Hi Sergey.
The frames are just bytes delineated by FFD8 and FFD9. We do have attach 
headers with Content-Length. I tried both sending headers and images as a 
single chunk and "first header then image". Both work with old server, but 
neither with new.
Unfortunately I can't send real code, but all it is is an mg_write. Like I 
said, if I send a single frame (with header) and then return, the browser 
client displays it just fine. When I send continuous frames, it pukes.

Greg

On Friday, August 29, 2014 10:47:30 AM UTC-7, Sergey Lyubka wrote:
>
> Hi Greg,
> you're mentioning frames - what frames are that?
>
> Are you talking about websocket frames?
> If not, do you set specify content length header?
>
> If you share your code (or high-level pseudo code), that could help too.
>
>
>
> On Fri, Aug 29, 2014 at 6:45 PM, gregory kosinovsky <[email protected] 
> <javascript:>> wrote:
>
>> Hi Sergey.
>>
>> The main problem I am trying to overcome is that the server does appear 
>> to block when multiple frames are sent to the server. If I return after 
>> sending 
>> a single frame, a client receives and displays that frame. However, if I 
>> send multiple frames, client gets nothing. I am just using mg_write -- 
>> which is what I used with the old version of the server without a problem. 
>> Using mg_send and mg_send_data has the same behavior.
>>
>> Greg
>>
>>
>> On Thursday, August 28, 2014 11:02:26 PM UTC-7, Sergey Lyubka wrote:
>>
>>> Hi Gregory!
>>> I believe you're talking about mg_send_data() function, not mg_send().
>>> mg_send() returns either 0 or the 3rd parameter, number of bytes
>>> mg_send_data() returns the size of output IO buffer.
>>>
>>> Both functions do not send data to the socket. mg_printf() and 
>>> mg_printf_data() are the same in this respect by the way.
>>>
>>> These functions only append the data to the output buffer, and mongoose 
>>> sends that data later on. That is by design, and the reason is that these 
>>> functions must not block the execution. Mongoose is asynchronous and 
>>> non-blocking.
>>>
>>> Hope that helps.
>>>
>>>
>>> On Fri, Aug 29, 2014 at 2:15 AM, gregory kosinovsky <[email protected]> 
>>> wrote:
>>>
>>>> It appears that the problem I saw earlier comes from the fact that -- 
>>>> for multiple calls to mg_write -- the last return value reflects the SUM 
>>>> of 
>>>> all previously written sizes. This indicates that the server is not 
>>>> sending 
>>>> the data as it arrives, but is buffering it. What would be the cause of 
>>>> such behavior and how would this be fixed?
>>>>
>>>> Thank you,
>>>>
>>>> Greg
>>>>
>>>> -- 
>>>> You received this message because you are subscribed to the Google 
>>>> Groups "mongoose-users" group.
>>>> To unsubscribe from this group and stop receiving emails from it, send 
>>>> an email to [email protected].
>>>> To post to this group, send email to [email protected].
>>>>
>>>> Visit this group at http://groups.google.com/group/mongoose-users.
>>>> For more options, visit https://groups.google.com/d/optout.
>>>>
>>>
>>>  -- 
>> You received this message because you are subscribed to the Google Groups 
>> "mongoose-users" group.
>> To unsubscribe from this group and stop receiving emails from it, send an 
>> email to [email protected] <javascript:>.
>> To post to this group, send email to [email protected] 
>> <javascript:>.
>> Visit this group at http://groups.google.com/group/mongoose-users.
>> For more options, visit https://groups.google.com/d/optout.
>>
>
>

-- 
You received this message because you are subscribed to the Google Groups 
"mongoose-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To post to this group, send email to [email protected].
Visit this group at http://groups.google.com/group/mongoose-users.
For more options, visit https://groups.google.com/d/optout.

Reply via email to