Status of Travel Mapping

[Jim]

One more for tonight: routes in the mapview.php pages are now color-coded by system.  A few of the colors could probably use some improvement, but it's another step, anyway.  Enjoy.

[oscar]

One more for tonight: routes in the mapview.php pages are now color-coded by system.  A few of the colors could probably use some improvement, but it's another step, anyway.  Enjoy.

The color-coding does improve contrast, and is nice.

But the underlying Google Maps image still creates clutter, that detracts from a view of traveled routes. Ultimately, for the maps shown in user stats pages for regions/subregions/localities, can we get something like the old CHM map display, that displays only traveled routes (with route numbers added), untraveled routes (heavily faded, but also with route numbers), borders, and coastlines, against a white/pale pastel background? Of course, Google Maps is for now taking care of the borders and coastlines, so we'd need to render those in some other fashion (as CHM did, though rather roughly for most of the coastlines) if we don't have a map as the background.

[vdeane]

One more for tonight: routes in the mapview.php pages are now color-coded by system.  A few of the colors could probably use some improvement, but it's another step, anyway.  Enjoy.

Oddly enough, I was just about to suggest that.

It would seem that the big hole at this point with the new HB is the lack of an index page.

[SSOWorld]

One more for tonight: routes in the mapview.php pages are now color-coded by system.  A few of the colors could probably use some improvement, but it's another step, anyway.  Enjoy.

Oddly enough, I was just about to suggest that.

It would seem that the big hole at this point with the new HB is the lack of an index page.

For software in development, I'm sure it's on his To-Do list.  Its something that the front-end designer can do as well.

[Jim]

It would seem that the big hole at this point with the new HB is the lack of an index page.

For software in development, I'm sure it's on his To-Do list.  Its something that the front-end designer can do as well.

Yes, I'm hoping someone else will help out with those kinds of things.  I am aiming to get a bunch of low-level building blocks in place that we (as a group) can put into a nice site.

[Jim]

One more for tonight: routes in the mapview.php pages are now color-coded by system.  A few of the colors could probably use some improvement, but it's another step, anyway.  Enjoy.

The color-coding does improve contrast, and is nice.

But the underlying Google Maps image still creates clutter, that detracts from a view of traveled routes. Ultimately, for the maps shown in user stats pages for regions/subregions/localities, can we get something like the old CHM map display, that displays only traveled routes (with route numbers added), untraveled routes (heavily faded, but also with route numbers), borders, and coastlines, against a white/pale pastel background? Of course, Google Maps is for now taking care of the borders and coastlines, so we'd need to render those in some other fashion (as CHM did, though rather roughly for most of the coastlines) if we don't have a map as the background.

Yes, I hope and intend that we'll have those at some point.  For my own priorities, I'm thinking of getting the mechanism by which stats can be generated down soon.  Anyone who'd like to try to get the non-Google maps sooner than it comes to the top of my list, the PHP code should be similar, but instead of adding things to a Google Map, they'd be drawn on the static map.  We have C++ code from yakra that shows how to manage those maps in GitHub, as is all of my code to date.

EDIT: The boundaries and coastlines *are* in GitHub, *but not* my version of the DB.  That will need to happen before those maps can be generated in a nice way.

[Jim]

The .list files I've been working with are now in a new GitHub repository TravelMapping/UserData .  I'd like a few of you who have some experience with GitHub or are willing to learn to clone the repository, update or add a (i.e., your own) .list file, commit and push it back.  I think this would create a pull request for me, as sole owner (for now) of the GitHub master of that repository, which I could then act on to get your update in, which I could then pull down to my own local clone used for site updates.  It's totally possible I have the workflow wrong here or that it's not a good way to do it, but I'd like to figure out a way that one or more others could take over accepting and putting .list files into the repository.

[vdeane]

The .list files I've been working with are now in a new GitHub repository TravelMapping/UserData .  I'd like a few of you who have some experience with GitHub or are willing to learn to clone the repository, update or add a (i.e., your own) .list file, commit and push it back.  I think this would create a pull request for me, as sole owner (for now) of the GitHub master of that repository, which I could then act on to get your update in, which I could then pull down to my own local clone used for site updates.  It's totally possible I have the workflow wrong here or that it's not a good way to do it, but I'd like to figure out a way that one or more others could take over accepting and putting .list files into the repository.

I tried to do that a little while ago and I'm getting this error.  Of course, I'm pretty rusty with git, so it's possible I'm doing something wrong.

