Upd: lib.cvsclient is not in incubator-netbeans.
I searched for a private class from that library, but couldn't find it:
https://github.com/search?q=org%3Aapache+CheckedInResponse.java&type=Code&utf8=%E2%9C%93
On 16.02.2018 15:19, Basin Ilya wrote:
> Hi Emilian.
>> I wonder, don't we have it already somewhere? Or in some library?
> I searched everywhere.
>
>> We can go all the way to Java 8
> This always surprises me. Now I can understand when an en-user app requires
> the latest java, but a library? IMO, things like Apache commons should run on
> two older versions
> of java.
>
>> Please make a GitHub PR
> Where's the official Github mirror?
>
>
> On 15.02.2018 18:20, Emilian Bold wrote:
>> The first patch is short and seems correct, the second a bit large to digest
>> so quickly.
>>
>> The idea of a generic inputstream with a timeout is cool. (Although, I
>> wonder, don't we have it already somewhere? Or in some library?)
>>
>>> I also suppose that we want to keep the java 6 level.
>>
>> We can go all the way to Java 8, no need to remain at Java 6. I guess the
>> existing code is just old(er).
>>
>> Please make a GitHub PR, it might be simpler to receive feedback.
>>
>> --emi
>>
>> -------- Original Message --------
>> On 15 February 2018 4:42 PM, Basin Ilya <basini...@gmail.com> wrote:
>>
>>> Here's the proposed patch.
>>> Changed lines formatted with Netbeans 8.2 default formatter. Copyright
>>> header not added to the new file.
>>>
>>>> Hi.
>>>> In case of wrong setup a cvs log command may print nothing. This makes the
>>>> LoggedDataInputStream#read() method in the Netbeans CVS client loop
>>>> forever. See this SO
>>>> question for details:
>>>> https://stackoverflow.com/questions/48741854/inputstream-wrapper-with-actual-read-in-a-dedicated-thread
>>>> There's a bug mentioning sleep() inside LoggedDataInputStream:
>>>> https://netbeans.org/bugzilla/show_bug.cgi?id=254761
>>>> There are two types of input streams the LoggedDataInputStream works with.
>>>> If it's a server connection, the stream is a SocketInputStream. If it's a
>>>> local connection, the
>>>> stream is a read side of a pipe returned by Process.getInputStream().
>>>> I suppose that we want to keep the Thread.interrupted() check in
>>>> LoggedDataInputStream, but using available() is definitely not the right
>>>> way to check for premature EOF.
>>>> I also suppose that we want to keep the java 6 level.
>>>> I want to but remove the sleep() and available() calls from
>>>> LoggedDataInputStream and instead catch the SocketTimeoutException and
>>>> check the interrupted status on a timeout.
>>>> In case of a socket we will set the socket timeout to 100ms.
>>>> In case of a pipe we will perform reading in a dedicated thread. See the
>>>> SO question above.
>>>>
>>>
>>>
>>>
>>> To unsubscribe, e-mail: dev-unsubscr...@netbeans.incubator.apache.org
>>> For additional commands, e-mail: dev-h...@netbeans.incubator.apache.org
>>>
>>> For further information about the NetBeans mailing lists, visit:
>>> https://cwiki.apache.org/confluence/display/NETBEANS/Mailing+lists
>>>
>>>
>>
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail: dev-unsubscr...@netbeans.incubator.apache.org
>> For additional commands, e-mail: dev-h...@netbeans.incubator.apache.org
>>
>> For further information about the NetBeans mailing lists, visit:
>> https://cwiki.apache.org/confluence/display/NETBEANS/Mailing+lists
>>
>>
>>
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@netbeans.incubator.apache.org
For additional commands, e-mail: dev-h...@netbeans.incubator.apache.org
For further information about the NetBeans mailing lists, visit:
https://cwiki.apache.org/confluence/display/NETBEANS/Mailing+lists
