Status of Travel Mapping

[Jim]

Is there a way to display a specific road in the maps?  Say I wanted to just look at what I'd driven on I-5.  Could I add something similar to CHM's "r=ca.i005"?

Yes, that's still there in the draft highway browser, which was the first of the maps produced in the new project.  For example:

http://www.teresco.org/~terescoj/travelmapping/hbtest/?r=ny.ny008&u=terescoj

[Purgatory On Wheels]

Ah ok.  I was leaving the mapview.php in the address and thought you might have disabled this.  Thanks.

[Jim]

At some point, all of the functionality will be organized into a coherent site, so you won't have to remember URLs from this discussion board...

[yakra]

One suggestion for hbtest:
Presently the waypoint numbering starts at 0. I'd prefer it starting at 1 as on the old site. That way, waypoint numbers would correspond with line numbers in the text file, making it easier to look up & edit points in gedit or notepad++, etc.

[Jim]

One suggestion for hbtest:
Presently the waypoint numbering starts at 0. I'd prefer it starting at 1 as on the old site. That way, waypoint numbers would correspond with line numbers in the text file, making it easier to look up & edit points in gedit or notepad++, etc.

Done, but barely tested.  Hopefully I didn't break anything else (seems unlikely that the change I made would).

[Jim]

My site update script switched over to use the new hwy_data folder for its highway data tonight.  I had to make a bunch of minor fixes (some missing files, files in incorrect locations, etc.) to the data but that's all committed and pushed, and the script is producing the same output as far as I can tell.  I'm updating the DB now, should be live in a few minutes.  Please let me know if you notice anything has broken with this switch.

At this point, we should not need to touch chm_final -- we'll retain it for historical purposes and for reference, but from this point forward, we'll be diverging from CHM's highway data.

I'd like to have a couple of experienced highway data maintainers who have fixes to put in fork a copy of the HighwayData repository and try to make a couple changes, then commit to your fork and when you're ready, submit a pull request.  Let's start it small for the moment, and once we know the process works, we can get going on a full-scale update to fix up existing active systems with changes and corrections that have accumulated since last year.

[yakra]

Shouldn't we hold off on diverging from legacy CHM data for a bit longer, until there's more of an active user base, and an equivalent of labelsinuse?

[mapcat]

Also, the Wisconsin problem I emailed you about last week still hasn't been resolved.  I found one error (US53BusSol has the same mileage as US53BusSup) but suspect there's another one, since the data is still ~3 miles off, and am continuing to look.  Details in this post.

Edit: added link to details.

[Jim]

Shouldn't we hold off on diverging from legacy CHM data for a bit longer, until there's more of an active user base, and an equivalent of labelsinuse?

I'm ready to take in as many users as we wish, but do we have a way to invite them in other than this forum?  I suppose we could grab everyone's list files from CHM.  I think we'd end up with many out-of-date and unmaintained files, so I'm hesitant.

We do have the last points in use data from CHM still available, and the new system has the much smaller subset generated by our 32 users in log file form.

[mapcat]

As the changes and corrections are made, would this also be a good time to explore redefining what makes a route worthy of being listed?

Examples that come to mind: adding New York's Cross County Parkway and Kentucky's Hal Rogers Parkway, among others, to US Select Named Freeways.  The part of the NYST's Berkshire Connector that isn't I-90 also ought to be considered.  IIRC, the argument against these was that they were either too short, or (in the case of the Hal Rogers Pkwy) mostly 2-lane.

[bejacob]

As the changes and corrections are made, would this also be a good time to explore redefining what makes a route worthy of being listed?

Examples that come to mind: adding New York's Cross County Parkway and Kentucky's Hal Rogers Parkway, among others, to US Select Named Freeways.  The part of the NYST's Berkshire Connector that isn't I-90 also ought to be considered.  IIRC, the argument against these was that they were either too short, or (in the case of the Hal Rogers Pkwy) mostly 2-lane.

I have long been curious why the Hal Rogers Parkway wasn't included when all the other Kentucky Parkways were on the list. It is a fairly major route in that part of the state despite being mostly a 2-lane road.

[english si]

Shouldn't we hold off on diverging from legacy CHM data for a bit longer, until there's more of an active user base, and an equivalent of labelsinuse?

Obviously there's the question of pulling in .list files and whether we do it. A wider user base will be harder to attract with outdated data, and just creates more annoyed people when their list file has errors due to genuine can't-help-but-break-it changes that might be out there - on top of annoyed people who want to clinch a route they drove on months ago, but can't until the data is updated.

Also, surely (given we're using .wpt, rather than .wpt2), we have the older labels in use in use that is correct but for the few changes made to .list files with our small test userbase?