Code Select Expand


vdeane@QueenElsa ~/TravelMapping/UserData/list_files $ git push -u origin masterUsername for 'https://github.com': vdeane
Password for 'https://vdeane@github.com':
remote: Permission to TravelMapping/UserData.git denied to vdeane.
fatal: unable to access 'https://github.com/TravelMapping/UserData/': The requested URL returned error: 403


[sammi]

The .list files I've been working with are now in a new GitHub repository TravelMapping/UserData .  I'd like a few of you who have some experience with GitHub or are willing to learn to clone the repository, update or add a (i.e., your own) .list file, commit and push it back.  I think this would create a pull request for me, as sole owner (for now) of the GitHub master of that repository, which I could then act on to get your update in, which I could then pull down to my own local clone used for site updates.  It's totally possible I have the workflow wrong here or that it's not a good way to do it, but I'd like to figure out a way that one or more others could take over accepting and putting .list files into the repository.

I tried to do that a little while ago and I'm getting this error.  Of course, I'm pretty rusty with git, so it's possible I'm doing something wrong.

Code Select Expand


vdeane@QueenElsa ~/TravelMapping/UserData/list_files $ git push -u origin masterUsername for 'https://github.com': vdeane
Password for 'https://vdeane@github.com':
remote: Permission to TravelMapping/UserData.git denied to vdeane.
fatal: unable to access 'https://github.com/TravelMapping/UserData/': The requested URL returned error: 403


You don't have push rights to the main repo. Fork it to, say, vdeane/UserData, then push to that repo instead, then make a pull request when you want it merged into the main repo. I did the same thing to sammdot/TravelMapping-Web and sammdot/TravelMapping-HighwayData.

Also, nice hostname.

[vdeane]

Having to create separate repositories for every single person who wants to do something sounds like a very cumbersome way to handle things.  When I worked in the computer field, the company I was at actually had everyone create separate branches for each change (though I didn't do that here because we don't have a convention nailed down on how to handle multiple users).

[Jim]

Having to create separate repositories for every single person who wants to do something sounds like a very cumbersome way to handle things.  When I worked in the computer field, the company I was at actually had everyone create separate branches for each change (though I didn't do that here because we don't have a convention nailed down on how to handle multiple users).

We definitely wouldn't do that.  I'll have more to say soon, as I've been experimenting with this.  For now, I'll say that I think we might want to set things up so we could allow people to email/upload/etc somewhere with no knowledge of GitHub, allow the more advanced user to manage his/her own list updates by maintaining their own fork and creating pull requests when they want a new list file imported (it's very possible this is more trouble than it's worth).  The main team of people responsible for user data maintenance would have course have the ability to push right to the master for themselves and your typical user who submitted new or updated list files outside of GitHub.

[Jim]

So anyway, here's what I'd like a few people with GitHub experience to try out:

Fork the TravelMapping/UserData repository into your account.  Make a change to your .list (I just added a comment at the top when I did mine).  Commit it to your fork.  Create a pull request to send it back to the master.  You can do all of this through the web interface if you'd like or use the other tools.

My experiment with this was based on my reading of https://help.github.com/articles/using-pull-requests/.

Again, I don't envision this as the mechanism we'd require all users to employ, but I'd like to see if it's a reasonable option for those who'd like to do it.

[SSOWorld]

So anyway, here's what I'd like a few people with GitHub experience to try out:

Fork the TravelMapping/UserData repository into your account.  Make a change to your .list (I just added a comment at the top when I did mine).  Commit it to your fork.  Create a pull request to send it back to the master.  You can do all of this through the web interface if you'd like or use the other tools.

My experiment with this was based on my reading of https://help.github.com/articles/using-pull-requests/.

Again, I don't envision this as the mechanism we'd require all users to employ, but I'd like to see if it's a reasonable option for those who'd like to do it.

Ok, I put a pull request in.  It was a little crazy trying to do it so don't get confused by the forking. (Forking git - git it? )

[Jim]

Ok, I put a pull request in.  It was a little crazy trying to do it so don't get confused by the forking. (Forking git - git it? )

Thanks.  I believe I've pulled your changes in, and next time I update the DB this list file's contents should be reflected.  It's definitely not an intuitive process, and probably not the way most people will choose to update their lists...  But, it is doable!

[SSOWorld]

you did, but I had a stray "A" in one of the lines - made the change

[mapcat]

I *think* I finally successfully submitted a pull request for a file named mapcatOLD.list using the web interface.

[Highway63]

One more for tonight: routes in the mapview.php pages are now color-coded by system.  A few of the colors could probably use some improvement, but it's another step, anyway.  Enjoy.

Is this different than the maps in the /hbtest/ directory? I'm still only seeing red lines.

Are the maps going to be overlaid on Google when the user checks each page, or will the maps be featureless except for shapelines (as it is on the CHM site)?

[Jim]

Is this different than the maps in the /hbtest/ directory? I'm still only seeing red lines.

