Avee's Place

Simple thoughts in a complex life

Monthly Archives: November 2009

I’m Waving!

Woohoo! Google invited me to join Google Wave Preview! Now I just need to find people to wave with.. 😉

Advertisements

Snow Leopard Unibody MBP Battery Life Test #2 : The Apple Test

Work has been crazy these few days, there’s this deadline that we’re trying to make in the office so I was having trouble allocating enough time for the perfect Apple environment test.

Moreover, my machine also has the notorious auto-sleep problem. A problem in which the notebook does not want to sleep by itself when it is left idle, even though the Energy Saver preference has been set properly. The manual sleep function works just fine though. This irritates me so much – you have to close the lid to preserve battery – that it made really want to format the machine as soon as possible. And the second battery test would not be valid if I formatted the machine.

Nevertheless I did a partial battery test using the Apple scenario. I did the test for only approximately half of the battery capacity. Here is the details of what I did. I charged the battery to a full 100%, and then used the notebook the way Apple tested them, by browsing some websites and writing some documents, the screen brightness was at 50% and so was the keyboard brightness. I used Firefox for browsing and OpenOffice.org to edit the document.

The result : my battery was down by 45% in 1 hour and 52 minutes. The average battery usage is 0.40178% per minute, which means 100% battery life would approximately equal to 249 minutes or 4 hours and 9 minutes of usage.

Hmm, that didn’t differ much than the result of the first test.. Maybe the running background processes are the ones really eating up the battery. And it’s possible that they are the ones responsible for the MBP’s auto-sleep problem. Well, we’ll see about that in the next test, after I format the notebook.

Keanehan Speedy : Speed That You Can Trust

Speed That You Can TrustHampir satu tahun yang lalu kantor kami mengganti provider koneksi internet kami dengan TelkomSpeedy. Alasan utamanya adalah biaya. Dulu waktu kami menggunakan dedicated line dari salah satu provider ternama, biaya yang harus kami keluarkan adalah sekitar 3,5 juta rupiah per bulannya. Padahal, kecepatan up/down speednya hanya 128 kbps. Dengan menggunakan Speedy, kami dapat memangkas biaya tersebut lebih dari 50% dan kami mendapatkan kecepatan yang jauh lebih cepat. Kami menggunakan 2 buah line masing-masing 1 Mbps.

Namun seperti kata pepatah bule, “you get what you pay for”, ternyata menggunakan Speedy banyak masalahnya. Misalnya saja kami tidak dapat meng-upgrade ke paket langganan kami ke paket 2 Mbps karena menurut pihak Telkom, line telpon di daerah kantor kami tidak bagus kualitasnya, selain itu ada juga masalah kecepatan koneksi yang tidak stabil, juga gangguan koneksi yang respon penanganannya tidak secepat dedicated line provider. Tapi ya memang semua itu sudah kami perhitungkan sewaktu kami memutuskan untuk pindah.

Dari seluruh permasalahan yang muncul, ada satu permasalahan yang betul-betul menarik bagi saya. Semenjak kami menggunakan Speedy, kami sering mengalami gangguan koneksi putus-nyambung, terutama pada saat Maghrib dan ketika hujan turun. Berbagai usaha kami lakukan untuk mencari penyebabnya, kami memperbaiki perkabelan telpon di dalam kantor kami, kami panggil pihak Telkom beberapa kali untuk diagnosa, kami coba konfigurasi ulang modem, tapi problem koneksi selalu berulang.

Petunjuk yang kami temukan dari proses yang melelahkan tersebut adalah bahwa Line Attenuation di tempat kami cukup besar, sehingga sinyal yang diterima oleh modem tidak bagus. Hal ini berakibat kepada modem selalu mengalami lock-loss. Dari proses tersebut juga satu-satunya hal yang berhubungan dengan hal tersebut yang kami belum benahi adalah modemnya. Namun waktu itu kami tidak memperhitungkan kemungkinan justru modemlah penyebabnya. Hal ini karena ketika kami waktu itu menggunakan dua merk modem yang terbaik yaitu Linksys dan D-Link yang harganya lebih dari 700 ribu satunya, model termahal yang ada di toko tempat kami belanja. Tentu saja dengan asumsi termahal == terbaik.

