Current data is from a terrible web scraper, would be nice to have the source data.

Also the suffixes like "/ 001" look ugly and seem to be used as a workaround for unique naming. Would be good to return it to the format "Lecture A" instead of "Lecture01 /001"

Received the following email:

After finalizing my timetable on Safari, clicking on the “Download as .ics” button results in this error message:
Safari can't open the page "data:blob:http://anutimetable.com/5fe43ace-ccc6-4284-823c-edc2678e5a8d”. The error is: “The operation couldn’t be completed. (kCFErrorDomainCFNetwork error –10.)” (kCFErrorDomainCFNetwork:-10)

I will have a look at this on the weekend, however this issue has existed since the start of the project so I suspect it will be another dead end.

This is an unlikely one since it would be a major change to the data source and scraper. Comment from Stalkerspace:

I like the alternative ANU timetable builder, however I can't seem to view the weeks that classes are on or off. Is there some extra function to do this, or should I dig through the ics file?

Received an email

I used Firefox and it was able to successfully download an ics file, but the ics file cannot be imported into the Calendar app for Mac. I tried several times but it still can't be imported nor read by the Cal app.

Empty cells appear when tutorials which clash with others are selected, when the width of the column is reduced when all clashes in a day are removed, disfiguring the table. Can be fixed with a page refresh.

COMP2100/6442

If you select a lecture, it removes all other lectures, even though they shouldn't be.

This is because the datasource has them all named as lecture 1

Some classes start at 7am (laws2244)

This is a problem since the earliest time shown is 8am, but at the same time, adding more hours will make the rest of site harder to use

Ideally, the only times should show the minimum amount of hours required to show the whole timetable for that user

For example COMP2310

Just got a message from someone saying the site wasn't working, turns out they were entering "STAT 7003" instead of "STAT7003"

Maybe have a cookie feature so returning users don't have to re-enter their courses.

There would also need to be a "clear courses" button.

I'm using Chrome Version 55.0.2883.95 (64-bit).

Before deleting:

After deleting:

When entering COMP4630, I get the error 'course not found'. But the lecture times for this course are up on the official ANU timetable.

Seems like ANU is changing the timetable multiple times a day because of o-week, can you do more frequent updates @samyex6?

In some browsers it creates strange behavior, I think people would prefer to just have their courses appear instantly

ICS files puts all dates into Semester 1 when imported to a calendar application, should put them into the correct dates, in the second half of the year.

For example, see SCOM1001, if you select one tutorial it does not get rid of the rest.

This may be due to an error in the data itself, or the scraper.

because we're currently using item.id to identify tutorials, but they might be changed after data updates, need to change it to something more reliable.

Looks like it's made to be hosted at ANU, is there a link somewhere that people can access it remotely for updates?

Using angularJS will be much easier to implement UI components.

Received the following email:

It would be pretty dope if you could rename the classes after you have selected them

For COMP3420 three lectures, you are only able to choose one of them( if you choose one the rest of them will disappear like choosing tutorials)

And COMP2100 Lab is 1 hour long not 2 hours.

btw: anutimetable.com is 1000 times better than the timetable system provided by ANU

From email:

When building a timetable with math1013, it seems to pick up the lecture times for the current semester and the previous semester. MATH1013 (and MATH1014) are run in both semesters.

For example, if I type in "com" I don't get any COMP courses, and when I type in "comp" only 3 of the courses are COMP courses.

Ideally the filtering would assume that the user is typing a course code at first, if they type a space or anything that does not match the form /(^[A-Za-z]{1,4}\d{0,4})/gm then assume they are typing the name of the course rather than the code.

For example "ANTH8006_S1 Social Analysis & Community Politics (Advanced)"

Should display the the user as "ANTH8006 - Social Analysis & Community Politics (Advanced)"

The actual value may need to be stored elsewhere to assist with mapping the selected value to the data

Not sure if /data/timetable.json is being used at all now with the updated data method, ideally we should just delete it as it is a large file (an attack vector for DDoS)

With the new scraper working great, people shouldn't need to load their own JSON

The button should be removed as it will just confuse non-tech savvy users

For example ENGN3230 is missing from the list, but available in the ANU system

Need to investigate:

• Any throttling ANU may be enforcing
• What happens to the scraper when a 200 response error page comes back from ANU

When there is only one class for a certain group left, the choose button shouldn't be visible.
The structure now does not store how many equivalent classes a group has, need to deal with this.