I have updates that I sent to Tim a year ago and I'm just itching to finally get them published to the wider world.

[english si]

Examples that come to mind: adding New York's Cross County Parkway and Kentucky's Hal Rogers Parkway, among others, to US Select Named Freeways.

Or better yet: NY Parkways and KY Parkways respectively, rather than ussnf.

[yakra]

One thing I've wanted to see is breaking NY & KY Parkways and OK Turnpikes out into their own systems.

Didn't occur to me till just now though, that if it were just NY Parkways, that would exclude the Berkshire Connector. And that part of the Thruway that's neither I-87 nor I-90, if we want to get that nitty-gritty about it. Might be worthwhile to change the definition a bit...

[english si]

NYST (inc Berkshire Connector) would surely (like the NY section of the GSP - which isn't a New York Parkway) just simply be in 'USA Named Freeways', not forming part of the specific set, but fine in the grab-bag?

[bejacob]

Shouldn't we hold off on diverging from legacy CHM data for a bit longer, until there's more of an active user base, and an equivalent of labelsinuse?

Obviously there's the question of pulling in .list files and whether we do it. A wider user base will be harder to attract with outdated data, and just creates more annoyed people when their list file has errors due to genuine can't-help-but-break-it changes that might be out there - on top of annoyed people who want to clinch a route they drove on months ago, but can't until the data is updated.

Also, surely (given we're using .wpt, rather than .wpt2), we have the older labels in use in use that is correct but for the few changes made to .list files with our small test userbase?

I have updates that I sent to Tim a year ago and I'm just itching to finally get them published to the wider world.

32 users is a large enough sample on which to test changes to the system. The work so far seems to have replicated CHM well, though there are still issues from time to time. Now that most of the underlying functionality is working it makes sense to move beyond CHM. Errors can still be identified and addressed. Gathering more users from CHM can be done at any time. Those of us here already (especially less technical folks like me) can continue to provide feedback and potentially ask the stupid simple questions when we see things we don't understand.

I know that in addition to new systems and new/revised waypoints, there is still the web presence to develop. I would guess many CHM users will not be ready to use TM until more of that is in place. I've been able to figure out much of what I want using the stats in the log file and the map overlays. When pages like this http://cmap.m-plex.com/stat/region.php?u=bejacob&c=usa&rg=oh&du=mi&sort=ra are ready (whether similar or completely redesigned), I think more CHM users will want to participate.

As for the next steps, I say "go for it!"

[Bickendan]

I also wouldn't mind seeing the Oregon side of WA 433 (the Lewis and Clark Bridge), the Bridge of the Gods and the Hood River Bridge added, as they are the only substantial Columbia River crossings. And if the ferry portions of Washington's highways are included, that would include the Oregon side of WA 409 (I like how OSM flags WA 409 on Westport Ferry Rd in Westport, OR, lol).

[intelati49]

Shouldn't we hold off on diverging from legacy CHM data for a bit longer, until there's more of an active user base, and an equivalent of labelsinuse?

Obviously there's the question of pulling in .list files and whether we do it. A wider user base will be harder to attract with outdated data, and just creates more annoyed people when their list file has errors due to genuine can't-help-but-break-it changes that might be out there - on top of annoyed people who want to clinch a route they drove on months ago, but can't until the data is updated.

Also, surely (given we're using .wpt, rather than .wpt2), we have the older labels in use in use that is correct but for the few changes made to .list files with our small test userbase?

I have updates that I sent to Tim a year ago and I'm just itching to finally get them published to the wider world.

32 users is a large enough sample on which to test changes to the system. The work so far seems to have replicated CHM well, though there are still issues from time to time. Now that most of the underlying functionality is working it makes sense to move beyond CHM. Errors can still be identified and addressed. Gathering more users from CHM can be done at any time. Those of us here already (especially less technical folks like me) can continue to provide feedback and potentially ask the stupid simple questions when we see things we don't understand.

I know that in addition to new systems and new/revised waypoints, there is still the web presence to develop. I would guess many CHM users will not be ready to use TM until more of that is in place. I've been able to figure out much of what I want using the stats in the log file and the map overlays. When pages like this http://cmap.m-plex.com/stat/region.php?u=bejacob&c=usa&rg=oh&du=mi&sort=ra are ready (whether similar or completely redesigned), I think more CHM users will want to participate.

As for the next steps, I say "go for it!"

I have been meaning to do this even before this project, where do I sign up?

