Web based management for RF
Bulkan Evcimen
I saw this tweet by Pekka the other day http://twitter.com/pekkaklarck/status/4938111943
and I was thinking we might start a discussion here.
Cheers
Pekka Klärck
>
> I saw this tweet by Pekka the other day http://twitter.com/pekkaklarck/status/4938111943
> and I was thinking we might start a discussion here.
Starting a discussion here is a good idea. Notice also that my tweet
was a reply to this question by @utws:
http://twitter.com/utws/status/4912653174
A wiki based solution ought to be relatively simple to implement using
some existing wiki engine as a base. I haven't studied options too
closely but MoinMoin (http://moinmo.in) would probably be a strong
candidate because it supports plugins, reStructuredText and remote
editing via XML-RPC, and it happens to be implemented with Python.
Somehow integrating RF to Fitnesse wiki might also be an option, but
I'm not sure it would be any easier than using any other wiki. The
main problem with any wiki based solution (incl. Fitnesse) is that you
edit the test data using plain text format in one big text box.
Maintaining such data can turn out to be really problematic -- that's
exactly why we are developing RIDE.
Another idea is be implementing a real web based editor that actually
understands RF test data. Basic functionality could probably
implemented using plain HTML but with Ajax and other Web 2.0
technologies it ought to be possible to implement real "WebRIDE". Such
a system could have all the same features as stand-alone RIDE, but a
web based system would be easier to maintain, truly platform
independent, and could also store all the data in one place. The
possibilities are pretty much unlimited.
How other people like these ideas and the overall idea of having a web
based system for creating/managing/running RF tests? If there's enough
interest a separate project could be started to make such a system a
reality. The NSN sponsored core development team doesn't currently
have time to take an active role in such a project, but I can
personally promise to help with design and perhaps also with
development.
Cheers,
.peke
--
Agile Tester/Developer/Consultant :: http://eliga.fi
Lead Developer of Robot Framework :: http://robotframework.org
Bulkan Evcimen
On Oct 22, 9:05 am, Pekka Klärck <p...@iki.fi> wrote:
> 2009/10/21 Bulkan Evcimen <bul...@gmail.com>:
>
>
>
>
> Starting a discussion here is a good idea. Notice also that my tweet
> was a reply to this question by @utws:http://twitter.com/utws/status/4912653174
>
> A wiki based solution ought to be relatively simple to implement using
> some existing wiki engine as a base. I haven't studied options too
> closely but MoinMoin (http://moinmo.in) would probably be a strong
> candidate because it supports plugins, reStructuredText and remote
> editing via XML-RPC, and it happens to be implemented with Python.
> Somehow integrating RF to Fitnesse wiki might also be an option, but
> I'm not sure it would be any easier than using any other wiki. The
> main problem with any wiki based solution (incl. Fitnesse) is that you
> edit the test data using plain text format in one big text box.
> Maintaining such data can turn out to be really problematic -- that's
> exactly why we are developing RIDE.
approach.
>
> Another idea is be implementing a real web based editor that actually
> understands RF test data. Basic functionality could probably
> implemented using plain HTML but with Ajax and other Web 2.0
> technologies it ought to be possible to implement real "WebRIDE". Such
> a system could have all the same features as stand-alone RIDE, but a
> web based system would be easier to maintain, truly platform
> independent, and could also store all the data in one place. The
> possibilities are pretty much unlimited.
RF test data
we could possibly use this in this in WebRIDE.
>
> How other people like these ideas and the overall idea of having a web
> based system for creating/managing/running RF tests? If there's enough
> interest a separate project could be started to make such a system a
> reality. The NSN sponsored core development team doesn't currently
> have time to take an active role in such a project, but I can
> personally promise to help with design and perhaps also with
> development.
common location
which I can maintain. Where as I need to check to see if and tell
everyone about new
versions of RIDE.
Pekka Klärck
> On Oct 22, 9:05 am, Pekka Klärck <p...@iki.fi> wrote:
>>
>> Another idea is be implementing a real web based editor that actually
>> understands RF test data. Basic functionality could probably
>> implemented using plain HTML but with Ajax and other Web 2.0
>> technologies it ought to be possible to implement real "WebRIDE". Such
>> a system could have all the same features as stand-alone RIDE, but a
>> web based system would be easier to maintain, truly platform
>> independent, and could also store all the data in one place. The
>> possibilities are pretty much unlimited.