Both the single-route browser and the map overlay viewer (for lack of a better name) are in the the hbtest directory.

A sample browser URL, showing a single route, u= is optional:

http://www.teresco.org/~terescoj/travelmapping/hbtest/?r=ny.ny119&u=terescoj

A sample map overlay URL, multiple routes by system(s) with sys= or region(s) with rg=, u= should be specified:

http://www.teresco.org/~terescoj/travelmapping/hbtest/mapview.php?rg=NY&u=terescoj

Are the maps going to be overlaid on Google when the user checks each page, or will the maps be featureless except for shapelines (as it is on the CHM site)?

I hope we'll eventually support both.  I can't promise I'll implement the CHM-style maps soon.  I intend to do it, but won't be upset at all if someone beats me to it.  There's a lot with error detection and stats generation that I am treating with higher priority.

[TAG42481]

I'd love to be able to view my LIST file as well.  I hadn't even heard of GitHub until this thread, sorry.  Thanks for the help - so glad to see that people who know what they're doing are taking up the charge for this (as opposed to people like me who don't know enough to contribute anything)!

http://cmap.m-plex.com/list/tag42481.list

[sipes23]

It's possible that I used the one-eyed babysitter to give me the space to figure out how to push my updates to the repository. (And yeah, I feel dumb. It's as browser driven as I could want. School won't start soon enough.)

[Jim]

I'd love to be able to view my LIST file as well.  I hadn't even heard of GitHub until this thread, sorry.  Thanks for the help - so glad to see that people who know what they're doing are taking up the charge for this (as opposed to people like me who don't know enough to contribute anything)!

http://cmap.m-plex.com/list/tag42481.list

Not a problem - I'm sure we will never require anyone to deal with GitHub to be able to use the new site.  I've added yours to my collection and you should be able to see your maps within a few hours, or as soon as I have a chance to run another site update.

[Jim]

It's possible that I used the one-eyed babysitter to give me the space to figure out how to push my updates to the repository. (And yeah, I feel dumb. It's as browser driven as I could want. School won't start soon enough.)

It worked.

[bejacob]

I love being able to zoom in on the maps. Already I've found a few mistakes I've made (false positives). One instance where I selected the wrong waypoint showed up on GA I-20 (I input 113 instead of 114). I go onto the freeway from US129/US441 and without being able to zoom in on the map, I would never have noticed my error, as 113 is a valid waypoint (just not the correct one). Zoom into this image where US129/441 & I-20 meet to see what I mean
http://www.teresco.org/~terescoj/travelmapping/hbtest/mapview.php?rg=GA&u=bejacob. I'm not sure I would have discovered it without the underlying map and the ability to zoom. I'm still finding mistakes, and with new travel coming up this weekend, I hope to fix my .list file soon, maybe by July 4th.

In any case, what I'm seeing so far is amazing.

What I (as a non-technical user) will be looking for as things move forward are the summary stats like these http://cmap.m-plex.com/stat/travsummary.php?u=bejacob. At this point I'm less concerned with the accuracy of the mileage than I am with the number of highways traveled/clinched. Those of you who have been a part of CHM much longer doubtless would like to see some of this as well. I'm sure you all have your own wishlist 

I have no idea what level of effort it will take to generate this, but based on what has happened already, it looks to just be a matter of time. As an avid travel/county collector who came late to the CHM party, I can't thank all of you enough for what you are doing. I spent countless hours putting together my .list file and did manage to get it uploaded in November, but never go to make corrections or add new routes before the site when dormant.

Keep up the great work. As more is developed, let me know through this forum. I will be happy to provide feedback and comments with somewhat of an outsider perspective.

Thanks again,
Brian

[Jim]

What I (as a non-technical user) will be looking for as things move forward are the summary stats like these http://cmap.m-plex.com/stat/travsummary.php?u=bejacob.

Me too!  It's what I'm working on next.  The main complication is accounting correctly for concurrencies so segments aren't counted multiple times in summary.  For example, take a segment that carries two interstate routes, a US route, and a state route.  It should count just once in overall and regional stats, once each for system stats for I-, US, and the state's system, and once in each route's stats.  There's also the issue that connected routes have not yet been included, which will be necessary for computing things like how much of I-90, across all states, has someone traveled.

[Wadsteckel]

- I'm happy to bring in .list files of others who might want to be able to browse their travels this way and the next time I run a DB update, they'll be there.

Hi Jim,

(Quoting Arnold Horshack) Ooooh! Ooooh!  Can I volunteer?  :-)

Just came home for a weekend, updated my travels file, then started reading the forum to try to catch up on the progress you guys are all making.  Happy to see progress, and am excitied to see a return of a functioning map site.  Let me know if I can send you my file, Jim.

Many thanks,
-Ed S