracks, but what people seem to love about it is the mix of "its alive"+ "it has quality content".
Sometimes, the "budgeteer" and the "community cheer-up-expert" get into each others way, too - its a bit hard to stir fun & time intensive projects like a collective video, while claiming there is too little budget to keep the servers running.

15th: It takes someone to substitute the experts
If you have 14 people working in a team, then you better be aware that 1.5 of them will either be sick or on holiday, or feeling that a five day working week was enough.
If you really want 14 people on the job on any working day of the year, you better hire 17,
and if you want them to be present seven days a week instead of five, better hire 22.
Oh, and if you'd want to make sure they are not only present during 8 business hours, but ideally 24/7, triple that once more and you'll need a recruiting manager extra next...
wikiloops does not need a 24/7 "budgeteer" or "SEO expert", the point I'm trying to make here is that "we have someone taking care of that" is not equivalent to "we have enough manpower to take the full potential". Right now, I'm mostly substituting myself around the named 14 tasks - leading to wikiloops being as you know it.

16th Let's not forget the maintenance man
He who does the data backups, downloads gigabites of music every week and saves it to external backup drives, he who copies database tables and makes sure your precious tracks and coomments are not lost in hurricanes or cyber attacks. Thank god we never needed a full backup restore so far, but not having our maintenance dude would leave wikiloops totally "pants down" in such an event.


Well, I guess that was enough of a read for you, and it ate a quarter of my day, too.
Maybe looking at the list you can relate to why I sometimes feel that trying to cope with all those tasks as a single person is hard to achieve, and maybe now you'll feel different about the need to have some funds for such a project - I do not know.
I offered to do all that with a budget of 40k US dollars in 2017 which was still very unrealistic (=too low) if you tried to hire the named experts from the free market, and those who have witnessed the support campaign at the end of 2016 will remember we (or, the "budgeteer" to blame him right out) were not able to scratch together 30% of that sum at that time.
If wikiloops is still out there today, than that is due to my reluctance to give up, to Wades financial backup and to roundabout 300 individuals who supported wikiloops by becoming a supporting member or simply donating.
I hope this thread might light up some thoughts, there is still so much potential that has not been reached out for, and as a genuine collaboration community, we could make use of some swarm-intelligence here to keep the project around the way we like it.
I am looking foreward to your comments and thoughts, and should there be interest / need to discuss a singled out aspect from my named list of tasks, I'd suggest we move such discussions to separate threads early to avoid a big unreadable mess here.
All the best from wikiloops headquarters, and thanks again for your interest :)

posted on #2
User Avatar Member
Posts: 22
Joined: 24.05.14
yes, you are right with all steps. mayby i do the part with the drinks on the next meeting:D. you can ask me for everything you see me;). nice wishes:W
posted on #3
User Avatar Supporter
Posts: 479
Joined: 27.02.15
A fantastic breakdown and insight into not just what keeps Wikiloops ticking but also the vast number of technical challenges it takes to make it what it is.

It's all too easy these days for people to assume things on the internet 'just work' and carry on 'just working'. And they get annoyed when it doesn't work. Until AI truly turns up and turns us all into batteries ;) , computers are still only as clever as the people who program them. That means not just making stuff work, but foreseeing potential user patterns and second-guessing how they would use it. The same goes for security - you can only block the attack vectors that you can foresee happening. Someone, one day, will find a workaround. And that goes for Google as much as it does Wikiloops. Google just have vastly more resources to throw at a situation!

In my experience, everything works great when you test it but real-world users have a habit of finding new and unique ways of breaking your beloved project! Especially when there's 50k of them!

$40k is not a lot of money to run a project like this and we should be thankful that Dick is able to wear so many feathers in his cap to keep costs down. And also the many volunteers who've offered their experience and skills to make up for the fact that Dick cannot do it all. That $40k, in the commercial world, would just about pay for a contract developer for 3-4 months to do one of the apps...

As an aside, if you need a hand with iOS (I thought you had someone on it), Dick, drop me a line.
posted on #4
User Avatar Supporter
Posts: 374
Joined: 25.03.12
Since i have got a fiber connection to my house (295 Mb/s at time of writing) I could volunteer as backup and maintenance.
Pure fingerstyle
wikiloops online jamsessions are brought to you with friendly support by:
Marva from Wales

"I'm inspired by Wikiloops, what a wonderful supportive and creative community!"

wikiloops uses Cookies and processes data in compliance with the GDPR,
as stated in our data privacy policy ..