Vielen Dank!
 

Cockpitvisit, awesome tool, thank you very much! This is so incredibly useful!

Here are my comments and things I stumbled across during a few days of use.

I understand that your data sources do not differentiate between code-shares and flights by subsidiaries. That is the only real drawback I found so far. I normally want to include flights from subsidiaries/regionals but usually exclude code-share flights such as SYD-JNB operated by Quantas when I limit the network to Star Alliance, for example. Well, I understand it's one minor limitation that it looks like we'll have to live with.

The setting to include regionals and code-shares in the results is very important. At the same time the full meaning and scope of this term will not be clear to the uninitiated (like me). Most will not bother to read this thread (I'm surprised about your patience with all these "give me a Mac version", "how do you zoom without a mouse", etc.). May I still suggest to change the tooltip for the menu entry to "Show codeshares and regionals (recommended)", adding the underlined part? I think it's safer to have it on than off as default for people who don't fully understand what it actually does.

For the benefit of those who actually do read this thread, I will point out three important lessons I learned when I felt there were routes missing in the database. First is to activate above setting and see if that brings up the flights. Second, if that is not sufficient, make sure the subsidiaries are not operating under their own code (JQ and 3K, BL or AK and FD, D7 come to mind). Third, make sure it's not a seasonal route. For example, I thought that NH SDJ-OKA, NZ SYD-RAR and NZ NAN-CHC among others were missing. Well, they're not, it's just that those routes are not operational for the current month, June 2011.

Speaking about seasonals, I still find them to be very important pieces of information. I remember that you previously turned down the request to include several months worth of route data, citing the amount of data involved as the reason. I suppose that's the amount of data you need to process, not the size of the routes.dat file? It's my understanding that you currently process one month of data. Would it be possible to add 2 weeks worth of data each about three and six months out? I think that should take care of the seasonal routes without a six-to-12-fold increase in overhead for you.

The following routes ARE indeed missing and I was wondering if there is a way for you to add them; extended long-hauls. NZ1 LHR-LAX-AKL or UA846 EZE-IAD-SFO come to mind. These kinds of connections can be very beneficial for award bookings since they consume only one segment despite the actual two flights (who share the same flight number).

One very minor let-down was that custom airline groupings are reset on restart of the program and switching the sorting of the airlines in the drop-down box. The latter one is a bit unexpected.

PZ needs to be added to the Star Alliance members in alliances.dat

Is this still an issue? I think it would be fairly straight-forward, all entries are unique keys in a database. I'll assume that under normal circumstances flights need only be added (after all, you did find the routes in a current timetable). "sort routes-from-scraping.dat routes-manually-added.dat | uniq > routes.dat" would prepare the new routes.dat for distribution. It would still be fairly easy even if there was a need to delete or change routes from the ones you automatically collect. Let me know if I can help you with this. Maybe this is the right way to tackle the extended long-hauls if collecting them automatically is too cumbersome?

The latest version mistakenly shows Comair (MN) as having an extensive route network within Indonesia.

Good idea, will do, thanks!

Not possible now, sorry. Maybe I can add seasonals for Star Alliance and oneworld only though.

The tool displays nonstop flights only, this is why NZ1 is broken into 2 parts. Direct flights that are not nonstop are intentionally filtered out. If you want a route map with bookable flight numbers rather than actual nonstop flights, I would suggest you to have a look at the online Star Alliance RTW booking tool.

As soon as you allow people to add their own routes, someone needs to take care of removing and maintaining them as well. This isn't likely to happen, and I certainly don't have time to do it either. What usually happens is that someone finds the tool, notices that his favorite airline is missing/incorrect and adds/fixes it per hand. After that, the user moves on and we are left with data that isn't updated automatically and that no one checks or maintains. Even worse are manual corrections to automatically gathered data - since they have to be checked and reapplied after every data collection.

Since the quality of the automatically gathered data is relatively good, I think it is a lesser evil to simply live with the errors/omissions in this data, rather than try augmenting it per hand. Of course, everyone is welcome to create his own routes.dat with improved data.

Cumbersome are not extended long-hauls of major airlines (most flyertalkers know them by heart anyway), but regional routes of smaller airlines not belonging to any alliance.

Thanks for noticing. The routes in Indonesia actually belong to Merapi (MNA). Will be fixed in the next update.

The Pensacola airport in Florida misses a lot of flights, apparently. Maybe you want to have a look at it? According to the tool only Delta and AirTran fly there, but it looks like three *A carriers fly there as well and they do so this month.

Star Alliance routes to PNS seem to show up correctly if you select Star Alliance and enable codeshares.

Oh, indeed. The tooltip for the airport itself does not include the regionals. This non-distinction between regionals and codeshares is a major stepping stone :(

Great job! Though I can spot two incorrect airport coordinates:

DUR -29.6144 31.1164 Durban King Shaka International Apt

NBS 42.0881 127.5489 Chang Bai Shan

Also, it appears that HHA represents the Changsha airport while CSX represents the city of Changsha itself. ZGC represents the Lanzhou airport while LHW represents the city itself. This might need correction for posterity.

Another thing, why does the tool show flights in Libya at the moment when all civil aviation is banned at this moment?

I eagerly await a further update!

My speculation would be that the scheduled flights were still published at the time the data was gathered. These updates are not daily to include the latest political turmoil and why should they? That would actually be detrimental, IMHO.

FWIW, just came across another interesting flight. AB7341 from POP to VRA. Trying to book the flight with Air Berlin proved to be impossible. I figure the flight is really MUC-POP-VRA, but POP-VRA vv is not on offer (I'm sure AB would like to offer seats). This creates a bit of an inconsistency in the data, but I suppose it's not possible for cockpitvisit to deal with this automatically.

This is an awesome tool!

As always I will use this as a tool and not as the definitive end all be all authority on routes flown. I think people who demand that plus continuous constant updates are asking a little too much.

This is great just to play around with and think about where I want to go next!

I think I have found a route that is not included in the data when it should be. Please have a look if there is a systematic problem (I understand and have sympathy for your decision to only fix systematic problems).

British Midland has direct flights between ALA and LHR a couple of times a week in both directions. At other times they operate as one-stop flights under the same number (BD991 and BD992).

The data is partly based on published timetables and partly on reservation systems, so if flights still appear there (even with the availability zeroed out), they will appear in ARM.

Air Berlin does fly POP-VRA, it just doesn't sell tickets on that route (no 5th freedom rights). Since the route is being flown, it is displayed in the tool.

The purpose of the tool is to show nonstop scheduled passenger flights, so it does not exclude routes that are not bookable, and does not include multi-segment flights under a single flight number.

The current data corresponds to the second week of June. According to the Star Alliance timetable, there are no nonstop BD flights between ALA and LHR during that week. So it appears correct to me.

Hm, that explains why the flight is not there. IMVHO it doesn't make it correct. It's only an explanation of the systematics. You only search through one week's worth of data? Isn't it possible to add another week, best from a couple of month's down the road?

I hope you won't get mad at me for keeping on pointing out stuff I think is missing. It's my way of giving back, by helping a tiny bit to make the tool even greater. Today's candidate are all Sunwing flights. Thank you again for all the effort you put into this.

Hi there cockpitvisit. Is your code-extraction code capable of identifying the flight number(s) for a route - even partially? If so, would you be willing to publish that data as well? I think one could put that data to good use.

To give credit where it's due: the idea came to me after reading this openflights ticket.

No, sorry.