Baidu SDK Leak, Nokia pCell Test, Debian LSB Drop & Spark vs Hadoop: Key Insights

Baidu SDK Security Vulnerability

Recent reports claim that Baidu's SDK contains a serious security flaw that allows attackers to upload and execute malicious programs on Android devices, effectively creating a backdoor. Users have complained that Baidu's response to the issue is slow and ineffective.

pswyjz: The bug is actually a backdoor; Baidu's fix took months and still doesn't resolve the problem. Hongri: My phone's antivirus flags Baidu Input and Tieba as worm infections, leading me to uninstall Baidu security tools. Jersey: Baidu's market share is shrinking as competitors like 360 gain users.

Nokia Plans pCell Wireless Test for Next Year

Nokia Networks announced that it will test a new wireless data system called pCell, invented by Artemis. Unlike traditional cell networks that share bandwidth among all devices, pCell creates a private signal for each device, allocating the full network bandwidth regardless of user density, potentially eliminating congestion in crowded areas.

Solosick: Similar technology has been used in large venues before; it's not entirely new. 王楠w_n: Is Nokia trying to monopolize bandwidth with this "black technology"?

Debian Abandons Linux Standard Base Compatibility

The Linux Standard Base (LSB) defines ABI standards to improve compatibility across distributions. Debian, the largest Linux distro, decided to drop LSB support, keeping only the lsb-base and lsb-release packages, citing high maintenance cost and low benefit. This raises concerns that Linux distributions may become increasingly fragmented.

EeeLo: Debian's stability is good, but the split is already happening; money shouldn't drive technology. lsstarboy: Linux development resembles a society needing a unifying belief; without it, fragmentation continues.

Will Spark Replace Hadoop?

As big‑data technologies mature, many enterprises face tool selection challenges among Hadoop, Spark, Hive, HBase, Storm, Kafka, Flume, etc. Spark is praised for high performance, developer productivity, and fault tolerance, but criticized for stability, scheduling, and Scala learning curve. Opinions vary on whether Spark will supplant Hadoop.

yybmsrs: Spark stores intermediate data in memory, writes to disk only when needed, and offers a richer abstraction than MapReduce; it will replace MR but not the entire Hadoop ecosystem. wenhq: Spark can run on HDFS or other storage and integrates with YARN; it won't replace Hadoop entirely.