Small fixes

[jvrsantacruz]

Hello there,

We're using multi-mechanize quite successfully at the company I'm working in, Floqq [1].
Sadly, there were a couple of annoyances that kept our set up from being totally stable.

Actually, we were using the software as a part of a low-traffic simulation, where, sometimes, not much data its left to be plotted after the tests. In that situation, the results.py side of the application was failing, dying with very simple IndexError exceptions.

Besides that, we use a barely complex set-up for this simulation, that includes python files within the test_scripts directory that are not intended to be used as testing scripts, and were rejected by the validator, which made the application burst with an exception. This behaviour was also changed to a simple warning in the command's output.

We really appreciate this fine piece of software you made and we're successfully and happily using it.

Kind regards,
Javier

[1] www.floqq.com

[cgoldberg]

thanks Javier,

I have been super busy, but will review your updates soon, and
hopefully pull your branch.

I was away from doing performance work for a bit, but have been
actively using multi-mechanize again.

just curious, have you run current trunk much (of m-m), or just using
the PyPI version? I merged a significant change in script loading a
month or so back that I haven't test fully yet... wondering how it
behaves.

thanks again!

-Corey

On Wed, Jul 25, 2012 at 3:05 PM, Javier Santacruz
[email protected]
wrote:

Hello there,

We're using multi-mechanize quite successfully at the company I'm working in, Floqq [1].
Sadly, there were a couple of annoyances that kept our set up from being totally stable.

Actually, we were using the software as a part of a low-traffic simulation, where, sometimes, not much data its left to be plotted after the tests. In that situation, the results.py side of the application was failing, dying with very simple IndexError exceptions.

Besides that, we use a barely complex set-up for this simulation, that includes python files within the test_scripts directory that are not intended to be used as testing scripts, and were rejected by the validator, which made the application burst with an exception. This behaviour was also changed to a simple warning in the command's output.

We really appreciate this fine piece of software you made and we're successfully and happily using it.

Kind regards,
Javier

[1] www.floqq.com

You can merge this Pull Request by running:

git pull https://github.com/jvrsantacruz/multi-mechanize master

Or you can view, comment on it, or merge it online at:

#17

-- Commit Summary --

• Avoids calling min/max with an empty sequence
• Fixes IndexError when splitting empty points sets
• Ignore invalid scripts instead of firing an exception
• Changes string path composing with os.path.join
• Avoid repeated code in the progress bar

-- File Changes --

M multimechanize/core.py (4)
M multimechanize/results.py (23)
M multimechanize/script_loader.py (3)
M multimechanize/utilities/run.py (17)

-- Patch Links --

https://github.com/cgoldberg/multi-mechanize/pull/17.patch
https://github.com/cgoldberg/multi-mechanize/pull/17.diff

---

Reply to this email directly or view it on GitHub:
#17

[jvrsantacruz]

Hi Corey,

The definitive set up we have working it's based on the code cloned from github.
I used the PyPI version at first, before getting hands in the code.

I did see the change you're referring to, and it seems to be working just fine AFAIK.

[dustinwhittle]

Any update on this?

[cgoldberg]

is anything here still valid, or has it diverged too much?