Hingga suatu saat salah satu modem kami bermasalah dan kami coba tukar dengan modem TP-Link standar yang didapat ketika pendaftaran, ternyata hasilnya koneksi stabil! Tidak terjadi lock-loss. Di saat-saat koneksi Linksys seharusnya sudah putus, modem TP-Link tetap terkoneksi, hanya pingnya saja yang menjadi sangat besar, bisa mencapai 10,000 ms!

Setelah mengetahui hal itu saya segera membuat pengajuan kepada kantor untuk membeli satu buah lagi modem TP-Link untuk mengganti modem di line lainnya. Sampai sekarang koneksi internet di kantor kami relatif stabil, meskipun koneksinya sering menjadi lambat ketika ada gangguan. FYI, harga modem TP-Link hanya sekitar 250rb.

Hal ini mengingatkan saya pada seorang kenalan dari Australia yang kena penyakit perut (saya lupa penyakit apa, disentri atau tipus ya) ketika makan bala-bala. Memang hidup di Indonesia ada seninya tersendiri.

Snow Leopard Unibody MBP Battery Life Test #1 : Real-World

I’ve finished the first of four items planned on the battery life test, the real-world usage test.

I do the test while working in the office. Since I am a developer working in Windows platform, I used the notebook to remotely connect to my development machine and see how long it would last.

I started with 100% battery, programs running were Firefox, Apple Mail, Adium, iTunes, TextEdit, and Microsoft Remote Desktop Connection. I didn’t check the display brightness level, but I assume that it was at 100% all the time because I have a big glass window behind my desk and I live in the tropics. I was using AirPort to connect to the network, and Bluetooth was disabled.

I did all the browsing, emailing, chatting, listening to music in OS X, and only coding on the remote Windows machine. I also did some downloading in iTunes. At lunch break, I streamed a movie for about twenty to thirty minutes to watch while eating. The machine went to sleep several times while I was briefly away from the desk.

I think this test could be described as “moderate wireless usage”. The Remote Desktop Connection constantly streamed some data during the test period, occasionally wireless usage went high when I downloaded something or when I streamed a movie. The CPU usage were not intensive.

The result, I got to work for 3 hours and 58 minutes, before the battery was completely drained (0%) and the machine went to hibernate by itself.

Next up is the Apple Environment test.

Snow Leopard Unibody MacBook Pro Battery Life Test

up to 7 hrs of wireless productivityApparently, my post about my suggestion on fixing the battery life problem in Snow Leopard has been generating enough traffic to make me feel compelled to do a battery life test. So here goes.

First, a short description of my machine :

  • It is a 13″ MacBook Pro, with 2.26 GHz processor and 2 GB of RAM. Logically, faster processor would consume more power for a given time and the same applies to larger amount of RAM.
  • I upgraded to Snow Leopard via the upgrade path, as advised by Apple (I read this on gizmodo.com).
  • It had all sorts of programs installed, the Applications folder had 70+ items and I also installed programs that runs as a service like SynergyKM, Growl.

I think this machine qualifies for a real world test machine.

The target of this experiment is to try to verify Apple’s claim about the notebook’s battery life, and to compare it to real-world usage scenarios. Apple claims that the battery would last up to 7 hours of “wireless productivity” (8 in 17″ MBPs). As usual, in every claim there’s the fine print. You can see in this page, in the fine print, 7 hours is achieved in a specific working condition. The keywords are in bold.

13-inch MacBook Pro testing conducted by Apple in May 2009 using preproduction 2.53GHz Intel Core 2 Duo-based MacBook Pro units. 15-inch MacBook Pro testing conducted by Apple in May 2009 using preproduction 2.8GHz Intel Core 2 Duo-based MacBook Pro units with a Better Battery Life setting. 17-inch MacBook Pro testing conducted by Apple in May 2009 using preproduction 2.8GHz Intel Core 2 Duo-based MacBook Pro units with a Better Battery Life setting. Battery life depends on configuration and use. See www.apple.com/batteries for more information. The wireless productivity test measures battery life by wirelessly browsing various websites and editing text in a word processing document with display brightness set to 50%.

There are several experiments that I plan to do. In order :

  1. Real-world usage test.
  2. The Apple Environment test.
  3. Format and do a clean install of Snow Leopard.
  4. Repeat item #1.
  5. Repeat item #2.
  6. Write a conclusion.

For anyone interested, I will update this post with links to each item, and I will try to do at least one test in a week.