>
> This is what I am leaning towards. RIDE already has code to parse the
> RF test data
> we could possibly use this in this in WebRIDE.
RIDE actually uses Robot Framework's parsing modules directly and in
the forthcoming RF 2.2 the target is to make these modules easier to
use from outside. It would also be possible to move some other RIDE
logic (e.g. keyword completion) either into RF core or into separate
"BaseRIDE" package. Reimplementing the wheel is not a good idea.
>> How other people like these ideas and the overall idea of having a web
>> based system for creating/managing/running RF tests? If there's enough
>> interest a separate project could be started to make such a system a
>> reality. The NSN sponsored core development team doesn't currently
>> have time to take an active role in such a project, but I can
>> personally promise to help with design and perhaps also with
>> development.
>
> I like the idea of having a web based system because there will one
> common location
> which I can maintain. Where as I need to check to see if and tell
> everyone about new
> versions of RIDE.
Yep, this is a very good point. Anyone else than Bulkan who's
interested about this?
Daniel Posada
pi
Interface!
On Oct 22, 9:22 pm, Daniel Posada <daniel.pos...@gmail.com> wrote:
> I'm very interested in the web editor! I think another great advantage is
> that it will make editing RF test files easier/more accessible for business
> analysts, QA engineers, and others who might otherwise find it inconvenient
> to use the raw files. That can be especially true if the raw files are
> stored in a source control repository which is not commonly used by those
> groups.
>
Elisabeth
Elisabeth
utws
a option for maintaining the chain: requirements - use cases -
testcases - testresults is also a big deal :-)
would work something like Bronime from the selenium team ???
cheers
Uwe
Gmail
what I would expect this would be. I would expect to see something
like RIDE, but saving to the central server. Then, some way of
exposing those tests (likely via http) to a locally run version of RF.
I may be completely off, but...
- Tim
Janne Härkönen
I decided to take some action and created a project, RFWeb, available
in Google Code [1] (surprise, surprise)
There's no content yet, however a project allows sharing prototype
code and ideas easily. So anyone interested in committing anything
(ideas in the wiki, code in the repository etc), please apply for a
project membership through me.
One of the first things that needs to be decided is the technology
stack to be used. Personally I'd prefer to use Django + jQuery, former
because I am somewhat familiar with it and latter because I have
wanted to learn it for a long time (and this seems like a perfect
opportunity). Since I am not sure how much I can actually contribute,
my opinion obviously matters little. Are there anyone willing to work
on the project having some opinions on this?
The other important thing is to figure out what to actually implement.
This requires having the problems that the system supposedly solves to
be spelled out. There are some in the air already, like editing,
downloading and executing but these would need to be fleshed out a bit
to be able to actually implement anything. What would be the most
important problem that you'd like this system to solve?
your turn now,
__janne
[1] http://code.google.com/p/rfweb/
Bulkan
On Wed, Oct 21, 2009 at 5:00 AM, Bulkan Evcimen <bul...@gmail.com> wrote:Hello,
>
> Hi all,
>
> I saw this tweet by Pekka the other day http://twitter.com/pekkaklarck/status/4938111943
> and I was thinking we might start a discussion here.
I decided to take some action and created a project, RFWeb, available
in Google Code [1] (surprise, surprise)
There's no content yet, however a project allows sharing prototype
code and ideas easily. So anyone interested in committing anything
(ideas in the wiki, code in the repository etc), please apply for a
project membership through me.
One of the first things that needs to be decided is the technology
stack to be used. Personally I'd prefer to use Django + jQuery, former
I would prefer Django + jQuery aswell.
because I am somewhat familiar with it and latter because I have
wanted to learn it for a long time (and this seems like a perfect
opportunity). Since I am not sure how much I can actually contribute,
my opinion obviously matters little. Are there anyone willing to work
on the project having some opinions on this?
I did volunteer to work on this a while back now.
