| ducky | The title of this talk is "Midas: Open Standard for |
|---|---|
| Admin/Financial applications" | |
| ducky | these are the topics for this talk |
| ducky | 1. What is Midas? |
| ducky | 2. Justification |
| ducky | 3. Midas's design base |
| ducky | 3.1 Needings |
| ducky | 3.2 Module's definition |
| ducky | 3.3 Architecture definition |
| ducky | 3.4 Components |
| ducky | 3.5 Comunication's Protocol |
| ducky | 3.6 Database |
| ducky | 3.7 Design's layers |
| ducky | 3.8 From design to implementation |
| ducky | 4. Expect ations |
| ducky | 5. Questions |
| ducky | 1. What is Midas ? |
| ducky | Midas is an open set of tecnical specifications which define a |
| generic model oriented to the Financial/Administrative Software | |
| implementation initially for comercial/or service's enterprises. | |
| ducky | Is a documentation repository which pretend a standard for the |
| FAS (Finnancial Accounting Software )tools design. | |
| ducky | The project's purpose is to encourage multiple implementations |
| or different modules written in different languages. | |
| ducky | In short, Midas ITS NOT A FINNANCIAL ADMIN SOFTWARE, Midas is |
| an open design for the FAS implementation or components of the same in | |
| different languages. | |
| ducky | 2. Justification. |
| ducky | - The existant free projects were designed specifically for |
| some countries or at least they present functionality differences (and | |
| resources comsumption) related to local needings (we already verified out). | |
| ducky | - The project's documentation about their design is very poor |
| (at least when we began the research about in the art state). Adicionally in | |
| some of the existant | |
| MJesus | jaime startlog |
| jaime | MJesus: Estoy logueando el canal #redes en [http://felix.zapto. |
| org/jaime/redes.html] | |
| ducky | ok |
| ducky | projects the information in the persistency 's mechanisms |
| (database), table's names and fields were found codified which make very | |
| complex to understand its structure and its respective relations between the | |
| tables. | |
| ducky | GNUe's case: we found extensive documentation (in English) |
| about its design but no any concrete implementation which could give us any | |
| light. Adicional, our initial idea was trying to be simple and concrete, | |
| because GNUe was very complex to assume. | |
| ducky | 3. Design's bases |
| ducky | 3.1 Needs |
| ducky | Thinking about the enviroment which it was Midas designed for, |
| we found what in Colombia, y a Latinoamerica's big part, exist serious | |
| limitations related to | |
| ducky | their infrastructure in telecomunications (common dialup |
| lines), which the digital information must be trasmited. | |
| ducky | In the case of many enterprises which have many branches or |
| have sub-enterprises of different geographical places, and their connection | |
| needing is imperative. | |
| ducky | The FAS software we can found in the actuality in the our local |
| marketing its not | |
| ducky | designed to work in efficent way under the given conditions |
| before (dialup | |
| ducky | lines / limited bandwith) |
| ducky | In our case (Colombia), in these moments theres no any free |
| solution adaptable | |
| ducky | to the enterprises needs and no any commercial solutions at low |
| price are very | |
| ducky | limited. |
| ducky | On the another side, the functionally viable solutions, are |
| commercial and their | |
| ducky | prices are too high, wich in some cases can beat the actives of |
| an small | |
| ducky | enterprise, which is absurd. |
| ducky | Some groups, have been customizing toos, like SqlLedger, (a web |
| solution, | |
| ducky | s/toos/tools/g |
| ducky | under the client/server model, where every data processing, is |
| performed at the | |
| ducky | server side), which on Intranet based enviroments, (LANs at |
| 10-100 mb/s), seems | |
| ducky | to work very well. |
| ducky | In these moments we know about some proyects at national |
| level, which are | |
| ducky | interested in fill this empty, what in present time we know |
| about free SAF | |
| ducky | software. |
| ducky | nevertheles, under our particular vision, in the SqlLedger |
| case, we didnt use to | |
| ducky | share the fact about any kind of client's request have to |
| travel, to the server | |
| ducky | once and another time |
| ducky | including data have been already transmited in early operations |
| ducky | We havent done the test, but we would like to see sqlledger |
| working | |
| ducky | trough a dialup line |
| ducky | doing an excercise where the enterprise, have an 500 |
| processe/services stock | |
| ducky | Aditionally, what would happen if i have a supermarket, with 30 |
| terminals, | |
| ducky | (486) |
| ducky | and i wanna implement a FAS system? |
| ducky | if i choose sqllegger, ill be in the duty, a graphical |
| environment | |
| ducky | in order to open a web browser |
| ducky | (sqlledger supports text-based browsers) |
| ducky | yea..we already know |
| ducky | but a text-based browser for the girl who is at the box? |
| ducky | "use LTSP and problem solved" |
| ducky | ..would say some people.. |
| ducky | but and what happen if we wanna free the central server |
| ducky | of the load of the X clients |
| ducky | its clear, in a LTSP scheme |
| ducky | concentrates the all the processing in a single point : the |
| server | |
| ducky | aditionally, we sum the resources, used by the FAS system |
| ducky | remember, we are talking about 30 online selling points |
| ducky | why cant i use terminal mode clients (ncurses,etc)? |
| ducky | if i would like the database in another server |
| ducky | ? |
| ducky | or if i already have particular database motor and i wanna use |
| in the implementation | |
| ducky | ? |
| ducky | could it be possible to offer a platform- indepent system? |
| ducky | programming language indepent |
| evanginee | does anybody have a log of the talk/discussion so far? |
| evanginee | on Midas? |
| ducky | why not to offer an FAS architecture, which let clients, and |
| servers | |
| ducky | evanginee, [http://felix.zapto.org/jaime/redes.html] |
| ducky | :) |
| evanginee | excellent, thanks ducky |
| ducky | joining to the infrastructure |
| ducky | and enteprise's needs? |
| ducky | 3.2 Module Definition |
| ducky | Analysing this needs, we began to delimit, a design what would |
| be compilant with the requests, of the existant demand | |
| ducky | in conectivity terms |
| ducky | aivability |
| ducky | hardware aivability |
| ducky | and funcionality |
| ducky | (we based on open develoment model) |
| ducky | Initially the design is composed of the following |
| ducky | modules |
| ducky | in the functional part |
| ducky | Accounting module: |
| ` | holas |
| ` | este server no se supone q debia ser japones? |
| fernand0 | no |
| ` | ah |
| ` | jo |
| ` | y de q es este server? |
| ducky | it starts on the suposed, of what the accouting should be the |
| result of the transactions done, by the software in some time | |
| ducky | This module has the capability to generate, the necesary reports |
| ducky | in order to achieve |
| ducky | the right administration on the enterprise |
| ducky | among of them, balances |
| ducky | GyP(lost, and gains) |
| ducky | Box flow |
| ducky | Official books (Major, diary, summary) |
| ducky | Stock module: |
| ducky | in charge of the management, of the client's state |
| ducky | and accounts |
| ducky | which lets client's identification |
| viva UMEET!) | |
| ducky | wholesalers |
| ducky | retailers |
| ducky | and making of their respectives stadistic resports |
| ducky | or account's status reports |
The Organizing Comittee