New version in test + blog move

The 4th version of 7Timer! has been put in test phase today (July 13). For details, please check http://y234.cn/?p=5492.

As part of the upgrade, this blog will be deprecated starting July 13, 2011, while the blog and the old archives will be preserved. The regular updates about the operation of 7Timer! will be posted on my blog instead.

7Timer domain switch to 7timer.com

Starts from Mar. 11, 2011, 7Timer begins to use 7timer.com as its primary domain. The old domain, 7timer.y234.cn, will be maintained, but users are encouraged to use the new one.

A super convenient tool to retrieve AP and APE forecast

My colleague at Guangdong Southern Cross Star Association (GSCSA), littlePig (Luo Zhenyu), just notified me that he has written a script that can allow users to click only once for AP/APE forecast. This is a wonderful tool since I have been headache over the site pointing issue for a long time. The tool is available at http://www.crux-gd.org/other/7Timer_gmap.html . The tool seems works well on all browser besides IE 7.0- and FireFox -- probably it is a problem at Google side.

I will continuously researching the idea of creating a user-friendly site database (I still need to study AJAX yet). The site searching function at 7Timer is far from good, I know.

New product released plus bug fixes

I am happy to announce that 7Timer releases new product after a 2-year pause. This new product, namely "APanel Extend" (APE), is now available for public testing.

To respond concerns on the difficulties on using APanel, APE is aiming to provide a user-friendly forecast. Many APanel elements are removed or simplified. Rainfall forecast is added to allow daily using. Also, forecast lead time is extended from72-hour to 192-hour.

Meanwhile I carry out fixes of two major bugs of the system. The bugs had caused the transparency and darkness forecast in APanel cannot be display correctly for western hemisphere users, plus some displaying error for users between 0.103W-0.000W. My apologize to all these users for a belated response to these bugs. The fixes should be effective approximately at 16:30 UTC, Jan 3.

Please note that APE is still an alpha version. Some planned designs are not yet implied as I will depart on Jan 5 for travel. Any comments or suggestions will be very welcome (by posting here or emailing me).

I will have a couple of time this spring and should have some time on 7Timer development. Stay tuned for good news!

Bug fixed

APanel drawing bug for south hemisphere and regions between 179.75E-180.00E has been fixed. Thanks to Johnson Lau for pointing out.

Great changes today

Sorry for a five-month silence, as I have to work with papers and exams -- they are essential to my future so I hope you can understand. But after all, I manage to get a week on 7Timer. Today I am glad to announce that great changes are ready, and this may be the most important upgrade for 7Timer since it borned in 2005.

First of all, I am happy to let you know that 7Timer finds a new home at Shanghai! You may recall that 7Timer was placed at Cosmostation in Beijing since it borned in 2005. Cosmostation is a server that kindly provided by the National Astronomical Observatories of China and maintained by Dr. Chenzhou Cui, to provide a free and stable web host for amateur astronomers in China. To-date, it has operated for almost nine years and served hundreds of amateur astronomers to build their own tiny home over Internet. However, as 7Timer is a bold creature, it eats up a lot of resources on Cosmostation, and I did occasionary feel sorry for Dr. Cui as well as other Cosmostation users about this over these years. This summer, Dr. Qing Lin and Mr. Haiming Tang from Shanghai Astronomical Observatory, Chinese Academy of Sciences, agree to provide a new server for 7Timer. The server is for operation of 7Timer only, so it is able to provide resources to conduct normal operation as well as some numerical astro-meteorlogical experiment, so today the transfer from Beijing to Shanghai is completed. I would like to express my big thankness to Dr. Cui for his long-term support of the 7Timer project, which is essential to the project's grow; to all users of Cosmostation for their supports; and to Dr. Lin and Mr. Tang for providing an excellent new server.