The other important thing is to figure out what to actually implement.
This requires having the problems that the system supposedly solves to
be spelled out. There are some in the air already, like editing,
downloading and executing but these would need to be fleshed out a bit
to be able to actually implement anything. What would be the most
important problem that you'd like this system to solve?
your turn now,
__janne
[1] http://code.google.com/p/rfweb/
We probably need to put all the ideas on what rfweb should do on the project page and start fleshing it out more.
Janne Härkönen
> I would prefer Django + jQuery aswell.
>
If you have Python, Django, Robot Framework and RIDE installed, you
can do a checkout and run python manage.py runserver to get it
running.
The data is read from the 'test' directory and the system supports
files only directly in that directory and only in txt format. RIDE
modules are used for parsing the data. What seems promising that there
is currently grand total of 62 lines of code, including some empty
lines and HTML templates, and you can actually get changes written to
the disk :)
As we have something concrete, it might be easier to find out the
things that should be added so that the editor would be usable.
thanks,
__janne
Bulkan
On Thu, Nov 19, 2009 at 12:25 AM, Bulkan <bul...@gmail.com> wrote:I did a *very* (and I mean **very**) crude proof a concept and committed it.
> I would prefer Django + jQuery aswell.
>
If you have Python, Django, Robot Framework and RIDE installed, you
can do a checkout and run python manage.py runserver to get it
running.
I'm trying to get this to work but for some reason when I clone the RIDE source repository I'm not getting the "plugins" directory, that views.py is importing
http://bulkan-evcimen.com
Pekka Klärck
> I'm trying to get this to work but for some reason when I clone the RIDE
> source repository I'm not getting the "plugins" directory, that views.py is
> importing
relatively big module reorganization - that RFWeb hasn't probably keep
up. I believe Janne already gave you commit rights for this project so
if you are able to fix the problem please just commit a fix. Otherwise
I expect Janne to fix it at after he's recovered from his company's
Xmas party on Friday.
Bulkan
2009/11/27 Bulkan <bul...@gmail.com>:
>There has recently been so much changes on RIDE - including a
> I'm trying to get this to work but for some reason when I clone the RIDE
> source repository I'm not getting the "plugins" directory, that views.py is
> importing
relatively big module reorganization - that RFWeb hasn't probably keep
up. I believe Janne already gave you commit rights for this project so
if you are able to fix the problem please just commit a fix. Otherwise
I expect Janne to fix it at after he's recovered from his company's
Xmas party on Friday.
I've fixed the problem now, but I can't push my changes (I don't think i've been given commit access for rfweb).
Bulkan
Never mind, I was able to push in my fix.
Antsam
Tatu Aalto
Ugh
This topic arises now and then to surface, but I don't recall anyone actually announcing or have started a project which would provide a silver bullet to this problem, with one exception (to the later part). There is rf-hub [1] and it provides you a web service to search keywords. I think Bryan's original idea was to extend the rf-hub to run tests too, but I am not aware has he made progress on that direction.
Your original question did actually contain two needs. Able to edit RF data and run the test from web service. The former need is a good idea and now and then there is a discussion to replace RIDE with a such service. But I don't recall anyone saying that they would have build a such a service. For the later need, community, me including, seems to be using a CI. Example Jenkins has a nice plugin to display Robot Framework results. And to be able to run tests, one would need install all software to some machines, have a queue to wait free machine and so on, which really fast starts to sound like a CI.
All this ranting doesn't mean that one should not stop looking for a solution to the problem, I am looking too, but in my humble opinion, the problem can not be solved by one silver bullet. To ease the pain to write and run tests requires many solutions and taking one small step at the time will help to reach the solution.
-Tatu
[1] https://github.com/boakley/robotframework-hub
You received this message because you are subscribed to the Google Groups "robotframework-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to robotframework-u...@googlegroups.com.
To post to this group, send email to robotframe...@googlegroups.com.
Visit this group at https://groups.google.com/group/robotframework-users.
For more options, visit https://groups.google.com/d/optout.
Jessica
在 2009年10月21日星期三 UTC+8上午11:00:46，Bulkan-Savun Evcimen写道：