Anton Litvinenko's dashboard

Anton Litvinenko replied on July 14, 2009 16:05 to the problem "API Retrieve User Token query curl example" in Pivotal Labs:

Could you please explain or point at some docs that explain how one'd configure curl to use ssl?

the -l option didn't change anything for me - i still see the empty string. I tried the -k flag that tells curl to accept any certificates - but no luck.

I also have tried to implement the integration in Java, but there i get the 401 error and the response doesn't contain the "WWW-Authenticate" header, so apache's httpclient doesn't know how it should perform authentication and crashes with:


WARN [main] org.apache.http.impl.client.DefaultRequestDirector - Authentication error: Unable to respond to any of these challenges: {}

org.apache.http.client.HttpResponseException: Unauthorized
at org.apache.http.impl.client.BasicResponseHandler.handleResponse(BasicResponseHandler.java:71)


I also tried to make a direct URLConnection (in Java) but result still was 401.

Perhaps you can share User Token query implemented in any programming language (Java, Ruby, Python), so i could validate my code against that?

Anton Litvinenko reported a problem in Pivotal Labs on July 10, 2009 15:40:

API Retrieve User Token query curl example

Retrieve User Token query example with CURL from API guide (http://www.pivotaltracker.com/help/ap...) returns an empty string. Is it correct?

I have also tried curl -u user:pass -X GET https://www.pivotaltracker.com/servic... but result was the same.

Anton Litvinenko replied on December 24, 2008 07:53 to the question "Crawler runs successfully, but no analytics data is showing" in Programeter:

Hi Keith!

Does the crawler crash after that?

At the moment to me it seems that it should continue working after such exceptions. Usually they mean that parser we use is unable to parse some of the files.

Anton Litvinenko replied on December 16, 2008 17:06 to the question "Crawler runs successfully, but no analytics data is showing" in Programeter:

Good point. I am a bit scared that YEAR setting might provoke memory problems again (as your repository seems quite big).

But let's try YEAR setting if we will have troubles with that, then i would ask you to go back to QUARTER setting. Is that OK with you?

Anton Litvinenko replied on December 15, 2008 21:09 to the question "Crawler runs successfully, but no analytics data is showing" in Programeter:

Great! I am very happy to hear that :)

Anton Litvinenko replied on December 15, 2008 18:57 to the question "Crawler runs successfully, but no analytics data is showing" in Programeter:

This is problem on our side... I've checked the logs - our server crashed due to memory troubles. Could you please set "crawlingPeriod" to QUARTER in the repo.conf file.

Please ran the crawler once more after that (you don't need to reinstall anything)

I am really sorry for all this mess. As a "candy": can I offer programeter services for free for three months for your development team?

Anton Litvinenko replied on December 15, 2008 15:49 to the question "Crawler runs successfully, but no analytics data is showing" in Programeter:

yeap, our hosting service provider had a major outage. now all servers seems to be back to normal.

Anton Litvinenko replied on December 15, 2008 09:46 to the question "Crawler runs successfully, but no analytics data is showing" in Programeter:

Keith, please download the latest version of crawler and try using it. The new version is 3.8.5. This NPE should be gone and hopefully we will get your repository measured :)

Anton Litvinenko replied on December 13, 2008 17:05 to the question "Crawler runs successfully, but no analytics data is showing" in Programeter:

Yay! Last exception was result of my stupid mistake in the code. I have rushed so much to get the working version out, that forgot to properly run unit tests.

I will make a new release soon and notify you when it's ready!

Anton Litvinenko replied on December 13, 2008 08:20 to the question "Crawler runs successfully, but no analytics data is showing" in Programeter:

Yep, now logs show that you use version 3.8.4

My fix didn't work - I'll have to dig deeper.

Anton Litvinenko replied on December 12, 2008 18:52 to the question "Crawler runs successfully, but no analytics data is showing" in Programeter:

From the logs i am seeing that you are still using a 3.8.2 version (the previous one):

2008/12/12 12:58:53 Client 0abc4d49d27c09827e9579538311b6a593731e38 uses a crawler with version 3.8.2

(timezone is Eastern Standard Time)

Could you please re-check that? I suggest uninstalling the previous version of crawler (don't forget to back up repo.conf file) and then making a clean install of the latest version.

Anton Litvinenko replied on December 12, 2008 17:12 to the question "Crawler runs successfully, but no analytics data is showing" in Programeter:

Yes, this is possible, but that will affect the accuracy of analytics (the longer period we take - the more accurate analytics are).

Anyway, i tried to fix this issue by making sure that author of modification is always defined. The fix wasn't tested thoroughly because it is quite difficult to reproduce issue without knowing exact circumstances. If that wouldn't work then we will get back to your suggestion to crawl only last few months.

Please download the latest version (should be 3.8.4) from your account at http://my.programeter.com and try to continue crawling.

Btw, now you can see how far repository was crawled on the project analytics page (former team analytics page).

Anton Litvinenko replied on December 12, 2008 13:28 to the question "Crawler runs successfully, but no analytics data is showing" in Programeter:

I am checking on it. Very bizarre, cause at the moment it looks like some revisions in your repository don't have an author.

Anton Litvinenko replied on December 11, 2008 19:04 to the question "Crawler runs successfully, but no analytics data is showing" in Programeter:

not at the moment :( but, we are planning on a release tomorrow and that will include such feature.

So far, when crawler hits the end, it will issue a log statement like: "Repository [XXXXX] has been measured till the end: ..." and if you ran the crawler after that then third to last statement on the log will be ""Repository XXXXX was already crawled lately!"

At the moment i see that you have almost reached October 2006. So i would prognose that you need to ran crawler for 3-4 times more.

A comment on the question "Crawler runs successfully, but no analytics data is showing" in Programeter:

We are updating GUI so that you would able to follow crawling state as well – Anton Litvinenko, on December 11, 2008 17:54

Anton Litvinenko replied on December 11, 2008 17:24 to the question "Crawler runs successfully, but no analytics data is showing" in Programeter:

Hey! Your repository has been crawled till 2399 revision (which is somewhere in February 2006). So, you would need to run crawler few times more in order to get to December 2008.

While your repository is being crawled we will start preparing analytics for you!

Anton Litvinenko replied on December 11, 2008 14:32 to the question "Crawler runs successfully, but no analytics data is showing" in Programeter:

yes, crawler expects it to be mis-spelled and also it is case-sensitive. So the line with license key should be:
licenceKey=0abc4d49d27c09827e9579538311b6a593731e38

Could you fix that please and try running crawler once again?

Anton Litvinenko replied on December 11, 2008 14:08 to the question "Crawler runs successfully, but no analytics data is showing" in Programeter:

Hi Keith!

Could you please send me (either here or to twitter) contents of your *.repo.conf file (with substituted repository path and username/password).

In logs on our side i see that crawling result come without the license key specified.

Thanks!

Anton Litvinenko replied on October 29, 2008 17:05 to the idea "Download doesn't work" in Programeter:

Thanks for letting us know!

The issue should be fixed now. Please try again

Anton Litvinenko replied on August 07, 2008 09:11 to the idea "tracking of "project management" web apps" in RescueTime:

Hey! I am also interested in separate tracking time for particular pages on the same server