Thinking a border of some sort to make the groupings of the classes more obvious.

Two entries for an identical tutorial don't necessary have the same id.

E.g.:

{"info": "Workshop01/ 006", "name": "SCOM1001", "hour": 11, "id": 2784, "location": "Green Couch Room", "day": "fri"},
{"info": "Workshop01/ 006", "name": "SCOM1001", "hour": 12, "id": 2785, "location": "Green Couch Room", "day": "fri"},

Which will cause an issue of treating these two as different classes.
Therefore I added another identifier fgroup (item.name + item.info) as a comparing element.

Do you think the table will look better after adding border-space?

.table {
    border-collapse: separate;
    border-spacing: 2px;
}

The semester has already started people already got their basic timetable figured out so there's no point of adding stuff piece by piece, V2 aims to migrate to the angularJS framework and the bunch of bugs & improvements.
Updates: #50 #60 #64 #69 #78
Possible updates: #65 #83 #84 Mobile-friendly

Welcome to post any ideas/solutions.

For example ACST4031_S1 Actuarial Control Cycle 1

I think it would be a lot clearer if the lesson style was applied to the timeslot rather than the lesson container.

Currently scraping timetable off-campus is ridiculously slow, will need to improve it by batching.

If a course has too many tutorials to choose from, the timetable will look nasty and difficult to decide.
So I'm planning to add a feature which will give you the best solution.

This selector only gives solutions that don't have collision on tutorials, which means lectures can have the same timeslot with tutorials.

As noted in the known issues, some classes go for 90 minutes and with the current system, this would lead to clashes.

Some changes in the last couple of days e.g. COMP4660, COMP4670

Hi Jan,

I've been poking around with the ANU Timetable Builder for a couple days and I just noticed an incident where the data on the timetable does not match up with that provided by the university. I'm not sure what the cause of it is yet but thought I'd alert you ASAP.

I observed that with ECON1101, there are three lessons displaying the info of "Lecture1 / 1" on anutimetable.com where two of these are actually supposed to read "Lecture2 / 2" and "Lecture3 / 2" respectively, according to the data at http://timetabling.anu.edu.au/sws2017/

I figure that it's probable that similar discrepancies exist for other courses.

The main problem with this is that it entirely throws a spanner in the 'choosing' of lessons because these lessons are inaccurately grouped together.

I noticed that when iterating through every element in timetableData["ECON1101"] and logging each lesson info, the dud "Lecture1 / 1" lessons were not even included in the data set (ie. only the two Monday lectures, one Tuesday lecture, and Thursday morning lectures were logged). I don't fully understand the implementation of the timetable or how they could be missing from timetableData["ECON1101"] and yet still displayed on the timetable so perhaps this part of my investigation is a red herring.

tl;dr I've attached an image of the discrepancy

Cheers,

If your web scraper is run manually, it can be easy for it to get out of date if they do updates without you noticing (all the time). @bshlgrs had this problem last year and got an email from ANU Timetabling when they noticed his was out of date. It'll also solve the issue of when you haven't updated your information when a new semester begins. :)

It's not something I can add myself without figuring out your data sources or there'd be a PR attached. :)

Received the following email:

Whilst I was making my timetable I noticed some options didn’t have the ‘choose’ feature despite them being duplicates. This is the case for the BIOL1008 Large Group Teaching and all the STAT1008 lectures.

Every time ANU modifies the initial form the web scraper POST data is invalidated - if the scraper could scrape this data from the form and then pass it into subsequent requests it would not require so much human intervention.

I've noticed one semi-regular problem with some events leaking over into the wrong day columns:

Seems very particular to the order that I add my courses, though. I tried adding the same courses in a different order and I didn't get the above bug but instead got the following, also problematic result:

(PS I've been playing around with the timetabler for a few days and I'm absolutely loving it. I've been trying to implement a few of my own improvements locally that may or may not come through. Y'all are doing great work!)

change option 'choose' for draggable? for example https://github.com/alanylc/CoUST

I really like your application

All times on courses (That I've seen: ENGN1217 ENGN1218 COMP1730 MATH1014) are displayed 1 hour earlier on the timetable.

http://anutimetable.com/ is showing up as the 4th result for me in Google, would be funny if we could get it to number 1

Requested via email

I think people would prefer something that automatically saved when they added a course - Clear should also clear local storage and cookies.

After that's done the Save and Flush features can be removed. Also the instructions at the bottom can be removed since it should be simple enough with just the two buttons.