#it-stuff

http://dollerkrams.de/?p=78

http://dollerkrams.de/?p=70

Beware, don't be stupid! Jangan pernah install plugins firefox ini di komputer bersama, di warnet misalnya. Cookie-cookie anda nanti juga bisa ikut disinkronisasi :))

Semenjak 2 minggu terakhir, coverage 3G operator seluler yang saya gunakan di daerah kos2an saya di Guru Mughni hilang. Terus terang, internet sudah merupakan hal wajib bagi saya. Walaupun mendapatkan unrestricted akses di kantor, tetapi di kos saya tetap membutuhkan akses internet untuk misalnya browsing situs-situs yang 'kurang pas' kalau di akses di kantor, seperti friendster, forum2 buat cari software terbaru, baik PC maupun Touch Dual, ataupun untuk, istilah Den Besar, content :)) Tetapi hal utama yang paling penting adalah, saya tidak bisa lagi nonton pertandingan Liverpool live (sudah sebulan terakhir nonton the Reds via streaming), karena bandwidth GPRS tidak cukup. Jadi, hilangnya coverage 3G cukup mengganggu saya. GPRS biasa tidak mencukupi speednya, bahkan di Guru Mughni EDGE pun tidak ada. Oleh karena itu, saya mencoba mencari solusi sementara, yaitu pindah ke provider lain :p Untuk itu, saya membutuhkan modem 3G/HSDPA lagi, karena tidak mungkin menjadikan Touch Dual saya sebagai modem, secara nomor utama mesti selalu aktif. Dalam pikiran saya : "Keluar duit lagi neh". Lalu saya teringat, saya memiliki modem pinjaman (hehehe), sayang sekali modem tersebut di lock hanya untuk jaringan operator tertentu saja. Modem-nya modem buatan Huawei, yaitu model E220. Setelah browsing beberapa lama, akhirnya saya menemukan cara buat unlock modem E220 supaya bisa dipakai di seluruh operator. Berikut langkah2 nya (disclaimer: Proses unlock modem Huawei E220 merupakan proses flashing ulang firmware dari modem. Oleh karena itu, langkah2 berikut memiliki resiko yang tinggi untuk dilakukan. Saya tidak bertanggungjawab atas apa yang terjadi pada modem anda. Saya sendiri telah mencoba, dan berhasil 100%. Pastikan anda mengikuti setiap langkah yang dijelaskan berikut untuk mencegah hal2 yang tidak diinginkan :D) 1. Pastikan aliran listrik ke PC anda 100% aman dan stabil. Disarankan menggunakan laptop, karena ada baterai untuk power cadangan seandainya listik down. 2. Keluarkan SIM card dari modem 3. Pastikan koneksi USB dari PC ke modem stabil. FYI, kabel data E220 sangat rentak rusak atau flicker koneksi 4. Download file upgrade firmware ini, dan jalankan. Ikuti instruksi yang muncul di layar. Modem anda akan connect/disconnect dengan sendirinya. Ini wajar, karena dalam proses flashing ulang, akan terjadi beberapa kali restart modem. 5. Download software upgrade ini, dan eksekusi juga. Ikuti juga instruksi yang akan muncul. 6. Uninstall software modem yang sebelumnya di install. 7. Setelah berhasil, disconnect, dan connect lagi modem anda 8. Install kembali software modem bawaan dari modem. Nama software yang baru adalah Mobile Partner, menggantikan Mobile Connect 9. That's it!! Anda tinggal mengkonfigurasikan koneksi APN sesuai dengan operator yang anda gunakan.

