I have to say, I am also getting a lot of these same errors as previous posts, all too often: I only purchased a day or two ago, but have perhaps 50% of the time unavailable so far for the
The issue is, you just purchased it very recently, so you don't have any historic data about its reliability, and are forming your opinion on the availability of these last one/two days, which by chance seems to coincide with an unusually higher unavailability of the geonames server.
Just as an example, if we discovered (just an assumption, I'm not giving real figures) that the geonames server is down 1 day every month, the real availability would be 97%, but for an user that bought it *that* day, it might seems like 0%...
So, I'd say one should at least wait some time and getting used to it, before forming an opionion about its reliability.
I would like to add that, our Trial doesn't have any expiration date. This is *particularly* important for a product were we can't full guarantee to be always usable, also because server availability is a lot dependent by the user location compared to the geonames server. This because, when XPOI says the server can't be reached, it's not always because the server is entirely down, but in some case it simply hasn't replied in a give amount of time, either because of the load, or just because a specific problem which might not be the same for all users. The geonames.org servers are based in Germany, which should be very good for european users, but maybe not as good for users overseas.
If you are selling a product that is a service mashup between FSX and various data sources on the web, I feel - for my own part at least - that you are accountable for the union of those web services to be successful, since that is the mashup-product you are selling.
Well, this is highly debatable. Yes, we are accustomed to free web browsers right now, but think about a payware web browser (I think Open used to be payware for many years).
You might decide to pay for the browser, even if there are free alternatives, because that one gives you a specific way of accessing data (which is not provided by the browser developers), which might look more convenient to you, comparing to the free ones.
So, what we are selling to you, it's just a different way to access the same data that you might be otherwise able to access for free from your web browser, by simply visiting geonames.org
I am writing with only one goal - that you consider how the vision of XPOI can not be hinderd by say a 50% chance of your users finding it "working" on a given flight. Either different mashup web services if geonames can't deliver consistently, or offline database or something...
That exactly the main issue. And, if you think about it, it would be solved and it would be in our best interest to solve it, if we really had to concede that the geonames server is not reliable enough for our needs.
And why ?
Because we have a TRIAL, that is not time limited in the sense it doesn't have an expiration date, so any user has the chance to fully assess the geonames availability AND because if we REALLY had a constant 50% availability of the server, we'll simply start losing sales, because lots of first-time users that simply checks the trial, would have an high chance of installing and discovering that it wouldn't work...
So, it's clear that we couldn't stay too long in a situation were the geonames server is not working most of the time, because this would automatically impact sales.
That was always our main theory about Trial version, and how they are the best user protection: a program with a Trial HAS to be good, because any problem will eventually be apparent from the Trial, so every problem is a potential sale lost. Beware of products without a Trial...
Yes, in this specific case, the issue is particularly complex to solve, because the main selling point of XPOI is also a potential weakness. Because of the vast amount of data, there aren't many alternatives to using an open source of data, a commercial provider would probably too expensive and make the project unattractively priced to the user, and it would be simply impossible for us to develop our own version of such database that would able to compete in scope with what geonames and wikipedia are offering, considering they are the work of thousand of users.
An offline database was considered, but it would add several issues, starting from bandwidth cost, because the installer wouldn't obviously be only 18MB, but it would go into hundreds of MB, with GB figures if you add also the wikipedia text.
And, an offline database wouldn't be updated so easily, which is another major selling feature for XPOI.
Found an error ? Your local area has something wrong/misplaced/obsolete that only locals might know ? You can fix it, and it would benefit immediately all other XPOI users AND for all the geographical community using that data. Which of course is valid in reverse as well, because as soon *anyone* in the geographical community updates the geonames server OR the wikipedia, any XPOI user get the fix in real-time too. Which would't happen if we were going to use a commercial data provider (only fixes from them would be possible) and it wouldn't be as large if we were to start our own geographical database, because the only contributors would be our XPOI users, compared to everybody else, that will continue to post fixes to geonames and wikipedia, because these are the sites most of the user posting fixes will go to.
The only thing that we found to be probably feasible, is to have our own database, but synchronized, let's say every week or so, with the geonames one. It wouldn't be as cool as now, with real-time updates, but might be still usable.
But it wouldn't be easy anyway, because we would have to mantain a server, so we might end up managing OUR OWN down-times as well, because is not so easy to estimate how big the load would be, without trying it first. For example, we don't have any idea how the users settings are, which has an impact on the bandwith required. Do we have all users flying with all POIs on ? Do we have half user flying with everything on, and half flying at 10% POI density ? Do they fly for 5 minutes with XPOI enabled ? Do they make a whole 4-hours cross country flight with everything on ? In which countries ?
There are so many variables, that is really difficult to estimate how feasible a project of having our own geonames server would be, without really trying. And what if we discover that, to setup a server that would work better than geonames, we would require to spend so much in bandwidth, that we couldn't even recover the costs with XPOI sales ?
So, I'd say: let's give it some time to use it, and try to judge it's reliability over a longer period of time, like a month or so, this should give a more precise estimate on the server availability.
If we really end up finding the geonames server would be too unreliable to be used, it would be clearly apparent in the Trial version, so it's surely in our best interest to make it work, one way or the other, if we want to continue selling it.