On the very first runs today, you may notice that 7Timer is times faster: the time it requires to perform a normal 384-hour package extraction and figure production is reduced from ~5-hour about one year ago to approximately 90 minutes. Also you may notice that the resolution is times higher: the resolution can be as high as 14km now, over one time higher than the old one (35km). The major part of this fantastic change is contributed by Mr. Johnson Lau (will be Dr. soon from Hong Kong, who also credited my deepest thankness. He spent days on writing an excellent distributed data processing system for 7Timer that enable the latter to process different kind of products almost in real-time with the data release at NCEP with ease, also he saves pretty much of my time by helping me on bug searches. Without Mr. Lau's effort, this fantastic change will not be possible, or at least will require me a pretty long time to do it. It is very fair to say that 7Timer benefits A LOT from his wonderful design!

Okay, I will summarize the new changes here. Ah, I am sorry that there are no new products (yeh, it is my word that there will be new product this year, but probably we have to wait until next year!) But I believe the changes below are exciting enough!

• Install a new data processing system, which significantly reduce data process time (from ~3h to ~1.5h).
• July's GFS model upgrade effective in 7Timer products, some forecast fields have resolution as high as 14km, including cloud cover, CWF, temperature, snow depth, and wind. High resolution forecast is extent to 192h; for forecast posted 192h, higher resolution forecasts are available (0.6deg or 1.0deg, comparing to the 2.5deg before today).
• New seeing model installed. The performance of the cloud cover forecast, as well as this new seeing model, is discussed on a paper that will be published on a scientific journal shortly (PASP, 123). For a preprint version, see http://adsabs.harvard.edu/abs/2010arXiv1011.3863Y.
• ``Altitude correction'' function removed as study has shown that it does not have positive effect to improve the accuracy.
• Some other misc changes (bug fixing, etc).

APanel service resumed

APanel service has been back into normal.

Emergency fix of APanel

A serious bug of the new APanel was just discovered, this bug has caused ALL forecast released since July 19 to be UNRELIABLE, and all forecast data has been removed.

I have make emergency fix of the program and things should be normal on the next update (2612Z update, should be available around 16h UT today).

My apologize goes to those who were misguided the bad forecast over the past week, and I will keep an eye if the next runs is okay for use.

About the service problems on July 18-24

Many users found that the 7Timer! service stopped updating since July 18 -- this is due to both the server and scripts upgrading, the service has been resumed on July 22. The server at the National Astronomical Observatories had been replaced with a better (and thus faster) one; while I am pleased to announce that with the precious help from Dr. Johnson Lau from Hong Kong, 7Timer! is now running in distributed-processing mode, which enable us to produce the output with less time. I will continue to revise the script code to further reduce the data process time.

But however I later received emails pointing about that the strange and unaccurate output of APanel, which allow me to discover a function switch from PHP 5.2 to 5.3, the function switch had produced unaccurate APanel output. A fix of the APanel script has been installed and will take effect since the next update (around 5h UTC today).

Thank you again for your support of 7Timer!

Server upgrade, bad news, and good news

You may have notice that 7Timer! has been running with frequent errors since July 11. I have to make apologize for this -- I intent to do a power-on upgrade of the server, however the process took much longer than usual and generate many unexpected problems, so the upgrade is not done yet. After working for two days, I barely got things back into normal - it is running okay at this moment, only with occasionary dataset missing at the Weachart section, but this is only a temporal solution.

To begin with, I am pleased to announce that 7Timer! has received a bond support from the Shanghai Astronomical Observatory, Chinese Academy of Sciences (http://www.astron.ac.cn), they kindly provided an excellent new server, which is intended to run 7Timer! applications ONLY.

However, due to some technical issues, the web application will remain at where it is placed at this moment (at a shared server at National Astronomical Observatories, Beijing). Now, with the generous help from Mr. Johnson Lau, I am working to make both servers working in a distributed ways, which can significantly speed up the process time. I hope the upgrade process can be completely done in a week. By then I will place another announcement here, and before that it is possible for other bad runs to occur.

7Timer! also received assistances from Shanghai Astronomical Observatory through out the year 2010 for new product development, so development of products such as "easyAP", "openAP" is now back on-schedule. I hope easyAP can be available for test by the end of August.

Thanks to all for your support of 7Timer!, your support is the key factor to make the project possible.