You are right. I misunderstood the issue. I read the thread again I thought
this was a py3 only issue. Will look into this shortly.
On Tuesday, 3 December 2019 08:26:15 UTC-8, Leonel Câmara wrote:
>
> py4web does sessions in redis by default (if you have redis). Just saying.
>>
>
> Massimo, this
>
> py4web does sessions in redis by default (if you have redis). Just saying.
>
Massimo, this really doesn't help. Web2py had working redis sessions and
right now, in master, it doesn't. Web2py should not break existing apps
just because there's py4web, there's no way people can all migrate t
On Thursday, 28 November 2019 23:21:39 UTC, Massimo Di Pierro wrote:
>
> py4web does sessions in redis by default (if you have redis). Just saying.
>
Yep and it looks promising but I don't have the guts to migrate a mature
site with tens of thousands of users on to it yet - a bit too bleeding edg
py4web does sessions in redis by default (if you have redis). Just saying.
On Thursday, 28 November 2019 14:56:37 UTC-8, Jose C wrote:
>
>
>> Any feel for how many people are using redis for sessions?
>>
>> +1 on on redis for sessions. Use it for multiple apps although they're
> all older web2py
>
>
> Any feel for how many people are using redis for sessions?
>
> +1 on on redis for sessions. Use it for multiple apps although they're
all older web2py versions and python 2 setups. Upgrading one to python 3
and web2py 2.18.5 which is when the redis session functionality broke.
Jim's fi
I submitted one quite some time ago for Python 2 and python redis 2 and 3,
but not for Python 3. The fix I proposed here doesn't not work for Python
2.While I'd prefer to put sessions in redis, they work perfectly well
in the database. Redis works just fine (any version) as long as you're
Did I drop the ball on this one? I do not see a PR
On Thursday, 21 November 2019 08:22:58 UTC-8, Jim S wrote:
>
> FWIW - I gave up on waiting. Converted my sessions to use the database.
>
> I use redis in other parts of my app within web2py and am having no
> troubles there.
>
> -Jim
>
> On Fr
FWIW - I gave up on waiting. Converted my sessions to use the database.
I use redis in other parts of my app within web2py and am having no
troubles there.
-Jim
On Friday, November 15, 2019 at 1:45:25 PM UTC-6, Jim S wrote:
>
>
> Do you know much about sessions? Are the values all supposed
Do you know much about sessions? Are the values all supposed to be
binary? The .decode on the
if to_native(rtn['unique_key']) == to_native(self.unique_key):
was to convert it from binary to test against the string stored in
self.unique_key.
Also, this fix isn't working in my python 2 insta
I haven tried it, but I think the best fix is not to make the decode work
by putting the obviously missing parenthesis. The fix is simply to remove
the .decode, because to_native should take care of it, so this:
rtn = {to_native(k.decode): v for k, v in self.db.r_server.hgetall(key).
items()}
W
Nice catch, Jim the .decode() on line 192. It appears to be working as
expected now (at least in my use case).
I am going to continue testing session functionality and will report if any
further issues arise.
--
Resources:
- http://web2py.com
- http://web2py.com/book (Documentation)
- ht
I just messed with this a little bit and got past the login with python3,
web2py and redis.
Here is the code, lines 185 thru 192 of redis_session.py:
rtn = {to_native(k.decode()): v for k, v in self.db.r_server.
hgetall(key).items()}
if rtn:
if self.unique
Just wondering if there has been any further work on this.
I'm working on converting all to python 3 but this is keeping us from
moving right now.
-Jim
On Wednesday, November 6, 2019 at 5:34:16 PM UTC-6, Jim S wrote:
>
> Same here. I've reported and submitted a fix but it was rejected. It
>
Same here. I've reported and submitted a fix but it was rejected. It
worked for all of our use cases.
On Wednesday, November 6, 2019 at 9:19:07 AM UTC-6, Leonel Câmara wrote:
>
> We also got this bug when we tried to upgrade to the latest web2py and had
> to rollback. We were using the old 2.X
On Wednesday, 6 November 2019 15:19:07 UTC, Leonel Câmara wrote:
>
> We also got this bug when we tried to upgrade to the latest web2py and had
> to rollback. We were using the old 2.X pyredis and then tried with 3.X you
> get the same problem with both. Please report this issue in github, if no
We also got this bug when we tried to upgrade to the latest web2py and had
to rollback. We were using the old 2.X pyredis and then tried with 3.X you
get the same problem with both. Please report this issue in github, if no
one fixes it we will probably end up fixing it as we also need this fix.
So it appears to be a bug... adding () after the decode on line 185 of
redis_session.py appears to resolve the cause of this traceback.
However another issue has cropped up... every page hit results in a new
session key being created which means they're unusable. Changing sessions
from redis
17 matches
Mail list logo
