ReadTimeout while trying to upload captions. about vimeo.py

[greedo]

@samcarecho this error is coming from urllib3 and as you noticed we are not re-catching and wrapping the error in our own library.

If you contact us directly through vimeo.com/help/contact and share your exact request with runscope I may be able to help you further.

[samcarecho]

Sorry @greedo I will not sign-up of a paid third party service, even though it offers a 30 day free trial and spend time going through it's documentation just for your convenience. I'm really busy. I would ratter just share the requests logs in a plain txt file.

If I can't get help or you can't help me I will be forced to strongly advise the non-profit to which I'm working as a volunteer to avoid Vimeo. I will make my case showing to them all the 500 code errors the Vimeo servers are returning, some of which are returned as "unknown error".

I opened the issue hopping for a mutual effort, where both of us would work together to reach the non-profit goal and identify possible errors within the Vimeo API. I even sent a message to the e-mail you have shared publicly on your GitHub account. To which so far I received no reply.

Regarding the captions upload, I can't wait and I have already solved the problem using server-side JavaScript and cron jobs. I have set uploads to sleep for 15 seconds after each server error and then retry, and to sleep for 10 seconds after a successful upload. This is working and mitigating server errors , but it is taking a painful long time.

The captions uploads is an emergency and just the tip of the iceberg.
I'm planing and implementing for the non-profit, as a volunteer, a captions translation management system that will make use of both YouTube and Vimeo APIs, for publishing purposes.

[Dashron]

The pricing plans on runscope.com are new. We were not aware of the change.

We tend to send users through runscope because it provides 2 powerful features.

1. It's an http proxy (like wireshark or charles), so we get the exact values your code sends and receives instead of example code which may work different from system to system.
2. The requests are sharable, editable and repeatable.

The vast majority of developers we interact with do not have the understanding or ability to use charles or wireshark to identify their http logs, so runscope was our go-to solution. If you are comfortable with other http proxies we are more than happy to receive text documents containing your entire request and response cycle.

Additionally @greedo asked to move to email because we should test the exact access token you are using, and verify that there are no conflicts there. We do not want your access token exposed to the world.

If you would like to continue conversation here we are more than happy, just verify with @greedo via email, or some private contact which access token you are using along with the previously mentioned request logs.

[Dashron]

After some more review, Runscope is still 100% free for a single user, making up to 10k requests per month. You can find that on their pricing and plans page: https://www.runscope.com/pricing-and-plans.

The plan structure is definitely confusing, but they are by far the best http proxy I have ever used. I recommend you give them a shot, but again we are more than happy to receive text logs from some other tool.

[samcarecho]

Thanks Aaron for your attention.

When I visited the Runscope website, the only option available has the 30
day free trial, as per this link:
https://www.runscope.com/signup

Please, don't take me wrong. Runscope seems to be an amazing service. I
just do not have, right now, the time to go through it's documentation.
As soon as I get some free time I will take a deep look on it's
documentation, as their service seems to be something that will probably
easy my life.
For now, I will dump the HTTP communication using Wireshark and send it to
either you or @greedo.

What e-mail address should I send the dump to?
I just have the public @greedo address, which is not a @vimeo.com address.

Cheers,
Sam

On Tue, Jan 27, 2015 at 3:32 PM, Aaron Hedges [email protected]
wrote:

After some more review, Runscope is still 100% free for a single user,
making up to 10k requests per month. You can find that on their pricing and
plans page: https://www.runscope.com/pricing-and-plans.

It's definitely confusing, but they are by far the best http proxy I have
ever used. I recommend you give them a shot, but again we are more than
happy to receive text logs from some other tool.

—
Reply to this email directly or view it on GitHub
#54 (comment).

[greedo]

@samcarecho you can send it to me - [email protected]

I'll look into it.

[samcarecho]

Thank you @greedo, I will send you the dump.

[greedo]

@samcarecho we made some changes on our captions backend that should solve this problem. It might be best to build in some retry logic by catching that urllib3 error.

Feel free to reopen this issue again if you have problems again.

[samcarecho]

@greedo I will test it again as soon as possible.
Thanks