I'm one of the fans of open source Database Management System (DBMS), called MySQL. The main reasons why I like MySQL are: free, easy to use, stable and powerful. Some people will doubt about The latter twos, but indeed MySQL is stable and powerful. I use MySQL a lot in company, not because my company cannot afford commercial DBMS, but simply because MySQL is really a fast and dependable solution, as my department often assigned to do projects that have tight deadline. But still MySQL, yet powerful, is not like commercial DBMS like Oracle. MySQL needed to be used properly if you want to get maximum performance (although commercial DBMS too :p no I mean in simplest things). Take Indexes for example. In MySQL too we have indexes, but maybe some people haven't notice how MySQL handles indexes. To start learning about MySQL indexes, let's see below query select * from car where color='red' If you have around 100 records in car table, it's fine not to have index. But how about if you have around millions of records, or maybe billions?. Index is a must. General rule in database world is that you should have index on column which you have query state that column in where condition. Like above example, you should index on color column, moreover if you do the query a lot, and the table itself has lot of records. The other rule we should have index on column, is the column which involves in join query (mostly, in foreign key to other table). select * from car a inner join company b on a.id_company=b.id Then, you should have index on id_company column too. Wait wait wait... What you mean by 'should have index on id_company column too' ?? It means, we should have two indexes, one for color column, and the other one for id_company. Simple like that :D ? Of course no, there someother considerations on MySQL indexes. There will be also compound index, that is index in more than one column. Please note that is different the previous case (two single indexes vs compound index) - Two single indexes == We have two indexes, that is one for column A, and the other for column B (like the previous case) - One compound index == We have one index, that is applied to two or more columns, let's say column A and column B Got the differences?? Let say you have query: 1. select * from car where color='red' 2. select * from car where color='red' and id_company=14 3. select * from car where color='red' and id_company=14 and production_year=2000 A. Indexed column(s) : [color] Let's say you have index only on column color. Then the index will apply on three queries above, but only part applied to Query 2 and Query 3. Why? Because the indexed column is only color. Thus, in Query 2 case, MySQL still has to do full table scan (after applying index color) to find records that satisfy where id_company=14 statement Conclusion for Indexed column A: Query 1 --> full index use Query 2 --> part index use Query 3 --> part index use B. Indexed column(s) : [color, id_company] (compound index) Same as above analogy, the index will be fully used in Query 1 and Query 2. But for Query 3, the index will only partly use to find records, as column production_year is not indexed. There maybe will come a question regarding whether Query 1 will benefit from index color and id_company. There answer is yes, because the query will use the first part of compound index, that column color to find the records Conclusion for Indexed column B: Query 1 --> full index use Query 2 --> full index use Query 3 --> part index use C. Indexed column(s) : [color], [id_company] (two single indexes) For Query 1, surely this case will benefit from the two single indexes. But how about Query 2? Yes it will also benefit, but not as good as B. Indexed column case. Why? Because the MySQL can only use one kind of index per query. As for case C, we have two indexes, then only one index will be used. This is the one of the limitations for MySQL DBMS. The same rule apply to Query 3. Conclusion for Indexed column C: Query 1 --> full index use Query 2 --> part index use Query 3 --> part index use D. Indexed column(s) : [color, id_company, production_year] (compound indexes) This index Case, is the best solution for the queries stated above. All queries will utilize the index that Case D has. Conclusion for Indexed column D: Query 1 --> full index use Query 2 --> full index use Query 3 --> full index use Fiuhhh, so far we have Case D as the best solution. But how about we have another query to challenge?? 4. select * from car where production_year=2000 The answer? If you only have compund index on [color, id_company, production_year], then to solve query, MySQL needs to do full table scan!!! Why?? Because production_year is the last part of the compound index (the order does matter!!!). So MySQL cannot use the index. It's different from Query 1 : select * from car where color='red'. The color is the first part of compound index, and it can utilized. So the solution?? You will need additional index on column production_year. So, will need to have indexes like this - [color, id_company, production_year] - [production_year] Index on modifier function used in where The last lesson (cieh :p) that we need to learn from this session, is that MySQL is not smart enough to apply the index on modifier function used in where statement. Let's say the car table has another column, called sale_date. The sale_date is datetime SQL type. You have single index on column sale_date. . Then you have SQL query like below - select * from car where year(sale_date)='2006' Will the sale_date index be used? The answer is no!! MySQL is not smart enough to do that. The solution? Never use the modifier function on where statement if you want to make use of index. So, change the SQL query into - select * from car where sale_date between '2006-01-01' and '2006-12-31' Then the index will be fully utilized :)

to be continued ...

Thanks god, one of my tasks has been done :) I was assigned a task, to make system that can send email containing privacy data. Because of IT security policy in my company, I need to encrypt the data first. The problem is that, the receipient of the email should not have additional program, to decrypt the data. So we concluded that we can use zip compression with encryption, as current available operating systems should support ZIP format directly, although the encryption in standard ZIP format can be broken easily. At first, I thought the ZIP compression and encryption in simple ... So I just looking around the web find PHP library for that.... (yess,, it must be opensource). Nope ... i failed to my one. Then, I moved up to find out how execute zip command in Solaris, the operating system of my system. But it tells that the installed version doesnt support encryption. So I try reinstalled the supported one, but also failed as the documentation is very bad (because too old) I think. Besides, I dont have root privilege. I started to desperate ... But somehow, after browsing again, came to my mind the opensource 7zip. So I just browse the sourcecode of 7zip command line program, and compiled it (need no root privilege). And fortunately, it works!!!! Thanks God ;)