July 1, 2009 at 06:24We read with fingers - 2
I'm not sure if this is interesting to the Habralyudam, but I decided to talk about what was done on www.Braille.ru , including with your invaluable help (for example, in the comments on my first topic on this topic).
decoding in title
I must say right away that not everything that you advised was done, but after much deliberation and debate with a colleague, we came to the conclusion that it is better to do as we did. All in order.
Probably, you should immediately write about what goals were pursued by releasing the second version of the converter. Among them:
1) to leave the JS library, for cross-browser display of Braille (there are still some problems with IE, but they are in the process of solving, and in general, they do not spoil the situation);
2) rewrite the kernel in order to include the rudiments of the API, which will be implemented later, to speed up the work and further simplify the addition of support for new languages;
3) make the font not only cross-browser, but also clear;
Here are three general ideas that guided us.
1. We still left typeface JS library.
Here are two screenshots, for comparison (pictures are clickable): the
first is floomby.ru/content/UoUkkV0aAk - a screenshot of the old version of the converter, it is no longer in the public;
the second - floomby.ru/content/23Lz4kI1ku - a screenshot of the new version of the converter, the one that is now on the site;
The number of requests and time decreased. And for the same phrase, the difference in page weight is almost 60 times.
The method for this was chosen quite extreme, and it carries several problems that have not yet reached the hands of, but which are solvable. If anyone is interested, I ask you to convert a phrase into the source code of the page or just look at the screenshot - floomby.ru/content/QrXo6vFuqE .
To make it possible to copy the converted text, a switch is made (arrow to the right of the text).
2. The core has been rewritten.
The new core is morally and physically ready to support the API, convenient output to the same PDF, and it works faster. Plus, the beginnings of reverse conversion and several mobile services (well, not necessarily mobile, but it will be more convenient from a mobile phone).
3. Clear font.
Someone in the comments on the previous topic correctly noted that the font is blurry even with a small increase. We tried to resolve this issue. That's what happened - floomby.ru/content/IPn74u6W0S - quite clearly, it seems.
4. Other.
With the wonderful design proposed by phenomen , due to the fact that the look of the site itself has changed a bit after abandoning JS. It is unfortunate that as soon as the appearance is settled down, I will try to ask to update the layout, I hope it will not.
There is a lot of work to do with punctuation marks and various dashes, minuses and more. The fact is that Braille is divided into two alphabets, six-point, and eight-point.
Six points cover the vast majority of the interests of blind people, therefore we use only it. There are no long dashes in the books for the blind, etc. specific characters. All this will need to be translated into something, into which - it is not yet clear. The All-Russian Association of the Blind gave a pair of telephones, work is ongoing on them, there will be meetings the other day.
Oh yes, the eight-point alphabet is used for the exact sciences. And as an impression of going to VOS, the Braille display works with both six and eight points.
Thanks again for the tips and bug reports.
decoding in title
I must say right away that not everything that you advised was done, but after much deliberation and debate with a colleague, we came to the conclusion that it is better to do as we did. All in order.
Probably, you should immediately write about what goals were pursued by releasing the second version of the converter. Among them:
1) to leave the JS library, for cross-browser display of Braille (there are still some problems with IE, but they are in the process of solving, and in general, they do not spoil the situation);
2) rewrite the kernel in order to include the rudiments of the API, which will be implemented later, to speed up the work and further simplify the addition of support for new languages;
3) make the font not only cross-browser, but also clear;
Here are three general ideas that guided us.
1. We still left typeface JS library.
Here are two screenshots, for comparison (pictures are clickable): the
first is floomby.ru/content/UoUkkV0aAk - a screenshot of the old version of the converter, it is no longer in the public;
the second - floomby.ru/content/23Lz4kI1ku - a screenshot of the new version of the converter, the one that is now on the site;
The number of requests and time decreased. And for the same phrase, the difference in page weight is almost 60 times.
The method for this was chosen quite extreme, and it carries several problems that have not yet reached the hands of, but which are solvable. If anyone is interested, I ask you to convert a phrase into the source code of the page or just look at the screenshot - floomby.ru/content/QrXo6vFuqE .
To make it possible to copy the converted text, a switch is made (arrow to the right of the text).
2. The core has been rewritten.
The new core is morally and physically ready to support the API, convenient output to the same PDF, and it works faster. Plus, the beginnings of reverse conversion and several mobile services (well, not necessarily mobile, but it will be more convenient from a mobile phone).
3. Clear font.
Someone in the comments on the previous topic correctly noted that the font is blurry even with a small increase. We tried to resolve this issue. That's what happened - floomby.ru/content/IPn74u6W0S - quite clearly, it seems.
4. Other.
With the wonderful design proposed by phenomen , due to the fact that the look of the site itself has changed a bit after abandoning JS. It is unfortunate that as soon as the appearance is settled down, I will try to ask to update the layout, I hope it will not.
There is a lot of work to do with punctuation marks and various dashes, minuses and more. The fact is that Braille is divided into two alphabets, six-point, and eight-point.
Six points cover the vast majority of the interests of blind people, therefore we use only it. There are no long dashes in the books for the blind, etc. specific characters. All this will need to be translated into something, into which - it is not yet clear. The All-Russian Association of the Blind gave a pair of telephones, work is ongoing on them, there will be meetings the other day.
Oh yes, the eight-point alphabet is used for the exact sciences. And as an impression of going to VOS, the Braille display works with both six and eight points.
Thanks again for the tips and bug reports.