[Jim]

I'll be offline a lot the next few weeks, and my time for the project will be much more limited when I get back.  I'd like to focus my efforts on helping others get to the point where they can work on the web front end and get going on highway data updating and later, new development.  I'd also like to get a few people to volunteer to take .list file updates from those who would prefer to keep their hands out of GitHub, and get them into GitHub on their behalf.  For now, I think just accepting them by email to some alias (I'm willing to set one up), with some sort of upload system to be developed later to replace that process.  I should be able to run occasional updates to my preliminary site a few times a week while I'm away, probably daily when I return.  I hope we can transition fairly soon to a situation where multiple people can run updates.  Anyone should be able to have a copy of the repositories and with Python can generate the .sql and .log files.  I think only one other has done that so far, and I'd like to get more people up to speed with how to do that.

I've mentioned before that I'm not likely to do much on the front-end web side.  I hope there are others more qualified than I am to do that part.  I will try to encourage some students to work on it, but I don't want to promise anything.  I think almost everything we need to replicate CHM is in the DB now, so it's a matter of writing PHP and JS code along the lines of what you see in the mapview.php and chmviewerfunc3.js in the Web repository to generate other maps and lots of stats.

[SSOWorld]

As a note: I've been able to get PHP code working to draw the lines of the CHM style maps.  However, I have not been able to get the clinched status to work yet.  I haven't published it yet, but it is in a fork in my personal directory in GitHub.  I'm trying to figure out how to filter out clinched points in the database. to properly draw the lines.  At this point, the lines do plot the routes, but the clinched lines saturate and draw all over.

https://github.com/ssoworld/Web for the code.  EDIT: VERY Prototype

EDIT 2: Need to change numbering scheme to start at 1.  Won't work until then.

NOTE: haven't done anything for just over a week - so DB may not be accurate.  Caution.

[Rothman]

I'm not a programmer by any definition, but just wanted to pop up to say that I hope this effort comes to fruition.  I'm sure there are a "lot" of other people like me lurking on the sidelines whose eyes sort of glaze over the technical details that are just waiting for a new site to come online so they can restart tracking their clinches.

Thanks to all that have been involved so far!

[Jim]

As a note: I've been able to get PHP code working to draw the lines of the CHM style maps.  However, I have not been able to get the clinched status to work yet.  I haven't published it yet, but it is in a fork in my personal directory in GitHub.  I'm trying to figure out how to filter out clinched points in the database. to properly draw the lines.  At this point, the lines do plot the routes, but the clinched lines saturate and draw all over.

https://github.com/ssoworld/Web for the code.  EDIT: VERY Prototype

NOTE: haven't done anything for just over a week - so DB may not be accurate.  Caution.

I owe you an email about this - I'll get back to you before I leave.

[bejacob]

I'm not a programmer by any definition, but just wanted to pop up to say that I hope this effort comes to fruition.  I'm sure there are a "lot" of other people like me lurking on the sidelines whose eyes sort of glaze over the technical details that are just waiting for a new site to come online so they can restart tracking their clinches.

Thanks to all that have been involved so far!

I'm very much in your situation. I've been sending updates of my list to Jim as I make them for inclusion each time he updates the database. So far things look great. Lots still to happen, but I will add my thanks to everyone working on the project. From my perspective, I submit my file and Jim waves a magic wand which produces a log, stats, and maps. Then I have to figure out why there is an extra line here or there. 

Again, kudos to all of you who understand what goes on inside the black box. To many of us, it's still magic. 

[rickmastfan67]

As the changes and corrections are made, would this also be a good time to explore redefining what makes a route worthy of being listed?

Examples that come to mind: adding New York's Cross County Parkway and Kentucky's Hal Rogers Parkway, among others, to US Select Named Freeways.  The part of the NYST's Berkshire Connector that isn't I-90 also ought to be considered.  IIRC, the argument against these was that they were either too short, or (in the case of the Hal Rogers Pkwy) mostly 2-lane.

I have long been curious why the Hal Rogers Parkway wasn't included when all the other Kentucky Parkways were on the list. It is a fairly major route in that part of the state despite being mostly a 2-lane road.

Because it wasn't a full freeway.  That was the sticking point for Tim.  I had the file done at that time when the other Parkways went online, but Tim didn't want to add it because of that.

[Highway63]

I know nothing about PHP or GitHub, but I can send my .list file and my states' folders (if necessary) to those in need of data if necessary.

One suggestion for hbtest:
Presently the waypoint numbering starts at 0. I'd prefer it starting at 1 as on the old site. That way, waypoint numbers would correspond with line numbers in the text file, making it easier to look up & edit points in gedit or notepad++, etc.

Waypoints for interstates started at 0 at the state line. (Hence the "0 999" point entries.)